Infrastructure protectionVirtual guard detects real-time leaks in water, oil-, or gas pipes

Often, water, gas, or oil distribution networks suffer from leaks in storage tanks, pumping failures, or illegal tapping. In order to prevent losses which typically result, Cristina Verde Rodarte, researcher at the Institute of Engineering of the Autonomous National University of Mexico (UNAM), designed a virtual guard that immediately detects abnormalities in any type of duct.

The software is called VIVIUNAM, and it performs logical deductions in real time, allowing the identification of the type of failure which has occurred and the source of the problem, thus preventing waste of time and resources required by digging or manually searching for the problem throughout the pipeline.

An Investigación y Desarrollo release reports that the virtual vigilante works with an algorithm, which by the laws of physics and the application of a mathematical model of fluid mechanics, calculates a series of data indicating what is the behavior of gas, water, or oil pipelines in normal operating conditions. These, in turn, are compared with the record of pressure measurements within the pipe and the difference between these measurements indicate the presence of leakage.

When the results from the mathematical model do not match the automatically recorded measurements, it is an indication that an error or an abnormal event is happening: for example, the pressure sensor may be disconnected, a leak is present, an illegal connection or a disturbance that alters the behavior of fluids within the pipeline.

Ph.D. in Electrical Engineering by the University of Duisburg in Germany, Verde Rodarte, said that the chemical and oil industries, and general processes involving fluid transport systems should have automatic, safe, and efficient monitoring in order to prevent accidents with highly volatile fluids or pollutants which have a negative impact on the economy, society, and the environment. She proposes to place the VIVIUNAM system in control distribution networks to report the presence of disturbances, and develop adequate contingency plans.

In various networks, such as the Cutzamala system, pumping plant Xotepingo (Mexico City), or distribution networks Pemex (Mexican Oil), pipelines have pressure and fluid gauges, but the data is only used for administrative purposes, rather than also use it to look for leaks so the problem could be corrected, noted Cristina Verde.

The release notes that an experiment was conducted in which leakage was emulated in a pilot pipeline of 200 meters with a diameter of 10.4 inches. The pipe was equipped with pressure and expense sensors. VIVIUNAM was automatically operated from a laptop, and the algorithms developed by scholars of II-UNAM proved effective.

The technology used to diagnose leakage is based on mass balances, measuring expense, pressure, and temperature at the ends of a pipeline without laterals. Calibration for each model is performed according to the topology and physical properties of the fluid and duct in question. Verde says this technique is economical because it does not require additional sensors to those already available in distribution pipelines. The software only requires data from the distribution network for the algorithm to work, she says.