OutBack's newest innovation in balance-of-system components
Jim, thanks again for your comments, as per the post to Julie; I am first going to transfer my Critical loads to a seperate Distribution board or FW/AC to clarify. I will then be able to give you the exact loads to assit with the final design if you won't mind.
OKÔÇª here we go. The baseline net requirement is 70 kWh/week, or 10 kWh/day x 5 days/week.
Allowing for 90% inverter efficiency, the battery bank will need to supply 11.1 kWh/day, or 48 V x 231 Ah. The battery bank size could be anywhere from twice the average daily energy requirement to about six times the average daily energy requirement.
Using the former 2X model, you might have to depend on the grid to regularly top off the batteries. You could set the invertersÔÇÖ ÔÇ£Grid UseÔÇØ mode to do this overnight. Accordingly, the 48 V x ~460 Ah battery bank would still be viable, although, at the estimated equivalent of ~125 50% discharges per year, the batteries would likely have to be replaced about every three or four 4 years.
Applying the latter 6X model, there should be a balance between energy use and recharging from the Sun, and grid use for recharging should be minimal. A 48 V x ~1,400 Ah would be appropriate. Big flooded-cell batteries from Rolls/Surrette or Absolyte AGM batteries from GNB would work, and long battery life (>10 years) should be expected.
PretoriaÔÇÖs latitude is ~25 degrees South. A typical fixed-tilt angle for a north-facing array in your location would be 25 degrees. However, 40 degree tilt angle would increase winter energy production potential. This increased angle would also improve array cooling and self-cleaning. The steeper angle would somewhat reduce summer energy production potential, but the longer days would likely compensate somewhat.
Insolation data indicates that Pretoria receives a daily average of the equivalent of 3.8 hours/day of ÔÇ£fullÔÇØ Sun. This data is for a horizontal flat plate collector and can be improved if the array is tilted north, as discussed above. For comparison purposes, see the data for Brownsville, Texas, which is located at ~26 degrees North.
http://rredc.nrel.gov/solar/old_data/ns ... /12919.txt
A winter net of 10 kWh/day would require spec-based generation of 10 kWh/day / (88% x 97% x 98% x 80% x 90%) = 16.6 kWh/day. Assuming a winter average of 5 hours/day of ÔÇ£fullÔÇØ Sun on a north-facing array tilted up at ~40 degrees, the PV array spec would be 3,320 W STC.
Configuring a PV array for a 48 V battery bank and for the controllerÔÇÖs limits can be a challenge. At one end, itÔÇÖs important to deliver a high enough array voltage in the summer to be able to drive a flooded-cell battery bank to its EQ target of ~62 V or so. Because PV array voltage drops when the array gets hot (up to ~35 C above ambient), and also allowing for voltage losses in the wiring and the controller, the arrayÔÇÖs STC Vmp spec should be around 83 V.
At the other end, the current crop of MPPT charge controllers generally have an absolute maximum input voltage spec of 150 VDC, and the MX60ÔÇÖs operational limit is ~141 VDC. Because winter temperatures can cause the array STC Voc to increase, a correction factor must be applied. Using the US National Electrical CodeÔÇÖs correction factor of 113% for temperatures in the range of -1C to -10C, the arrayÔÇÖs STC Voc must not exceed 141 V / 113% = 124 V
So, the challenge is to configure an array with an STC Vmp of ~83 V, and an STC Voc of ~124 V. One minimum option would be 25 each Kyocera KC-130 modules wired 5 x 5 for an array rated at 3,250 W STC. A single MX60 controller could handle this array for a ÔÇ£48 V systemÔÇØ.
A more conservative option would be 18 each Sharp 200 W modules configured 3 x 6 for 3,600 W STC. Although a single MX60 controller could probably handle this array, itÔÇÖs officially limited to 3,200 W STC. Accordingly, a new FLEXmax 80 controller is probably warranted.
In summary, I recommend you adjust your system BOM and weÔÇÖll then take a fresh look at it.
Jim / crewzer