Irradiance

The amount of solar power you can generate depends on a few factors.  Most important are latitude (affecting the angle of sun and length of daylight) and the orientation of the panels.  Since I live in the Northern Hemisphere, it is best for my panels to be angled toward south (technically, south-southwest but the difference is minimal). For optimal year-round generation, the angle of the panels could be changed seasonally (steeper angles in winter when the sun is lower, and flatter angles in the summer when it is higher).  However, the expense of installing adjustable angle panels was not worth the difference, especially when my roof angle favored summer generation and my time-of-use rates made the summer generation the most lucrative to my pocketbook.

The technical term for this potential power is "irradiance".  It essentially equates to an equivalent number of hours of direct sunlight on the panels, making further calculations easy without the need for more complex math.

You can find a multitude of irradiance calculators on the internet, some giving you data down to the day based on exact latitude and other factors.  (It's just spherical trigonometry and math equations, but then you have to factor in average cloud cover, etc.)  I chose to get monthly predictions using the calculator at Solar Electricity Handbook, entering data for my city, and southerly roof orientation.   Unfortunately most of the 17 degree "best angle" on my roof is already taken up by a solar-thermal pool heater so I had to use a flatter 6 degree angle for 65% of my panels.  I used the calculator interpolating between the "best summer performance" 21 degree angle and "flat" to conservatively predict the performance assuming all my panels were at 6 degrees.   The "best summer" actually isn't the best for my latitude, however.  12.5 degrees is apparently appropriate for my 36 degree latitude, so I figure I'm doing pretty good within a few degrees on either side of that number.

My system went "live" on Jan 5.  The irradiance figure I arrived at for my installation, for January, was 3.19.  Multiplying the by my system size (10 kW) and the number of generating days (27) my calculations predict 861 kWh of energy.  Actual generation from the installed power meter?  1.09 MWh (exceeded prediction by 27%).  Looks like I'm generating at least as well as my model predicted, maybe better.