Acceleration-Based Multihop Wireless Sensor Network for Leak Detection in Pipeline Networks



Leakage in liquid distribution pipeline networks is a significant problem in Nigeria, which not only cause immense amount of liquid lost but serious environmental and health hazards. For this reason, accurate detection and localization of leakages has become one of the most focused areas of research.

Phase One – Completed 2017

  • It has been proven theoretically that variation in pipe surface vibration is proportional to the variation in pressure and flow rate. From this model, it is assumed that acceleration events picked up by sensors on a pipeline will give a good approximation to the pressure and flow rate in the interior of the pipeline. Since leaks typically correspond to instantaneous observable decrease in pressure and flow rate, it can be assumed from the model that observable decrease in surface vibration are leak indicators.
  • Project Description

In the first phase of this research, an outdoor multi-hop experimental testbed of wireless sensor network (WSN) measuring 128m x 110m which incorporate six (6) custom-built Micro-Electro-Mechanical Systems (MEMS) based wireless sensor nodes for non-invasive measurement of induced vibration was developed for pipeline monitoring.

The MEMS accelerometer-based wireless sensor nodes were used for measuring flow-induced vibration on the surface of a pipe and to determine the change in water pressure caused by rupture and fluid leak. Experimental investigation of the relationship between flow-induced vibration and the pressure fluctuations was conducted using the Testbed. Valves were installed and adjusted manually to simulate some normal events on a pipeline network such as valve opening or closing, switching on and off the pumps and to simulate pipeline damage. Measurements of vibration were performed in pipe sections of a water-filled loop subjected to a wide range of flow rates. Analysis of the measured flow-induced vibration data showed that there is a non-linear but proportional relationship between the water flow rate and flow induced vibration. Therefore, water flow rate monitoring can be transformed into acceleration-monitoring of the pipe surface. The latter is a significantly more economical alternative due to the use of less expensive sensors such as MEMS or other acceleration sensors.

Phase Two – Field Test

This phase will require the construction of more thirty (30) numbers of The MEMS accelerometer-based wireless sensor nodes and the extension of the experimental testbed to cover the water distribution network covering the hostels in Nnamdi Azikiwe University, Awka campus.

In the field test, changes in induced pipeline surface vibration due to scheduled events such as valve opening and closure by operators, switching on and off of pumps, will be measured and reported over a period of at least 12 months. Characterization of “normal” events in the pipeline will lay the ground work for monitoring pipeline network for “true” leaks.

The field test will involve collection and analysis of large amount of data with the aim of collecting enough data to analyze the signature of the scheduled events. By determining the unique signatures of scheduled events in the pipeline, one would be able to determine with some accuracy what to expect if there is an unexpected rupture along the pipeline.

The research team comprises the following academics from the department of Electronic & Computer Engineering:

  1. Engr. Prof. V.E. Idigo – Research Partner
  2. Engr. A.C.O. Azubogu (PhD) – Research Partner
  3. Engr. Dr. C.O. Ohaneme – Research Partner
  4. Engr. Gerald Nwalozie – Research Partner/PhD candidate
  5. Three M.Eng candidates