As mentioned before, my initial planning had changed somewhat and I was now looking at installing a battery back-up system to help out during the the increasing load-shedding periods. We were getting shut down on an almost daily basis for 2 to 3 hours at a time, usually during the most inconvenient times: dinner time, children’s homework time, sitting down and watching TV time. You get the picture.
The search for the most appropriate system was on. Note that I say “most appropriate” and not “best”. I can’t afford the “best” system, but I can find something that is most appropriate to our needs and budget. If the system is designed and planned properly, upgrading any component at a later date would be a fairly simple “plug and play” exercise.
Choosing a Solar Inverter
Choosing a solar inverter turned out to be a minefield of options:
- Off grid
- Grid tied
As I want to stay connected to the grid, off-grid isn’t an option. This would require a fairly large solar array and very large battery bank. Completely out of the question for my needs. I could use a small system as load-shedding back-up, but then I’m limited when I want to add solar power for the house.
A grid tied system would fit my requirements, but most systems are way out of my budget. Firstly, a standard grid-tied inverter only operates when there is utility power (and sunshine). The traditional systems allow for grid feedback, which isn’t really an option if you want to do it legally. Most of our meters aren’t geared for feedback, and neither are the utilities in terms of providing credits for energy sent back to them. To allow the grid tied system to work during load-shedding, you would need an additional grid-tie limiter to allow the inverter to work in island mode. In order to provide battery backup, you require an additional solar charge controller and a battery inverter. All this leads to a fairly complex and expensive system. The plus is that it is usually tried and tested equipment that gets used with plenty of options in terms of monitoring, scalability, etc.
Next option would be a bi-directional inverter. This is essentially the same as the above, but without the need for the second battery inverter. Again, way too pricy for my pocket.
The final option for me was a hybrid inverter. It ticks most of the boxes regarding what I need from the system:
- Pure sine wave output
- Solar input
- Grid input
- Battery input
- Built-in MPPT charge controller
- Depending on the type, option to feed back to the grid when utility is active or to switch to island mode during grid failure.
- Compact size
After a lot of searching and researching I stumbled across a supplier in Bloemfontein who imports inverters and sells at a very good price. They were offering a 5kVA Hybrid inverter at around R11,500 when I found them. Take a look at Ecotrades Solar Power Systems for more details.
The inverter they offer is the Axpert MKS 5kVA unit from Voltronic Power in Taiwan.
- Pure sine wave inverter
- Built-in MPPT solar charge controller
- Selectable input voltage range for home appliances and personal computers
- Selectable charging current based on applications
- Configurable AC/Solar input priority via LCD setting
- Compatible to mains voltage or generator power
- Auto restart while AC is recovering
- Overload and short circuit protection
- Smart battery charger design for optimized battery performance
- Cold start function
- Parallel operation with up to 6 units only available for 4KVA/5KVA
Having looked at my load profile previously, I have determined that I could safely reduce my maximum demand at any time during the day to less than 4kW with an average load of around 1,5 – 2.5kW. That would make the 5KVA Axpert MKS perfect for my needs. It also offers a short duration overload capability of up to 8kW.
The hybrid inverter also works perfectly for my backwards planning of starting with battery back-up rather than solar generation. With the addition of a battery bank, the inverter can be set up as a UPS with solar panels being added later.
The other nice thing is the ability to parallel connect the units – up to six in parallel to give a capacity of 30kW. This is way more than 99% of household systems would ever need.
The maximum connected PV input is 3kW which suits me perfectly as I don’t have space for more panels than that on my roof and is sufficient to run the house during the day as well as charging batteries when needed.
My estimate is that once the system is fully installed, I will be able to generate an average of around 14kWh/day. That’s about half of my current consumption which still includes an electric geyser and 750W pool pump that runs for 4hours a day. Just those two pieces of equipment account for about 12kWh a day.
Choosing a Battery System
The next step was to get a set of batteries for the back-up system. Yet another minefield!
The choices are staggering, Flooded Lead Acid, Sealed Lead Acid, Lead Crystal, AGM, Gel, Lithium Iron, 2V, 6V, 12V….
Once again, my choice was dictated by budget and suitability of purpose. I wouldn’t be needing the batteries for anything other than occasional back-up at this stage, so anything more than a decent deep or semi-deep cycle sealed lead acid battery would be a bit of an overkill.
For normal periods of load-shedding I would need around 100Ah at 48V to power the house (excluding the electric geyser and pool of course). This would mean a bank of 8 x 12V100Ah batteries connected 4 in series to get 100Ah at 48V with another string of 4 batteries connected in parallel with the first to get 200Ah at 48V. Allowing for a maximum of 50% depth of discharge (DoD) I would get the required 100Ah.
Based on an average load of 2kW, this would give me just over 2-hours of operation. Night-time loads (in periods when load-shedding is most annoying) are usually less than 1kW, so the system should be able to provide back-up for close to 5-hours.
I have however decided to start with half of the final planned battery bank and bought 4 x 12V/102Ah Sealed Lead Acid semi-deep cycle batteries to start with. I managed to find batteries going for R1,600 a piece and snapped them up. As soon as funds become available again, I’ll add the next four batteries.
The batteries are Excis FMF102’s made by First National Battery, and should be good for the next 3 to 5 years. Once these have reached their expected lifespan I’ll start the battery research again. By that time I expect the choices to be even more overwhelming!
The cost of the planned system so-far is:
Gas Stove: R 13,400
Hybrid Inverter: R 12,000 (delivered, not installed)
102Ah/48V Battery Bank: R 6,400 (delivered, not installed)
Cost so far: R 31,800
The next step is to start with the installation and rewiring required to get the system operational. Stay tuned for the installation.