Ultrasonic Air Leak Detection: An Investigation to Improve Accuracy of Leak Rate Estimation

Abstract

A compressed air test rig was designed to develop new procedures and improve the reliability of results that are obtained when conducting a leak survey using an Ultrasonic leak detector such as the UE Systems Ultraprobe.

A test rig was designed that allowed the pressure of air in a compressed air system to be controlled by a regulator. This allowed experiments to be conducted that were able to recreate leak situations in a controlled environment. The pressure of the air through the test rig was measured both at the supply end and at the proximity of the leak site, while the volume flow rate was measured at the supply end of the rig.

A number of leak geometries were examined, with compressed air being passed through open ended tubing and also discs with different leak geometries, some round holes and some rectangular. Initial studies were also carried out on flange leaks and pinpricks and slits in lengths of tubing. These were omitted from the study at this stage to allow further experimentation to be conducted in both areas.

The experimentation was carried out using an Ultraprobe 9000 leak detector which was positioned at a set distance from the leak at a series of angles to the flow. The ultrasound level was measured at each point and compared with the pressure and volume flow rate of the air in the system.

The results showed that the ideal angle to ensure the maximum level of ultrasound is at 30 to the axis of the leak. While the optimum distance for ensuring a consistent level of ultrasound is 150mm from the leak.

The length of any air lines branching from the main distribution network is shown to be an important factor when quantifying the volumetric flow rate of air from an open ended tube or tubing with a significant orifice in it. The pressure drop in a 1m length of tubing was shown to be approximately 50%, and if you consider that often the flow rate being used has been obtained using the outside diameter of the tubing rather than the inside diameter this can become a considerable over estimation of leak rate.

The geometry of a regular shaped orifice, such as a round, or rectangular hole was shown to have little to no effect on the flow rate through it for a constant area. However a coefficient of discharge to account for imperfections in the flow was developed for round and rectangular geometries, these were 0.74 for a round hole and 0.79 for a rectangular hole. These correction factors in tandem with the length effect factor for tubing and the improvements to the measurement procedure, allow a higher degree of accuracy to be obtained when conducting a leak survey.