低频电磁探测技术在煤层气富集区的应用

A field experiment of low-frequency magnetotelluric detection was designed and conducted in Hudi, Qinshui Basin. Based on the processed multi-source electromagnetic data of this survey, the geological evaluation and coalbed parameter analysis are carried out in order to detect the coalbed me-thane (CBM).The following understandings are obtained: A. Different lithologic layers and coalbeds can be effectively identified on super-low frequency (SLF) electromagnetic data; B. Collapse pillars and faults can be extracted from 2D apparent resistivity inversion of audio frequency electromagnetic (AMT) data and they are consistent with that of seismic data. And subtle beds are also identified; C. The anomaly center depth, anomaly scope, and its average magnitude of SLF data have good correlation respectively with the buried depth, thickness, and gas content of coal reservoirs. The prediction accuracy of gas content is 87.33% with a coalbed methane evaluation model supported by vector machines, and the average error is only 0.17m ³ /t. Our research proves that the low-frequency magnetotelluric detection is effective in the CBM exploration.