Pressure Modelling: Where it all goes wrong
Gareth S. Yardley1 and Brent Couzens2
1Shell UK Exploration and Production, Aberdeen AB12 3FY, UK
2Shell International Exploration and Production Inc., P O Box 481, Houston, Texas 77001-0481, USA
Shell uses basin modelling software for a range of applications including charge prediction, reservoir quality prediction and pressure prediction. When used for pore pressure prediction, basin modelling is viewed as one of a suite of complimentary tools. Other commonly used tools are seismic velocity analysis, pressure cell/geological modelling and real time monitoring of drilling operations. Each pressure prediction methodology has its pros and cons and an integrated workflow using the relevant tools for a particular location is recommended practice to yield the most robust prediction possible.These techniques work successfully in many areas, particularly high sedimentation rate, clastic dominated deltaic environments where disequilibrium compaction is the dominant cause of overpressure.As exploration moves into more challenging geologic environments and data-constrained frontier settings, conventional approaches to pore and fracture pressure prediction often work less well. In many of these settings, assumptions inherent in the modelling are violated (e.g. rocks are at maximum historical stress, the principal stress direction is vertical, the velocity signal reflects the stress state of the mudrocks, reservoirs and juxtaposed mudrocks are in pressure equilibrium). Challenging environments include high pressure / high temperature settings in clastic basins and/or carbonate-dominated or mixed lithology settings, where vertical effective stress-based rock property models do not sufficiently describe the diagenetic processes that control porosity and permeability evolution. Additional challenges are found in compressive settings and uplifted sections, where we lack quantitative models for rock property evolution accounting for complex stress histories and non-vertical stresses. In many of these settings, limited or poor quality calibration and seismic data and an incomplete description of the geologic setting further complicate predictions. This presentation will show examples of our recent experience of pore pressure prediction in these challenging areas and we will highlight problems with our current basin modelling tools and rock property models.These represent current industry problems. The challenge is out there for industry and academia alike to tackle these problems from both the development of suitable software tools and the quantitative understanding of subsurface processes such as rock property evolution.
AAPG Search and Discover Article #90066©2007 AAPG Hedberg Conference, The Hague, The Netherlands