Narada Battery Charging Axpert Settings

[Kalito]

I have setup new batteries for my solar system, Narada 12NDT200 AGMs. The attached spec sheet gives the following key:

• Max Operating temp: 55degC
• Recommended Operating temp: 25degC
• Max Charging current: 50A
• Bulk charge range: 56.4 - 57.6V
• Float charge: 54V

There is no temp coefficient so I have assumed -30mV/degC suggesting that I need to reduce the lower Bulk and float by 0.6V as the ambient temp is generally 25-30degC this rainy season and hence setup the Axpert as follows:

Charging Current: 20A (10% of battery capacity)

• Float: 53.4V
• Bulk: 55.8V

What are your comments?

12NDT200_NARADA.PDF

[Chris Hobson]

I like your thinking and would charge at the voltages as you have suggested. These Naradas are rated at max charging current of 0.25C so you could charge at 50A. Some AGM manufacturers want you to charge at this high rate. No one has managed to explain to me why. With a 20A charging rate you may not have enough "solar day" (depending on discharge). I also had 260 Ah AGMs and ended up charging them at 30A which was fine. I started charging at 20A but found  that 30A suited my "load mix" better.

 

Remember to adjust your charging voltages back up for your cool season.

[Kalito]

Thanx for your indight. All this thinking is a glimpse into your inspiration - lol.

I take your good point on the charging. My load is fairly low under 1kW. I am adding two banks for low DoD. I should be effectively off-grid. 

Battery balancer next 

[Chris Hobson]

4 minutes ago, Kalito said:

. Battery balancer next 

Speak to @Chris-R

[Kalito]

1. How does a double 48V battery bank operate in terms of runtime? Is it double the single bank?

2. Let us say I want to keep a 30% DoD, from the battery spec sheet using Voltage/SOC curve its roughly 50.5V (12.625V/battery). My float charge is at 53.4V. Looking at the battery spec sheet on watts per cell of 50W at terminal voltage per cell of 1.75V the batteries should run for 8hrs. On a two bank system as is the case do I expect to see 16hrs or close?

3. This morning was a good test day with overcast skies and rain. The Inverter is set to SOL priority so it drew extra juice from the batteries. The draw was 6A (fridges running) and the battery voltage at 51.2V this later improved to 51.3V as the amps dropped to 3A and finally to 52.2V as the rain stopped and enough solar was being generated for the load and the bank was being charged with excess. Solar was charging at 5A with a panels voltage of 104V. The inverter was pulling 688W (load at 329W). I also observed what I know to be the day’s highest load when the 1HP water pump ran. The load was 1.44kW and extra juice was extracted from the batteries at 8A (solar at this point was maxing at 1.09kW). Is it correct to assume that at this draw the bank can run about a day? The spec sheet shows 24hrs at 8.7A with terminal voltage of 1.85V/cell.

[___]

On 12/20/2017 at 12:10 PM, Kalito said:

How does a double 48V battery bank operate in terms of runtime? Is it double the single bank?

Two strings in parallel? Depends on the kinds of loads you were doing. If you did about a C10 rate on the single string, then on two strings you're now doing a C20 rate. At C20 the combined battery has more usable capacity than at C10. So technically it will be a little bit more than twice... but it depends. If you were already doing low discharge rates, you'll get about double.

Parallel strings are not a problem on discharge. They contribute proportionally to their inherent capacities. It's on recharge that you run into trouble (one overcharges and the other undercharges).

 

[Kalito]

21 hours ago, plonkster said:

Two strings in parallel? Depends on the kinds of loads you were doing. If you did about a C10 rate on the single string, then on two strings you're now doing a C20 rate. At C20 the combined battery has more usable capacity than at C10. So technically it will be a little bit more than twice... but it depends. If you were already doing low discharge rates, you'll get about double.

Thanks that makes sense. Indeed, its a 48V system so two strings of 4x200Ah each.

I did a test drive and switched off utility to test run-time. By the time solar generation stops the batteries will have some juice eaten into so around 52V. So I am off-grid and I had a load of around 300W. The interesting observation is that the Mecer battery icon shows half the bars. The Inverter has been set to 49V as the low batt level. The inverter showed the amp draw was around 5A. I wokeup to a low battery alarm about 05:30hrs. this is a period of about 11hrs. What is interesting though is that the display showed about 50.1V. The question is therefore why it was alarming at that level when it is set to alarm at 49V. The default inverter setting is 42V I think. The inverter actually switched off supply. Solar was charging and the batt level was 50.2V thereabouts. The Inverter return voltage is set to 55V. Perhaps my setting should be SBU and not SOL. @Chris Hobson you want to take a stab at this observation as to why its going to low batt alarm at 50.5 which is like 30% SOC?

Solar generation starts about 05:30hrs though under 100W but by 7am it was doing 1kW and the charger was charging at 13A which later increased to 18A. I am really utilising my 12x array.

21 hours ago, plonkster said:

Parallel strings are not a problem on discharge. They contribute proportionally to their inherent capacities. It's on recharge that you run into trouble (one overcharges and the other undercharges).

@plonkster any suggestions on mitigating the under/overcharging?

[Chris Hobson]

4 minutes ago, Kalito said:

@Chris Hobson you want to take a stab at this observation as to why its going to low batt alarm at 50.5 which is like 30% SOC?

Your low battery alarm sounds 2 or 3 V about what ever you have as your Low Dc Cut off voltage (Program 29). My recommendation is to leave this at the  default value of 42V. Yes 42V and your batteries are dead but changing this setting plays havoc with the SOC of the Axpert. If you find it annoying then switch the alarms off and leave Program 29 at some useful value. Since the SOC determination of the Axpert is so poor I have handed control of switching the load on and off to a BMV 702 battery monitor.

50.5V is hardy discharged. It is somewhere between 80 and 100% SOC.

 

[___]

1 hour ago, Chris Hobson said:

My recommendation is to leave this at the  default value of 42V.

The battery low setting is a safety setting. It is supposed to be deliberately low, somewhere between 10.5V and 11V for each 12V block. This kicks in when things are so far past gone that you have to switch off immediately to avoid damage. A battery at 49V is healthy and there is no reason to man your battle stations.

Estimating your SOC based on voltage is error prone. It depends on your load, the age of the batteries, and even ambient temperature to some extent. What I've done in the past is to record the voltage where the batteries are just before the inverter goes to passthru, as well as where it settles a few hours afterwards. Those (open-circuit... or almost) values can then be looked up against the tables the manufacturer gives you, to get an estimated SoC, and you can then build your own sort of map so that you know, for example, that "at 500W, my batteries pull down to 50V, but that is really equivalent to 52V open circuit, and therefore 70%" (I sucked those numbers from my thumb). And then with some testing and experience you can pick the right voltage to go back to the grid, so that at the usual load you do at that time of the day the voltage will roughly correspond to another open circuit voltage which again roughly corresponds to the SoC you are interested in. In other words, it is very inaccurate.

And then as your batteries age, they will draw down lower, and the entire SoC mapping exercise will need to be done again.

(Boy am I glad I have a setup with proper non-hacked-on SoC).

2 hours ago, Kalito said:

any suggestions on mitigating the under/overcharging?

This is somewhat of a holy grail thing. If you tell us... well don't tell us... just sell that thing and make lots of money.

That is to say... no... such a device doesn't exist. When you have two parallel paths, current takes the path of least resistance, or more specifically, if I can quote some old poetry: Wide is the gate and broad is the way that leads to the left battery bank, and many electrons go that way, narrow is the gate and difficult the way to the right battery bank and few electrons find it...

The problem is, the better batteries suck up most of the charge and overcharge (leading to faster wear), the worse batteries take less of the charge (and wear faster due to sulphation). The only solution at the moment is to make sure you always slightly overcharge the combined bank to ensure that the weaker string (whichever one it is) gets enough charge. It shortens the life of the other one a little, but overall that givers you more time than the alternative (where the weaker string sulphates).

The only way to balance currents is to artificially modify the resistance of the paths with extremely low resistance values. Trouble is the range is somewhat problematic, they have to be larger than the small resistances of something like a current shunt, but still small enough not to cause silly voltage drops under high currents. And then you need some kind of regulatory machinery... it comes down to whether it is worth it to go to all that complexity... or just do what you should have done in the first place: Don't have parallel strings, or have as few as possible :-)

So basically, if you have two strings, make sure the total cable lengths are the same, the batteries are a similar age, they live in similar conditions, slightly overcharge always to be sure... and then just live with the slight reduction in battery life. Last paper I read on that says you could get as much as 90% of the original life out of the combined bank.

However... what really makes or breaks it is how deep you discharge it. Shallow discharges and parallel strings...

Well, just read the paper yourself.

 

[Kalito]

 

2 hours ago, Chris Hobson said:

Your low battery alarm sounds 2 or 3 V about what ever you have as your Low Dc Cut off voltage (Program 29). My recommendation is to leave this at the  default value of 42V. Yes 42V and your batteries are dead but changing this setting plays havoc with the SOC of the Axpert. If you find it annoying then switch the alarms off and leave Program 29 at some useful value. Since the SOC determination of the Axpert is so poor I have handed control of switching the load on and off to a BMV 702 battery monitor.

50.5V is hardy discharged. It is somewhere between 80 and 100% SOC.

Oh pure genius! Prog 29 is set to 48V. I now understand missed the confusion of Axpert SOC when Prog 29 is customised. I am learning lots   Seeing that I am unlikely to hit 42 under current setup I can take to 42V to make Axpert happy

You have brought up another issue I am confused about. - The BMV. There is a local distributor of this: http://www.midnitesolar.com/productPhoto.php?product_ID=444&productCatName=Battery Accessories&productCat_ID=39&sortOrder=2&act=p no clue about the monitors appreciate suggestion of what is recommended in my setup to enhance life of batt. 

So your BMV monitors SOC? What do you do for balancing?

[Kalito]

On 2017/12/18 at 7:58 PM, Kalito said:

I have setup new batteries for my solar system, Narada 12NDT200 AGMs. The attached spec sheet gives the following key:

• Max Operating temp: 55degC
• Recommended Operating temp: 25degC
• Max Charging current: 50A
• Bulk charge range: 56.4 - 57.6V
• Float charge: 54V

There is no temp coefficient so I have assumed -30mV/degC suggesting that I need to reduce the lower Bulk and float by 0.6V as the ambient temp is generally 25-30degC this rainy season and hence setup the Axpert as follows:

Charging Current: 20A (10% of battery capacity)

• Float: 53.4V
• Bulk: 55.8V

What are your comments?

12NDT200_NARADA.PDF

I have been away from the site for some time and forgot to correct this post. Actually the average temp compensation would be -3mV/Cell/Degree giving us -0.036V compensation for 5 degree increase in temperature above 25 degrees:

• Float: 53.9V
• Bulk: 57.6V

@Chris Hobson I have embraced SBU and saved alot on power over here I do not know why I was on SOL. The system runs ok drawing at most 12A at night (I can not resist the 25A with a 10,000BTU Samsung AC on for 1hr at night) and fully utilises the PV array during the day to run daytime loads and charge the bank. I have seen the charging current hit 33A from about mid morning. Absorption is indicated as 4hrs. Does this mean this must be sustained for that period?

[Coulomb]

7 hours ago, Kalito said:

Absorption is indicated as 4hrs. Does this mean this must be sustained for that period?

Not continuously, you just need to rack up 4 hours total. That seems a long time. 

With unpatched firmware, any time over the float voltage counts, but that is wrong. With the patched firmware, only time near the absorb / CV setting voltage counts towards your absorb time, as it should. This is the second charge bug.

With factory firmware, you get this bug if using timed absorb, but the premature float bug if using automatic (not timed) absorb. So you can't get your battery properly charged either way. 

[Kalito]

This makes alot of sense thanks. I am upgrading the firmware today with 73c. 

Attached the batt charging curves.

[Chris Hobson]

18 hours ago, Kalito said:

 

@Chris Hobson I have embraced SBU and saved alot on power over here I do not know why I was on SOL. The system runs ok drawing at most 12A at night (I can not resist the 25A with a 10,000BTU Samsung AC on for 1hr at night) and fully utilises the PV array during the day to run daytime loads and charge the bank. I have seen the charging current hit 33A from about mid morning. Absorption is indicated as 4hrs. Does this mean this must be sustained for that period?

It depends what you want to  achieve. These are standby batteries and so will not last with heavy cycling. With SBU you will make you system much more efficient.  You have noticed you are fully utilising your PV run loads and charging the bank. This is how I operate. In you case it does come with a cost in that your batteries would last longer in standby mode. You need to choose "Do I save my batteries for power outages or do I cycle them save more electricity and have a shorter lifespan for the batteries and replace them with batteries more suitable for my application when they start failing".

[Coulomb]

On 2017/12/22 at 7:16 PM, Kalito said:

 

@plonkster suggestions on mitigating the under/overcharging?

There is one easy technique for two strings: use diagonal connections. So take the positive to the inverter from one string of batteries, and negative to the inverter from the other string. That way, the voltage drops tend to even out, and you don't need ridiculously thick cables. 

For four strings, you can do a sort of hierarchical version of this. Connect each pair of strings diagonally, and diagonally each string within its pair.

[___]

6 hours ago, Coulomb said:

diagonal connections

Even that only helps while the batteries fairly new are more or less identical in behaviour. The moment one of them starts to stray and the internal resistance of that one cell/battery gets closer to the cable resistance (the one you so carefully balanced), then all bets are off, that one string is perpetually undercharged.

I have thought about it many times. The traditional way to force current sharing is to put a resistor in both lines, one that creates a voltage drop that dwarves the one caused by the internal resistance, but of course you don't want to do this with batteries, that would be ridiculous. Ideally you want an adjustable high-current low-resistance device so you can steer the current. The more I thought about it, the more I realised what you really need is independent operation (each bank has its own charger, and inverter, paralleling on the AC side) and before that went much further I got new batteries and that was that :-)

[Kalito]

8 hours ago, Chris Hobson said:

You need to choose "Do I save my batteries for power outages or do I cycle them save more electricity and have a shorter lifespan for the batteries and replace them with batteries more suitable for my application when they start failing".

Indeed the choice is save some bucks but use the “Life-line” tariff (200kWh) priced at a generous $0.01/kWh currently. 5-10kwh daily comes from utility for the gysers, microwave. So failing batt is not a biggie at the moment as the road leads to Lithium.

Speaking of which, aren’t Pylontechs not recommended for cycle use? I remember reading somewhere to that effect. Whats your experience having migrated from LA to them?

[Abdul Basit]

I have Narada 48NPFC100 Battery  when i connect inverter it trips and shut down then i restart it then same happens after repeating this process about 4 or 5 time it starts working 

Experts tell me the reason and solutions please

[Coulomb]

4 hours ago, Abdul Basit said:

when i connect inverter it trips and shut down...

Is there a fault code?

Or does it just blank the display and drop all the relays?