Charging matters

[Bush Man]

Good evening to all!

I hope that in time I'll learn enough so that I can contribute as well.

With lots of input from the forum, my off-grid system on a farm in Limpopo as commissioned in September 2016 is as follows:

1. 6 x Enersol 255W panels in two strings of 3 each, installed at a 10% north facing angle.  

2. Microcare 3 String combiner box (currently using only 2 strings)

3. Axpert 5KVA inverter - Float at 54V, Bulk at 56.4V, cut-out at 48V, Max charging current 40Amps (being 13% of 300Ah). 

4. 8 x Omnipower 180 Batteries, which are actually 150Ah at C/10, giving me a 48V 300Ah system.

Load is as follows:

1. Fridge freezer - 600W per 24h  (as tested over 3 days in December)

2. Two bar fridges on occasion - jointly 600W per 24h (as tested over 3 days in December)

3. Microwave 1500 Watts - only occasionally for a short period 

4. Dishwasher 1100 Wh per cycle - only on sunny mornings to draw directly off the panels. 

5. A couple of LED globes for lighting

5 800W pressure pump occasionally when the gravity feed water system isn't operational.

The place is mostly used on weekends only, with only the fridge-freezer running permanently.

 

I have 2 questions to start off with:

1. The maximum charging current is set at 40 Amps as per above.  How many watts would that be?

2. For the first couple of months the system was working as expected.  It charged at around 56.4V, and stabilised at 54V before mid day.  It was very rare to see it below 50V. In December it was necessary once to allow the system to go down to 46V with the pressure pump working overtime during an extended overcast spell.  Since then I have rarely seen a charging voltage above 54V, it typically stabilises at around 52V and it is often below 50V.  The Axpert settings are still as before. Could such a single deep discharge damage the batteries, or should I look for something else?

Will appreciate your comments!

 

 

[Don]

Hi Bush Man, welcome to the forum. 

Answer to your first question = 40 amps charging is in the region of 2000 Watts, depending on the voltage. 

   - In my personal opinion, it is not the one time in December that your pressure pump ran over an extended period that caused damage to your system. You have been hammering your batteries from day one.

   - You have insufficient number of panels to fully charge your batteries on a daily basis. 

   - You have insufficient battery capacity to cater for your load.

You have 6 x 255 Watt panels = 1530 Watt. On a clear day without clouds you would be lucky to get say 90% peak performance out of your panels = 1377 Watt. The panels will struggle to charge your batteries, even without any load. You need about 6 hours of sun, just to charge your batteries. If you have a few cloudy days in a row, it will be even worse.

Secondly, even with just running the fridge freezer at 600 Watts, that load is too much to cater for your batteries to run 12 hours from when the sun sets until the next morning. Then you add in the load from the pressure pump and the dishwasher and the bar fridges. You will struggle just to even fully charge your batteries during the day with only the fridge running.

You have 300 Ah batteries. Remember, you don't want to drain them more than say 70% state of charge. That means you only have 90 Ah available for about 12 hours while there is no sun. That means you can only draw 7.5 amps out of your batteries during that time. Your fridge freezer alone is drawing 12 amps.

You have set your charging current to 40 amps, you can set it to 60 amps, it will make no difference, as you can only charge your batteries at 25 amps at best with your panels (with the fridge in idle state), as you are off grid with no AC grid power available.

[Bush Man]

Thanks, Don, much appreciated!

My thinking in setting up the system was as follows:

The Fridge-Freezer (A++ Rating) draws 600 Watts over a 24 hour period (226kWh/year), and the two bar fridges together do the same.  1200 Watt-hours are thus withdrawn over a 24 hour period.

The panels can put back say 1300 watts for 6 hours, i.e. 7800 Watt-hours.  My reasoning was that 1200 Watt-hours daily draw as opposed to 7800 Watt-hours daily supply should leave more than sufficient spare capacity for the other occasional uses, particularly as the Axpert allows the dishwasher (used on sunny mornings only) to be supplied directly from the panels.

Where lies the error in my reasoning?

 

[Chris Hobson]

1 hour ago, Bush Man said:

The panels can put back say 1300 watts for 6 hours, i.e. 7800 Watt-hours.  My reasoning was that 1200 Watt-hours daily draw as opposed to 7800 Watt-hours daily supply should leave more than sufficient spare capacity for the other occasional uses, particularly as the Axpert allows the dishwasher (used on sunny mornings only) to be supplied directly from the panels.

Where lies the error in my reasoning?

 

Hi Bushman

Welcome

ON PV Output (on which 3/4 of the systems are Axperts - therefore good to compare to) the average panel efficiency is about 3kWh/kW. So with 1530W of panels your system will produce an average of 4590W. If your install is good with few inefficiencies you would perhaps get about 5500W so well less than the theoretical values used as a yard stick within the industry.

The second thing to bear in mind that Amps in to your battery is not Amps out - there is a 15% inefficiency. I have seen figures as high as 30% but 15% is a common figure quoted.

I think you meant to say 600Wh for the fridge freezer. It does not look like your loads are too high for your battery bank. You have roughly 4000Wh of storage if you use 90Ah but Don is right your battery bank appears to be being chronically undercharged.On any day that your dishwasher is used you have an energy deficit and I suspect on the days that the lift pump is used too.

My remedy

1. Get hold of a gennie and give your system a good charge until the middle LED on your inverter stays on and stays on for 3 hours or so. That will get you batteries into better shape than they are now. (Just as an aside the Axpert firmware 52.30 is problematic with gennies. Scroll through your Axpert LCD display and U1 is your firmware version).
2. Add one more string of panels (preferably of the same make and size). This will ensure that your batteries float properly.

[Don]

2 hours ago, Bush Man said:

The Fridge-Freezer (A++ Rating) draws 600 Watts over a 24 hour period

That was a bit confusing to me, because watt is an instantaneous reading, not over time. So I read it as: if you take a reading any time during a 24 hour period, the Fridge-Freezer will be drawing 600 watts of power, which is a lot of power for a A++ Fridge-Freezer.

2 hours ago, Bush Man said:

(226kWh/year)

That makes more sense to me, as kWh is a measurement over time.

[Don]

14 minutes ago, Chris Hobson said:

I think you meant to say 600Wh for the fridge freezer.

Thank you Chris, I came to the same conclusion if you look at the annual consumption of 226 kWh, 620 Wh per day. 

[Don]

Bush Man, I would also recommend that you determine your base load of your system over a few days and not just look at one or two items like a Fridge-Freezer. You might have a TV, DSTV decoder, Laptop and charger, ADSL modem, cell phone chargers, 1 or 2 security lights on during the night, wife=hairdryer, etc. All these "insignificant" items add up at the end of the day. 

[Guest]

3 hours ago, Bush Man said:

The Fridge-Freezer (A++ Rating) draws 600 Watts over a 24 hour period (226kWh/year), and the two bar fridges together do the same.  1200 Watt-hours are thus withdrawn over a 24 hour period.

That is a average of 25w for the fridge and 50w for the 2 bar fridges = 1800Wh per day.

400w of panels is needed to recharge the batteries in that time in 5.5 sun hours.

Taking the 8 x Omnipower 180 Batteries, at face value, i.e. c20 of 180ah for the calc, you need min 300Ah bank to power the load for 24hours with one days backup to 20% DOD. 

The rest of the load is negligible as the microwave is in minutes, the dishwasher daytime if used and the pump occasionally.

To discharge batteries for weekend use with a week to recharge is perfectly acceptable. 

What am I missing guys?

 

[Chris Hobson]

1 hour ago, The Terrible Triplett said:

What am I missing guys?

Hi TTT

I don't like the multiply by 5.5 calculation

There are many more systems on PV Output now and I would say 3 to 3.5kWh/kW is realistic.

I think Don is correct in that there are loads on the system not being taken into account. If the loads are truly as described the battery voltage should have recovered by now. Omnipower are good VRLA batteries and one deep discharge should have done no long term harm.

Dishwasher takes a big chunk out of your daily production.

[Don]

I think @PaulF007 will be the best to comment as he has a very similar setup. He is also off-grid, has 1 Axpert 5kVa inverter, 360 Ah battery bank and has a very low base around 150 Watts. The main difference is that Paul has 12 panel (3000 watt), which is close to double the capacity that Bush Man has. I am sure on cloudy days, even Paul will struggle to get his batteries fully charged, not sure?

I just have a feeling that if you have a few cloudy days and you add to that the pressure pump running at the same time, Bush Man will end up with a deficit as @Chris Hobsonalluded to. Say he can only get them to say 85-95% SOC by end of day when the sun sets, he has a problem. The next day it might be totally overcast and rain the whole day. That will just kill the batteries.

2 hours ago, Chris Hobson said:

Get hold of a gennie and give your system a good charge until the middle LED on your inverter stays on and stays on for 3 hours or so. That will get you batteries into better shape than they are now.

That is the only way you will be able to get them fully charged again. 

[DeepBass9]

I was running with 1800W of panels off grid for 18 months, with a similar setup and it was fine, but generator backup is essential as you are pushing the limit of what the system can do. A few hours generator charging on less than sunny days makes a world of difference. Once the SOC gets behind, its takes full sun to get it back to where it should be. Also I was not running fridges overnight, just 12 hours during the day. I have 300AH @ 48V batteries. My batteries would get quite well charged, but would not get to float too often. I have now upgraded to 3kW of panels and they float every day now pretty much, but the generator is still there for completely overcast days.

[PaulF007]

@Bush Man as the guys said your base load is quite important especially when you have a fairly small system. I would not even look at the dishwasher with that amount of panels and even with my 3kw array on cloudy days we use as little as possible to assist with the charging. 
For me the most important part was to log my data (Emoncms works perfectly) and check exactly what is running when. At one stage I had everything off and would sit with TV and look at the live data and switch one thing at a time on and note the load. Then I would decide whether it is useful to have it running 24/7 or not. It is amazing what every little thing you add does to your base load. 

Seeing that you only use the system on weekends there should be nou reason that your batts should not be full by the weakened but the one thing I am missing is the fact that there is no proper battery monitor so you don't really have an idea as to what your batteries went through and I for one would recomend looking at a BMV.
Also I would say you need to get the batts tested and make sure you don't have maybe one bad cell  as that could also draw down your bank.

 

This image is just to show how much difference there is between the BMV and the Inverter's reading of the state of charge

This was on of the bad weather days when that bats did not get to full charge but the inverter said it was full. Some else on the forum there is a whole tread on the subject.

Also I agree with @Chris Hobson that either get more panels or a genny to charge the batts full if need be.

[Bush Man]

Thanks, guys - a wealth of information as always!

Another string of 3 panels is definitely an option, and not that expensive.

Before I go there, however, I would like to see if I can't remove the potential culprit from the equation, i.e the pressure pump.  The normal water gravity supply system is also solar dependant (not my solar).  The result is that it often runs dry on overcast days, at which point I need to fall back on my backup storage and the pressure pump.  Being overcast, this puts strain on the batteries, causing a deeper discharge than originally intended, with the batteries then struggling to recover.  I'll see if I can move my backup storage up the slope to create my own gravity feed, and do away with the pressure pump.

A point of interest, during overcast weather the six panels still generate 200-300W worth of PV input for most of the day according to the Axpert reading.  Taken over 6 hours, this equates to 1200 - 1800Wh, which should be more than sufficient to run at least the fridge freezer and some LED lights indefinitely if I'm not mistaken?

Don mentioned in his first post that in order to get to 40 Amps charging current I would need to get to around 2000 Watt of panel input.  How is this calculated?

[Don]

1 hour ago, Bush Man said:

Don mentioned in his first post that in order to get to 40 Amps charging current I would need to get to around 2000 Watt of panel input.  How is this calculated?

W = Volts x amps

W = 50 x 40

    = 2000 watts

I just used 50 volts as a round figure. You have a 48 volt system, guys would normally use that figure. It is still around 2000 watts. Then if you try and cater for all the inefficiencies in the system, you probably need to add another 25%. Then you are looking at 2500 watts of panels to get to charge your batteries at 40 amps, with no load on the system. If you know what your base load is and add that as well, you will probably get closer to 3000 watt panel requirement.

[Bush Man]

Before I forget, the batteries are indoors in a well ventilated cupboard.  Also during December I fitted a contact switch 12V alarm to the cupboard door.  Apart from the contact switch it consists of a 12V siren and a small electronic timer for the siren (see attached).  I connected it directly onto one of the batteries.  Could this have any significant impact on the batteries or recharging?

Alarm Timer Instructions.pdf

[DeepBass9]

11 minutes ago, Bush Man said:

A point of interest, during overcast weather the six panels still generate 200-300W worth of PV input for most of the day according to the Axpert reading.  Taken over 6 hours, this equates to 1200 - 1800Wh, which should be more than sufficient to run at least the fridge freezer and some LED lights indefinitely if I'm not mistaken?

Yes it probably will, but you are not doing any charging at the same time, so the next night will kill your batteries.

In that case I run my generator of a few hours in the morning to get the SOC up, and then maybe at night for an hour to top it up. 

[superdiy]

49 minutes ago, Bush Man said:

Before I forget, the batteries are indoors in a well ventilated cupboard.  Also during December I fitted a contact switch 12V alarm to the cupboard door.  Apart from the contact switch it consists of a 12V siren and a small electronic timer for the siren (see attached).  I connected it directly onto one of the batteries.  Could this have any significant impact on the batteries or recharging?

Alarm Timer Instructions.pdf

It is really hard to tell, because the timer's specs does not include any consumption information. Best would be to either contact the manufacturer or to measure it with a multimeter. I doubt that it will draw more than 50mA, which is approximately 0,6W and even with a constant draw of around 1W, it should not be a problem.

[___]

2 hours ago, Bush Man said:

I connected it directly onto one of the batteries.

I stand to be corrected, but given our long conversation about balance in battery banks, I suspect that even a smallish load could be detrimental to the battery it is connected to. During charging it acts as a bypass shunt, and during discharge it means your one battery is always a couple of watt-hour lower than the other one in the morning. I don't know, this might be nothing to worry about, but I would be concerned about how it adds up over time. If you have a balancer on there as well it might not matter, the balancer should cancel out some of that during the day, but my gut feeling is still that it would be better to use a DC-DC buck converter to step it down to 12V. Once you have that it also opens up other avenues for low-power devices, such as 12V 5W LED downlights :-)

[Bush Man]

Thanks, will switch off the alarm for a week or two in any event and see if it makes a difference.

My last question for now.  On a typical sunny day last year while charging the batteries mid morning I almost never saw the Axpert PV Input reading exceed 900 Watts, despite the 1500 odd watt panel array.  As per Don's input, the current Axpert settings allow up to approx 40 Amps/2000 Watts into the batteries.  Why then only 900 Watts on the reading?  I have since seen it peak at 1400W once when I switched on the microwave as well, so all the panels are connected and available.

[superdiy]

1 hour ago, plonkster said:

I stand to be corrected, but given our long conversation about balance in battery banks, I suspect that even a smallish load could be detrimental to the battery it is connected to.

I totally agree with the balancing issue. I just tried to explain that the expected consumption of the "alarm circuit" should not be a reason for concern if you look at the bigger picture of the total daily consumption. I also agree that would be better though to not connect it to one specific battery, but rather via something like a step-down smps fed by the entire battery bank.

[___]

31 minutes ago, Bush Man said:

I have since seen it peak at 1400W once when I switched on the microwave as well,

If the batteries only need 900W to reach absorption voltage, only 900W will be used. Lead cells charge slowly, especially as they reach >=85% SoC. That's my guess :-)

I'm actually playing this game as well to some extent. I changed my system in December to feed the difference into the rest of the house, and I now have my SoC limit set to 95% (ie I literally use only the top 5%). This works very well most days, because there is always a washing machine or a tumble dryer or a dish washer or SOMETHING running. But when we were on holiday I had it set to 65%, using 35% of the battery. Otherwise the battery fills up by 9AM and then throttles down... :-)

[___]

1 minute ago, superdiy said:

I totally agree with the balancing issue. I just tried to explain that the expected consumption of the "alarm circuit" should not be a reason for concern if you look at the bigger picture of the total daily consumption. I also agree that would be better though to not connect it to one specific battery, but rather via something like a step-down smps fed by the entire battery bank.

I agree with you there, should not be a concern re overall consumption.

[Bush Man]

1 hour ago, plonkster said:

If the batteries only need 900W to reach absorption voltage, only 900W will be used.

Would it then make sense to add additional panels if the batteries can only use 900W of the existing 1500W panels?

[DeepBass9]

Probably an a xpert limitation. You could run the panels through a separate m ppt.

[Guest]

1 hour ago, Bush Man said:

Would it then make sense to add additional panels if the batteries can only use 900W of the existing 1500W panels?

Therein lies the golden question: What is the load and for how long must it be powered. That is for batteries and panels.

Then you add the daytime load you want to power on top of that, check the inverter maximums, and gooi more panels on the roof for that load.

But then you are over powered for battery charging ... and the games begin.

Luckily it all has been done before here.

 

8 hours ago, Chris Hobson said:

I don't like the multiply by 5.5 calculation

The 5.5 is the average good sunlight hours in South Africa over 1 year.