Sequential multi-parametric inversion of controlled-source electromagnetic data for imaging conductive and susceptible structures

Title
Sequential multi-parametric inversion of controlled-source electromagnetic data for imaging conductive and susceptible structures
Author
Kyubo Noh
Advisor(s)
Joongmoo Byun
Issue Date
2018-02
Publisher
한양대학교
Degree
Doctor
Abstract
Controlled-source electromagnetic (CSEM) method has been widely used to image subsurface structures of electromagnetic (EM) properties, but mainly restricted to the electrical conductivity. In this dissertation, multi-parametric sequential inversion techniques were developed to image not only conductivity structure but also susceptibility one. In order to suggest the sequential inversion process, the three-dimensional (3D) forward modeling algorithm was developed based on the edge finite element method and then a series of response analysis processes was carried out using the modeling algorithm. Through the response analysis, the frequency and phase characteristics of the responses from each physical property in general CSEM circumstances (including transmitter frequency, survey geometries, and values of physical properties) were systematically investigated. Based on knowledge from the response analysis, the sequential inversion was set to iterate the conductivity inversion using out-of-phase data and the susceptibility inversion using in-phase data. In addition, the possibility of a linear magnetostatic interpretation was discussed for the susceptible anomalies of less than 0.1 SI. The inversion process was realized by developing two inversion frameworks using 1) a voxel-based parameterization and 2) both voxel- and model-based parameterizations. For the voxel-based inversion, two workflows were suggested: 1) using EM kernel for both conductivity and susceptibility inversion and 2) using EM and linear magnetostatic kernel, respectively, for conductivity and susceptibility inversion. Inverting real airborne data sets using both inversion workflows, similar results were obtained but the computing time was dramatically saved with the second workflow. For the model-based parameterization in the second inversion framework, I developed a particle mapping (PM) method which optimizes the location of each particle using a priori information on the susceptibility-volume product as a constraint. This PM method was applied to the imaging of magnetically enhanced proppants for hydraulic fracturing monitoring. In addition, the new regularization method which adopts a fuzzy clustering method was suggested to incorporate geometries of fracture sets derived from microseismic interpretation. Conducting numerical experiments for two models simulating fully and partially propped fracture sets, the PM inversion provided clear and intrinsically interpretable imaging results for the propped volume.
URI
http://www.dcollection.net/handler/hanyang/000000105901http://repository.hanyang.ac.kr/handle/20.500.11754/68454
Appears in Collections:
GRADUATE SCHOOL[S](대학원) > EARTH RESOURCES AND ENVIRONMENTAL ENGINEERING(자원환경공학과) > Theses (Ph.D.)
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE