Irradiance Forecasting for Grid Integration
SOLAR RESOURCE ASSESSMENT AND FORECASTING Kleissl Solar Resource Assessment and Forecasting Lab Abstract
Determining Irradiance using Satellites
Photovoltaic (PV) arrays and other solar energy conversion systems are capable of generating large amounts of electric power, but they produce variable output. Solar resource assessment (How much solar radiation can be typically expected?) and forecasting (How much solar radiation can be expected in the next hour or next day?) are critical to expanding the penetration of solar power on the electric grid. Large grid penetration of energy harvested from the sun requires accurate, site specific estimates of power output for transmission and distribution planning and economic reasons. Satellite remote sensing models, numerical weather prediction, and ground sensors are used to quantify the “solar resource” and provide solar forecasts for time horizons from 10 minutes to 72 hours. At UCSD we are applying all of these techniques to address issues of variability inherent in solar energy conversion systems. Our research group is working towards integrating theses methods as well as developing new approaches to estimate and forecast the solar resource.
The single largest factor affecting the amount of solar radiation arriving at the surface of the earth is cloudiness. Geostationary weather satellites directly image clouds on a global scale and can be used to estimate and map the amount of cloud cover over large geographic regions (Fig. 1). Maps of surface irradiance can be generated directly from these cloud maps (Figs. 1 and 2). Satellite derived irradiance data can be used to determine the best location for new solar power plants and to perform long term economic analyses of new and existing systems.
MA
Figure 5: Map of the forecast GHI [W m-2] in March at midday from the North American Mesoscale model (NAM).
Irradiance Forecasting for Grid Integration Figure 3: (left) 2.5 x 2.5 mile map showing locations of DEMROES solar power resource stations (green markers), solar PV installations (blue), sky imager and ceilometer (red). (right) 1 second GHI during a clear day (Nov 8) and a partly cloudy day (Nov 9).
As the number of installed solar energy systems increases worldwide, accurate, high resolution (temporal and spatial) solar irradiance forecasts are of critical importance for facilitating the penetration of solar power on the electric grid. Utilities and independent system operators use this information to manage generation and distribution. In our lab we are working to improve existing forecasting technology (Fig. 5) and developing novel approaches to accurately forecast irradiance on 10 minute time horizons (Fig. 6).
Figure 1: Geographical distribution of the cloud index [-] (top) and the global horizontal irradiance (GHI) [W m-2] (bottom) derived from GOES-W satellite data using the Perez et al. (2002) model over 235 x 241km.
Figure 4: Effects of rooftop PV on building heat transfer. (left) Thermal infrared image of the underside of the Powell Structures Laboratory roof showing decreased roof temperatures beneath a PV array, which implies a reduction in building cooling load. (right) Diurnal cycle of temperature variations on the underside of the roof. The lowest temperatures are observed beneath the tilted PV array. Figure 2: (left) Map of the mean annual GHI for California. The data were obtained from the National Solar Radiation Database (NSRDB, Perez et al.,2002). (right) Modeling the optimum tilt angle (tilt from horizontal) for fixed-mount PV systems facilitates optimal design and maximizes the output of PV systems.
Figure 6: (left) A total sky imager captures a 360° hemispherical photograph of the sky every 30 seconds. (right) Advanced image processing techniques are used to identify clouds and track their speed and direction. This information can be used to predict the exact time when individual clouds will shade a PV array. Acknowledgments Funding was provided by the U.S. Department of Energy and the California Energy Commission through the California Solar Energy Collaborative.
References 1. Perez, R., Ineichen, P., Moore, K., Kmiecik, M., Chain, C., George, R., Vignola, F. 2002. A new operational satellite-to-irradiance model. Solar Energy 73(5), 307-317. 2. North American Mesoscale Model (NAM). Accessed 04/01/10
Determining Irradiance using Satellites
Photovoltaic (PV) arrays and other solar energy conversion systems are capable of generating large amounts of electric power, but they produce variable output. Solar resource assessment (How much solar radiation can be typically expected?) and forecasting (How much solar radiation can be expected in the next hour or next day?) are critical to expanding the penetration of solar power on the electric grid. Large grid penetration of energy harvested from the sun requires accurate, site specific estimates of power output for transmission and distribution planning and economic reasons. Satellite remote sensing models, numerical weather prediction, and ground sensors are used to quantify the “solar resource” and provide solar forecasts for time horizons from 10 minutes to 72 hours. At UCSD we are applying all of these techniques to address issues of variability inherent in solar energy conversion systems. Our research group is working towards integrating theses methods as well as developing new approaches to estimate and forecast the solar resource.
The single largest factor affecting the amount of solar radiation arriving at the surface of the earth is cloudiness. Geostationary weather satellites directly image clouds on a global scale and can be used to estimate and map the amount of cloud cover over large geographic regions (Fig. 1). Maps of surface irradiance can be generated directly from these cloud maps (Figs. 1 and 2). Satellite derived irradiance data can be used to determine the best location for new solar power plants and to perform long term economic analyses of new and existing systems.
MA
Figure 5: Map of the forecast GHI [W m-2] in March at midday from the North American Mesoscale model (NAM).
Irradiance Forecasting for Grid Integration Figure 3: (left) 2.5 x 2.5 mile map showing locations of DEMROES solar power resource stations (green markers), solar PV installations (blue), sky imager and ceilometer (red). (right) 1 second GHI during a clear day (Nov 8) and a partly cloudy day (Nov 9).
As the number of installed solar energy systems increases worldwide, accurate, high resolution (temporal and spatial) solar irradiance forecasts are of critical importance for facilitating the penetration of solar power on the electric grid. Utilities and independent system operators use this information to manage generation and distribution. In our lab we are working to improve existing forecasting technology (Fig. 5) and developing novel approaches to accurately forecast irradiance on 10 minute time horizons (Fig. 6).
Figure 1: Geographical distribution of the cloud index [-] (top) and the global horizontal irradiance (GHI) [W m-2] (bottom) derived from GOES-W satellite data using the Perez et al. (2002) model over 235 x 241km.
Figure 4: Effects of rooftop PV on building heat transfer. (left) Thermal infrared image of the underside of the Powell Structures Laboratory roof showing decreased roof temperatures beneath a PV array, which implies a reduction in building cooling load. (right) Diurnal cycle of temperature variations on the underside of the roof. The lowest temperatures are observed beneath the tilted PV array. Figure 2: (left) Map of the mean annual GHI for California. The data were obtained from the National Solar Radiation Database (NSRDB, Perez et al.,2002). (right) Modeling the optimum tilt angle (tilt from horizontal) for fixed-mount PV systems facilitates optimal design and maximizes the output of PV systems.
Figure 6: (left) A total sky imager captures a 360° hemispherical photograph of the sky every 30 seconds. (right) Advanced image processing techniques are used to identify clouds and track their speed and direction. This information can be used to predict the exact time when individual clouds will shade a PV array. Acknowledgments Funding was provided by the U.S. Department of Energy and the California Energy Commission through the California Solar Energy Collaborative.
References 1. Perez, R., Ineichen, P., Moore, K., Kmiecik, M., Chain, C., George, R., Vignola, F. 2002. A new operational satellite-to-irradiance model. Solar Energy 73(5), 307-317. 2. North American Mesoscale Model (NAM). Accessed 04/01/10
