Implementation of Data Processing and Automated Algorithm Based Fault Detection for Solar Thermal Systems
Abstract
Solar thermal systems are designed to function for about 25 years, but faults and malfunctions are likely to occur at a certain time causing reduced solar gains and economic losses. Due to the fact that usually a backup heater is installed to guarantee the supply of the heat demand, faults and malfunctions of the solar thermal systems are often not recognized or only with a delay. Hence, fault detection is an important issue. Because a manual monitoring would be time and cost intensive, an automated detection method is needed. In this paper an automated algorithm based fault detection is presented. The data transfer and storage is done via data emails, which are written into a server based database guaranteeing the usage of different structures of data files. Data can then be accessed everywhere, so that the data analysis can be handled on different computers. This leads to the possibility to use several computers at once in order to increase the performance of the fault detection. The fault detection itself is carried out by functions (algorithms) analyzing the data for deviations from normal operating conditions. The structure of the database and the different levels of algorithms used for the fault detection as well as an example implemented on a monitored solar thermal system are presented in this paper.