A novel formation model for electromagnetic telemetry including surface and innovative casing antenna system


Authors

Qiuzhao Dong (Weatherford International) | Jiefu Chen (University of Houston)

Publisher

SEG - Society of Exploration Geophysicists

Publication Date

October 16, 2016

Source

2016 SEG International Exposition and Annual Meeting, 16-21 October, Dallas, Texas

Paper ID

SEG-2016-13957149


Abstract

EM telemetry such as gap-based, cable-based and casing based have been introduced and applied in well logging field. The interpretation could be based on numerical model such as FEM method, transmission line method etc. or simple data transmission EM method such as Soulier proposed (Soulier, 1993). Numerical methods are more accurate especially for oil based mud but need longer computation time and more parameters; the Soulier method (approximated analytical method) is quick and simple but less accurate especially for oil based mud. All these interpretations need the preconditioned layer formation. The measured formation resistivity varies much and it is too time-consuming and not realistic if each measured point is seen as one layer especially for numerical model.

An efficient layered average method is proposed to solve the layer separation problem in EM telemetry application. The method employs the median filter to remove too noisy points. The averaging resistivity of each layer is approximated based on the EM propagation along a coaxial cable acting as a wave guide, it takes consideration of the current leaking rate to formation, and it also automatically includes the effect of the thin conductive layer. Layer separation is decided by a cost function which uses the sensitivity to formation resistivity. The criterion is also optimized with regard of the accuracy and computation complexity. Numerical samples are presented to prove the concept and the robust of the method. Finally some field examples are given to demonstrate the behaviors of the electromagnetic telemetry system and show the efficiency of the layer separation method.