What is the ideal "tilt" for a south-facing panel?
It depends. Doing preliminary runs with our simulation software (still in the development and testing phase) it appears that no simple "rule of thumb" is completely adequate. Our software takes into account both direct sunlight and diffuse "skylight " and the "typical" weather, for a given location, using the TMY3 (Typical Meteorological Year) database, which constructs a "typical year," on an hour-by-hour basis, drawing on comparisons over a range of many years. Here are the results for our "tilt analysis" for three different locations:
The rule of thumb is to aim the face of the panels 15° - 18° above the celestial equator. This is equivalent to tilting the panel relative to the ground by an amount equal to the latitude minus that number of degrees. However, for Miami, the magic number is 0°, for direct sunlight, and only 5° to 10° if you account for direct + diffuse light. The latitude of Miami is between 25° and 26°.
Barstow is in California's "high desert" with lots of direct sun. Note first of all the larger total energy production. Barstow is somewhat farther north at about 35°. For direct sun, +5° is optimal, whereas for direct+diffuse light, +10° is optimal, still short of the standard 15° rule of thumb.
Seattle is between 47° and 48° north and it is known for its drizzly weather. Note that a large fraction of the total energy comes from diffuse light. Solar panels see the greatest amount of diffuse skylight when they face straight up, so on balance, the optimum aiming angle for this location is 20° or even 25° above the celestial equator.
How much do you lose by mounting panels in non-optimum orientations?
This is also a question without a simple, direct answer. Here is the answer for my own home in Strathmore, CA:
(Note that the graphs on this page do not include the origin.)
The bar on the left represents the actual installed configuration: south-facing panels aimed 15° above the celestial equator. (15° is in fact the optimal offset angle for this location.) We have a north-south roofline and have mounted the angled panels shown on the home page. The middle bar of this graph represents the energy expected from panels mounted flat on the east-facing roof. The third bar, for comparison, is what might be expected on a flat roof with the panels facing straight up. There is an added benefit to tilting the panels, by the way, that is not taken into account (yet) by our software. Solar panels are temperature sensitive. They operate at higher efficiency if they are cooler. Air circulation around the panels is therefore an advantage for tilted panels over panels that are mounted flat against a hot surface.
Programs that will estimate your energy production are available on various websites. These results were generated by our own in-house software. One advantage of writing one's own program is that it can be configured at will to answer many different questions. The two questions on this page illustrate some of the functionality of the software in its present form. Tools like this can help us answer your questions about your particular site and ultimately can save you money and improve your energy production.