1. How does an ultrasound camera work for leak detection?

An ultrasound camera detects high-frequency sound waves generated at leak points by escaping compressed air, escaping process gases or the inflow of ambient air in vacuum applications. A microphone array uses the microphone signals to perform the calculation of the direction of the sound source and superimposes the result as a coloured heatmap onto a video image.

The method operates passively: the camera does not require a test gas such as helium or formier gas, but instead detects the acoustic energy generated by turbulence at the leak site. This enables leaks to be localised whilst the plant is in operation, without the need to shut down large sections of the plant, dismantle equipment or treat them with leak detection spray.

The technical benefit lies in the combination of acoustic localisation and visual identification. Users not only see that ultrasonic activity is present, but can also identify in the camera image which component, valve, hose, connection or section of pipe is generating the sound pressure. This reduces search time, prevents misinterpretations and simplifies the documentation of detected leaks.

Example: Localising an ultrasonic source on a hard-to-reach hall ceiling. The hotspot marks the acoustic source in the camera image.

In industrial leak detection, the distinction between audible sound and ultrasound is crucial. Most production noises fall within the audible, low-frequency range. Leaks in compressed air and gas pipes generate broadband turbulence with high-frequency sound components at small leak openings. Large leaks often generate audible sound as well; however, in production environments, this is frequently masked by machine noise.

The ultrasound generated by a leak is not caused by the gas itself, but by the turbulence that occurs as the gas escapes. The pressure difference, the geometry of the leak and the shape of the outlet opening determine the sound intensity and the frequency spectrum of the generated ultrasonic waves.

Higher frequencies offer better spatial resolution due to their shorter wavelength. However, as the frequency increases, atmospheric attenuation also increases, thereby reducing the achievable detection range.

The choice of measurement frequency therefore represents a trade-off between resolution, range and interference suppression. In many industrial applications, a frequency range around 40 kHz has proved to be favourable. In environments subject to interference, however, a different frequency window may yield better results.

Further reading: Leak detection despite interference sources and FFT

For CS INSTRUMENTS, compressed air is the primary application for ultrasonic leak detection. The reason for this is the direct link between leaks, consumption, compressor load and energy costs. Any unplanned air loss increases demand on the consumer side. If the leak is rectified, less air needs to be generated, dried and treated.

Finding leaks is the first step. Verifying the actual savings through consumption measurements taken before and after the repair makes this a robust efficiency measure. The ultrasound camera, LD 500, Leak Reporter, flow measurement and monitoring complement one another in the tasks of localisation, documentation, prioritisation and validation.

The acoustic map is generated using beamforming. The microphone array detects the incoming ultrasonic signals at defined positions. The system performs a calculation to determine the probable position of the sound source based on the differences in travel time and phase. A sound pressure value is determined for each pixel and superimposed on the camera image as a heatmap.

The heatmap is not a photograph of the leak, but a visualisation of the calculated sound source. By superimposing it onto the camera image, the hotspot can be clearly attributed to the component causing it.

Further reading: Beamforming explained simply

The detection range of an ultrasound camera is not determined by a fixed value. It depends primarily on the size of the leak, system pressure, frequency range, environmental interference, viewing angle and atmospheric attenuation. Large leaks can be detected from a significantly greater distance than micro-leaks.

Acoustically, it is not the type of gas itself that is detected, but rather the ultrasonic waves generated by the gas escaping. Whether a leak can be detected depends primarily on the pressure difference, the geometry of the leak, the flow rate, the measurement distance and the environmental conditions. The type of gas, on the other hand, is mainly relevant for the economic, safety-related and quality-related assessment of the leak.

Many production noises fall within the audible range. However, plant components can also generate ultrasonic signals, for example pneumatic actuators, conveyor systems, presses, friction points, blow-off nozzles, partial discharges or other leaks. Using FFT analysis and the selection of a suitable frequency window, such sources of interference can be identified and specifically filtered out.

Reflections occur on hard surfaces such as glass, Plexiglas, concrete, sheet metal or smooth machine casings. A hotspot on a wall, floor or other surface may therefore also be caused by an echo. The most important step in the test is to change the viewing angle: a real sound source remains localised at the component causing it, whilst a reflection shifts or disappears.

Further reading: Leak detection despite interference sources and FFT