Robust solar radiation datasets are key to competitive solar project financing. Financiers generally consider solar resources to be stable from year to year compared to other renewable resources, and view solar resource miscalculations when assessing them as the biggest risk in solar projects. Lenders and credit rating agencies will therefore require validation of solar resource datasets to predict electricity production and revenue for each project site. The variability of solar resources exhibited by historical solar data, as well as the accuracy of the dataset, both play an important role in predicting future performance probabilities. They also affect financial contracts that may be associated with the project.
While an increasing number of dedicated solar resource datasets are being compiled and sold on the market, most datasets are derived from publicly available data. Most of the new datasets are based on models from satellite imagery and validated with ground-based measurements. Although the content of the review is related to newer commercial datasets.
In order to build a reliable and profitable dataset, it is important to understand the variability of solar energy resources and the uncertainty in different data components. This article will analyze two widely used datasets, the US National Solar Radiation Database (NSRDB) jointly launched by the National Renewable Energy Laboratory (NREL) and the US Sandia National Laboratory, and the Canadian Weather Service provided by Environment Canada. Resources and Engineering Dataset (CWEEDS). In discussing the data in the dataset, this article will describe the method of obtaining the data, as well as the Typical Meteorological Year (TMY) data file compiled on the basis of the dataset. While TMY data files may be suitable for initial evaluation, they generally do not constitute a profitable dataset. Specific examples will be provided to illustrate the limited value of TMY data files and to explain why long-term datasets that create these data files must be utilized.
With the increasing use of satellite-derived data in resource assessment, especially in developing countries, this article will discuss such data. Since developing countries do not have long-term ground-based measurement datasets, or have very limited datasets available, satellite imagery data needs to be used in resource assessments. In addition, the data in the NSRDB database from 1998 to 2005 in the article comes from a model using satellite imagery. The paper also examines ground-based measured irradiance data values and their accuracy, and illustrates the importance of linking measured data to long-term data sets. Finally, it describes the establishment and use of profitable datasets through the NSRDB database and other available datasets, and summarizes the main characteristics of profitable datasets, uncertainties and their impact on the determination of financing terms.
