An Efficient Solution for Water Supply Networks

In a computer implementation, a water network model is represented by a set of network data (values of parameters) and a set of equations (physical operation of the model). The fundamental model is formulated using the laws of physics. Subsequently, different models can be derived by mathematical manipulation. For water network models three physical laws are employed: flow continuity, head-loss continuity and component equations (head/flow law). Industry and academia have made a significant investment in the research and development of computer algorithms for the design, modelling and control of water supply networks. The industrial use of many of these algorithms is commonplace and they are complex procedures used to solve systems of equations. Therefore, powerful numerical algorithms used to solve the model equations are crucial in terms of their computational complexity i.e. requirement of average number of iterations for finding a solution with a required accuracy. Along with physically simplified networks representing certain aspects of reality, neglecting other aspects with a smaller number of components than the original network, effective solution methods for water supply networks become common practice. To do this, first we must form a mathematical model representing the physical system; second, we need to set up an efficient solution implementation which is specific to the equation systems and finally we must visaliuze and post-process the results.