A team at the Technical University of Madrid (Universidad Politécnica de Madrid, UPM) have created a free software program that simulates the functioning and predicts the energy production of photovoltaic systems.
Photovoltaics systems today are predominantly connected to electrical networks, and produce power which varies from a few kilowatts for small systems installed in houses or buildings, to various hundreds of megawatts, for large photovoltaic plants.
The researchers from the Solar Energy Institute (IES) at UPM developed a model that allows the prediction of the energy production of photovoltaic systems starting from minimum meteorological information.
They then integrated the model into a simulation program for photovoltaic systems, which has been shown to be useful for research and educational purposes, and for engineers and photovoltaic system professionals.
According to the researchers, their software allows predictions of the energy produced by the photovoltaic systems during the beginning of projects, allowing optimization of their design, maximise the production of energy, and assess the economic viability of proposed photovoltaic installations.
While photovoltaic simulation programs are not new, most engineers today use software that require, as input data, time series of solar radiation and ambient temperature from sources like state meteorological agencies or satellite databases.
However, these time series are not always available for the site of a project and so many engineers use indeed a synthetic series of data that are produced from a reduced set of meteorological data.
In contrast, the IES-UPM model allows engineers to predict the production of energy of a photovoltaic system with only two out of twelve monthly values of solar radiation: the transparency of the atmosphere (termed the “Linke turbidity”) and the ambient temperature.
The team has compared their model with time series from over 200 weather stations, and found less than two per cent differences.
The IESPRO model and program is able to simulate different photovoltaic systems based on different technologies, assembled on static structures (located on the ground or roof). It can also take into account solar tracking and back-tracking, where the photovoltaic generators stop pointing to the sun in order to avoid mutual shadows between followers.
The system is able to simulate with data series from any time resolution (between seconds and hours) and any length of time. It can be thus be used to predict the instantaneous power generated by the photovoltaic system in real time, or its energy performance in the medium and long term.
The latest development sees the model available to the general public, with a version for the Internet called SISIF. Users can download the source code, written in PHP language, and includes all the models and algorithms of IESPRO.
So far SISIFO only simulates systems connected to an electrical network, but in the future, it will be able to simulate other photovoltaic applications, such as water pumping systems, photovoltaic-diesel hybrid systems, ventilation and air conditioning systems, or autonomous systems for the electrical supply in isolated places of the network.