Avaliação da operacionalidade de uma E.T.E. com reator UASB utilizando rede neural no processo de gestão ambiental

Abstract

With the increasing commitment of companies and people with the environment question, this subject becomes door of entrance for many commercial transactions, what it becomes the implantation Environmental Management Standards (EMS) a something almost with priority. This work aims evaluate and establishes parameters of control to reactor U.A.S.B. (Upflow Anaerobic Sludge Blanket) of the industrials Wastewater Treatment Unit (W.T.U.) of one Pig s Slaughterhouse, through of Artificial Neural Networks (ANN), adjusting the same to the specifications of an EMS, Control data of the W.T.U., such as COD (Chemical Oxygen Demand), SS (Settling Solids) and Temperature of reactor UASB have been used for the training of the ANN and posterior simulation to evaluate the behavior of system UASB. It can be evidenced a great fluctuation in the parameters of entrance and exit of effluent caused by the daily fluctuation in the number of abated animals, causing in some days overload in the system and other load lack. It can be concluded that the ANN showed be a powerful tool to evaluate the processes foreseen in a EMS, since data that are gotten in the control of the same can constantly be used in stages of the training and retrain of the ANN, resulting in better foresees to prevent possible situations of disconformities or ambient damages.