Modeling and locating underground water pipe leak with microseismic data

Traditional pipeline leak locating methods require that geophones have to be, placed on the pipe wall. While if the exact location of the pipeline is unknown, the leaks may not be identified accurately. To solve this problem, considering the characteristics of pipeline leak, a continuous random seismic source model is proposed and geological models are established. Based on the two dimensional (2D) viscoacoustic equations and the staggered grid finite-difference (FD) algorithm, the microseismic wave field generated by a leaking pipe is modeled. Cross-correlation analysis and the simulated annealing (SA) algorithm are employed to obtain the time difference and the leak location. Analysis and discussions of the effects of number of recorded traces, survey layout, and offset and trace interval on the accuracy of the estimated location are also conducted. Simulation and data field experiment results indicate that: (1) A continuous random source can realistically represent the leak microseismic wave field in a simulation using 2D viscoacoustic equations and staggered grid FD algorithm. (2) For the leak microseismic wave field, the cross-correlation method is effective for calculating time difference of the direct wave relative to the reference trace. However, outside the refraction blind zone, accuracy of the time difference is reduced by the effects of refracted wave. (3) The SA algorithm based upon time difference, helps to identify the leak location effectively, even in the presence of noise. Estimation of the horizontal distance is more accurate than that of the depth, and the locating errors increase with increasing number of traces and offset. Moreover, in the refraction blind zone, trace interval has almost no impact on the accuracy of the location estimate. And the symmetrical array provides a higher estimate accuracy than the asymmetrical array. (4) The acquisition method of time difference based on the microseismic theory and SA algorithm has a great potential for locating underground pipelines leak from an array located on the ground surface. (C) 2016 Elsevier B.V. All rights reserved.