Comparison Testing of a Solar System with CSTG and DST Methodologies
Abstract
The Solar Thermal Testing Laboratory of CENER performs outdoor efficiency tests for factory-made solar systems according to international standard ISO 9459-2, using the CSTG method, as well as ISO standard 9459-5, using the DST method. The first method (CSTG for "Collector and System Testing Group", also called Input-output method) consists of three different parts: one part for determining mixing in the storage tank during draw-off, another part for determining daily system performance, and the last part for the determination of storage tank heat losses. The efficiency test in the DST method (also called Dynamic method) consists in different test sequences with different system behaviors: S-Sol for characterizing the collector array performance at high efficiencies, S- Store for characterizing store heat losses and collector array performance at low efficiencies and S-Aux for determining the heat losses and the volume fraction of the auxiliary heated portion of the store. In both methods the result is a characterization of the solar system thermal behavior and then a long-term performance prediction. For the long term performance prediction, the thermal output energy of the solar system (Q) and solar fraction fsol at on different reference locations and for different load volumes are calculated. In this study we tested two solar thermosyphon systems according to both methods. The purpose of the paper is to show the results of the measuring output energy (Qmed) compared to the modelized output (Qmod) and to analyze the system characterization obtained for each methodology. Then we will compare the long-term prediction results obtained for those two solar thermosyphon systems using both methods, as done in Carvalho et al. (2000) and Kaloudis (2010) et al. We will analyze the causes of the maximum differences between both test methods for in the new results as well as in the context of the literature results. A new approach which includes the uncertainty derived of applying the literature conversion factors is proposed.