Probabilistic Prediction Method of Formation Pore Pressure

Abstract

ABSTRACT Formation pore pressure plays a critical role in petroleum drilling and is an indispensable basic parameter for casing program design and mud weight optimization in petroleum drilling. Formation pore pressure can be predicted using seismic, logging and mud log data. However, due to the concealment and complexity of deep formations and the inherent errors in seismic, logging and mud log data, it is always difficult to accurately predict formation pore pressure. In order to quantitatively determine the uncertainty of the formation pore pressure, a probabilistic pore pressure prediction method has been proposed. First, the basic effective stress and dc-exponent methods for pore pressure prediction were introduced. Second, the data processing procedure with sliding window treatment was presented, the statistical characteristics of Eaton index and normal compaction trend (NCT) were investigated, and the quantitative prediction of uncertain formation pore pressure was performed using the Monte Carlo simulation. Finally, the case study of the X-1 well in the Sichuan Basin was conducted, the uncertain pore pressure profile was calculated and compared with measured data, effective stress method and dc-exponent method. The results showed that the pore pressure increases with the increase of Eaton index, while it decreases with the increase of K value, so the influences of uncertain Eaton index and NCT cannot be ignored. The case study showed that all measured pore pressure data are within the pore pressure profile at the 5% to 95% confidence interval, but the conventional prediction methods cannot accurately predict the pore pressure in all cases. The present method integrates both logging and mud log data, the confidence interval pore pressure profile is more consistent with the actual situation, and it can provide significant support for casing program design and mud weight optimization in petroleum drilling.