Words of wisdom: “In theory, Imp * Vmp = panel Watts. There's an awful lot in the real world that can affect that.” – Cariboocoot1)
For various reasons it will be a rare event when your panels put out their full rated power. One reason for this is the power needs somewhere to go; unless you are in Bulk stage or are running big loads full power isn't needed so full power isn't generated.
The other big reason is that power ratings are derived from testing under specific lab conditions.2) The standard lab test is at:
A situation where you might get the panel's rated power (or even a bit more!) would be when the sun is directly overhead on a cold, clear day at high altitude. Least power would be produced on a hot, overcast day when the sun is low on the horizon.
Because the difference between lab and actual conditions is so large, some manufacturers also publish NOCT4) specs, a derated (lowered) set of specs which might or might not be more indicative of what you will see in your use case. NOCT is another tool in the toolbox, not gospel truth.
In practical terms, it's common to see a maximum of 75% of STC under good conditions, more under great conditions, and much less under poor solar conditions. And about half of that in winter.
The location and season both have profound effects on solar harvest. For example, Phoenix AZ has over 8.5x the amount of solar power available in summer as Seattle WA in the winter. For any given location, winter harvest will be about half of summer harvest.
You can estimate solar harvest using models like PVwatts.
Solar panels do not push power; they respond to the solar charge controller's demand as it tries to meet your present needs. When those needs are minimal (battery already charged, small loads) the controller and panels will be loafing.
Related: is my solar working?
Due to their design, PWM and shunt charge controllers will very rarely allow the panel to run at max output for given conditions. The lower the battery voltage (Vbatt) the lower the panel voltage (Vpanel), therefore the lower the power output.5) The output can be increased somewhat by tweaking battery voltages higher.
MPPT controllers can run the panel at max output when needed, but are much more expensive.
Perhaps counterintuitively, setups with ample panel may see lower peak outputs than smaller setups. This occurs because the overpaneled systems can get the bank charged before local solar noon when maximal harvest can be observed. This is a feature, not a bug. Overpaneled systems are built to meet needs under less-than-optimal conditions. If you want to see Big Numbers, start a huge load at local solar noon under good solar conditions.
Long runs of wire between the panel and controller can result in losses that make the panel appear to be putting out less power. In reality, the lost power has been converted to heat6) in the wiring. Solutions:
Solar panels are dark in color and get very hot. Unfortunately, voltage (and therefore power) output decreases as panel temperature increases. This is the reason an air gap between the panels and the camper's roof is recommended to allow cooling airflow.
The data below, derived from this calculator, show power from a 100W mono/poly panel dropping off as ambient temps rise:
Ambient temp in F | Ambient temp in C | derated power |
---|---|---|
32F | 0.0C | 97w |
40F | 4.4C | 95w |
50F | 10.C | 92w |
60F | 15.6C | 90w |
70F | 21.1C | 87w |
80F | 26.7C | 84w |
90F | 32.2C | 82w |
100F | 37.8C | 79w |
110F | 43.3C | 76w |
So a snowbird who “chases 60” will be losing ~10% of panel output during the warmest part of the day. Snowbirds chasing 70 will be losing ~13% of panel output.
Note: sub-freezing temperatures can push Voc/Vmp above their rated values. If you will use solar in sub-freezing conditions leave plenty Voc headroom in your solar charge controller spec.
Note: that radiated heat from the underside of panels can raise temperatures inside the camper.
The sun will climb in the sky until it reaches its highest point for the day (local solar noon), then will start dropping again. This affects how much power you can harvest:
For a given latitude and time of day the sun's location in the sky is calculable and can give you the cosine of solar zenith angle (“cosine” hereafter).7). You can use the cosine to understand how much power your system might put out.
Examples: if you have 200w of panels, your mppt controller typically yields 83% after derating, and the calculated cosine is .70 then you might expect ~116w in clear conditions at that time in that location. (200 x .83 x .7 = 116.2).
Another way of thinking about this is that panel ratings are given for 1,000w/meter2. At that time and place only 700w/meter2 land on the panel. (1000 x .70 = 700)
Tip: you can work backwards from cosine calculations and observed power harvests to find your system's overall efficiency in different conditions.
All other things being equal, peak harvest will happen at local solar noon8) when the sun is highest in the sky.
This leads to a paradox: some systems with relatively large arrays will have completed Bulk charging before solar noon and so do not have the chance to demonstrate full output. In this scenario a smaller array could theoretically show higher peak output power but could not produce more overall Wh over the course of the day than the larger array.
Insolation9) is the the amount of solar power reaching the panels. This can be affected by:
Even the altitude and type of clouds can affect harvest:
in addition to total sky cover, cloud type is a significant factor in determining the reduction in solar radiation. In particular high, thin cirriform clouds are significantly less effective in reducing solar radiation than are lower, thicker clouds.14)
Based on lux measurements, we can estimate how sky clarity/brightness affects the amount of power theoretically available to the panel. Takeaways: overcast skies greatly reduce output, and there is no meaningful power available at sunrise/sunset.
Also see these anecdotal observations.
Latitude affects both the angle (incidence) of sun to the panels and, to a lesser degree, seasonal hours of daylight. Greater latitudes (closer to the north pole for the US) will have lower overall insolation averages than lesser latitudes. They will have more extreme variation in insolation between summer and winter. The most striking example of this is when those regions have 24hr sun in summer and 24hr night in winter.
Poor insolation affects panel amps (Ipanel) radically but panel volts (Vpanel) stay stable until insolation is very low (like ⇐20%).
This means any power generated at very low levels of insolation15) will likely be trivial. Increasing panel Voc to try to get more power in marginal conditions may not be effective. Consider overpaneling instead.
this subsection has been moved here
Insolation maps attempt to combine the effects of the variables above to estimate hours of full sun16) equivalent (FSE) per day for large areas. This can be helpful when planning moves around the country.
Areas with atmospheric extremes will be outliers when compared against other locations at their latitude. Oregon and Washing, for example are low insolation outliers because of the famously drizzly weather. Phoenix and the southwest are high insolation outliers because of an unusual percentage of sunny days.
You can multiply your panels' temperature derated output by the hours of full sun equivalent to get an idea of the maximum harvest you can expect from the panels.
Example:
Partial shade in good sun can have a drastic effect on panel output. This is because panels are made of strings of individual solar cells; having some strings turned “on” and some “off” (to prevent reversing current into the shaded cells) can result in dramatic power reduction.
This effect can be attenuated somewhat by panel design (bypass diodes), parallel rather than serial panel connections when using PWM controllers, use of amorphous (thin-film) panels, and the use serial rather than parallel panel connections of MPPT controllers.
Some conditions can cause an effective magnification of solar power, the opposite of shade in a way. This is usually caused by reflection of previously-uncaptured light onto the panels – the panels are receiving both normal direct light and the additional reflected light at the same time. Since panels are rated at a lab standard 1000w/meter2 , this multiplication of available light can cause the panels to make more than their rated power, and can cause current and/or voltage to rise beyond rated specs.
Examples:
Insolation figures are given for flat surfaces18). This means when the sun is low in the sky19) you may be able to increase solar yield by tilting the panels toward the sun. In effect the panel is more perpendicular to the sun so it can capture maximal solar energy.
Pro:
Con:
Optimum aiming involves both tracking the suns elevation above the horizon (zenith) and tracking the sun's east-to-west travel (azimuth). Doing both can increase yield ~30% but requires frequent repositioning throughout the day.
Leaving panel tilt in a reasonable default is more common and has milder yield improvments. In this approach the panel is aimed at true21) south22) at a particular angle from perpendicular23) depending on time of year and latitude.
Tilt angles calculated by solarpaneltilt.com:
30 degrees latitude is near the southern border of the US; 45 degrees latitude is near the northern border.
Further reading:
According to lab testing, dusty panels cause a derating of 5%-6%.
Note: these differences exist but are of little practical effect.
Since the latter point is counterintuitive:
at visible wavelengths, overcasts are far from spectrally neutral transmitters of the daylight incident on their tops. Colorimetric analyses show that overcasts make daylight bluer and that the amount of bluing increases with cloud optical depth. Simulations using the radiative-transfer model MODTRAN4 help explain the observed bluing: multiple scattering within optically thick clouds greatly enhances spectrally selective absorption by water droplets.24)
Note: visible light is roughly 400-750nm.25). Below that is UV and above that is infrared.