A New Approach to Estimate from Monitored Demand Data the Limit of the Coverage of Electricity Demand through Photovoltaics in Large Electricity Grids.

In a traditional large electricity grid without storage, there is a limit to the maximum photovoltaic energy that can be consumed as the demand and generation may not match, either in magnitude or in time. This paper aims to provide a new method to estimate the limit of the coverage of electricity demand by photovoltaics in large electricity grids. This new method eliminates the random and the periodic variability over time as it is based either on the load duration curve for demand and the output duration curve for PV generation. We will assume there is no energy storage or inter-network exchanges. Moreover, conditions for the best scenario for photovoltaics are provided in order to estimate the upper limit: photovoltaic overgeneration is not considered and a complete system flexibility is assumed. The knowledge of this limit will manage to provide not only a reference for the planning of the energy sector but also to analyze the viability of the integration of future photovoltaic projects in the electrical system. In order to illustrate the method, several large electricity grids have been analysed in order to determine the aforementioned limit. Values between 19.3% and 29.9% have been obtained.

A New Architecture Based on IoT and Machine Learning Paradigms in Photovoltaic Systems to Nowcast Output Energy.

The classic models used to predict the behavior of photovoltaic systems, which are based on the physical process of the solar cell, are limited to defining the analytical equation to obtain its electrical parameter. In this paper, we evaluate several machine learning models to nowcast the behavior and energy production of a photovoltaic (PV) system in conjunction with ambient data provided by IoT environmental devices. We have evaluated the estimation of output power generation by human-crafted features with multiple temporal windows and deep learning approaches to obtain comparative results regarding the analytical models of PV systems in terms of error metrics and learning time. The ambient data and ground truth of energy production have been collected in a photovoltaic system with IoT capabilities developed within the Opera Digital Platform under the UniVer Project, which has been deployed for 20 years in the Campus of the University of Jaén (Spain). Machine learning models offer improved results compared with the state-of-the-art analytical model, with significant differences in learning time and performance. The use of multiple temporal windows is shown as a suitable tool for modeling temporal features to improve performance.