The Relationship between GHI and Power in Kuwait

Understanding the relationship between observed GHI and measured solar power output is crucially important for energy utility companies. An accurate forecast of expected solar power output supports effective power resource management, avoiding brownouts and blackouts. It also supports significant power cost savings. The Kuwait Institute of Scientific Research (KISR) has requested the National Center for Atmospheric Research (NCAR) to produce accurate power forecasts for the new Shagaya Renewable Energy Park in western Kuwait. KISR will utilize this information for power resource planning and grid stability. In order to accomplish this task, NCAR will be producing regular forecasts of GHI, and will then convert the forecasted GHI to outputted power using a machine learning model.

In order to set up an accurate machine learning model, one must have a solid understanding of how GHI and other factors relate to power. To gain such an understanding, an in-depth analysis of solar datasets from Shagaya was performed. This analysis revealed several factors that influence the relationship between observed GHI and measured power output. Some of these factors include dust, time of day, cleaning and maintenance schedules for both the solar panels and pyranometers, precise physical configuration details, and meteorological variables.

This presentation will highlight some of the most important factors that affect the GHI-to-power conversion at Shagaya. It will show how these factors change the relationship between GHI and power, and how machine learning models can be configured to produce more accurate power forecasts. It will also highlight the importance of consistent maintenance of the GHI sensors and the solar panels that generate power.