Representation of a Thermosiphon System Via Neural Networks Considering Installation Parameters 1 2 1 1 ~22 L. E. Zairatel, E. M. Pereira , L. A. R. Oliveira', V. P. Gill, T. R. A. Santos', B. M. Nogueiral, M. A. Rodrigues UNA University; 2 Energy Researches Group (GREEN) Brasil. Abstract- The research of alternative forms of energy production became more important in a context where the natural resources are scarce. In this sense, thermosiphon systems have been developed as an alternative way of energy economy for the water heating process using a renewable energy Cool water tank source: the sun. A thermosiphon system is greatly influenced by several parameters: the ambient temperature (Tamb), the input OLtPLt water temperature (Tin), the solar irradiance (G), the flow rate warm (in), the inclination of the solar collector (I), the height of the Storage tank water water storage tank (H) and mainly by the manufacturing process. Inp Nowadays, there are interests in the development of analytical water Solar collector models that consider parameters of installation such as: height of the water storage tank and inclination of the solar collector. These analytical models can be complex and non-linear. In the last decades, ANN (i.e. Artificial Neural Networks) have been used to represent many kinds of industrial processes, dealing F _ with the complexity and non-linearity of them. Moreover, ANN Fig. 1. Schematic diagram of thermosiphon system. are capable to deal with manufacturing aspects unconsidered by the analytical models but that are important to determine the The performance of a thermosiphon system has been efficiency of the real thermosiphon system. In this work, ANN investigated, both experimentally and analytically, by have been proposed as a new alternative to represent numerous researches [1-8]. The efficiency of this kind of thermosiphon system considering the different parameters systems can be calculated through the equation: related to the efficiency. A trained ANN can eliminate the necessity of new laboratory experiments for real and new mc (T -T T conditions of installation. p out in1 Keywords- Artificial Neural Networks, Thermosiphon System, GA ( ) Solar Energy. extern I. INTRODUCTION where: ri is the thermal efficiency, mh is the flow rate, cp is the water heat capacity, Tout is the water output temperature, Tin is Due to the hydrographic basins decreasing situation around the water input temperature, G is the solar irradiance and the world, associated with the constant population increasing, Aextern is the area of the collector. a new reality can be noticed where the traditional way of In the last years, ANN have been proposed as a powerful energy producing (by hydroelectric power plants) may not be computational tool. ANN have several advantages compared sufficient. Therefore, new ways of energy producing are to other techniques such as: their good performance when necessary and, among these, solar energy systems are an dealing with non-linear problems, their generalizing capacity alternative. and the short processing time that can be reached when trained Solar energy systems, specifically solar water heaters, have nets are in operation. The performance of this kind of systems considerable importance as substitutes of traditional electrical depends of its components, geometrical characteristics, systems. An example of water heating system is called manufacture process and installation conditions involving the thermosiphon, the most widely used of all solar energy solar collector inclination and the water storage tank height. thermal convention devices. Thermosiphon systems have In references [4,5] the representation of thermosiphon competitive cost compared to the conventional energy systems systems through ANN (i.e. Artificial Neural Networks) is available Intthe whole world. discussed, although installation parameters were not Fig. 1 contains a schematic diagram of a thermosiphon considered. The main objective of this work is to represent system, where the most important component is the collector the thermosiphon system considering installation parameters. plate or solar collector. In this work is proposed a neural representation for the thermosiphon system considering the installation parameters. This representation permits to consider manufacture aspects, I1-4244-0457-6/06/$20.OO ©2006 I EEE