Effects of User-Induced Transients on a Service Line: Preliminary Results from WEL (Perugia, Italy) ^†

Abstract

The integrity of water service lines (SLs), crucial components in water distribution networks, is potentially threatened by transients generated by users’ activity. This paper presents the results of laboratory tests carried out at the Water Engineering Laboratory of the University of Perugia (Italy) to investigate the effects of user behavior on an SL. In particular, transients due to changes in the discharge in the SL, simulating the activation of domestic devices, are investigated. The analysis of the effect of such maneuvers provides insights that can assist water utility managers in the design, installation and maintenance of SLs.

1. Introduction

Water service lines (SLs) are crucial components of water distribution networks (WDNs), connecting private plumbing systems to the main pipes. Field experience reveals that SLs are often structurally vulnerable elements, being frequently affected by leakages that can be undetected for long periods of time. The large number of breaks in SLs is traditionally attributed to their less accurate installation and their larger number compared to the main pipes. Another cause of the failure of SLs is often ascribed to the pressure regime. However, as proven by recent studies, fast pressure changes due to transients generated by users’ activity (i.e., changes in water consumption) can induce additional stress on the SLs and the main pipes of WDNs [1,2].

The central role of the SL as the connection element between the point of transient generation (i.e., the user and its private plumbing) and the main pipes of WDNs is still a topic not fully investigated. In particular, experimental tests are still needed in order to appropriately capture their transient behavior. Accordingly, the aim of this paper is to highlight the mechanisms behind the propagation of pressure waves in SLs by means of laboratory tests carried out at the Water Engineering Laboratory (WEL) of the University of Perugia (Italy). More specifically, the main aim of the executed laboratory tests was to investigate the effects of the user-induced pressure transients on the SL. From this perspective, pressure data, collected at a high frequency (i.e., 2048 Hz), were acquired at several sections of the system during transients generated by a pneumatic solenoid valve, simulating the user’s activity.

2. Experimental Setup

The experimental setup at the WEL has been designed in order to simulate a real low- density polyethylene (LDPE) SL supplying the plumbing system.

To simulate the SL (blue segment in Figure 1), a 12-m-long DN40 LDPE pipe (internal diameter D = 32 mm and wall thickness e = 4 mm) is installed and supplied through a tee junction by a 120-m-long DN110 high-density polyethylene (HDPE) pipe (D = 93.3 mm and e = 8.1 mm). A piezometric head of approximately 20 m is provided by a pressurized tank. The SL supplies an end branch of multilayer material pipes, reproducing the user’s plumbing system. Such a material, indicated as PEX/AL/PEX, consists of a layer of aluminum (AL) between two layers of cross-linked polyethylene (PEX). The SL is connected to a 20-m-long DN20 pipe (D = 16 mm and e = 2 mm) installed in series with a 10-m-long DN16 pipe (D = 12 mm and e = 2 mm). This main line has four 2-m-long DN16 branches.

To simulate the end uses, a pneumatic solenoid valve in series with a ball valve is installed at the downstream end of each branch (Figure 1). The ball valve allows the simulation of different values of the discharge, Q, while the solenoid valve generates controlled, repeatable, and fast transients. This allows us to generate sharp pressure waves in order to identify the interaction mechanisms of pressure waves with system elements.

3. Transient Tests

The main aim of the laboratory tests is to investigate the effects on both the SL and the main pipes of the WDN, in terms of the overpressure generated by changes in the users’ consumption.

For the layout described in Section 2, opening and closing maneuvers have been executed at the end-use V, for different values of Q (=0.22 L/s, 0.20 L/s, 0.15 L/s, 0.10 L/s and 0.05 L/s), by adjusting the opening degree of the ball valve. To allow the pressure in the system to stabilize, the opening and closing maneuvers are performed with a sufficiently long time interval (Δt = 120 s) between two successive maneuvers.

4. Results

Figure 2 shows the pressure signals at the measurement sections after a closing maneuver executed at V with Q = 0.22 L/s. The main feature of such signals is the large values, highly unexpected (≈120 m), achieved by pressure at almost all monitored sections of the plumbing system. In particular, the maximum value occurs at section U₄ (≈135 m). To point out such extreme values and the differences between the sections, for all considered values of Q, the values of the dimensionless overpressure, η,

$$\eta ={\displaystyle \frac{{\left(H_{max}-H_{min}\right)}_i}{{\left(H_{max}-H_{min}\right)}_{\mathrm{U}4}}}·100 i=\mathrm{M},\mathrm{C},\mathrm{A},{\mathrm{S}}_1$$

(1)

are shown in Figure 3. In this figure, s is the distance of the section from U₄, where the pressure waves are generated; in particular (Figure 1), section M is located 12 m away from U₄, C at 32 m, A at 33 m, and S₁ at 81 m.

It is worth noting that the dimensionless overpressure $\eta$ does not depend strongly on the discharge. The initial pressure variation at the maneuvering section is dissipated by the propagation of the pressure wave in the system. In particular, very minor dissipation occurs in the first 12 m, with the pressure variation dampening from 100% to 95%. However, when it passes through the DN20 branch that connects the flowmeter to the internal piping system, larger dissipation occurs, with the pressure variation reaching 16% of the initial pressure variation. In the branch connecting the flowmeter to the WDN, this pressure variation reaches 14%. This variation is further reduced by this branch: when it reaches section S₁ inside the WDN, it is reduced to 2%. Despite this significant mitigation, the pressure waves reach values of 30 m in the SL and 18 m in the main pipes of the network (Figure 2). In other words, the end-use maneuvers induce pressure variations even tens of meters upstream.

5. Conclusions

This study presents the results of preliminary laboratory tests executed at the Water Engineering Laboratory of the University of Perugia (Italy) to investigate the effects of end-user maneuvers on an SL and on the main pipe of the WDN. The analysis of the experimental data reveals that the waves generated by the maneuvers of the end users cause very large values of overpressure at the end user. Such overpressure gradually dampens toward the network. This work leaves room for numerous future developments, such as the possibility of keeping the configuration unchanged but replacing the SL with pipes of different diameters and different materials.