Life Cycle Performance Evaluation of an Air-Based Solar Thermal System

Abstract

Performance at the McKay center after 22 years of operation was characterized through experimental measurements of component parameters. Effects of the component parameters on systems level performance was evaluated using numerical simulation tools (TRNSYS) calibrated against actual operation of the building. Calibration of the collector arrays indicated degradations of 16 percent in Fr(??) and 19 percent in FrUL on average for the two arrays installed at the site. Thermal losses from the pebble beds were characterized by conductances of 7.5 and 3.0 W/m2-C for the two beds, which far exceed the design conductance of 0.65 W/m2-C. Flow distributions within the beds were highly non-uniform with an average 3:1 flow ratio across the pebble beds. Effects of component degradations did not directly translate into similar systems level degradations. Annual simulations performed under typical meteorological year (TMY2) conditions indicated that the high bed losses and non-uniform flow have resulted in a 1 percentage point decrease in the solar fraction. Degradations in the collector array had a more significant effect on system performance, lowering the solar fraction by 12 percentage points. Together, physical deteriorations of the solar components have lowed the solar fraction from an optimum of 45 percent to 32 percent. The single largest effect on system performance was not attributable to a physical component, but to the control logic that governs system operation. Faulty operation of an air damper within the system and improper location of a control sensor were responsible for the low solar fraction of 15 percent representative of how the system was found at the inception of this study. In-situ testing methods and novel calibration techniques are described that result in minimal cost and interference of system operation. Systems level calibrations employing daily integrated energy comparisons are also discussed along with the sensitivity of the model to load dynamics. Possible methods of fault detection for solar thermals systems and a minimal instrumentation package are presented that may prevent similar faults from compromising the performance of present and future solar installations.