Does PVWatts model hourly wind speed?

[nerdralph]

I've been testing a number of design tools such as Sunny Design, Ginlong Solis design and PVWatts. PVWatts estimates a surprisingly low amount of clipping even at 1.3:1 DC:AC ratio, with system losses set at 5% and 97% inverter efficiency. The location is Halifax, Nova Scotia.
When I downloaded the hourly data, the wind speed column is all 0, and the predicted temperatures seem consistent with no wind.
PVWattsHourly.png

With no wind, the system temperatures are higher, and therefore the predicted array output is lower. Under totals, wind speed shows as 17m/s, so maybe it is using the annual average rather than hourly? In case it is an issue with the international data, I tried Banor, ME, and it also shows zeros for hourly wind speed.

[J.P.M.]

PVWatts uses one of several weather databases that all include mostly modeled and a few measured weather variables, including wind velocities and vectors. Not much of any of trhe data bases contains much in the way of actual, measured weather data, but the data bases PVWatts uses do have a lot of modeled wind vector data.

Consult NREL for details and its National Solar Radiation Data Base (NSRDB) or the PVWatts help screens for more info. on weather data bases and remember that PVWatts and other models are just that: models used for system design and not predictors of short term performance performance.

BTW, I'd bump the system loss parameter to something close to 10 %. The model default is 14 % but many users have and continue to be finding that the 14 % leads to results that are more conservative than actual measured data seems to show.

Read all the PVWatts help/info screens, read the TMY manual and look at the NSRDB for more info and education.
In modeling, GIGO applies.