Comparing Solar Panels


Works in theory! Practice? That's something else
Efficiency is only relative to the physical size of a solar panel; that is a 200W panel produces 200W at STC regardless of its efficiency or size.

So, if all panels of an equal wattage produce the same power... how do you compare solar panels other than price, warranty, & physical size?

First, it seems there's a few break/make factors:
You can detect salt from the pacific ocean in South Carolina. Fortunately for most of us you only need to worry about it if you're within a mile of a salt coastline, live somewhere they're still salting the roads, a desert area, a salt pan, well... you get the idea. For this, look for the IEC 61701 rating if it applies to you.

IEC 62716 (Ammonia corrosion testing) may be important to you as modules that are installed onto livestock farms and greenhouses are subjected to particular environmental conditions (ammonia from fertilizer) that may be very hostile to regular solar panels.

You can Calculate Pounds/sqft force from wind speed by: PSF = 0.00249 MPH^2 and 1 pascal (Pa) = 0.021 pounds per sq. foot (psf)
So, at first, I thought that from a manufacture's panel datasheet I could assume a wind rating of 2400 Pa is rated for 145 mph winds, 5400 Pa is rated at 208 MPH.
But not so, the tilt and the attachment points (e.g., in an HVHZ you can add a third rail) of the panel are uber important. See Basics of Wind Loading.

Roofs have "classes": A (e.g., concrete tile, fairly fireproof), B, C, and unrated (e.g., cedar shake roof). Originally all panels were class C, and so they came up with a subdividing class I, II, III. Then it changed again to types 1 through 17. This is all really confusing to me. Still is, but try SolarABC's explanation.

Some local ordinances require an anti-reflection coatings (ARC) on panels. Bare silicon has a high surface reflection of over 30%. The reflection is reduced by texturing and by applying ARC coatings to the surface.

Once you've eliminated panels using the above factors above, it's time to crunch the economics. I think the general way to do this is to calculate all the watts the array could generate over it's warranty life; one year at a time, then multiply by the cost of power accounting for inflation in that year, sum them all, then divide by the cost of the array. Or do what I did, ask SAM.

Effect of Temperature Coefficient (Pmax)
Standard test conditions are at 25 degrees C, and the datasheet will have a value for Temperature Coefficient (Pmax). I've seen them vary from -0.25 to -0.43 % / °C . Basically, this is the percent power lost from the panel for each degree above 25, or gained below 25. NREL and some panel datasheets have NOCT data, which provides the number at 20 degrees C with a 1 m/s breeze. NREL may also have hourly temperature data for your location for the last few years.

During operational hours, Panel temperature is higher than ambient temperature: Tcell = Tair + (NOCT-20)/80*s; where S = insolation in mW/cm2.

So how big a difference does this make? You could use the meteorological data from NREL and calculate it all... but I was too lazy. Instead, I created a hypothetical ~8 kW system in SAM and using 11 cents per kWh calculated the costs of the lost power lost over its lifetime.
Pmax         Annual Output     delta from -.25   Lifetime Output  loss        $/life
-0.25         11,088
-0.30          9,498                     1590            39,750               4,372
-0.35          8,635                     2453            61,325              $6,745
-0.40          7,737                     3351            83,775              $9,215

Does that mean you should go for the lowest pMax?
Nope! It'll depend on your local daytime temperatures; for a cold climate that would reverse. But, if you live where it's usually above 25 C from 9 am to 5 pm on the roof, then a low Pmax might be for you.

Effect of Panel Degradation
Most panels offer a 1st-year loss followed by linear decline. For example, Biku panels will generate 80% of their power at 30 years. Sunpower will generate 94% of its rated power in 25 years. Panasonic HIT panels also have good life characteristics.

Unfortunately, SAM doesn't seem to factor that in automatically, you have to manually input the numbers if the Lifetime screen. It also won't account for the big first-year loss. So, let's say we use a 7.25kW that provides about the power needed. I thought this would have a bigger impact, but it turned out to only be less than ~$60/yr.

loss/yr%     Total $saved@ 2.5% inflation    Notes
   0             52,000                      Hypothetical
.3586            49,542                      Panasonic HIT
.6               48,108                      Q CELLS 325 WATT 85% @ yr 25

Where the diodes are can make a difference for partial shading.

Consider the two panels to the right. The left is a standard panel, the right a typical
120 half-cut cell panel.

In the left panel, snow blocking the bottom would negatively impact the entire panel.
In the right panel, snow on the bottom only affects the bottom half of the panel. A
leaf shading the left columns of the left panel doesn't impact the rightmost
columns. Similarly, a leaf on the lower left column of a 120 half-cut cell panel affects
the panel even less.

Shade should be avoided and can invalidate the warranty on some panels. Where
shading is severe microinverters, optimizers, or the Maxim chip may be warranted.

Monocrystalline tend to slightly outperform polycrystalline in low-light conditions
(e.g., cloudy days), but technology is always changing.​

Grade (A,B,C,D)
Usually, panels from the top-tier providers are all Grade A cells. Most people should stay away from Grade C & D cells as they are off-spec and you can't know what you're getting. Grade A & B cells are both within spec, the difference is that Grade B cells have visible defects. (ref)

Other Considerations
Finally, there are some other factors that may or may not apply to you. For example, if you live in harsh conditions double-glass (goes by a variety of names) typically have 3x better resistance to environmental factors. If you're planning on using a string inverter with panels in series and parallel, consider what happens in a decade - will you be able to get a replacement with matching voltages/current?

Where to get Solar Panel Data
The California GO database is the best listing of consistent and complete information I've come across
Last edited: