Development Of An Intelligent Water Blending System For Irrigation Of Crops With Various Salinity Tolerance And Aquaculture

The application of Artificial Intelligence (AI) systems in decision-making and intelligent control systems has recently gained attention of researchers. One such application is to optimize the water quality and distribution, and to ensure reliable water supply for different consumers. Irrigation is among the important water consumers due to the large amount required to supply the increasing needs for agriculture, and due to the crop yieldsalinity tolerance. AI methods such as goal programming have been used for irrigation scheduling and stochastic goal programming for modeling of future water consumption needs. Water blending in pipes has also been addressed to balance the salinity of irrigation water. Desalination plants use different methods of desalination, which usually produce pure water, but they are expensive. In most cases the desalination plant is integrated with a blending system to blend the pure water with other sources of water for balancing the ingredients, including the salinity, to be suitable for human use and to increase the volume of water. In a typical arid agricultural area, there will be abundant low quality ground water and little quantities of good quality water. There is a need for water blending systems suited for smaller farming communities in arid areas such that more water is made available for crop irrigation depending on the salinity tolerance and also water for aquaculture or livestock. The aim of this work was to propose an artificial intelligence solution to connect many tanks in a network topology, where each tank supplies water with a specific salinity tolerance. The water from two source tanks (one saline groundwater, and the other fresh water) is mixed inside the sink tanks to provide the required salinity in each tank and consequently reduce the fresh water consumption. A mathematical model for water blending was developed to simulate mixing water in a network of tanks. Genetic algorithm (GA) was used as a search engine to find the optimized solution for the amount of water needed to be transferred from one tank to another to balance the salinity that ensure the minimum usage of fresh water. Two cases were simulated involving two source tanks and four sink tanks with various salinity tolerances. One case was for crop irrigation and the other for aquaculture. Laboratory calibrations on the results produced by the GA indicate less than 10% error between simulated and measured EC of the blended water. Further simulation results showed that blending water with different salinities in a network of connected tanks can balance the salinity of each tank according to the crop salinity-tolerance data extracted from FAO reports. The blending system allows the salinity level that minimizes the use of good quality water while the crops can still attain 100% yield potential. This is achieved when sink tanks are connected to each other and GA is used to determine the volume of intertank water transfers. The intelligent water blending system developed in this study provides a mechanism to extend the blending unit to produce water with different salinity levels to meet different standards for use in irrigation or aquaculture. This system will help water managers make better use of various water sources to produce more water for expanding agriculture, aquaculture or industrial use in arid areas.

Preview PDF FK_2008_50.pdf Download (2MB)

Actions (login required)

View Item