L. Martin1*, L. F. Zarzalejo1, J. Polo1, A. Navarro1, R. Marchante2

1 CIEMAT, Department of Energy, Av. Complutense n°22, Madrid, 28040, Spain

2 IrSOLaV, Calle Santiago Grisolia (PTM) 2, Tres Cantos, 28045, Madrid, Spain

Corresponding Author, luis. martin@ciemat. es

Abstract

Solar energy is gaining huge relevance due to non sustainable current energetic model based on fossil fuels. In the case of solar technology to produce electricity, the integration of power generated into electric grid presents new horizons, such as the estimation of short-range electricity generation to optimize its management; avoiding situations of load reduction and anticipating supplying problems. In this work, a methodology to predict values of solar energy is proposed based on half daily differences of solar radiation lost component time series and qualitative predictions of sky conditions. The dataset used belongs to the radiometric station of Spanish National Weather Service (AEMet) sited in Madrid. Prediction methodology used is artificial neural networks. Skill score of models fed with qualitative prediction as input and without is compared with persistence.

Keywords: Energy meteorology, solar radiation, solar radiation forecasting, solar radiation time series properties, artificial neural network

1. Introduction

Although considerable effort has been done to make use of solar energy efficiently from industrial revolution, expecting fossil fuels would run out in the future, only minimal resources have been directed towards forecasting incoming energy at ground level [1]. However, the necessity to have forecasting models which could optimize the integration of solar thermal power and photovoltaic into electric grid within different sources of electric power generation will grow up as they gain recognition as an energetic resource in the near future.

Photovoltaic and solar thermoelectric power are main sources of solar energy for electricity generation. Currently the potential market is huge and its development is being supported by agreements in Kyoto protocol and by progressive series of regulations regarding green energy (feed-in tariff) established in several countries like Spain and Germany [2]. In the case of Spain, current legislation (Royal Decree 436/2004, 12th of March) allows to minimize investment risks to promoters and to contribute opening up great perspectives to solar energy development.

Energy stock market participation is regulated by two basic rules: on the one hand it is necessary to predict the amount of energy which will be produced, up to 72 hours before, and on the other hand deviations of energy produced compared to programmed one are strongly penalized.

In this work a methodology to predict half daily values of solar energy is proposed, with temporal horizon up to 72 hours. Artificial neural networks techniques are used to predict global solar irradiance values. The prediction is done directly over the differences on solar irradiance measured consecutively. This transformation is done to have a stationary variable with a probability distribution similar to Gaussian distribution.

Forecasting half day values is a first step to make hourly predictions, which is the resolution demanded by legislation. Besides half daily values are of great importance for the operation and energy production programming of concentrating solar thermal power plants which has an storage system.