ERA

Download the full-sized PDF of Multivariate Stochastic Seismic Inversion with Adaptive SamplingDownload the full-sized PDF

Analytics

Share

Permanent link (DOI): https://doi.org/10.7939/R3W669M8X

Download

Export to: EndNote  |  Zotero  |  Mendeley

Communities

This file is in the following communities:

Graduate Studies and Research, Faculty of

Collections

This file is in the following collections:

Theses and Dissertations

Multivariate Stochastic Seismic Inversion with Adaptive Sampling Open Access

Descriptions

Other title
Subject/Keyword
Stochastic Inversion
Geostatistical Modeling
Multivariate
Type of item
Thesis
Degree grantor
University of Alberta
Author or creator
Hadavand-Siri, Maryam
Supervisor and department
Clayton Deutsch, Civil and Environmental Engineering
Examining committee member and department
Boisvert, Jeff (Civil and Environmental Engineering)
Bentley, Larry (Geoscience, University of Calgary)
Sacchi, Mauricio ( Physics, Geophysics)
Leung, Juliana (Civil and Environmental Engineering)
Askari-Nasab, Hooman (Civil and Environmental Engineering)
Department
Department of Civil and Environmental Engineering
Specialization
Mining Engineering
Date accepted
2017-03-23T15:32:13Z
Graduation date
2017-06:Spring 2017
Degree
Doctor of Philosophy
Degree level
Doctoral
Abstract
Reservoir management tools use geostatistical models to make better decisions and improve hydrocarbon recovery. Developing improved numerical modeling techniques that lead to more accurate and precise geostatistical models will improve flow forecasting, hydrocarbon recovery, reservoir management and contribute to more responsible and sustainable management of our natural resources. The conventional stochastic seismic inversion techniques integrates different source of data to obtain high quality geostatistical models. These techniques initially provide acoustic impedance models that match well and seismic data. Such acoustic impedance models are related to reservoir physical properties via rock physics models. These techniques aim to reproduce the data within the quality of data. However, there is no guarantee that the final petrophysical models will reproduce original seismic data. Fidelity with the original seismic data will be reduced due to an element of randomness at each step. To overcome this issue a new approach is proposed and developed. The new approach proposes to simulate multiple reservoir physical properties simultaneously. This research study presents a research toward a fully coupled categorical - multivariate continuous reservoir modeling in stochastic inversion context with Petro-Elastic Model and convolution. The application of multivariate geostatistical techniques would improve conventional stochastic inversion approaches. The flexibility of using variogram based techniques and multi point statistic methods to model complex geological features further improves the stochastic inversion approach. The multivariate stochastic inversion approach combines a trace by trace (column wise) adaptive sampling algorithm with multivariate geostatistical techniques to pick the best physical properties of reservoir that match the actual seismic data. The adaptive sampling method uses an acceptance-rejection approach to condition geostatistical models to well and seismic data. This technique samples the realizations inside the space of uncertainty. The number of realizations attempted is changed based on the size of space of uncertainty. In this study, a general framework is presented to calculate the size of the space of uncertainty. This becomes practically relevant when rejection sampling approaches are being used to condition geostatistical models as in the case of stochastic inversion. The size of the space of uncertainty is shown to be the product of exponential entropy values. This is corroborated from information theory, but the application of this in presence of spatial correlation and conditioning data is new. Modeling multiple reservoir properties simultaneously through the close integration of seismic inversion and multivariate geostatistical techniques leads to high resolution reservoir property models that are suitable for improved reservoir management. A case study with realistic data set is developed to compare the results of multivariate stochastic inversion approach with conventional stochastic method.
Language
English
DOI
doi:10.7939/R3W669M8X
Rights
This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for the purpose of private, scholarly or scientific research. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
Citation for previous publication

File Details

Date Uploaded
Date Modified
2017-03-23T21:32:15.740+00:00
Audit Status
Audits have not yet been run on this file.
Characterization
File format: pdf (PDF/A)
Mime type: application/pdf
File size: 12394828
Last modified: 2017:06:13 12:26:44-06:00
Filename: Hadavand-Siri_Maryam_201703_PhD.pdf
Original checksum: 8eaab3c31b9fb9fd76f9ad10b947e270
Activity of users you follow
User Activity Date