Shadow camera system for the generation of solar irradiance maps

Abstract

International audienceHighly spatially and temporally resolved solar irradiance maps are of special interest for predicting ramp rates and for optimizing operations in solar power plants. Irradiance maps with lead times between 0 and up to 30 min can be generated using all-sky imager based nowcasting systems or with shadow camera systems. Shadow cameras provide photos of the ground taken from an elevated position below the clouds. In this publication, we present a shadow camera system, which provides spatially resolved Direct Normal Irradiance (DNI), Global Horizontal Irradiance (GHI) and Global Tilted Irradiance (GTI) maps. To the best of our knowledge, this is the first time a shadow camera system is achieved. Its generated irradiance maps have two purposes: (1) The shadow camera system is already used to derive spatial averages to benchmark all-sky imager based nowcasting systems. (2) Shadow camera systems can potentially provide spatial irradiance maps for plant operations and may act as nowcasting systems. The presented shadow camera system consists of six cameras taking photos from the top of an 87 m tower and is located at the Plataforma Solar de Almería in southern Spain. Out of six photos, an ortho-normalized image (orthoimage) is calculated. The orthoimage under evaluation is compared with two reference orthoimages. Out of the three orthoimages and one additional pyranometer and pyrheliometer, spatially resolved irradiance maps (DNI, GHI, GTI) are derived. In contrast to satellites, the shadow camera system uses shadows to obtain irradiance maps and achieves higher spatial and temporal resolutions. he preliminary validation of the shadow camera system, conducted in detail on two example days (2015-09-18, 2015-09-19) with 911 one-minute averages, shows deviations between 4.2% and 16.7% root mean squared errors (RMSE), 1.6% and 7.5% mean absolute errors (MAE) and standard deviations between 4.2% and 15.4% for DNI maps calculated with the derived approach. The GHI maps show deviations below 10% RMSE, between 2.1% and 7.1% MAE and standard deviations between 3.2% and 7.9%. Three more days (2016-05-11, 2016-09-01, 2016-12-09) are evaluated, briefly presented and show similar deviations. These deviations are similar or below all-sky imager based nowcasts for lead time zero minutes. The deviations are small for photometrically uncalibrated, low-cost and off-the-shelf surveillance cameras, which is achieved by a segmentation approach

Validation of spatially resolved all sky imager derived DNI nowcasts

“…To the best of our knowledge, this is the first time nowcasted two dimensional irradiance maps are compared with spatially resolved reference irradiance maps. The reference system is briefly described here and in more detail in [4]. In this publication, a thorough validation of the reference system including additional improvements is presented.…”

“…2b): The field aggregated persistence error metrics do not rise significantly with increasing lead times as the considered temporal and spatial variabilities are similar for most situations. This effect is further discussed in [4].…”

“…1a). A detailed validation of the reference system is given in [4] and preparatory works regarding the reference system are presented in [5]. In this paper, the focus lays on the methodology to validate and benchmark nowcasting systems.…”

Applications of a shadow camera system for energy meteorology

“…The irradiances (DNI -Direct Normal Irradiance, GHI -Global Horizontal Irradiance, GTI -Global Tilted Irradiance) for the shaded areas are derived from pixel intensities of the current orthoimage relative to normalized pixel intensities of the reference orthoimages. The shadow camera system is presented and validated in Kuhn et al (2017a). Comparing pixels of the irradiance maps to corresponding ground measurements for one-minute temporal averages, the shadow camera system shows deviations of RMSE (DNI) between 4.2 and 16.7 % and RMSE (GHI) deviations below 10 %.…”

“…This approach is based on downward-facing cameras ("shadow cameras") that take images of the ground from an elevated position. In these images, the brightness of the ground as influenced by cloud shadows can be seen and converted to cloud shadow and irradiance maps (Kuhn et al, 2017a). Furthermore, the application of a shadow camera system for the validation of all-sky imager based nowcasting systems is discussed.…”