Home ] [ Up ] [ UCEP Feedback Page ] [ Table of Contents ] [ SEARCH ]



Performance of the GTAC PV System for 1996

The Georgia Tech Aquatic Center (GTAC) photovoltaic (PV) system was brought on-line in
late June, 1996. Starting in July 1996, the Georgia Tech University Center of Excellence
for Photovoltaic Research and Education (UCEP), with support from the U.S. Department of
Energy, Georgia Power Company, and Georgia Tech, has monitored the PV system’s performance
and produced reports of this performance each month. This page contains a summary of the
system’s performance during the period July-December, 1996.

  • AC Energy Production [Mwh]


Month Measured Predicted*
July 33.2 45.3
August 31.3 43.7
September 28.7 32.6
October 17.4 33.2
November 18.1 23.7
December 15.2 19.4

* The "predicted" AC energy production for
each month was found using the computer model PVFORM 3.3, with the TMY providing
metheorological data. Most of the difference between predicted and measured AC energy
outputs during this startup period for the system is due to system down time, necessitated
by various adjustments and repairs on the PV array. For example, in July 1996, a lightning
strike on the roof blew fuses on one of the seven source circuits on the roof. This was
detected relatively quickly (in about six days), but acquiring parts and making the
repairs took longer than expected, in part due to the Olympic security that was in place
at that time. This event explains most of the difference between predictions and
measurements in July 1996. Further repairs and correction of minor startup difficulties
were made in August, after the Olympics. In October, a plumbing failure led to a nine-day
shutdown of the PV system. Finally, in December, about 20% of the array was disconnected
during the last half of the month as part of an experiment being conducted by the UCEP.

  • Mean daylight-hours module temperature [oC]


Month Measured Predicted*
July 42.4 37.4
August 42.1 38.4
September 36.4 32.2
October 31.4 26.8
November 19.6 18.6
December 17.1 12.0

* The measured mean daylight-hours module temperatures are typically
a few degrees higher than the predicted ones. We believe this is due at least in part to
the roof curvature, which impacts air flow above and beneath the array, and standing
seams, which may impede airflow behind the modules more than expected.

  • Daily average solar energy flux [MWh/m2/day]


Month Measured Predicted
July 5.52 5.97
August 4.86 5.76
September 4.19 4.43
October 4.20 4.29
November 2.76 3.11
December 2.58 2.44

  • Average sunlight-to-AC-power efficiency [%]


Month Measured
July 6.1
August 6.6
September 6.9
October 7.3
November 7.7
December 8.2

* This "average" efficiency was found using only the times
during which the system’s DC power was higher than 80 kW. For lower DC powers, sensor
inaccuracies cause the values to be suspect.