Characterization of the Solar Resource - A Brief Overview
Sunlight is the main driver of natural processes on Earth. However, the availability of the solar resource is not equally distributed across the surface of the planet and can vary considerably even across small distances. This is due in part because the availability of the solar irradiance on the ground is a function of many environmental variables, such as aerosol concentration, local microclimate, etc.
The design and operation of solar energy generation systems require the knowledge about the local solar resource, and the more accurate and the longer the solar data sets are available for the site of interest, the better. Vignola et al. (2012) point out that bankable solar irradiation data sets are crucial for the financing of solar energy generation projects and coupling ground-based solar irradiance data with satellite data for the same period plus years of historic irradiance data is one way of building such bankable datasets . In other words, multiple ways of measuring the solar resource are needed to properly characterize its availability to the location of interest for the development of solar energy generation projects. Reliable solar irradiance data is also a requirement for the production of reliable solar energy forecasts, important information for solar PV system owners, and grid operators. However, the availability of measured local solar irradiance data is often constrained by the cost of the radiometric systems, especially in the case of the most accurate ones . And when ground-based solar irradiance data is available close or at the site of interest, this data is frequently limited to recorded measurements of the global horizontal irradiance (GHI).
However, non-horizontal surfaces (e.g. PV arrays and vertical façades) cannot use GHI data directly, and so empirical correlations must be applied to transform GHI into the plane of array (POA) irradiance, reflecting the anisotropy of the solar resource largely for "clear sky" conditions. However, these empirical relations have their limitations, and frequently are an important source of uncertainty for forecasting the solar resource. Such uncertainty can be avoided by using more comprehensive solar radiometric systems, capable of providing data on the different components of the solar irradiance and/or measuring the irradiance at the POA. Hence to improve the assessment of the solar risk for decentralized photovoltaic (PV) power production, smart buildings, smart grids, and energy storage solutions, it is useful and why not necessary to have more detailed information about the solar resource at a specific site. In special, data on the direct normal irradiance (DNI), the component of the sunlight which carries most of the available solar energy, in addition to GHI would be especially welcomed.
In the USA high-quality solar irradiance data is available at a few localities around the country. For example, the National Oceanic & Atmospheric Administration (NOAA) operates and maintains the Surface Radiation Budget Network (SURFRAD) which aims to "support climate research with accurate, continuous, long-term measurements of the surface radiation budget over the United States". Measurement includes GHI, albedo, direct normal and diffuse horizontal irradiance, photosynthetically active radiation, UVB, spectral solar, and other meteorological parameters such as temperature and wind speed and direction. There are, however, only seven of these stations in the US.
NOAA SURFRAD station at Sioux Falls, SD. From left to right: Rotating Shadowband Radiometer, a photosynthetically active radiometer (PAR), a UV-B radiometer, and a pyranometer measuring GHI. Then a Pyrheliometer is shown mounted at a solar tracker. Source: NOAA
Open-Source Solar Irradiance Data
The Solar Ecology Research Collaborative developed two systems to measure the solar irradiance that are at the same time affordable and reliable: the Solar Irradiance Detector (SID) and the All-Seeing Eye (ASE).
Solar Irradiance Detector (SID) System - Working Prototype
All Seeing-Eye (ASE) Multipyranometer Radiometric System
The SID systems are designed to measure GHI and/or POA data. The system thus can be composed of one or two pyranometers while the ASE is developed to provide information on GHI, DNI, and DHI (and also POA, if needed). The main difference between the SID and the ASE is the number of pyranometers that the systems require (minimum of 1 sensor for the SID versus 5 for the ASE system). The ASE system also uses a machine-learning algorithm trained to return estimates of the DNI from the inputs of the five pyranometers (plus the calculated solar zenith angle). Both SID and ASE also measure the ambient air temperature and internal system temperature.
The SID and ASE systems are designed to be affordable, this way, to lower the cost of the system, we opted to use additive manufacturing to produce the parts needed to protect the electronics of the radiometric systems. The housings for the sensors and data logger were designed and printed using a combination of CAD software and fused deposition modeling (FDM), also known as 3D printing. Besides printing our parts, we also chose to use electronic components that are easy to configure and could be virtually assembled by everyone.
 Vignola, Frank, Cathy Grover, Nick Lemon, and Andrew McMahan. "Building a bankable solar radiation dataset." Solar Energy 86, no. 8 (2012): 2218-2229.
 V. Srikrishnan, G. S. Young, L. T. Witmer, and J. R. Brownson, “Using multi-pyranometer arrays and neural networks to estimate direct normal irradiance,”Solar Energy, vol. 119, pp. 531–542, 2015.