Complex Conductivity Tomography for the Identification of Filling Materials in a Karst Pipeline

• Published in: Pure and applied geophysics 2022-05, Vol.179 (5), p.1729-1748
• Main Authors: Liu, Wei, Zhou, Qiyou, Chang, Yanjun, Li, Zhihua, Gan, Fuping
• Format: Article
• Language: English
• Subjects: Buried pipes; Carbonates; Clay; Comparative analysis; Conductivity; Data collection; Detection; Earth and Environmental Science; Earth Sciences; Engineering geology; Field tests; Fillers; Geology; Geophysics/Geodesy; Hydrogeology; Identification; Identification methods; Induced polarization; Karst; Mathematical models; Methods; Model testing; Modelling; Parameter identification; Parameters; Sand; Submarine pipelines; Tomography; Water
• ISSN: 0033-4553; 1420-9136
• DOI: 10.1007/s00024-022-02991-4

The identification of filling materials in an underground karst pipeline is highly important in karst hydrogeology and karst engineering geology; however, there is no suitable surface detection method to identify the common filling materials (such as clay, sand, water) in a buried karst pipeline. This study aimed to identify pipeline materials using the spectral induced polarization (SIP) method on the surface. We carried out experiments in the laboratory and field under a carbonate environment to test the validity of various IP parameters obtained from complex conductivity tomography for identifying pipeline materials. The gradient array was used to collect data in the study, and the Cole–Cole model was introduced. The data were processed using a two-step inversion method. Firstly, the conductivity magnitude and the phase during each frequency are inverted, and later the parameters of the Cole–Cole model are inverted according to the phase spectrum. The results indicate the differences in spectral characteristics of the phase inversion section, the imaginary conductivity inversion section, the Cole–Cole model parameter inversion section, and the Cole–Cole model parameter values ( m , τ , c ) can be used to distinguish clay-filled, sand-filled or water-filled karst. Of particular note is the τ parameter of the Cole–Cole model; the τ value of the water-filled pipeline can reach tens of seconds and shows strong contrast with surroundings in both the pilot physical model test and field test. The water richness of the medium may be closely related to the τ value, which could be a key parameter in identifying whether a karst pipeline is filled with water. A comparative analysis of all IP parameters could greatly improve the recognition of pipeline filling materials in the field. This work proposes a SIP method for high-resolution detection of near-surface karst pipelines.