OK, time to update my Virtual Tracker concept with some data, and it shows that it does work as I expected - 2 parallel arrays in different orientations into a single MPPT charge controller works perfectly well, and is a viable alternative to a tracker, now that PV panel costs have fallen so far. I originally posted about the idea here: solar-wind-gear/topic1507-40.html#p25960
The idea is to place panels tilted up 60degrees, ie pointing 30 degrees above the horizon, facing east and west, in order to get a fairly constant output throughout the day.
Essentially double the number of panels can be used on a certain rated inverter, ie a 2.5kW inverter with 2.5kW of panels facing north, could accept 5kW of panels in the Virtual Tracker arrangement, with no risk of letting the magic smoke out.
In these days of minimal to zero FiT, at least in NSW, it is desirable to have an even PV output throughout the day, rather than a big peak around solar noon, in order to minimise the giving away of energy to power companies to sell at 50c/kWh or more during peak rate periods. They are making profits at your expense, so it is best avoided if at all possible.
I prepared this graph from data I collected on the 6th and 7th of January- 2 perfectly clear, hot days. Temperature profile throughout the day was not identical, but was close enough, and both maxima were just over 36C. Solar radiation graphs show identical profiles for both days. Panels used for the test were Indian-Pacific branded 250W monocrystalline panels, on the ground, leaning on an Aluminium adjustable height platform (the latest in high tech mounting systems!
) North facing panels were set to 31deg tilt, equal to my latitude, and in Virtual Tracker arrangement were tiited at 60deg from horizontal, facing east and west.
VT-North-facing.comparison.gif has kW and kWh for both days for north facing and my Virtual Tracker arrangement, plus the Virtual Tracker curve multiplied by 2 to show how it looks in comparison to the north facing array. The slight bulge in the VT profile for a couple of hours either side of solar noon at 13:02 was, I suspect, related to cooler panels- due to them pointing 60 degrees off-sun (@0.5 incident radiation), yet still having the same surface area for radiating heat away. The slight bump at the start of that was also due to panel cooling, when I washed some bird droppings off one panel.
The truncation of the curves before 6pm is due to my location in a valley, I lose the sun about 2 hours before sunset, which in this case handicaps the VT arrangement, so I've marked the difference at solar noon- 0.2kWh less from the VT arrangement than north-facing, which would equate to 0.4kWh down over the day for a location with clear E and W horizons. Clearly there would be minimal gains to be had in locations with restricted views to the east and west, but with clear views there is a real advantage to be had by using more panels in a Virtual Tracker arrangement when looking for flat output curves.