Acoustic leak detection sensor technologies rely on the fact that liquid and gas leaks are noisy. However, it’s not necessarily a noise that the human ear can detect, it can be in the ultrasonic region. Leaky water pipes are a significant environmental challenge – UN Sustainable Development Goal No. 6 states that clean (drinking) water is a scarce source that needs to be protected; aging infrastructure adds to the challenges. Leaking gas pipes can be dangerous as well as costly. In either case, acoustic leak detection can provide a solution.
当流体或气体出口时，会产生声噪声。与具有大约2 kHz的声学签名的液体相比，气体往往会产生更高的噪音，直至超声范围。
漏水管道
许多因素会影响泄漏的水管产生的噪声，包括管道直径，泄漏的大小和水压。此外，管道材料和管道周围的地面对噪声的传播方式产生了强烈影响。例如，与金属管相比，水管充当声能和塑料管的低通滤波器，特别是引入更高的阻尼水平。
泄漏产生的声能通过管壁，通过管道周围的土壤以及管道中的水传播。这些声音通常在岩石土壤和干燥土壤中传播更大的距离。在许多情况下，泄漏区域周围的水池的形成会使声音降至太安静，即使使用高度敏感的设备也无法听到。如前所述，气体往往会产生较高的频率声音，而与典型的漏水声音相比，这些声音可能会带来更远的声音。如果漏水引起的噪音太微弱而无法检测到，则可以将示踪气体注入管道中，以促进使用简单的声学听力棒泄漏的位置（图1）。
当今的许多水分配网络已在战略地点永久安装了听力设备。使用频率分析和互相关方法，永久性听力设备可以帮助检测泄漏的发生。频率分析测量并滤除干扰泄漏识别的环境噪声。
Cross correlation can be used to identify the exact location of a leak. It uses two sensors and the leak must be located between the sensors. In cross correlation, the time delay of the leak noise between the two sensors is combined with the speed of sound in the specific type and size of pipe to determine where the leak is between the sensors. Unfortunately, local factors such as soil condition can vary along the length of a pipe making it challenging to determine the exact speed of sound. Placing sensors closer together throughout the distribution pipeline is one way improve the performance of leak detection systems.
In a growing number of cases, acoustic noise sensors are being integrated into water meters. Meters are found at numerous and strategic locations and provide good coverage for leak detection. The meter is an integral part of the water distribution system and has a good mechanical coupling needed to listen for leaks. And smart meters have an integral wireless communications link that can deliver information about potential leaks in addition to information about water usage.
将泄漏检测整合到智能电表中时的主要挑战是低功耗。智能水表电池预计至少将持续15年。为了解决功耗的问题，使用了非常低的电源传感器，并且每天只能在仪表处进行一次声学噪声签名。
新管道材料带来新的挑战
当内部和外部压力试图均衡时造成的泄漏位置的湍流是噪声的来源，取决于管道材料以及其他因素。铜，钢管和铸铁管中的泄漏通常在500至1,500 Hz的范围内产生声音。PVC管中泄漏产生的声音范围通常在70 O 850 Hz范围内。
此外，与金属管相比，PVC管与水的耦合更强，这会导致噪声信号的显着衰减和泄漏噪声在PVC管中的传播不如金属管中的PVC管道传播。结果，由于需要更多的检测器，因此使用两个泄漏探测器以及声学相关性来定位泄漏并不那么简单或实用，并且在PVC管道的情况下，它们需要更近。
The relatively high rates of attenuation of leak noise propagating in PVC pipes (the ‘wave-speed’) and the variability in the speed at which it propagates both affect correlator performance. Accurate estimates of wave-speed are especially important when looking for leaks in PVC pipes. In most installations, the wave-speed is estimated from historical databases and determined from calculations made using assumed material properties and pipe geometry. The use of historical databases greatly limits the applicability of this correlating technique. More recently, a finite element method (FEM) has been developed using multiphysics software to calculate the wave-speed and wave attenuation values used in the correlator (Figure 3). The results of the FEM calculations were compared for two cases, in the first instance, the out-of-bracket excitation mechanism was a loudspeaker, in the second instance, the out-of-bracket excitation was a leak (Figure 3) the predictions using both out-of-bracket excitation mechanisms were in close agreement with actual systems.
超声传感器检测到气体泄漏
加压气管泄漏在25至100 kHz范围内发射超声波噪声。传统的气体泄漏探测器测量累积气体，并有滞后时间才能做出反应。超声波泄漏探测器可以在发生时立即识别泄漏，从而触发更快的警告。
For example, one stationary pole-mounted gas leak detector uses four ultrasound acoustic sensors to monitor wide areas for gas leaks (Figure 3). It can be used indoors or in outdoor environments and can withstand precipitation and wind and can identify leaks regardless of their location or gas stratification or dilution. Features of this system include:
- 对2至40米（7至130英尺）的有毒，可燃或惰性气体泄漏的瞬时反应
- 传感器没有运动部件，没有年龄或漂移，可以在不校准的情况下运行，并包括自动化自动测试以进行故障安全操作
- 工作温度范围为-40至85°C
用于识别工业压缩空气系统泄漏的手持仪器使用64个麦克风的矩阵作为以特定模式排列的声传感器，以及矩阵中间的可见摄像头，以提供扫描区域的图像（图4）。根据声源和仪器的相对位置，64个麦克风在略有不同的时间接收声音。麦芽管间时间差异用于定位声源，然后将声音叠加在相机所拍摄的图像上，以向操作员显示泄漏的确切位置。
Compressed air leaking from a pipe creates broadband noise across the audible and ultrasonic frequency ranges to 40 kHz and higher. Many industrial compress air leak sensor systems use narrowband ultrasonic sensors centered at about 40 kHz. The distance and angle between the sensor and the leak can impact the effectiveness of ultrasonic sensors. The use of narrowband ultrasonic sensors has limitations, including:
- Ultrasonic frequencies are strongly attenuated by atmospheric absorption.
- 测量的角度有很强的影响ound pressure level produced by a compressed air leak.
- 工业环境中常见的嘈杂环境会使窄带超声传感器的性能降低。
用可以在Audible和Ultrasonic范围中使用的宽带传感器代替窄带超声传感器可以补偿窄带传感器的局限性。扩展的频率范围增加了泄漏检测系统的质量和准确性。在压缩空气泄漏的情况下，可听见的频率范围具有最大的声压水平，而超声波范围内的声压水平则显着较低，使用窄带超声传感器进行检测，具有挑战性。
概括
如图所示，使用各种声学泄漏检测技术，可以准确的水和气体泄漏位置。当内部和外部压力试图均衡时造成的泄漏位置的湍流是噪音的来源，并且取决于许多因素。漏水往往会在几千期以下发出声音，并且可以使用窄带声传感器有效检测到，而气体泄漏会产生更广泛的频谱，该光谱延伸到超声波范围内，并且可以从使用宽带传感器进行检测中受益。
参考
关于超声波气泄漏探测器，艾默生
如何使用声学测量来定位漏水, Kamstrup
声学成像摄像机的泄漏率定量（LRQ）方法，Fluke
关于振荡声学在塑料水分配管道泄漏检测中的作用，国际结构动力会议
图片
图1：泄漏探测器
图2: International Conference on Structural Dynamics, Page 2, Figure 1
图3：艾默生
图4：2个产品图像的Fluke复合材料
Filed Under:Sensor Tips
