Automated monitoring, failure detection of key components, control strategies and self-learning controls of key components

Abstract

Recommendations for an automatic monitoring process of solar thermal systems are given - starting from sensors to data acquisition, data storage, computation of benchmarks and fault detection.

Efficient Gathering, Storing, Distributing and Validation of Data

“…IEA SHC Task 55 Fact sheet BD.3 [3] A fact sheet about automated monitoring, listing recommendations for monitoring solar-thermal plants. The report contains a list of recommended measurements, installation conditions, and sensor specifications.…”

“…For example, irradiance data is required to calculate the effective collector efficiency , estimate thermal power according to the collector Keymark equation (ISO 9806) , and carry out the Thermal Power Check as defined by ISO 24194 [9] and the Dynamic Collector Array Test [13]. For non-concentrating collectors, it is often sufficient to measure the hemispherical solar irradiance on the collector surface [3,8,9]. In contrast, direct normal irradiance (DNI) is required for concentrating collectors [8,9].…”

“…This section provides recommendations regarding data loggingtransforming the input of the measurement sensors into digital output and recording the results. The content of this section is mainly based on the IEA SHC Task 45 fact sheet [1,2], Task 55 fact sheet BD.3 [3], and lessons learned from developing the SunPeek software [20]. A summary is provided in Table 12.…”

“…For further processing, the time zone of the timestamps must be correct, for example, to compute the angle of incidence or run plausibility checks like "no irradiation at night." A lesson learned from the SunPeek project [20] is that storing data in UTC is recommended, as time zones can change 2 and can often lead to bugs 3 , especially in the case of timestamps with daylight saving. Thus, the recommendation is to use UTC for storing the data.…”

“…Resampling to equidistant timestamps for easier computation of key performance indicators and for merging data from multiple files. Typically done with linear spline interpolation [3], [18], [26]. Interpolation (short-term) Replaces short periods of NAN data with linearly interpolated values.…”