Research Area: Exploration

Seismic waveform inversion from fracture parameters

Project Number: 6263
Project Duration: 1. January 2013- 31. December 2015

Project Director: Morten Jakobsen


The goal of the project was to understand fractures in reservoirs and how to glean the fracture properties from seismic data.

Seismic data is recordings of waves that have travelled through the earth. These waves may originate from a natural earthquake or from artificial shots. In both cases will the wave be altered by the rock that it travels through and with theoretical models it is possible to invert for which conditions that have caused the recorded signal. Thus are we able to look inside rocks and the earth itself. Some of the conditions that alters the waves can be the type of rock that the wave travels through, it can be big faults or it can be small fractures or cracks in the rock that are filled with water or oil and gas. This project has focused on the latter. We call such small fractures sub-seismic. This means that the fractures are smaller than the wavelength that we study. In such conditions, we cannot see the individual fracture, but we can detect the gathered response of many. Although these fractures are small, they are very important to understand for those that search for fluids beneath our feet, as these fractures can contain much of the fluids we can extract and they can be crucial in the fluids ability to flow within the rock.

A seismic wave is generally an elastic wave. That means that many independent factors that determine the behaviour of the wave. In this project, we have used a rock physic model to reduce this number to only a few. Thus increased the calculation speed and made elastic wave inversion more accessible. Also by inverting only for fracture parameters, we do not only reduce the number of unknown model parameters, but also obtain a more stable and robust solution to the inverse problem. The most important advantage for this method from a geological point of view is that fracture parameters are much more relevant for interpreting the rock.

We have made the theoretical foundation and the first proof of concept of using a transport matrix for one fracture set to invert for fracture density. We have also investigated effects of pore fluid pressure communication between aligned fractures and pores. Our methods can in principle also deal with multiple fracture sets, and numerical experiments related to this can now easily be performed in future projects.

Pilskog, I. and Jakobsen, M., 2016. Fracture model based seismic waveform

inversion in HTI media: Effects of fracture-fracture interaction.

Journal of Geophysics and Engineering, under preparation.

2. Jakobsen, M. and Pilskog, I., 2016. Elastic Born inversion for fracture density

using a Gassmann-consistent model of fractured porous media.

Geophysical Journal International, under preparation.

Peer-reviewed conference papers

1. Jakobsen, M. and Pilskog, I., 2016. Gassmann-consistent Born inversion for

fracture density. Extended abstract, 78th EAGE meeting, Vienna (submitted).

2. Pilskog, I., Lopez, M. and Jakobsen, M., 2015. T-matrix FWI for fracture

parameters. Peer-reviewed extended abstract, 77th EAGE annual meeting,

Madrid, Spain.

3. Jakobsen, M., Pilskog, I., and Lopez, M., 2015. Generalized T-matrix approach to

seismic modeling in fractured reservoirs and related anisotropic systems. Peerreviewed

extended abstract, 77th EAGE Annual Meeting, Madrid, Spain.

4. Pilskog, I., Lopez, M. and Jakobsen, M., 2014. Linearized waveform inversion for

fracture density, extended abstract, 16th International Workshop on Seismic

Anisotropy (16IWSA), Natal, Brazil.

Other relevant publications (project director)

Jakobsen, M. and Ursin, B., 2015. Full waveform inversion in the frequency domain using direct iterative T-matrix methods. Journal of Geophysics and Engineering, 12, 400-418.

PostDoc: Ingjald Pilskog

E-mail: ingjald@pilskog.no
Videnskapsakademiets logo
VISTA Drammensveien 78 NO-0271 OSLO NORWAY Tel: +47 22 84 15 00 / +47 454 03 077 E-mail: vista@dnva.no