Part 1 The brief overview - Going Solar

The background to my project...

In South Africa we have a new word, its a verb that most people are growing to hate with a passion...

"Load Shedding"

This is as a result of dismal failure on the part of Eskom to maintain the power grid in the country, instead the management are raking in millions in bonuses while seemingly the workers are loafing and doing everything except their jobs.

We need to move on to plan B in our homes, there are a few options each with its own pro's and cons..

1. Uninterruptible power supplies (UPS) can be put onto each required appliance but the issue with this is that the UPS' are costly and only give a few minutes of backup power. Their intended design is really to provide power for a short period of time while power is shifted from grid to generator. Cost is about R 2000.00 for 1000 watts which lasts about 10 minutes.
2. Generators are fast becoming common in many houses, you can easily pick up a 7 KVA unit that will easily drive an entire house including the geyser. The problem with generators are the noise levels and fuel required to drive them. It also seems the city is now requiring residents to have a noise certificate, a electrical connection certificate and all required safety equipment for fuel storage. Unless you have UPS' on all your appliances you will still suffer the power down and power up when Eskom switches to load shedding.
3. Solar solutions are growing in popularity but the many unscrupulous players in the market are really taking Joe public for a ride. They all see this as a opportunity to make money...fast. If  you make the leap and request a quote for a solar solution, you need to be sitting when you read the quote. It will be anything between 100k and 500k. These solar solution providers are missing the most important point...what the customer actually needs.

I spent some time looking at the various solar options but the way the solution is provided just is not sustainable at all. All the solutions want to charge up a huge bank of batteries in the day that will run your house for the night, then you also need to take into account rainy days that have little or no sunlight. This amounts to a huge battery installation which chases up the cost.

What is actually needed in simple terms is a system that comprises, Solar panels, Batteries and Eskom. What you want to do is run on solar by day, after all, if the sun is out use it, it costs nothing. Then at night when there is no sun switch to the grid (Eskom), if Eskom does load shedding at night you can fail over to batteries and even supplement power with a generator.

Feasible Solar Solution details

The most important rule to remember with Solar solutions, Watts in = Watts out

People get very confused with solar panels, batteries and inverters, you are mixing 12 volt solar panels, 12 volt DC batteries and converting to 220 volt AC and then mixing in watts and KVA. Its all one big jumble in the end and before you know it you cant make head or tail of what's watt..

Unpacking the components

Solar panels are all Direct Current (DC) and come in 12 volt or 24 volt.

Deep cycle batteries are also DC and come in 12 volt or 24 volt.

Inverters takes DC in one side and give you 220 v AC out the other side, on inverters there is one important thing to look for, it MUST be pure sine wave. Dont even bother with modified sine waves.

There are a myriad of options as far as DC voltages go, you can build a system using 12 volt, 24 volt, 48 volt or 96 volt. (can actually go higher too)

The lower the DC voltage the simpler the system but the less the efficiency. Simpler = less costly.

So you need to play around with cost vs efficiency and decide what works for you.

KVA and Wattage ratings relate to each other and indicate how much power the system can provide.

Generally to get Watts from KVA you multiply by 0.8 so a 5 KVA inverter would mean it can provide 4000 Watts.

Any appliance will have a wattage rating e.g a kettle is about 2000 watts, so to accurately decide on exactly what size inverter you need you will have to list all the appliances you want to run and their wattage ratings.

Remember that you generally don't run all the appliances at once, and this is the critically important bit for sizing the system. Decide what are the main appliances you will run at the same time and size the system on that.

Where do I start

I have taken a different approach to Solar, the customer one. Most solar companies will tell you what you need to spend.

I think they need to ask the customer, what can you afford to buy. The system needs to be scaleable so that I can start off with the basic requirements and as I see the system is working for me I can add on "components" that will save me money.

For example:

I want to put together a system and just move across one or two appliances to start with. Lets choose fridge and freezer and my lights. (now I know you are saying lights...solar...bear with me..)

For this example I will choose a 5KVA inverter, simply because its the most scalable model, if I got to a point where I was using all 5 KVA I could add a second in parallel and have a 10 KVA system. This inverter I sourced from China, it is a MPPT Solar regulator and off grid inverter. (will explain that later)

The 5 KVA inverter has a DC battery requirement of 48 Volts DC.

So my system would be:

A single 5 KVA inverter

4 x 12 volt 100 Amp Hour deep cycle batteries

Solar panel array in a 48 volt configuration to provide as many watts as required. Remember my initial statement, Watts in = Watts out

Say for our example we have a fridge of 350 watt and a freezer of 400 watt, typically the calculation for fridges and  freezers is 8/24 cycle time, meaning that out of 24 hours the device runs for 8 hours.

Lets configure lights as 200 watts, as I said solar lights??? Well, we want the lights on "solar" so that they are connected to our solution. This way when we have load shedding we can run the lights off the inverter.

So our requirement is 350+400+200=950 Watts 

That would be the peak value, meaning if we had all the lights on and both the fridge and freezer running. Typically the fridge and freezer run on and off making up a total run time of 8 hours per day.

The solar panel array needs to be done in a 48 Volt configuration, so assume I can get 4 panels of 250 watts 12 volt each, connected in series this would give me a 48 Volt giving me 1000 watts.

My configured system now is as follows:

5KVA or 4000 Watt inverter, 48 volt 100 Amp/Hour battery pack and 1000 Watts of Solar panel.

By day the system will run directly off solar power and charge the batteries, if the demand for whatever reason goes above the 1000 watt the inverter will draw this from the batteries. If we now run into the evening and the sunlight drops to a level where my panels are no longer effective the system will switch to the grid. Power will now be drawn directly from the grid to supply my devices. If load shedding is now implemented the system will switch over and supply power from the batteries. According to the specifications this 5KVA inverter will supply 1000 watts of power for 268 minutes before it shuts down.

As I mentioned the Inverter I found is a combination unit, it will take solar panels in and using the MPPT technology will regulate and change my batteries. I said the system is off-grid meaning that it does not "sell" excess power back to the utility. Currently there is no formal agreement in place with the city for this yet. Personally I would not hold my breath.

Nuts and bolts

Lets get into the nitty gritty of it now, starting with the inverter.

I sourced a special inverter that gives me exactly what I am looking for, something to take in solar power, it must charge the batteries, it needs to connect to the grid and finally invert and supply 220 volts AC to the house.

I also required that it be scalable, if in a few years time I decide that I need to provide more power I need to be able to add on without having to replace the inverter. The box I found can do just that, up to 3 of them can be connected in parallel. So a single 5 KVA unit can be expanded to 15 KVA which is far more that most households require.

Before we can tap into the sun we need to see what sunlight is available, below is a link that shows average sunlight hours for Cape Town.

http://www.cape-town.climatemps.com/sunlight.php

We can see that summer is great but winter it dwindles down to just under 6 hours a day of sunlight and this at a low angle of 32 degrees.

Look at the following table below, this would be the average power you could generate from a solar panel array of 1000 Watts per day for the corresponding months:

Effectively though for our example you would in actual fact be able to run off solar for the time indicated under the sunlight column. What this does indicate though is the effective savings you would get per day/month.

Real costs

Typically an average household will use 24 units a day which averages out to about 800 units a month. This is roughly about R 1500.00 of electricity.

24 units a day is 24 KW/h and if you want to use averages and ignore peak it equates to about 1000 watt an hour.

Then we can also make the safe assumption that 800 units a month is 800 KW/h or 800 000 watts.

If 800 000 watts cost about R 1500.00 then we can say we are paying 0.1875 cents a watt or around R 45.00 a day.

So lets put this back into the above table with some costs and cost savings in.

Cost of the example system

Below is a table of costs for the various components, this excludes electrical connectivity of the solar system and house distribution board.

While R 24 000.00 may seem a hefty amount for the example of a fridge, freezer and a few lights we need to keep in mind that we designed the system for peak load. 

We could in theory add more appliances to the system as any extra requirement up to 4000 watts will be sourced from the battery if the solar panel cannot provide it.

If I wanted to add to the system to carry more load I will require nothing else but solar panels. If say we wanted to go the full route of 5 KVA we would need an additional 3000 watts of panel only. This would mean we would be able to run an entire house on solar for the daytime hours at the cost of 

R 50800.00

Lets plug 4000 watt into the cost analysis and see what effect it has on the daylight solar savings..

Suddenly we see that if you were using 4 units an hour (4000 watts) you would save just over 

R 22 000.00 a year on power, that means the system will pay itself off over 2 years.

The added benefit you receive is that whenever load shedding happens you will have some appliances that you will still be able to use. This time will vary on the load being drawn but 4000 watts will have about a 40 minute windows before the battery's in our example are depleted. This can be extended by doubling up the batteries, a cost of around R 5000.00. This will extend the battery time to about 90 minutes at full power being drawn. 

My opinion on this is to rather minimise electricity use during load shedding and keep with the minimum amount of batteries. These are the only components of the system that don't last, they have a life time of about 5 - 10 years.

The minimal set in our example of  100 Amp hours will last as follows:

Say for example if we had a 5 KVA system that only had 100 AH of battery we could still run on battery for 268 minutes if we were to just minimise the load to 1 KVA (800 Watts)

Here is another thought though... say we went back to the original example where we just ran solar panels of 1000 watt and configured our system to allow it to run on solar, then supplement solar with battery until the battery was depleted, we could extend the off grid period by 4 hours. (for a load of 800 watts and ignoring the 50% battery rule) 

(the 50% rule with solar batteries, they should never be depleted beyond 50% capacity else they start to bend the internal plates and become dead short)

Lets add 4 hours of "Battery running time to the savings"

Suddenly the annual savings go up to just over R 8341.28 so looking back at the Battery table I see that if I had 200 Amp Hours of battery I would get 613 minutes of backup time. That equates to 10 hours.

Plugging in the extra battery run time I see the savings now go up to R 12447.53 a year, so where the original system paid back R 5603.00 a year and the outlay was about R 238000.00 by using the battery we are suddenly in a position where the outlay is now about R 29 000.00 but the payback is 12 447.00 per year.

Over doing it

One thing we need to be mindful of (and what makes this particular solution so great) is that we need to design the system to produce as much as we need and no more. If we look back at the annual cost savings of  R5 603.00 keeping in mind that our system was specified for 1000 watts only based on the fact that we only required an average draw of 1000 watts, what happens is say we build a 2000 watt system?

The answer, not really much. Its basically wasted power, since we only require 1000 watts the other 1000 watts is just not used. Since cost savings is only on what you use, it is actually a waste in terms of that we needed to build that extra 1000 watts into the system but since its not used its not "paying for itself"
Lets put this extra 1000 watts into the costs...

Our input costs have increased to R 32800.00 but based on our usage we will still only save
R 5 603.00 per annum, this just means our repayment term just went from under 5 years to 6 years.
In the back of our mind we need to remember that the batteries will not last longer than 6 years so we may be in a situation where the solution has not repaid itself and we are already adding new costs to it.


Viable or not

Taking all the calculation above into account it does seem that going the solar route while still being connected to the Grid is a viable idea, going solar and being off the Grid is a whole new challenge because you need to start to factor in the odd week that the sun doesn't shine at all during winter.

I will take the plunge and publish each phase of the project as I go along with as many pictures and details as I can.

Keep checking back here for more....