One year's off grid solar production

[DeepBass9]

I am posting this to help with system sizing for people considering going off grid. This is one years monthly production figures. The system size is as follows:

Generation

3000W solar array, two separate arrays of 6x300W and 4x300W panels, through separate MPPTs. Fixed angle installation, 1800W at 30deg north, and 1200W essentially flat (roofs dictate the angles)

Diesel generator for backup.

Wind generator (I can't measure this but I guess it makes 1or 2 kwH per month depending on wind)

Solar geysers (no electrical backup)

Load

Household of 4, normal fridge, freezer, washing machine, computers etc etc. Water pumping from borehole, pool pump (2hrs in winter, 6hrs in summer).

Gas stove for cooking, wood fires for heating.

Storage

8x150Ah Trojans, currently 2.5 years old.

I have been measuring the montly production for a year now since I upgraded the system to 3000W. Note that the graph is more of consumption rather that production as the batteries float on most sunny days, so there is excess production mostly. If the weather is overcast I run the generator, rather than stress the batteries, especially in winter I run the generator for an hour or so in the morning to power toaster, microwave etc before the sun is up.

The seasonal effect is obvious on the graph, the pool pump is used more in summer which accounts for some of the extra power consumption, I'm not sure what all of the other extra consumption might be, I think we are getting less frugal with power use as time goes by and we are used to the system, so we vary from about 300kWh per month in winter, to almost 450kWh in summer. Solar efficiency varies from around 3 in winter to 4.5 in summer (kW produced/kW installed per day).

Diesel use peaks in May & June as would be expected, you often get overcast days in May.

So here's the figures, I hope someone will find this useful.

 

 

[___]

The pvoutput.org team page also gets you an idea of what kind of power people are making on a day, so far could not figure out how you look at historical data though.

I added myself in December, after finally writing some code to patch Venus into pvoutput. What's kind of interesting is how people on the east is initially ahead in the morning, and then the westerners catch up towards late afternoon. @The Terrible Triplett starts at the bottom every morning for example, with a very small and also off-grid system. Usually by late afternoon, in terms of efficiency, I end up in third or fourth place (hybrid inverters are awesome). @ebrsa pretty much beats everyone on a daily basis :-)

[Guest]

My system first re-charges last nights battery usage, as designed 6 years ago to ensure batteries are always ready ASAP after the last Eskom failure.

Once they reach the setting (90-94-95% SOC - varies between winter and summer), the inverter comes on powering between 600-800w, peak at 2.1kw, the rest of the day.

Also have 2 different sets of panels on 2 controllers on the same battery bank, so PVO in my case is not a true reflection of how efficient the system is, for all is drive by solar, no Eskom near it.

Every once in a long while, someone (looking around whilst typing) forgets a setting and the system goes, batts so low the inverter switches off.

Now THAT gets the juices stirred I tell you. Mine, the kids, the wife AND the batts, like in properly stirred. I just shout as hard back asking who did what and why ... whilst quietly setting things back to what they are supposed to be, grumbling all the time (for show) hoping no-one makes the link.   

Hey, come on, you all know the 'long haired house animals' are very dangerous when antagonized! Man's gotta hide as best he can.

[DeepBass9]

3 hours ago, plonkster said:

I added myself in December, after finally writing some code to patch Venus into pvoutput. What's kind of interesting is how people on the east is initially ahead in the morning, and then the westerners catch up towards late afternoon.

That's because SA is a single time zone so the sun gets up earlier in the east and later in the west. Well spotted!

I was trying to find a website with this data on, but when I installed my system I figured out that local noon here is 12:20. True noon at 12 is at 30 degrees east, around Pietermarizburg.

South africa stretches over about 15 degrees of longitude, so there should be a variance of about 1 hour between the extreme east and west of the country (15 deg * 24 hrs = 360 deg). so every degree you are west of 30E, noon will be 4 minutes later.

[DeepBass9]

Hmm, can also explain why people think that Capies are always a bit slower in the morning! 

[ebrsa]

@plonkster I am glad to hear that you think my setup works well but I see both you and @pilotfish are ahead of me on efficiency. My Raspberry Pi controlling the system with ICC is set to switch to batteries at 1:00 am and to grid at 17:00. My wife turns on heavy loads like dishwasher and stove some time after that. I have 6 panels facing north on one Axpert and 3 facing north--west on the other one. In winter I have to switch to batteries/solar later and back to grid earlier. Even in summer I manually switch to grid if there is not enough coming from the panels. Mostly my batteries don't go below 75% SOC and at 17:00 are usually fully charged. Even so I have set parameter 32 to charge at 58.4V for 90 minutes which happens in the early evening and charge current is around 4A for my 450AH 48V bank of T105's.

If you have any views on how I could optimise further, I would be glad to hear.

[___]

On 2/2/2018 at 5:53 PM, ebrsa said:

I see both you and @pilotfish are ahead of me on efficiency

I have no idea who that " Homepv Sonstrall Hights"" guy is.... except that he cannot spell, and he's either cheating or his array is bigger than declared... or he has a tracker  :-P

[Mark]

 Sonstrall wakes up when the sun hits the panels and switches his PC or starts his PVO feed... he then misses a good few hours power usage and therefore the efficiency is excellent.  Easy to do

[Chris Hobson]

Couple of others have a ±60W PV feed through the night 

[SilverNodashi]

On 2/1/2018 at 12:01 PM, DeepBass9 said:

That's because SA is a single time zone so the sun gets up earlier in the east and later in the west. Well spotted!

I was trying to find a website with this data on, but when I installed my system I figured out that local noon here is 12:20. True noon at 12 is at 30 degrees east, around Pietermarizburg.

South africa stretches over about 15 degrees of longitude, so there should be a variance of about 1 hour between the extreme east and west of the country (15 deg * 24 hrs = 360 deg). so every degree you are west of 30E, noon will be 4 minutes later.

Wasn't there some talks of adding daylight savings, and another timezone to South Africa?

[Chris Hobson]

4 minutes ago, SilverNodashi said:

Wasn't there some talks of adding daylight savings, and another timezone to South Africa?

Jip they had Plett time  where you go off into the forest partake in recreational pharmaceuticals and then it can be whatever time you want it to be. On a more serious note I think Plett time was an hour behind SAST but it never caught on. Interestingly China which is a huge country only has one time zone.

[Guest]

2 hours ago, SilverNodashi said:

Wasn't there some talks of adding daylight savings, and another timezone to South Africa?

Yes, the line was drawn right through a guys house, so he freaked, wrote to the papers asking how one part of the house can be an hour earlier than the other side ...

It was publsihed, big who wha about it ...till you saw ... 1 April 19XX :-) 

[PaulF007]

PVO - Genade , I had forgotten completely about that site. The funny thing is that the data is still going there! So I had a look and shoot I am 3rd on the list. He he not that it makes any difference , but boys will be boys  . 

 

[pilotfish]

I see that @Sidewinder is doing himself in - his system info in 12x 300w panels (3600w) but with a bit of dyslexia thrown in has listed his system size as 6300W. If that was corrected he would murder everyone on efficiency by miles!

...although the current output is 4437 watts at 10h44, so clearly not a 3600w panel???

[pilotfish]

7 minutes ago, pilotfish said:

his system info in 12x 300w panels (3600w)

Looking at the stats I'm guessing that the error is in the description 12x 300 which should read 21x 300, and so the system is 6300W as listed.

[ibiza]

I see his system contains 12x300+9x300!

[pilotfish]

16 minutes ago, ibiza said:

I see his system contains 12x300+9x300!

Ok, I was just looking at the pop-up when you hover over 6.300kW on the ladder which says 12x 300. I guess better info available elsewhere.

[Sidewinder]

Hi all,

Yes, my system is a bit temperamental at the moment.

I have 24 x 300W panels, but I have disabled 1 x 1800W string, as I'm preparing to add another V series inverter, thus feeding each MPPT with 1800W, instead of doing 3600W through 2 x MMPT's, like i've been doing for the last 6 months. I found that on very sunny days, when pushing over 6000W, it's temp rises very high, and the PV then starts oscillations.

Plus i want to have peace of mind pushing 8 kW through one inverter, when hairdryers etc come into play

So just waiting for a extra buss bar to split my 4 strings to 4 x  (fuses, isolator & surge protection). 

Hopefully all done by this weekend.

[Sidewinder]

On PVOutput, I have created 2 sets of strings, as 9 of my panels are mounted @ 18 degrees, while the other 15 are at 38 degrees, all facing almost north. Hence some of the info may be displayed off. As soon as I have installed the new inverter, I should be back to 7200W. 

Come to thing=k of it, I think I made a mistake. 7200 - 1800 = 5400W. No wonder the max I've been getting these last few days has been around that.

Will correct that.

[Mark]

@Sidewinder PVO gives you the option of creating a "Parent" System and then depending on how you choose to monitor the inverters you can publish each inverter to a separate child system.

The parent then consolidates the data and the children can be view separately.  A pain to set up but works well.

Regards

Mark 

[Sidewinder]

OK, so late this afternoon got it all sorted. 2nd Infini is up and running. Will do a more detailed report tomorrow morning.

[Sidewinder]

So started monitoring and checking that all values for the 4 x mppt's are ok. Everything seem ok, but Emoncms and pvoutput are only reporting 1 off the 4 pv strings. So i have adjusted pvo to suit temporarily.

As usual, Manie is on it. Thanks for the support.