PVWatts vs. CSI Calculator = Way Off

[cebury]

I ran PVWatts, which gave VERY different Estimate KWH results than the CSI De-Rating/Incentive Calculator.

PVWATTS Input has:
* AC rating (de-rated from value-input off the DC rating)
* Zip Code to gather SunHours (PV-Watts was sent 5.8 for my zip).
* Tilt & Azimuth

To attempt Identical Calculations:
* Set PVWatts DC-AC-Derate Factor so both AC ratings are the same between the two sites (so PVwatts = CSI-epbb ac-cec rating it calculated based on my panels+inverter).
* Same Zip code, Tilt (22 deg) and Azimuth (180) used on both sites.

The CSI site allows you to de-rate for shading in each month, but even when I factor that in PVWATTS is over 3000 annual kwh in excess.

Is PVWatts "averaging" for shading or something we're not able to see or adjust? I guess the CSI site could be doing the same thing for SunHours, but the rest of the input factors you either see what it's doing or can control.

[cebury]

Of course after I type that all up I think I found out why: PVWatts attempts a more thorough de-rating factor (info at: http://www.nrel.gov/rredc/pvwatts/ch...ers.html#dc2ac) for wiring,soiling, etc to default to .77. The value I entered was .86 to adjust it to the CSI site (which is AC-CEC, obviously not the same). The CSI De-Rating calculation gets closer by factoring in tilt, azimuth, and shading, but stops there. So by taking the CSI Rating then adjusting for the 6 factors that pvwatts is adjusting for (using defaults) that makes my proposed system DERATE = .71, below the default .77. Which I think? makes sense for fairly inefficient Yingli 235 panels.

That makes the two numbers only off by 200 annual kwh.