This 9.2 kWp photovoltaic (PV) array is located at Humboldt State University’s Telonicher Marine Laboratory in Trinidad, CA. The array was installed in 1990 and is situated approximately 150 meters from the ocean in a cool, marine environment. It played an integral role in the Schatz Solar Hydrogen Project, a renewable hydrogen demonstration project decommissioned in spring 2012. The primary function of the SERC array was to power an air compressor for the marine laboratory. Excess power produced by the array was shunted to a Teledyne Energy Systems ALTUS™ 20 electrolyzer to produce hydrogen fuel for a proton exchange membrane (PEM) fuel cell. Together, the PV array and the PEM fuel cell powered the air compressor 24 hours a day using renewable solar energy.
In 2001 the data acquisition and control system of the array was modernized. In 2006 the array was rewired from 24V to 48V operation and connected to a set of maximum power point trackers as part of a major rebuild. In 2011, as part of a maximum power point tracking hardware test, the array was again rewired for 192V operation and converted to a grid-intertie system, so that any excess power produced by the PV array was fed directly back into the grid.
The PV array consists of 192 ARCO M75 PV modules, a single-crystal silicon module with a rated power of 48W. The individual PV cells are laminated to tempered glass with ethylene vinyl acetate (EVA). The modules were originally configured in 12 independent subarrays, each consisting of 16 modules wired as 8 parallel strings of 2 modules each for nominal 24 Volt DC operation. In 2001 the system was rewired as 6 independent subarrays, each consisting of 32 modules wired in 4 sets of parallel strings of 8 modules in series pairs for 48 Volt DC operation. In 2011 the array was rewired as 12 subarrays, half consisting of 16 modules wired as a single string and the other half as 8 parallel strings, each consisting of 2 modules and a maximum power point tracking voltage converter. The array is tilted permanently at an angle of 30 degrees to the horizontal. One module was damaged in 1996 and replaced with a module with similar size and performance characteristics, a Siemens SM50-H. Two other modules were later replaced due to physical damage.
Prior to initial installation in 1990, SERC research engineer Jim Zoellick generated current-voltage (I-V) curves to describe the performance characteristics of each of the 192 modules. The purpose of Zoellick’s study was to investigate the effect of mismatch losses on power output from the array. The data indicated that mismatch losses were very small, averaging 0.1%, and that the actual field performance of the modules was lower than the nameplate rating. The pre-service data provided a valuable baseline for future analysis of the array. See Effects of Mismatch Losses in Photovoltaic Arrays (PDF; 446K).
Between September 2000 and June 2001, research engineers Antonio Reis and Nate Coleman generated I-V curves and performance parameters for the 191 original ARCO M75 PV modules to evaluate the changes in their performance after 11 years in the field. Each module was removed from the array and cleaned prior to testing. These new data were compared to the original 1990 data to determine the extent to which each module’s power production had degraded over time. Reis and Coleman found that maximum power output for the modules decreased by an average of 4.39%. This power loss was due mainly to reduced current output, believed to be caused in part by the following visually observable physical defects:
Reis and Coleman presented the results of the study, “Comparison of PV Module Performance Before and After 11 Years of Field Exposure,” (PDF; 275K) at the 29th IEEE Photovoltaics Specialists Conference in May 2002.
In 2010, SERC Engineers Mark Rocheleau, Marc Marshall, and Scott Rommel tested every module for the third time after twenty years of service. The degradation processes continued and appear to have accelerated over the intervening nine years. The average peak power had declined from the 2001 levels by an additional 12.4% to 33.43 W and ranged from 9.64 W to 37.95 W. Not only is the average peak power steadily falling, but the variation among the modules is also dramatically increasing. The impact of increased variation is to accentuate the mismatch among the modules, which in turn reduces the peak power output of the array as a whole.
Charles Chamberlin presented the results of the study, “Comparison of PV Module Performance before and after 11 and 20 Years of Field Exposure,” (PDF; 14.6MB) at the 37th IEEE Photovoltaics Specialists Conference in June 2011.