BROWN, PAUL, Department of Geological Sciences, University of Durham, UK and Fossil Fuels and Environmental Geochemistry (Postgraduate Institute), NRG, University of Newcastle, UK, RICHARD E. SWARBRICK Department of Geological Sciences, University of Durham, UK, and ANDREW C.APLIN, Fossil Fuels and Environmental Geochemistry (Postgraduate Institute), NRG, University of Newcastle, UK

Abstract: Estimation of Pore Pressure in Shales: How Useful are Shale Compaction Curves?

Direct measurement of pore fluid pressure by the Modular Dynamics Test or Repeat Formation Test tools in shales is impossible due to their low permeability. The use of shale compaction curves is thus the basis of several methods of pore fluid pressure estimation; pressure from seismic, from wireline and in basin modelling. All these methods require the definition of a normal compaction curve (NCC), or set of normal compaction curves for the shales. These curves are typically empirical, being based on regional experience or using calibration from soil mechanics experiments, but some are based on work in rock mechanics. If the sediments are young (less than 63 Myr old) and cool (less than 70 degrees C) it is assumed that any deviation of a porosity profile away from a NCC is due to the onset of abnormal pressure which can be readily estimated using an equivalent depth method. This is outlined in Figure 1. In many cases a local NCC is defined using the porosity profile from the upper section of the well which is assumed to be normally pressured. This assumption is not necessarily correct as the onset of abnormal pressure may occur at shallower depths than the wireline coverage of a well. This occurs commonly in deepwater. In this case, other methods are required to define the NCC, such as use of regional or generic compaction curves.

Shales are exceedingly variable in all of their properties.This variability further complicates the definition of shale NCC's as shale compaction characteristics vary considerably. Figure 2 shows the large range of compaction curves recently published. Of these, only Baldwin and Butler's (1985) curve is not originally based in soil mechanics; however, all are empirically derived.

Aplin et al (1995) showed that the compaction rate of many shales is dependent upon the fraction of clay-grade particles, those less than 2 micrometers in diameter, in the sediment. They used samples taken from the North Sea to calibrate their soil mechanics based model. This model has been studied, and a full error analysis involving a set of Monte Carlo simulations has been performed on a set of synthetic data fed into the models. The input to the model comprised porosity derived from a density log and clay fraction, assuming normal pressure. Figure 3 shows the resultant error distribution in the estimate of pore pressure for a single lithology. The error distributions of the model's input are what would be expected if tools were working well.The distribution of the output is skewed towards lower pressures and the average is less than the required value of pressure. This means that most estimates of pressure using this model are slightly more likely to be underestimates. If the input distributions were broader, the corresponding output distribution would also be broader and the average value is even smaller than actual value of pressure.

Normal compaction curves are a major tool for use in the estimation of shale pore fluid pressure.These curves must be used with caution and a knowledge of the magnitude of error in measurements, and estimates of input parameters to the models used are essential. The model investigated shows that a slight underestimate of pressure is likely to occur ranging, at.an equivalent burial depth of 1.5 km (4900 ft), from 0.5 MPa (73 psi) in good conditions with high quality data to over 2 MPa (290 psi) in poorer data quality conditions. This error analysis has significance in assessing estimates of pressure from basin modelling, as wireline estimates of porosity are commonly used to define NCC's in these models. In several packages, these are broadly based on the soil mechanics compaction curves and so a direct error analysis can be produced using these techniques.

References:

Aplin, A. C.,Y.Yang, and S. Hansen, 1995,Assessment of B, the compression coefficient of mudstones and its relationship with detailed lithology. Marine and Petroleum Geology, v. 12, p. 955-963
Baldwin, B. and C. 0. Butler, 1985, Compaction curves:AAPG Bulletin, v. 69, p. 622-626
Burland, J. B., 1990, On the compressibility and shear strength of natural clays: Geotechnique, v.40, p. 329-378
Dickinson, G., 1953, Geological aspect of abnormal reservoir pressures in Gulf Coast Louisiana: AAPG Bulletin, v. 37, p. 410-432
Mann, D. M. and A.S. Mackenzie, 1990, Prediction of pore fluid pressures in sedimentary basins: Marine and Petroleum Geology, v.7, p. 55-65

AAPG Search and Discovery Article #90923@1999 International Conference and Exhibition, Birmingham, England