Computer Science · Artificial Intelligence
Application of high resolution insolation data for photovoltaic system design analysis
See on the mapAbstract
It has been suggested that weather data recorded hourly in fluctuating conditions (e.g., partly cloudy days) can mask significant changes and may lead to erroneous estimates of photovoltaic energy production. To demonstrate this, the computer package PVFORM (version 3.2) was modified to base its estimates on 10-minute rather than hourly data. Both the original and modified programs were then used to analyze the same one year's worth of meteorological data from eastern Virginia. Estimates of both plane-of-array insolation and DC energy output were compared to the actual recorded values. The estimates based on 10-minute resolution were found to be consistently more accurate than their hourly counterparts. Based on the results obtained, it is recommended that if photovoltaic system performance predictions are required with an hourly resolution, only multiple-reading hourly input data should be used. If monthly (and to some extent daily) resolution is all that is required, then snapshot-type hourly data may give acceptable predictions.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>