Please use this identifier to cite or link to this item: http://localhost:8081/xmlui/handle/123456789/1204
Authors: Sharma, Ashok Kumar
Issue Date: 1989
Abstract: Water supply systems are the essential public services. Significant amount of money is being spent every year for providing safe drinking water facilities. The current design practice adopted by design engineers finds a workable solution disregarding the cost aspect. As the distribution system shares the major part of the water supply system cost, special attention is required for optimal design of distribution network. This will save huge sums of money. Water distribution systems are designed for a predecided time span called design period. For conducting an analytical study, an empirical equation for the determination of design period of a water distribution system has been developed. This equation is a function of life of pipes to be used, rate of increase in water demand and future discount factor. To analyse a water distribution system, the information about the primary loops forming pipes is essential. Theoretically it is possible to generate loop data from the pipe link-node information. An algorithm for the generation of loop data has been developed. Furthermore, an algorithm for the analysis of multiple input point networks has been developed which finds out pipe discharges, input point discharges and nodal pressure heads. The design of gravity flow water distribution systems is dependent upon the relative elevations of the nodal points of the network as the flow is sustained on account of gravitational energy. An algorithm for the design of gravity flow water distribution systems has been developed such that the design parameters are obtained by minimizing the system cost subjected (ii) to the nodal head constraints. Sequential linear programming has been used for the minimization of the system cost. An approximate equation has been developed for providing a starting solution. This starting solution being in the neighbourhood of the optimal solution, saves computational time. The selection of the economical pipe out of the commercially available sizes of different material and class depends upon its diameter and water pressure head acting upon it, the algorithm for the same has been developed. The cost function of the pumping systems besides the pipe cost includes pump cost, pumping cost and storage cost. The algorithm for the design of pumping systems has been developed such that the design parameters are obtained by minimizing the system cost subject to the nodal head cons traints. If a large water distribution system having multiple input points is decomposed or spilitted optimally into a number of small systems of single input point and then designed individually, the total computational time required is much less than the time required in designing the system as a whole. Moreover, this will eliminate present practice of dividing a large water distri bution system into small systems by designer's intuition. A methodology for optimal decomposition of a composite network of multiple input points has been developed. This saves considerable computation time and storage. The existing water distribution systems require reorganization as the water demand of the area covered by the system exceeds the design demand. An algorithm for the reorganization of the existing water distribution systems has been developed. It gives the parallel pipe diameters and the diameters of newly proposed pipe links. (iii) Some-times the increase in water demand is such that it is economical to improve the pumping capacity and pumping head only. Asimple expression for the determination of such an increase in water demand in terms of original demand of the existing system has been developed.
Other Identifiers: Ph.D
Research Supervisor/ Guide: Awamee, P. K.
metadata.dc.type: Doctoral Thesis
Appears in Collections:DOCTORAL THESES (Civil Engg)

Files in This Item:
File Description SizeFormat 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.