Improving Efficiency and Reliability in Water Distribution Systems
Abstract
This book contains the lectures given in the International Course "Improving efficiency and reliability in water supply systems", hosted and sponsored by the Menendez Pelayo International University (U.I.M.P.) and co-sponsored by Aguas de Valencia, the British Council and the EC Cornett and Erasmus programmes. The short course took place in Valencia (Spain) in November 1994, with an attendance of more than one hundred delegates. We must not only acknowedge and thank Dr. Joaquin Azagra, as UIMP Director, but also his collaborators D. Luis Moreno and Lidia Lopez for their support in the preparation of the Course and during the course taking place. UIMP sponsorship allowed us to assemble in Valencia an eminent cadre of lecturers coming from all over the world, that covered in an ordered and precise fashion some of the more relevant aspects on efficiency and reliability in water supply systems. We are very thankful to all these leading lecturers for their invaluable cooperation. The publication of this book and the Spanish edition as well, have been made possible thanks to the sponsorship of both Polytechnic University of Valencia througout its Chancellor, Justo Nieto, and Aguas de Valencia throughout its General Director Alvaro Aguirre. We must also thank Kluwer Academic Publishers and especially their Publisher Petra van Steenbergen for her assistance, careful presentation and production of the book.
Chapters (17)
Water supply systems modelling constitutes a basic tool for an adequate technical management of those systems. In practical applications, it is frequent to use different models depending on the problem to be analyzed: from the simplest one, the static model, to the most complex one, the elastic inertial model, better known as water hammer.
In this paper, general remarks of each model are explained, although the main emphasis is put on their conventional applications.
Relating to hydraulic infrastructures.
Supply system: connected set made of production works (intakes, wills, plants, adduction pipes), inputs from other systems and distribution network.
Distribution network: coherent set made of reservoirs, pumping stations, transport pipes and sections (distribution pipes, connections, valves, control pressure elements, measurement elements).
Relating to the consumed water.
Collected volume: volume taken from its natural environment through the fitting works (VC).
Produced volume: volume becoming from the production works to be delivered at the distribution network or to be used in recharging groundwater reservoirs (VP).
Volume delivered at distribution: volume as a result of algebraic addition of produced, imported and exported volumes (VD).
Service network volume: volume used in the network normal operation, including reservoir cleaning water, purges, pipes pouring away, etc..
Leakage volume: volume as a result of leakage due to the lack of watertightness.
Unprofitable volume: volume lost in operation incidents (overflows, etc.).
Supplied volume: volume got by the users at supply points (VS).
Registered volume: volume as a result of users measurement elements, including utility ones, gaugings, and valuations on broken and not read counters (VR).
Turnover volume: volume as a result of bills, usually different than registered ond due to the inclusion of valuations and minimums (VF).
The growing concern on improving the use of water in distribution systems, requires increasing the quality and the reliability in such systems. This paper presents a series of techniques aimed at reducing water losses and increasing the efficiency in the supply.
The proposed methodology covers the selection and optimization criterion of the strategies to be applied, in order to achieve the objectives in an optimum way. A hierarchy based on technical and economical criterion established in this way, such that with the help of mathematical models and the coordination with the rest of the management and exploitation functions of the supply, will allow the development of a program which reaches and maintains optimum working conditions.
It is also recommended to carry out a complete analysis of the hydraulic system, so that through its diagnosis we would be able to determine the causes of diminish safety and efficiency, and enable us to economically optimize the necessary measures to fulfil maximum acceptable risk constraints and maximum efficiency in the management of the energetic and water resources, these last concepts scarcering day by day.
To avoid unnacounted-for water in a water distribution network is always of a paramount importance ,but mainly in countries like Spain where water is a scarce resource.
In this paper a real leak detection plan with its main results , is presented.
Monitoring of water distribution systems has a number of benefits, not the least of which is its use to maintain an up-to-date calibration of the system and to detect leaks. A formal inverse calculation is to the best method of performing these functions. The steady state calculation for leak detection appears to be impractical due to the necessity of knowing accurately the pipe friction factors. The unsteady inverse problem is able to calibrate the system and perform the leak detection simultaneously. Its difficulty stems from a high sensitivity to data error and the need for an accurate forward simulation of the system. Sensitivity analysis is able to determine the necessary accuracy of data for a desired level of exactness of results. High order interpolation and Courant numbers near unity are necessary to keep the forward computation accurate.
It is well known that the use of variable speed pumps for operating pressurized water distribution systems provides many potential benefits. In this paper, several models for investigating the performance of water distribution systems with variable speed pumps are presented. The models were developed to determine the optimum speeds for operating VFD (variable frequency drive) pumps to meet stated objectives.
The magnitude of water losses due to leakage from a water supply network, and the dependence of leakage on network pressures, are discussed. A methodology is presented for the determination of optimal pressure reducing valve settings minimising excess pressure in a water supply network for varying consumer demands. The methodology takes fully into account the network hydraulic equations, including terms that relate leakage losses to network pressures. The linear theory method is used for the iterative linearisation of the nonlinear network equations, leading to the solution of a succession of linear programs until a PRV outlet head minimising network pressures is obtained. Computational results for an existing network are presented, demonstrating the applicability of the method.
The energy audit of water networks has to be focused on DIRECT and INDIRECT interconnected facts such as:
i -
Operational schedule of pumping stations and available storage to attend several demand scenarios;
ii -
Instalation of ACV’s — Automatic Control Valves — to avoid extreme pressure changes due to hourly demand variations influencing network leakages;
iii -
Customer water volume measurements: the measured water volume consumption is influenced by the domestic storage dimensions! Generally, if any, these storages are emergency ones operating at maximum level with small valve openings for water admission, resulting in:
a -
Errors in the water consumption measurement
b -
Dead water in the customers’ storages;
These facts, with illustrated examples are presented in the paper.
The essential requirements of pump scheduling to reduce energy costs are illustrated by reference to some related computer programs and activities. In each case there is a discussion of methodology and procedures, with applications taken from actual networks and systems. The readily available computer programs related to pump scheduling consist of:
GIPADS for evaluation of hydraulic and cost models of pumps.GINAS for calculation of pump energy costs under dynamic’ operations, with particular use for automated evaluation of optimal pump combinations, and validation of optimized pump schedules under real-time operations.
GIPOS/GIMPOS for evaluation of optimized pump schedules for some simple system structures.
GIPSOC for optimized pump scheduling and control of a specific multi-station, multi-reservoir system.
GIDAP for prediction of demand patterns to give advance knowledge of system loading. Finally an integrated application consisting of a real-time optimal pump scheduling scheme is described.
The two primary objectives in design and operation of water distribution systems are (i) minimisation of the cost of supplying water and (ii) maximisation of reliability in the supply of water. These two objectives are in conflict, however, in that increases in reliability of supply almost invariably lead to increases in cost. The problem of design and operation of water distribution networks is therefore multi-objective in nature. One of the major issues facing engineers and governments involved in the design, operation, and maintenance of water distribution system is how to reconcile these two objectives such that both cost and reliability are achieved at acceptable levels.
Over the past twenty years a major financial effort has been made to improve the water supply in Portugal. A significant evolution can be observed, with 77% of the population currently connected to public systems and a target of 95% by the year 2000. Supplying a high percentage of the population is however not sufficient by itself. The quality of the service provided to the users is a fundamental issue, particularly where continuity of supply, water quality and reliability are concerned. Recent surveys based on simplified criteria indicate that 40% of the population supplied are thought to enjoy a good service, 54% an adequate service, and 6% a poor service.
This paper summarises the causes and consequent effects of pipe failures in water distribution mains and then illustrates how some of these impacts can be modelled. Two modelling approaches are described. The first of these is based on a full network analysis of the failure event, its isolation, repair and restoration of normal service. This approach also includes the proper dependence of network flows and nodal outflows on network pressures and delivery head. The second method illustrated is an approximate technique which relies only on existing (normal operation) network flow information and which offers a way of gaining an initial indication of the impact of a pipe failure occurring at any specified point in the network.
The problem of optimum design of water distribution networks has been thoroughly discussed in the specialised literature and also hardly criticized for the lack of practical applicability of the proposed methodologies. The mathematical problem is certainly very complex and the final results depend greatly upon the data and other facts that are difficult to model.
During the past few years a great number of papers introducing the concept of system reliability in the planing of future rehabilitation, improvement and enlargement of networks, as an alternative to the simple idea of minimizing the investment and maintenance total costs have appeared. However, the proposed mathematical formulations often resulted in a more complex solution, difficult to apply.
In this paper, it is the authors intention to offer, under their knowledge, an overview about the different ways in which the problem of designing and improving hydraulic networks may be faced. We also will introduce new concepts based on heuristic criteria which will probably become more important in the future.
Hydraulic systems, including those for water or fuel distribution, are often designed on the basis of steady state operating conditions. During transient operations, however, pressures much higher than steady state values can develop. Transient operating conditions can result from a number of causes, including pumps stopping or starting, valves opening or closing, and system startup or shutdown. Some systems, such as aircraft fueling systems and fire protection sprinkler systems, always operate in the transient condition. Operation of check valves and air release/vacuum break valves can cause large pressure surges. Burst pressures may be exceeded or fatigue limits may be reached as the result of transient operations. This paper presents three case studies in which transient operating conditions are substantially more severe than those indicated by a steady state hydraulic analysis. These case studies are based on field data and experience with existing operating systems, and on systems for which preliminary designs are being developed. In some cases the pipes failed. In other cases surge control devices were required or operating procedures to reduce pressure surges had to be developed. The results of computer analyses of the various cases will be shown and, where possible, compared with field data.
A water distribution system that is properly designed and maintained should operate for an extended length of time with minimal unscheduled repairs and operational problems. Proper analysis, selection, operation and maintenance of the control devices, especially pumps and valves, is an important factor for satisfactory performance and long life of a piping system. For a valve, principles related to proper selection and operation will be demonstrated with examples. Several applications will be discussed to demonstrate the effect of design choices on the operation and life of the system. These examples will cover flow control, torque requirements, prevention of reverse flow, pressure reduction, cavitation control and transient control. Examples are also given to demonstrate proper selection of single and multiple pumps in series and in parallel. The discussion includes the role of future demands on the selection of the valves and pumps. When demand exceeds the original design conditions, or if significant changes are required in the operation of the system, each important control device should be analyzed to see if it can operate safely at the new conditions or determine if modification or replacement is required.
A new computer model is presented to predict the spatial and temporal distribution of constituents in a pipe network under slowly varying unsteady flow conditions. Unlike the other available models which use steady-state or extended-period simulation of steady flow conditions, the model uses a lumped-system approach to compute unsteady flow conditions and includes dispersion and decay of a constituent during travel in a pipe.
The pipe network is first analyzed to determine the initial steady-state conditions. The slowly varying flow conditions are then computed by numerically integrating the governing equations by an implicit-finite-difference scheme subject to the appropriate boundary conditions. The one-dimensional transport equation is used to calculate the constituent concentration over time during travel in a pipe assuming a first-order decay rate.
The transport equation is numerically solved in two steps by using two different finite-difference schemes. For pure advection, Warming-Kutler-Lomas (WKL) explicit scheme gives the best result while an explicit scheme is used to calculate diffusion and decay. Complete mixing is assumed at the pipe junctions. The model is verified by comparing the results with those of EPANET (Rossman, 1993) for two typical pipe networks.
There is growing interest in understanding the factors that influence changes in water quality in distribution systems. Water quality models provide a systematic method for studying these factors. The United States Environmental Protection Agency (USEPA) has developed a water quality model called EPANET that is widely used in the U.S. In addition the U.S. EPA is conducting a comprehensive research program to study water quality in networks that includes bench, pilot and field studies.
... For instance, the slow variation in discharge through a water distribution network was considered by treating the unsteadiness of flow along the pipeline system as a sequential steady-state situation. Such a scenario necessitates time-dependent boundaries for each iterative solution, thereby increasing the computational efforts required [2]. Additionally, external mass balance equations have been employed in extended-period simulation to examine the impacts of variations in water levels provided from reservoirs within a water distribution network. ...
Hydraulic analysis of water distribution networks stands as a paramount mathematical application within the realm of water resources engineering. The EPANET software, established by the U.S. Environmental Protection Agency, serves as a model for water flow within distribution systems and is employed in the design and analysis of distribution networks, along with addressing pressing issues in water distribution management, such as water quality investigation. The software's distinctiveness is attributed to its reliance on the Hybrid Node Loop (HNL), which accelerates the numerical resolution of problems with intersecting equations. In this study, the Tikrit University campus was selected as the site to examine the pressure and discharge behavior within the water distribution network during morning and evening pumping periods. It was observed that the pressure values exhibited a decrease in the university's north-western sections and buildings, particularly those distant from the pumping station, during the morning pumping periods. This pattern is attributable to the heightened demand in the colleges and faculty residential neighborhood, suggesting a correlation between high consumption and reduced pressure values. In the evening period, characterized by a diminished population of students and employees in the university's colleges, an elevated pressure level within the network was noted, likely due to the reduced demand level. However, demand remained high in the female students' dormitories and the faculty residential neighborhood. Interestingly, the male students' dormitory, being remote from the pumping station and supplied with water through the network's end following the faculty residential neighborhood, maintained low-pressure values across both periods. This study underscores the need for strategic planning in the design and management of water distribution networks, particularly in areas with variable consumption patterns and demand levels.
... The most obvious indication of the physical deterioration and failure of the pipe network is leakage. Pipe failures can be regarded as either persistent, progressive or sudden [11]. Leaks waste both money and a precious natural source, and they create a public health risk. ...
One of the major challenges facing our water utilities is the high level of Non-Revenue Water (NRW) in the distribution networks. This paper assessed the performance of current management practices by Syarikat Air Melaka Berhad (SAMB) to deal with NRW. Information and NRW management data (from 2012 to 2013) were gathered and analysed. Statistical methods were used to evaluate the effects of pipe length and number of connections of the distribution network to the leakage level; and to determine the causes of leakage (water loss). In 2014, Melaka’s NRW percentage was 21.4% as compared to the national average of 35.6%, which is the second lowest rate among the states in Malaysia. Results of the study revealed significant positive relationships between average MNF (L/s) with number of connections and pipe length, with the prediction model of average MNF (L/s)=-4.42+1.088*10-2(NC)+1.07*10-4(PL), R2= 73.19%. The results also indicated that in a compact and urbanized city like Melaka, number of connections in the network appears to be most influential to the average MNF (water loss) (shown by a strong positive relationship, r = 0.847) as compared to the less compact zone (such as Perak) where pipe length appears to be more influential.
... Consequently, we may consider a superposition of two waves emerging the two sources of disturbance. The following function depicts the envelope for pressure oscillation caused by two sources of disturbance: (6) where -pulsation of the modulation wave, ; ...
Numerical simulation and experimental laboratory measurements were developed in order to study the unsteady flow of water in a single loop distribution network. Transient movement results as a hydraulic system response to sudden valve maneuvers in a water supply network. Investigation on pressure variation was carried out on a representative loop, hydraulically similar to a real water single loop network aiming to correlate the extreme pressure values, frequency, and damping coefficient to the consumers' water demand and valves maneuver pattern. Theoretical and experimental results reveal the same extreme pressure values, but the laboratory recorded oscillations have a lower frequency and an increased damping ratio than the simulated ones.
... The analysis used a modeling software package EPANET, which enables modeling of hypothetical conditions of different pressures, leakages and solutions of supply lines (introduction of pressure booster pumps, swages etc.). The analysis gave consideration to the cases of water supply over a relatively flat terrain with a tank located at the specified elevation point and water supply over a terrain inclined in direction from the pump to the reservoir1234567891011121314151617181920. ...
Based on the acknowledged theoretical proofs and experiences in the modernization of water supply
systems, it can be concluded that out of the range of available measures to reduce leakages, the regulation
and/or reduction of operating pressures is being recommended as the most efficient method thereto. The known
proof-theoretical papers on water supply system leakage and the corresponding operating pressure losses
advocate thesis on possibility to reduce the level of operating pressures to the range of 3.5- 2.5 bars, in all parts
of the network where applicable, which is evidenced to be a sufficient level enabling the uninterrupted operations
of the large percent of the users. The justification of the previously stated attitude can be proved by means of
simple schemes of supply line which correspond to the possible states that can occur in actual operating modes
of the water supply system. An analysis was performed with focus on the location and role of the pressureboosting
pumping stations used to increase the pressure in achieving the new low-pressure state and their
respective influence on techno-economic parameters upon which a decision is made on economic viability of the
solution concerned.
... The analysis used a modeling software package EPANET, which enables modeling of hypothetical conditions of different pressures, leakages and solutions of supply lines (introduction of pressure booster pumps, swages etc.). The analysis gave consideration to the cases of water supply over a relatively flat terrain with a tank located at the specified elevation point and water supply over a terrain inclined in direction from the pump to the reservoir1234567891011121314151617181920. ...
Based on the acknowledged theoretical proofs and experiences in the modernization of water supply systems, it can be concluded that out of the range of available measures to reduce leakages, the regulation and/or reduction of operating pressures is being recommended as the most efficient method thereto. The known proof-theoretical papers on water supply system leakage and the corresponding operating pressure losses advocate thesis on possibility to reduce the level of operating pressures to the range of 3.5-2.5 bars, in all parts of the network where applicable, which is evidenced to be a sufficient level enabling the uninterrupted operations of the large percent of the users. The justification of the previously stated attitude can be proved by means of simple schemes of supply line which correspond to the possible states that can occur in actual operating modes of the water supply system. An analysis was performed with focus on the location and role of the pressure-boosting pumping stations used to increase the pressure in achieving the new low-pressure state and their respective influence on techno-economic parameters upon which a decision is made on economic viability of the solution concerned.
Here, recent developments in the key numerical approaches to water hammer modelling are summarized and critiqued. This paper summarizes one-dimensional modelling using the finite difference method (FDM), the method of characteristics (MOC), and especially the more recent finite volume method (FVM). The discussion is briefly extended to two-dimensional modelling, as well as to computational fluid dynamics (CFD) approaches. Finite volume methods are of particular note, since they approximate the governing partial differential equations (PDEs) in a volume integral form, thus intrinsically conserving mass and momentum fluxes. Accuracy in transient modelling is particularly important in certain (typically more nuanced) applications, including fault (leakage and blockage) detection. The FVM, first advanced using Godunov’s scheme, is preferred in cases where wave celerity evolves over time (e.g., due to the release of air) or due to spatial changes (e.g., due to changes in wall thickness). Both numerical and experimental studies demonstrate that the first-order Godunov’s scheme compares favourably with the MOC in terms of accuracy and computational speed; with further advances in the FVM schemes, it progressively achieves faster and more accurate codes. The current range of numerical methods is discussed and illustrated, including highlighting both their limitations and their advantages.
The difference between the volume of water delivered into a reticulation or distribution network and the volume which is legitimately consumed is termed unaccounted for water. the net water production is presumably the most accurate metered flow as it is generally metered at the source, for example, the purification works. On the other hand, consumption of water is frequently not metered so that the amount consumed is not easy to estimate. It also may not be consumed legitimately as there may be illegal connections into the distribution system.
The contemporary epoch of socio-economical globalization and the latent scarcity of natural and energy resources, associated to adverse climatic conditions periods, present new challenges to engineers in design preparation, works execution, operation and management of public utilities, particularly in the treatment, transportation and distribution of water.
Purpose of this paper is to present for one representative segment of the water distribution network -DMA (District Metered Area) results and benefits of pressure reduction and control and effects on pipes burst frequency. Statistical analysis covered time frame of the last 4 years (2 years before pressure control was implemented and 2 years with pressure control) in one specific DMA. This work is extensions of activities related with the pilot project for the DMA Knežija (16 km of pipelines) in the water distribution system of the city of Zagreb, capitol of Croatia, Europe [1].
Transient movement results as a hydraulic system response to sudden valve manoeuvres in a water supply network. Investigation on pressure variation was carried out on a representative loop of a pipe network. Both numerical simulation and experimental laboratory measurements were developed in order to validate the software Hammer for looped networks. Theoretical and experimental results reveal the same extreme pressure values, but the recorded oscillations have a lower frequency and an increased damping ratio than the simulated ones.
In the last decade, leakage has become a high priority to water, oil and sewage transport industries and to the regulators, in order to fulfil legal milestones, to increase the profits of the companies, to protect the environment and to use the best as possible the scarce natural resource that is water. Leakage reduction and control can only be achieved with the implementation of an integrated leakage control system that includes the identification and the characterisation of leakage in the hydraulic systems, leaks' detection and location, leaks' repair and, finally, the implementation of a continuous monitoring system.
The risk analysis represents a modern approach to determining the level of provision of drinking water supplies for the consumers
and the safety of the whole drinking water supply system. The authors present the most frequently used methods of risk analyses
of drinking water supply systems and address identification of qualitative as well as quantitative risks posed by the individual
system components, the evaluation methods and interpretation of results. What is also described here are the basic principles
of implementing the Hazard Analysis and Critical Control Points (HACCP) method and the possibility of transferring some experience
and methods from other fields of industry, which might be more advanced in this area. The presented methods are demonstrated
on a case study of a risk analysis concerning a selected section of a water supply system.
ResearchGate has not been able to resolve any references for this publication.