Full text loading...
-
Seismic Attributes and Advanced Computer Algorithm Method to Predict Formation Pore Pressure: Paleozoic Sediments of Northwest Saudi Arabia
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, IPTC 2013: International Petroleum Technology Conference, Mar 2013, cp-350-00190
Abstract
Oil and gas exploration professionals have long recognized the importance of predicting pore pressure before drilling wells. Pre-drill pore pressure estimation not only helps with drilling wells safely but also aids in the determination of formation fluids migration and seal integrity. With respect to hydrocarbon reservoirs, the appropriate drilling mud weight is directly related to the estimated pore pressure in the formation. If the mud weight is lower than the formation pressure, a blowout may occur and, conversely, if it is higher than the formation pressure, the formation may suffer irreparable damage due to the invasion of drilling fluids into the formation. A simple definition of pore pressure is the pressure of the pore space fluids in excess of the hydrostatic pressure. The cause of abnormal pore pressure includes quick burial of unconsolidated sediments before dewatering, hydrocarbon formation in the pore space of rocks, mineralization such as smectite-illite osmosis, tectonic movement of sediments and other geological processes that result in overpressure. Geoscientists and engineers have used various techniques to predict abnormally high formation pore pressure from seismic data. These techniques primarily employ the empirical relationship between seismic velocities and formation density. Consequently, many investigators have shown the sensitivity of seismic velocities to other factors such as lithology, gas and pore fluid content, which reduces the reliability of purely velocity-based methods in predicting abnormally high formation pressure (Young and Lepley, 2005). In this study we will focus on using an advanced pattern recognition computer algorithm, called Support Vector Machine (SVM) and attribute data from seismic and wells to the prediction of formation pore pressure in the lower Paleozoic Qalibah formation of northwest Saudi Arabia.