Integration of Rock Physics Modelling to Improve Pore Pressure Prediction in Unconventional Shales
Jakob Heller*1 and Venkatesh Anantharamu 1, 1. Ikon Science
Abstract
Accurate formation pressure (pore pressure) prediction is essential for executing a safe and cost-effective drilling program. Unexpected overpressure is a major cause of drilling hazards, costing the industry billions of dollars, and posing a huge potential risk for damage to the environment. Overpressure also has a significant impact on the ability to artificially fracture shale formations as well as increasing the production drive of liquid hydrocarbons and favor higher production rates.
The standard approach to predict pore pressure is to use petrophysical logs or seismic velocities and assume that shales are cool, homogeneous and that any overpressure is due to under compaction. It is also generally assumed that the stress field is extensional such that the rocks currently are at the maximum burial depth/effective stress.
It is often the assumption that standard pore pressure prediction techniques cannot be applied to onshore, unconventional plays due to their often complex burial/uplift and erosion history combined with extensive diagenesis and hydrocarbon generation. Variability in total organic content (TOC), Vclay, mineralogy, cracks, natural fracture patterns, and permeability can all influence the evolution of pore pressure and our ability to accurately estimate pore pressure. For instance, changes in TOC have a similar response in elastic properties to changes in porosity and may give a “false” impression of the presence of overpressure and/or magnitude of overpressure.
In this study, pore pressures in a well (Well A) from the Midland Basin, West Texas have been estimated in a regional context. First, any potential overpressure mechanisms were assessed through Vp-Rho cross plot analysis. Then pore pressure magnitudes in the study area were assessed and estimated using a regional derived DFIT pressure trend validated against reported closure pressure from wells in the study area as well as from publicly available pressure data. Lastly, shale pore pressures were predicted in well A from raw (Vp) log data as well as from Vclay and TOC corrected (Vp) log data using both a conventional pore pressure prediction technique (equivalent depth method) and a rock physics constrained model (Vernik, 2016).
This initial study highlights that 1) shale pressure predictions from sonic/Vp log data using conventional pore pressure prediction techniques and a rock physics constrained PPP model using an inelastic compaction modulus (Cm) of 21 MPa (comparable with overpressure mainly from loading) gives similar results and matches reservoir data points from previous studies in the Wolfcamp Formation and 2) there is a need for a separation of lithology and pore pressure effects on log response to avoid misinterpretation of pore pressure/overpressure. Hence the integration of rock physics workflows in pore pressure prediction is important.