I decided to choose a blank spot on the wall next to the existing Distribution board for my Inverter, it made sense to keep all the power stuff together in one place.
I bought a wooden workbench (those flat pack DIY ones) and used the lower shelf for my batteries, keeps them off the ground and out of the way. I used nice big welding cables to run from the batteries up to a 100 amp DC trip switch, from there it runs to the Inverter battery input.
New Inverter as it arrived
100 AMP DC Trip switch
Mounting the inverter was pretty easy, it "hangs" on 3 screws much like a painting does, you need to mount it eye height so reading the display is easy.
Installed Inverter
I ordered a solar panel switch box that terminates all the solar panels into "fuses", these feed into a DC trip switch that is the equivalent of the maximum current my inverter can handle.
This I mounted slightly above the inverter, so the cables from the panels drop into the switch box and from there drop to the Inverter solar input.
PV Panel Isolator Box
I needed to get an electrician out to make some changes, the existing DB was too small and messy, I wanted a split DB so that one half would feed all lights/appliances on solar and the other half would feed all lights/appliances on the grid.
A NOTE on this, I have not gone the whole hog and just put everything onto solar, I calculated what my average daily requirements are and then scaled the system according to that. To me it makes no sense designing a system that has an unreal ROI.
For instance to cater for an average house running everything you must be able to draw a peak load of about 9000 watts, this would be say a geyser, microwave, kettle and general appliances. This is in my design a triple system, each inverter can do 4000 watts so you need to have 3 in parallel do get enough and have spare to avoid tripping everything.
I decided that I am not catering for no sun periods, at night I revert to the grid. Since we are at work most days to me it then made sense to leave the kitchen appliances on the grid, so the oven, microwave, kettle and geyser stay on the grid DB. In most cases the kettle and microwave are used at night anyhow when there is no solar, no point in scaling they system to include them if its never going to pay itself back. The only draw back here is that if we have load shedding and I need to run any of these I will have to plug them over to an inverter fed plug.
The rest of the house I moved over to the DB being fed from the inverter. The exception to the pay back theory are the lights, I know lights are used at night when there is no solar but I put them on the inverter so that in the event of load shedding we still have lights.
Once all the connections were made the last thing to do was the solar panels. An array mounting kit could cost you upwards of R 20 000.00, I was not going to pay that. There are a few options depending on the roof type. If you have a corrugated roof its dead easy, you simply buy a length of aluminium angle and pop rivet one side to the panel sides, then use those roof screws to go straight into the trusses.
In my case its cement tiles, I made some brackets from hoop iron, pop riveted one end to the panel and screwed the other into the trusses or batten (whichever was easier)
I used an aluminium channel to hold the two Solar panels together, then riveted that to the hoop iron that was fixed to the trusses.
Below is the Aluminium channel used
Here are the two panels side by side being riveted to the aluminium channel.
I used nice big rivets so there was no chance they would pull through.
Here is one set of panels, these make up one array of 48 volts, they are 2 x 380 watt 24 volt panels.
Here is the one hoop iron bracket screwed into the batten and curved around the tile
Once curved over the tile it runs under the panel, bends up by 90 degrees and is riveted to the side of the panel.
This is the top side, towards the apex of the roof
A second on showing the top side of the panel towards the apex of the roof
Here is the mounted panel showing the bottom side of the panel. You will notice I do the 90 degree bend at the edge of the tile. This is done first to position the panel, the top "fixer" is then done to fit the panel length on the tiles.
I ran the wires under a tile into the roof and made my series connections in the roof and then ran from there down to my solar switch box using 4mm wire.
Once the wiring was complete I double checked the voltages before the trip switches in each case, 48 volts DC from the batteries, about 76 volts DC from the panels, 230 volts AC from the grid input.
Switch on
Power on was first the 100 Amp trip switch from the batteries, once that was on I started up the inverter, after that I switched on the solar array and finally I added the grid input power.
System powered on and running from Grid
Then I needed to configure the system, I loaded the software supplied and connected to the inverter unit via the supplied cable. The software makes it easy to set priorities, cut back voltages, overrides etc.
System disconnected from Grid but powered on Panels
Then it was/is a case of sit back and watch how it all works, Its been almost a month and there have been some very good days and some not so good days. Seeing that its solar...on a rainy day you can kiss any savings goodbye, but that is to be expected.
My average saving thus far has been 22%, the best day was 43% and the worst was 1%. The saving over last month thus far was 27.02%.
As the sum moves around the day are getting better and better, my solar geyser is also starting to work better and I think by October my power usage will have dropped by 50% of what it is in Winter.
I think by November when we usually switch off the solar geyser element we should be using about 90% less electricity than we do in winter.
I keep referring to winter because I used it as a peg in the ground, power use is very dynamic and there is no way to get 100% accurate usage figures, each day is different. I took the average winter units per day and made that my yardstick. My feeling is whatever I can save off that figure is good for now.
I have some stats from the past few years, I managed to put them into a graph format that shows how our power use increased year on year.
Power use for 2013 (No Solar Inverter)
The first part of the year we ran under 15 units a day average, from May it went above 15 and stayed well above that for the entire year.
Power use for 2014 (No Solar inverter but a some solar powered LED Lights)
Usage wach much more evenly spread but never dropped below the 15 units a day mark.
Power use for 2015 ( Solar inverter installed 15 August)
Usage started above 15 units a day and wet up again to the 25 units a day mark until mid October when the Inverter was installed, thereafter it dropped below 15 units a day.
Power use for 2016 (Running on Solar Inverter)
Usage has stayed below 15 units a day, this despite additional load being run on the hot days in the form of an air conditioner. I did add another 500 watts of solar panels for the air conditioner though.
If I assume constant use (just for the sake of argument) and say that summer months before the system was installed I was using 17 units a day in summer months and in winter months 24 units a day then rough calculations for average power use is as follows:
If I assume constant use (again for argument sake) of 11 units a day post system installation for the summer months and using the same jump of 7 units between summer and winter so assume this winter will be an average of 18 units a day, then the post install average power use is as follows:
This shows a 30% saving on power
I calculate it as follows, at a cost of R 1.79 per unit its an annual (estimated) saving of about
R 4000.00
This puts it very much in perspective that the ROI is just a little over 7 years.
Whether that is worth it or not I cannot say, we will have to see how much added value comes from the threat of load shedding in the next few years to see if the convenience factor plays any role is making the ROI seem better or not.
:-)
In the end it was a fantastic experience and I learnt a great deal in the process.
