The solar resource curve is different for each location, and will continue to change daily, quarterly, and yearly. Unfortunately, most of the proposed solar power plants do not have relevant specific data. Therefore, developers, financiers and project beneficiaries can only rely on published irradiance datasets. These proposed power station data are developed and published by government agencies. While a large number of proprietary datasets based on satellite imagery are readily available, they are often limited by record length.
Irradiance datasets typically contain global irradiance (global irradiance, or GHI), which is the sum of the direct irradiance and scattered horizontal irradiance (DHI) projected onto the horizontal plane by the solar normal direct irradiance (DNI). Datasets for electrical power generation also contain other key meteorological data, such as dry bulb temperature, wind speed and dew point temperature, as predicted accurately. However, for most solar systems, temperature and wind speed data are only second-level data compared to irradiance data.
While Concentrated Solar Power (CSP) systems rely only on DNI, photovoltaic (PV) systems typically use both DNI and DHL, and a conversion model must be employed to calculate the total irradiance (POA) data available on the PV system’s array panels . The conversion system includes all solar irradiance factors, in addition to other factors (depending on the model) that affect POA irradiance, including ground reflected irradiance, scattered irradiance around the sun, and horizon boosting effects. If available, the model will also account for system design parameters, including array orientation, tilt, and 1- and 2-axis tracking. Within the industry, various transformation models that are routinely used are well documented in the literature, notably the Hay and Davies model, the Perez model (Dufle and Beckman, 1991), although the Perez model is generally more complex, it has a horizon-enhancing ( horizon-brightening) function (which is ignored by the Hay and Davies models). The Hay and Davies model supports project investment based on actual data analysis of POA irradiance, thereby avoiding any uncertainties and potential risks that newer models often overestimate (although different kinds of models The difference is usually small).
Most historical datasets provide irradiance information on an average hourly basis. Hourly data can usually make adequate predictions about the average performance of a solar system, but high-resolution data is more suitable for accurately analyzing the system’s transient performance.
In order to support the rigorous evaluation and financing of projects, the most critical factor in assessing solar resource risk is the annual solar resource, usually expressed as total irradiance (kWh/m²) or daily average irradiance [kWh/(m².d)] , the detailed data refer to DNI data in CSP and POA radiation in photovoltaic power generation (PV). However, the different distribution of solar energy resources on a daily, monthly and quarterly basis will affect the profitability of the project.
Assessments of solar energy resources and their associated risks typically involve multiple data sets over different time periods. Data collection for repeated time periods helps improve the continuity and consistency of the dataset. A comprehensive risk assessment for financing must be based on a comprehensive and detailed data analysis, and any related inquiries and questions will help minimize the risk of solar resource project siting.