Using Artificial Neural Networks to Determine Significant Factors Affecting the Pricing of WPT Effluent for Industrial Uses in Isfahan

The evidence indicates increasing trend of use of municipal wastewater treatment effluent as an alternative source of water both in developed and developing countries. Proper pricing of this unconventional water is one of the most effective economic tools to encourage optimum use of fresh water resources. In this study, artificial neural network is employed to identify and assess the factors affecting effluent tariffs supplied to local industries in Isfahan region. Given the wide variety of factors involved in the ultimate value of wastewater traement plant effluent, an assortment of relevant factors has been considered in this study; the factors include the population served by the treatment plant, volume of effluent produced, maintenance, repair and replacement. costs of operating plants, topography, different water uses in the region, industrial wastewater collection fees, unit cost of pipe and fittings, and the volumes of water supplied from springs and aqueducts in the region. Neural network modeling is used as a tool to determine the significance of each factor for pricing effluent. Based on the available data and the neural network models, the effects of different model architectures with different intermediate layers and numbers of nodes in each layer on the price of wastewater were investigated to develop aand adopt a final neural network model. Results indicate that the proposed neural network model enjoys a high potential and has been well capable of determining the weights of the parameter affecting in pricing effluent. Based on the the results of this study, the factors with the greatest role in effluent pricing are unit cost of pipe and fittings, industrial use of water, and the costs of plant maintentance, repair and replacement.