Exploration of Pressure Driven Method in Hydraulic Modelling of Water Distribution Network

Abstract

A water supply network is assessed on its efficiency and capability to cater to the demand of consumers. There are two approaches for analysing a network viz. Demand Driven Approach (DDA) & Pressure Driven Approach (PDA). It assumes that the demands at all nodes are always satisfied, irrespective of the available nodal pressure. But many a times, in an operational network, certain situations such as pipe burst, pump failure, fire demand, supply shortage etc. arise. These situations cause pressure deficiency or even negative pressures. Such lower pressures cause water shortage at delivery points or sometimes even supply disruption. This is where DDA fails. PDA is comparatively a new approach. But it is capable of simulating pressure deficient conditions and analysing their effects on the network. Though, not for designing, but it can be efficiently used to simulate different conditions, analyse their effects and determine solutions to mitigate hazards. This paper compares DDA & PDA by analysing a looped network.

Introduction

I. INTRODUCTION

Water supply systems are built with the primary purpose of delivering water from source to consumer. A typical water supply system consists of a source, intake structure, pumps, water treatment plant, clean water storage reservoirs, pipe network, and other appurtenances like different kinds of valves, flow meters, pressure gauges etc. Once the water is filtered, treated and purified at treatment plants, it is pumped to elevated service reservoirs. Water distribution network includes transportation of water from service reservoirs to consumer ends. Further, the network is divided into number of zones depending upon population, consumer type, topology of the region, capacity of service reservoir. The zone is further divided into multiple District Metered Areas (DMAs). DMAs ensure equitable supply of water throughout the zone and better water management. The aim is to distribute water to all the consumers in equal quantities and in adequate pressure ranges. The performance evaluation of a network is done on the basis of its capacity to cater demands of consumers and the residual pressures at consumer ends. Hence, for efficient working of the network, proper layout and correct hydraulic design are very important. This process of designing a network is known as Hydraulic Modelling. It is comprised of engineering, mathematical and geo-spatial design of the network. An efficient hydraulic model allows minimum head losses in the network and hence maintains maximum pressures.

II.. OBJECTIVES OF PROJECT

1. To analyze a water distribution network in Water GEMs.
2. To study Demand Driven and Pressure Driven Approach and apply it on a benchmark network.

III. LITERATURE REVIEW

Various studies have been performed on analysis of water distribution networks and Pressure Driven Analysis (or Pressure Dependent Demands). Certain papers are related to the importance of Pressure dependent demand function in water network analysis and use of PDD function in software like water GEMS, water CAD, EPANET. In this report, literature review covers most of the papers regarding with design, analysis and research parameters of design of water distribution network and pressure dependent demand approach.

IV. METHODOLOGY

The study started with learning basics of Water Supply Engineering and Principles of  Hydraulics and their application.

Different methods and approaches for analyzing a network are studied. A looped network is   analyzed with 1 manual method i.e. Hardy-Cross Method and two softwares i.e. Water GEMS and Tal Tantra by IIT Bombay. The same network is then analysed by Demand Driven and Pressure Driven Approaches. For comparison of DDA and PDA, pressure deficient condition of pipe burst, and supply shortage are considered in the analysis, and its effects are studied. For simulation of various scenarios, WaterGEMS has been used.

V. CASE STUDY

In this project the case study – I network is chosen from “Modeling Pressure Deficient Water Distribution Networks in EPANET” 16th Conference on Water Distribution System Analysis, WDSA 2014, Procedia Engineering 89 (2014) 626 – 631, as shown in Fig 3.5. This network is used by many researchers and the data is available for this network. In this project work, Hardy- Cross method is used for analysis of this benchmark network. For software analysis, WaterGEMS Connect Edition Update 1 is used. WaterGEMS is based on Global gradient algorithm. The results of both the methods are compared for validation of the software. Following are the details of benchmark network.

The network consists of 13 nodes, 21 pipes and 2 sources. The total system demand is 0.874 m3/s.

 The network is simulated for following cases:

A. DDA – Normal Network (All pipes open)

1. Hardy- Cross Method
2. Water GEMS Analysis

Table 1. Details specifying characteristics of network pipes

Table 2. Details specifying characteristics of network junctions and sources

Table 3. Results of Flows in Hardy-Cross, & WaterGEMS (Flow values in m3/s)

Fig .2 Scatter plot for discharge in pipes in Hardy-Cross & WaterGEMS

 Fig 2. shows scatter plot for discharge in pipes in Hardy-Cross and WaterGEMS, the value of R2 is 0.9999 which implies that  values of flow by Hardy-Cross and WaterGEMS are almost matching.

B. Comparative Analysis

Comparative analysis is done for cases mentioned earlier. The results are shown in Table 3 and Fig 2. It shows comparative analysis between pressure heads and Flows at junctions for Normal Network and Pressure deficient network (Pipe-3 Burst), with DDA.

Table 4. Pressure head at junction

Remarks: It is the basic concept of DDA, that even if the pressure heads decrease at junctions, the actual flows at all junctions remain same as that of target demand. Fig. 3.17 show pressure heads in all three conditions. However, with drop in pressures, actual flows at junctions should also drop.

Table 3 and Fig. 2 show comparative analysis of pressure head and actual flows at junction in Pipe-3 burst condition, with DDA and PDA.                                       

Table 5. Comparison of DDA & PDA – Flows at junctions

REMARKS: It can be seen from above table 3 and Fig.5 that, Demand Driven Analysis is not capable of simulating pressure deficient conditions and it can lead to wrong results for such  situations.

Total demand in the network is 0.874 m3/s

1. In Pipe-3 Burst condition and Demand Driven Analysis, even when the pressure heads drop, the actual discharges at junctions = 0.874 m3/s. It implies that there is no demand shortage in pipe burst condition and all the demands are fully satisfied.

In Pipe-3 Burst condition and Pressure Driven Analysis, when the pressure heads at junctions drop, the discharges at

a. junctions also decrease. The actual discharges at junctions = 0.834 m3/s. It implies that there is shortage of 0.040 m3/s in pipe burst condition and the demands at pressures lower than reference pressures aren’t satisfied fully.

b. In Pipe-3 Burst condition, 7 junctions have pressure heads less than reference pressure and discharges less than their actual demands.

c. It implies that Pressure Driven Analysis is better than Demand Driven Analysis for simulating Pressure deficient conditions.

Conclusion

1) Demand Driven Analysis is good for designing a water distribution networks, but it fails to simulate and analyze various pressure deficient conditions in the network. 2) In this project work, the Hardy-Cross analysis and WaterGEMS analysis of a looped benchmark network, gave almost same results for flow in pipes. 3) Theoretically, Pressure Driven Analysis gives more realistic results in pressure deficient conditions, than Demand Driven Analysis. 4) On simulating pipe burst condition in Benchmark Network, around 85% junction’s pressure head dropped below Reference Pressure i.e. 12 m which causes shortage in flow by 0.040 m3/s. 5) From study, it can be inferred that Pressure Driven Analysis gives more realistic results in pipe burst situation, than Demand Driven Analysis i.e. Demand Driven Analysis failed to analyze this condition properly. 6) Further, from Pressure Driven Analysis of Benchmark Network, it can be inferred that Reference Pressure is an important parameter while analyzing network by Pressure Driven Approach.

References

[1] M. A. H. Abdy Sayyed, R. Gupta, T. T. Tanyimboh, (2014), “Modeling Pressure Deficient Water Distribution Networks in EPANET”, Procedia Engineering 89, 626 – 631. [2] Mrs. Vaidya Deepali R., Mali Sandip T.(2019), “Pressure driven approach in water distribution network analysis: A Review”, The International journal of analytical and experimental modal analysis. [3] Muley M. R., Kadu M. S., Dahasahasra S. V., (2016), “Managing disasters using Pressure Dependent Demand Analysis-Case study of Shirpur town”, Civil Engineering Infrastructures Journal, 49(2): 339 – 346. [4] P. Sivakumar, R. K. Prasad, (2014), “Simulation of Water Distribution Network under Pressure Deficient Condition”, Springer Science Business Media Dordrecht, 11269- 014-0677-0 [5] Rajesh Gupta, (2015), “History of Pressure dependent Analysis of Water Distribution Networks and its Applications”, World Environmental and water resources congress. [6] Tom Walski, Devin Blakley, Matthew Evans, Brian Whitman, (2016) “Verifying pressure dependent demand modelling”, XVIII International Conference on Water Distribution Systems Analysis. [7] Tom Walski, Zheng Yi Wu, Daniel Bowdler, (2002), “Efficient Pressure Dependent Demand Model for Large Water Distribution System Analysis”, 8th Annual International Symposium on Water Distribution System Analysis. [8] Zheng Wu, Rong H. Wang, Thomas M. Walski, Shao Y. Yang, Daniel Bowdler, Christopher C. Baggett, (2014), “Extended Global Gradient Algorithm for Pressure Dependent Water Distribution Analysis”, ASCE, 0733-9496, 135:1,13. [9] Zheng Yi Wu, Thomas Walski, (2006), “Pressure Dependent Hydraulic Modelling for Water Distribution Systems under Abnormal Conditions”, IWA World Water Congress and Exhibition.

Copyright

Copyright © 2022 Sonali Gaikwad, Dr. D. K. Rajmane, Dr. S. D. Khandekar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.