Use of Compressional and Shear Wave Velocity for Overpressure Detection

Abstract

Abstract New results are presented that show overpressure can be detected from surface seismic, cross-well, sonic logs and measurements ahead of the drill bit. By analyzing experimental data we show that in many room-dry rocks, the Poisson's ratio (PR) decreases with decreasing differential pressure (confining minus pore pressure). This means that in gas-saturated rocks, PR decreases with increasing pore pressure. We confirm the generality of the observed effect by theoretically reproducing it via effective medium modeling. These results may prove useful in the extraction of "overpressure atributes" from seismic data. Introduction and Problem Formulation Early and reliable detection of overpressure compartments is an important (perhaps the single most important) aspect required to improve deep drilling safety. Directly related to the overpressure problem is the shallow water flow (SWF) issue. Many deep-water wells in the Gulf of Mexico have encountered pressurized granular layers in the uppermost sediment section. The pressure was high enough to cause powerful flows of water and sand in the well bore, resulting in the loss of the well. Because the origin of these flow problems is generally within near-seafloor (shallow sediment section) layers, this pressurized-layer phenomenon has been labeled Shallow Water Flow. The challenge is to establish a reliable experimental and theoretical basis for detecting and predicting overpressure and SWF potential from seismic, cross-well, well logs, and ahead of the drill bit Typically, elastic-wave velocity in dry rock is measured in the laboratory by varying confining pressure while maintaining constant pore pressure. Because velocity reacts to the differential (confining minus pore) pressure (e.g., Wyllie et al., 1958), such data can be used to predict in-situ velocity variations in rock with gas due to pore pressure changes at constant overburden (Figure 1). Velocity at in-situ saturation conditions can be calculated from the dry-rock velocity using fluid substitution equations (e.g., Gassmann, 1951). Due to the large compressibility of gas, the in-situ velocity in rock with gas is very close to that in rock with air in the laboratory at the same differential pressure. The time scale of laboratory experiments is much smaller than the geologic time scale of overpressure development. Still, the laboratory experiments where pressure changes rapidly can be used to model the important transient (late-stage) overpressure mechanisms that are invoked when the pressure of the fluid in the rock mass is allowed to increase relative to hydrostatic through (a) aquathermal fluid expansion; (b) hydrocarbon source maturation and fluid expulsion; (c) clay diagenesis; (d) fluid pumping from deeper pressured intervals; and (e) decrease in overburden due to tectonic activity (Huffman, 1998). The decrease in the P-wave velocity with increasing pore pressure has been used for overpressure detection (e.g., Grauls et al., 1995; Moos and Zwart, 1998). However, velocity does not uniquely indicate pore pressure because it also depends, among other factors, on porosity, mineralogy, and texture of rock.