Characterization of Air Valves: Alternatives and Consequences

<table><tr><td><p dir="ltr">Air valves are crucial in water distribution systems, enabling controlled filling and emptying during start-up and shut-down, and releasing trapped air at key points. Yet, their characterization remains a challenge. Traditionally, valve selection relies only on matching valve size to pipeline diameter, disregarding real capacity. This practice often produces unreliable designs, aggravated by incomplete or inaccurate manufacturer data.</p><p dir="ltr">This study explores alternative methodologies for valve characterization, comparing them with conventional experimental testing. Three main approaches are examined. The first follows normative codes based on experimental tests. Two strategies exist: using large compressors or blowers to generate instantaneous flows, and discharging large volumes of compressed air. Both face major limitations, either due to huge storage requirements or excessive energy demand.</p><p dir="ltr">The second approach applies Computational Fluid Dynamics (CFD), which has become a promising tool to predict valve performance. Characteristic curves from simulations show strong agreement with experimental data at pressure differentials below 0.3 bar. However, CFD struggles to capture the precise moment of kinetic closure, reflecting the inherent difficulty of validating such transient phenomena.</p><p dir="ltr">The third method uses hydrodynamic similarity with water-based testing. Simple and practical for water industry technicians, this technique reproduces airflow conditions with water and delivers results consistent with direct air testing, especially under the low differential pressures typical of these valves.</p><p dir="ltr">In conclusion, correct valve selection requires evaluating both air admission and expulsion capacities, rather than relying solely on diameter. The comparative analysis of experimental methods, CFD simulations, and hydrodynamic similarity presented here provides valuable insights to improve valve design, testing protocols, and the overall reliability of water distribution networks.</p><p dir="ltr">This paper was presented at the 21st Computing and Control in the Water Industry Conference (CCWI 2025) at the University of Sheffield (1st - 3rd September 2025).</p></td></tr></table><p><br></p>