Experimental Investigation of Pressure Reducing Valves Under Transient Conditions: A Laboratory Study ^†

Abstract

Pressure-reducing valves (PRVs) play a key role in water distribution networks (WDNs), where they are employed to regulate pressure levels and mitigate leakage. The present work describes an experimental study designed to investigate the transient response of a PRV installed at the inlet of a laboratory-scale district metered area (DMA). The closure of a downstream valve along a service line reproduces a typical end-user action and generates a pressure surge propagating through the system. Two distinct operating scenarios were examined, and the analysis of the measured pressure signals offers a basis for discussing the effect of hydraulic transients on the network.

1. Introduction

Pressure management is widely recognized as one of the most cost-effective approaches for mitigating real water losses in distribution networks, with pressure-reducing valves (PRVs) representing essential components of this strategy [1,2,3,4,5]. When installed at the inlets of district metered areas (DMAs), PRVs regulate downstream pressure to a prescribed setpoint, regardless of upstream fluctuations—those, for example, induced by pump operations [6]—and variations in flow rate. While their steady-state hydraulic performance has been extensively documented, the transient response of PRVs remains an open research challenge. At the same time, user-driven maneuvers have been recognized as a significant source of pressure transients in distribution networks [7,8]. Laboratory investigations have highlighted that inappropriate valve sizing or operation under very low-flow conditions may induce self-excited oscillations and amplify water-hammer phenomena [9], a risk subsequently confirmed in field applications [10]. More recent studies have further advanced this understanding: in [11], unstable dynamics of piston-actuated PRVs at extremely low discharges are reported, whereas field-scale measurements by [12] demonstrated that PRVs may behave similarly to a dead end, reflecting pressure waves and interacting non-trivially with pipe-wall viscoelasticity.

In this context, the present work reports on a laboratory-scale experimental campaign specifically designed to investigate the dynamic response of a PRV under controlled hydraulic conditions.

2. Laboratory Setup and Tests

The experimental facility (Figure 1) consists of a laboratory-scale water distribution network (WDN) entirely constructed from high-density polyethylene (HDPE) pipes and supplied by a constant-head reservoir. The main supply line is a DN110 pipe, 43.17 m in length, with an internal diameter D = 93.3 mm and a wall thickness e = 8.1 mm, which connects to a two-loop network. The first loop comprises four DN75 pipes, each 100 m long (D = 63.8 mm, e = 5.6 mm), whereas the second loop shares one link with the first and consists of three DN50 pipes, also 100 m in length (D = 42.6 mm, e = 3.7 mm).

A DN25 service line (SL), 23.60 m in length (D = 20 mm, e = 3 mm), is connected to nodes 5, 6, and 7. Each SL branch terminates with a ball valve, used to simulate end-user withdrawals at demand points 5u, 6u, and 7u. A DN80 pilot-operated pressure-reducing valve (PRV) is installed along the main supply line to regulate downstream pressure.

The WDN is equipped with pressure transducers placed at multiple key locations to record pressure fluctuations during transient events, with a sampling frequency of 2048 Hz. System demand is continuously monitored by means of an electromagnetic flowmeter installed on the DN110 main line.

The analysis presented in this work focuses on the pressure transients generated by valve closures at end-users and on their effects on the pressure signal measured at section 32, located immediately downstream of the PRV. Two operating scenarios were defined to study the influence of the PRV’s control mode. In scenario #1 (non-regulating condition), the network was supplied by a pump delivering a head of 2 bar, resulting in an initial steady-state discharge of approximately Q₀ = 0.21 L/s. In this configuration, the PRV was kept fully open, thus behaving as a pass-through device. In scenario #2 (regulating condition), the system was instead fed by a pump operating at 5 bar, while the PRV was adjusted to regulate the downstream pressure to the same steady-state value observed at section 32 during scenario #1.

3. Results

Figure 2a–c shows the pressure signals recorded at section 32 during transients induced by the rapid closure of the end-user valves 5u, 6u, and 7u, respectively. For each case, the responses obtained under both operating scenarios are reported. A comparison of the two scenarios reveals that, irrespective of the end-user performing the closure, the first overpressure peak is essentially coincident. This observation confirms that the steady-state pressure at node 32 and the initial discharge Q₀ were identical in both configurations. However, following the initial pressure variations, the signals begin to diverge. In particular, under scenario #2, the pressure consistently rises above the corresponding trace of scenario #1, exceeding the setpoint imposed on the PRV when operating in regulating mode.

4. Conclusions

This study has presented preliminary results from laboratory tests conducted at the Water Engineering Laboratory of the University of Perugia. The experiments focused on user-induced transients in a complex water distribution network equipped with a pressure-reducing valve (PRV). Two operating scenarios were examined: in the first, the PRV was fully open and operating in a non-regulating mode, whereas in the second, it was set to regulate the downstream pressure.

The comparison between these two scenarios, based on pressure measurements taken immediately downstream of the PRV, highlights a critical aspect of its behavior. Specifically, when an end-user valve closure occurs, the downstream pressure in the regulating scenario tends to rise above the imposed setpoint, revealing limitations in the valve’s ability to maintain effective control under transient conditions.