Optimization of Energy in Water Distribution Systems by Effective Measurement and Control K. A. Steele INTRODUCTION
The optimization of a water distribution system on the basis of minimum specific energy consumption is dependent upon various parameters. In addition there are numerous secondary co nditions which must be satisified and constraints to be taken into account. The task is consequently a very complex one. The optimization system presented performs this task with the aid of a computer in on-line open loop, which determines the optimum conditions to which the plant shall be managed . A further development towards fully autom atic optimization using on-line closed loop is contemplated, as is a lso a development in the direction of an additional structure optimi zation when planning new plants. The system as described in this paper has been installed at the East Worcester Water Authority U .K . The function of a water di stribution system can be defined as the provision of an adequate water supply at minimum cost. The contribution of the measurement and control system towards performing this task consists of two main components: l) Choice and design of the appropriate measurement and control system required to balance resources in order to provide water where it is required and in sufficient quantities. 2) Optimization of the measurement and control system according to I) to achieve minimum specific energy costs, by determining the operating strategies required to attain this, but having a proper regard to the co nstraints imposed by the water distribution system itself. The techniques for I), in larger plants in connection with programmable contro l systems, process computers and telemetering systems, are already well established. Further enhancement and improvement will be achieved in the future by the development of technology related to measurement techniques , data transmission and display methods, i.e. work which is within the development program to which companies within Brown Boveri are committed. ¡ Optimization of the energy consumption to a minimum according to 2) can be achieved with equipm ent availab le at present, especially with the aid of process computers, by developing the appropriate software . The need to save energy need hardly be mentioned and everyone today is aware of the reasons for this . Quite apart from this, new in -
Keith A. Steele is the Utilities Sales Manager for Kent Process Control (Instruments and Systems) at Luton, U.K. 20
vestments can be reduced or even avoided, arid thu s.capital costs saved, by the more efficient use of existing resources.
variab le input data by means of analysis and computation to obtain output data. Systems for data acquisition and analysis of this type are already well established in the water in dustry as components of telemetry networks. Prediction
The data on which prediction of demand is based is made up from records of past consumption , which are modified in accordance with expected future domestic and industrial demands. Thus, long-term forecasts are obtained which are continuously modified and validated by current data. Mathematical Model of the Water Distribution Syste m
In essence the system model consists of a set of equations which express the behaviour of the water system, the latter comprising components such as pipes, valves, pump stations, reservoirs and water towers, etc. For each of these components there are characteristic equations which, as a rule , are expressed by the variables pressure head and flow in relationship to each other. Special features of the system components are given below: ,, p 1-1
Fig. I. Softwa re-configuration of the optimization system I = Input
III
0 = Output
Thus the object of optimizat ion is to firmly estab lish an operating strategy for the overall plant. Figure I illustrates the software system used with the computer in order to determ ine the operating strategy, the system comprising various functional units . Its main function s are: -The acquisition and analysis of current data with a furt her input from a source of statistical data . - Prediction of performance requirements based on statistical data, which are modified by development trends manifest in data acquired recently . -Simulation of the water distribution system by means of a mathematical model, whereby the effect of the various strategies in meeting constra ints and reducing operating costs may be determined. -Step-by-step optimization, which delivers, as a resul t, the operating program to be followed, i.e. the optimum operating strategy. DATA ACQUISITION SYSTEM
Data from the distribution plant is collected by thi s system and verified against other information in its own sub-systems. This improves the security of operation. In addition the sub-systems combine the
Zo
Fig. 2. Characteristic of fixed-speed centrifugal pump with throttling valve I 2 3 4
5 Hv
HR Hs Qmax
Q. ZQ
ml 11111
QH characteristic of the pump QH characteristic of the line Static head Hs Power P for driving the pump shaft Efficiency ~ of the pump Pressure difference across control valve ( = valve losses) Pressure difference in the line ( = line losses) Static head (yellow) Maximum flow Flow at operating point Possible range of adjustment of flow Valve losses D Static head Line losses
Pump Stations Pump head/flow rate' characteristics are typically nonlinear and a general form for a single fixed-speed pump is shown in Fig. 2. 'Flow rate Q, defined as the water vol ume per unit of time, e.g. in m' / s.
WATER