Improved Pore Pressure Prediction by Integrating Basin Modeling and Seismic Methods

Hans Martin Helset, Mikael Lüthje, and Ira Ojala
SINTEF Petroleum Research AS, Stavanger, Norway

Introduction
The objective of this work is to develop improved methods for predicting pore pressure from seismic data. Analysis of seismic data is combined with basin modeling techniques in order to obtain an accurate and robust prediction of pore pressures prior to drilling. As part of the project, velocity-depth trend models for sedimentary rocks have been developed and calibrated.

Seismic velocity data is often used for prediction of pore pressures prior to drilling. Standard methods include effective-depth methods and empirical methods (e.g. Eatons method). These methods have been used with fair success in areas like the Gulf of Mexico where sedimentation rates are fairly rapid. Applying these methods in other areas, e.g. the North Sea, has been more problematic. The prediction methods typically use relations between velocity and effective stress, and are designed for cases where compaction disequilibrium is the cause of overpressuring. In this work we include other pressure-generating mechanisms besides compaction disequilibrium when using seismic data for pressure prediction.

The velocity of seismic waves depends on the porosity and pore pressure and the burial history of the sediments. The relationship between seismic velocity and pore pressure is also related to the particular mechanism that generated the pore pressure. Hence, seismic velocities of the sediments depend both on the compaction state of the sediments and on diagenetic processes. Both compressional and shear velocity change with burial depth. Knowledge of the velocity-depth relations is important when interpreting seismic data. Information provided by basin modeling can help establish the correct velocity-pressure relation to use in the pore pressure analysis (Bowers 1995).

A large part of the sediment fill in sedimentary basins is made up of shales. A proper description of the shaly sequences is therefore important both for the interpretation of seismic data, and for a accurate estimation of the overburden properties.

Velocity- depth trends
Seismic velocity in shale depends both on porosity, mineral content and pore pressure. To be able to estimate pore pressures from velocity data, velocity-depth trends for normally pressured shales must be established. Shales compact both because of external stresses (mechanical compaction), and because of diagenetic processes (chemical compaction). The most import diagenetic process in shales is the smectite illitization.

It has proven problematic to establish general velocity-depth trends (Storvoll et al. 2005). Hence, we have taken a model-based approach in order to make predictions away from well-control. Rock Physics based models are used for calculating velocities at a given stress and diagenesis state (Holt and Fjær 2003). Burial and diagenesis history will be calculated using the basin model Pressim. Log data has been used to validate and calibrate the models.

Integrated pore pressure prediction
Traditionally, pore pressures in subsurface sediments have been predicted using either methods based on analysis of seismic velocity data or methods related to basin modeling. These two approaches have been used independently. Seismic methods typically use relations between velocity and compaction state (effective depth methods) or correlations directly between velocity and pore pressure (e.g. Eatons method). These methods may work if compaction disequilibrium is the only cause of overpressuring. In this case, the compaction state of the sediments corresponds to the effective stress level.

Other pressure generating mechanisms, termed “fluid expansion mechanisms” (Bowers 1995) or “secondary pressure” (Huffman 2001), include overpressuring from diagenetic processes, hydrocarbon maturation, aquathermal expansion, and up-dip transfer of reservoir pressures (lateral fluid flow). Diagenetic processes include both smectite illitization in shales and quartz cementation in sandstones. These processes may alter the effective stress without a significant change in porosity. A simple use of porosity-based methods may therefore give large errors in fluid pressure predictions.

Basin modeling (Pressim) can be used to analyze the loading and unloading history of the sediments, as well as impact of diagenetic reactions. The Pressim results will then help choosing the appropriate compaction and rock physics models to use when interpreting the seismic velocity data.

Results
Data from a North Sea well has been used to illustrate the modeling approach. The reservoir is located at around 4000 m depth. The overburden consists mainly of shales and claystones. The burial history and temperature history are calculated based on the lithostratigraphy for the well. The Rock Physics model is used for calculating the velocity-depth trend for hydrostatic pressure conditions. Both Vp, Vs and the Vp/Vs ratio for a hydrostatic pore pressure are displayed in Figure 1. These depth trends incorporate effects of varying lithology, compaction and clay diagenesis.

For this well only Vp was recorded in the sonic log. The log data is shown in Figure 2 (left panel) together with the predicted normal velocity trend. Deviations between the two curves can be attributed to overpressuring. The velocity model is used to optimize for overpressure in order to obtain a match between observed and predicted velocities. The resulting overpressure is plotted in Figure 2 (right panel). For comparison, the actual mud weights used are included. Predictions show that the overpressure in the interval between 1500 and 2000 m may be greater than expected during drilling. Also significant deviation between predicted and observed pressures is seen around 4000 m around the reservoir. The high overpressure in the reservoir is likely to be caused by a combination of lateral fluid flow and of quartz cementation in the sandstone. These processes give late build-up of overpressure that is not reflected in undercompaction of the shales. Calculations of unloading effects as well as fluid flow (not included in this example) are therefore needed to properly interpret velocity data and predict pore pressure at this depth.

References
Bowers, G. L. (1995). "Pore Pressure Estimation From Velocity Data: Accounting for Overpressure Mechanisms Besides Undercompaction." SPE Drilling & Completion(June 1995): 89-95.

Holt, R. M. and E. Fjær (2003). Wave velocities in shales - a rock physics model. EAGE 65th Conference & Exhibition, Stavanger, Norway.

Huffman, A. R. (2001). The Future of Pore-Pressure Prediction Using Geophysical Methods. Offshore Technology Conference, Houston, Tx.

Storvoll, V., K. Bjørlykke and N. H. Mondol (2005). "Velocity-depth trends in Mesozoic and Cenozoic sediments from the Norwegian Shelf." AAPG Bulletin 89(3 (March)): 359-381.

Figure 1. Normal velocity trend for a selected North Sea well