Sewer Condition and Blockage Detection Classification Using Novel Acoustic Instrumentation

在英国，下水道系统正在老化，监控不力，长约30万公里。由于城市化程度提高，环境监管更加严格，以及气候变化的后果是降雨事件更加频繁和强烈，该系统面临着越来越大的容量需求。水务管理局是水务行业的经济监管机构，它规定自来水公司有法律义务维护其下水道系统的结构和运行状况，并逐步减少下水道的洪水事件。2004年，Ofwat确定了5 700起下水道洪灾事件和另外11600处财产，年洪灾风险为10%。在这些个案中，约90%的水浸是由一段污水管阻塞所致。因此，监测管道是否有障碍物，然后迅速清除它们，可能会成为减少下水道泛滥的有效方案的重要组成部分。该项目的目的是开发新的声学技术(传感器和软件)，以生产可用于现场下水道的接近市场的原型，以快速和客观地测量管道内状况并识别堵塞和损坏。下水道监测目前仅限于解读闭路电视画面或使用光线勘测。这些方法需要插入带有摄像头的移动手推车，并沿管道部分向上移动，采集图像，然后手动检查图像，并根据标准的下水道修复手册文件对缺陷/障碍物进行分类。与英国下水道运营商的讨论表明，他们每5年就会对大约2%的下水道网络进行CCTV调查。该项目将开发一种替代的快速方法，用于客观分析下水道状况和定位堵塞。布拉德福德大学最近的研究(EPSRC赠款EP/D058589/1)证明，可以在实验室管道中使用声学检测方法测量管道堵塞面积、裂缝程度、水位和横向连接位置。通过这种方式，可以在空气中的检修孔之间确定下水道的管道状况。该声学装置的关键部件是一个小型多传感器阵列和先进的实时信号处理算法，可以克服沙井环境中环境噪声和混响的影响。将这项技术转让给商业部门的关键障碍是：(I)建造本质安全、坚固耐用的仪器；(Ii)获取足够的现场数据，向潜在用户清楚地证明，声波传感器可以可靠地识别下水道中的缺陷和障碍物；(Iii)证明声波技术可以提供与传统闭路电视方法兼容的数据，并可以映射到英国自来水公司使用的现有污水状况分类和污水数据库中。实地收集的数据将对确定潜在利益攸关方的工业供资和商业利益的性质和范围至关重要。调查人员正在联系四家支持这项工作的商业组织：Richard Long Associates、Mouchel Parkman、约克郡水务公司和泰晤士水务公司。人们普遍认为，闭路电视测量每100米长度需要2到4个小时进行测量，图像分析的时间也差不多。根据出入限制的不同，这一成本在每直线米2至40之间。声学测量可能会将测量和分析时间减少到每100米数十分钟。可以实现显著的效率。这将使运营商能够更频繁地对其网络进行更多的调查，从而能够制定更好的监测战略，及早识别缺陷，特别是堵塞，从而减少洪水事件和更好地进行计划维护。

In the UK, the sewer system is ageing, poorly monitored and around 300,000 km long. The system is subject to increasing capacity demands because of increased urbanisation, more stringent environmental regulation and the consequences of climate change in the form of more frequent and intense rainfall events. OFWAT, the economic regulator for the water industry, imposes a legal duty on water companies to maintain the structural and operational conditions of their sewer systems and also to progressively reduce flooding incidents from sewers. In 2004 OFWAT identified 5700 sewer flooding incidents and an additional 11600 properties with a 10% annual risk of flooding. In approximately 90% of these cases, flooding was caused by an obstruction in a section of sewer pipe. Consequently, monitoring pipes for obstructions and then rapidly removing them could form an important part of an effective programme to reduce sewer flooding. The aim of this project is to develop novel acoustic technology (sensor and software) to produce a near market prototype that can be used in a live sewer to measure rapidly and objectively in-pipe condition and identify blockages and damage. Sewer monitoring is currently limited to the interpretation of CCTV pictures or the use of LightLine surveys. These methods require a mobile trolley with camera to be inserted and travel up a pipe section acquiring images which are then manually examined and defects/obstructions classified according to the standard Sewer Rehabilitation Manual document. Discussions with sewer operators in the UK indicate that they CCTV survey around 2% of their networks every 5 years. This project will develop an alternative fast method for analysing objectively the condition of a sewer and locating blockages. Recent research at the University of Bradford (EPSRC grant EP/D058589/1) has proved that the area of pipe blockage, extent of cracks, water level and positions of lateral connections can be measured in the laboratory pipes using an acoustic method of inspection. In this way the pipe condition of a sewer could be determined between manholes in the airborne regime. The key element of this acoustic device is a small multi-sensor array and advanced, real-time signal processing algorithm which overcome the effects of ambient noise and reverberation in the manhole environment. The key barriers to transferring this technology to the commercial sector are: (i) construction of an intrinsically safe, robust instrument; (ii) acquisition of sufficient field data to demonstrate clearly to potential users that acoustic sensors can reliably identify defects and obstructions in sewers (iii) demonstration that the acoustic technology can provide data that is compatible with the conventional CCTV methods and can be mapped onto the existing sewer condition classifications and sewer databases used by UK water companies. The data collected in the field will be critical in determining the nature and extent of the industrial funding and commercial interest from potential stakeholders. The investigators are in contact with four commercial organisations who support this work: Richard Long Associates, Mouchel Parkman, Yorkshire Water Services and Thames Water. It is generally agreed that CCTV surveys take between 2 and 4 hours per 100m length for measurement with a similar time for image analysis. This costs between 2 and 40 per linear metre depending on the access constraints. Acoustic measurement potentially reduces the measurement and analysis time to tens of minutes per 100m. Significant efficiencies can be achieved. This would enable operators to survey more of their network more frequently so would allow for better monitoring strategies to be developed, the earlier identification of defects and especially blockages leading to fewer flooding incidents and better planned maintenance.