Brief recommendations with regard to policy and management are made with regard to these items. It is concluded that the approach suggested wiII provide a basis for the successful application of computers as a strategic resource in the professional practice, which should give rise to increases in both productivity and profitability.
Computer-based optimization of water distribution networks is not yet standard engineering practice. The major reason for this is the inability of most optimization techniques to deal with the difficulties encountered when optimizing complex, real-life networks. In this paper an 'exhaustive enumeration' technique for the optimization of pipe networks is presented. The flexibility of the technique and how it deals with realistic constraints in pipe network design are discussed. One of the features of the technique is that it can be used to produce not only a single optimum solution, but also a list of other non-inferior, or Pareto, optimal solutions. The concept of Pareto optimality, and its use as a powerlul planning tool, is explained by means of an example. A tank sizing technique can also easily be incorporated into the optimization procedure, and the same example is used to illustrate that the tank sizing, and pipe sizing problems are interrelated, owing to the trade-off that exists between pipe cost and storage cost.
A system has been developed that provides a solution for the management and engineering analysis and design of water supply systems. The engineering software is fully integrated within a graphical package and a database. It includes pipeflow analysis, dynamic simulation and cost optimization of pipe networks. The integrated approach provides not only a powerful engineering tool but furthermore a management system in terms of a powerful database, data capture facilities and report generation.
The paper deals with the status quo and how this is likely to change in the coming decade as regards computer implementation. The design-detailing process is looked at from different angles such as the steps the process follows, the persons involved, the methods used and the formats used for schedules and sketches. Time and cost implications are addressed. From this, steps that can be automated by using computers are extracted and suggestions are made as to how these should be practically applied in an integrated environment. A practical example is presented.
The paper gives an overview of so-called 'expert systems', what lhey are, how they are selecled and developed and how they acquire their knowledge. Reference is also made to systems currently belng developed in South Africa.
An integrated llood management system (IFMS) is being developed incorporating the recent advances in computational hydraulics, computer graphics and data handling. It has live major components: engineering calculations, graphics, database, 'goodness-of-fit' statistical methods and a cost optimization component. The engineering calculation programs include urban and rural drainage, backwater computations, nood rouling, sediment transport and quality control. The IFMS identifies the trend in the nineties of integration of presently separated activities such as planning, implementation, operation and maintenance, as well as management support.
Computer models are indispensable for the detection of environmental changes because of the large data sets and the complexity of the environmental processes. Unfortunately many of the processes related to the water environment cannot be modelled successlully, but computer-generated graphical presentations can provide valuable insight.