Formation Evaluation of Magnetic Resonance Logging-While-Drilling Data Recorded at a High Rate of Penetration

Abstract Magnetic resonance (MR) logging-while-drilling (LWD) data have numerous petrophysical applications for reservoir characterization. Under certain conditions, a high rate of penetration (ROP) can affect the data quality of MR LWD T2 distributions, compromising some of the estimated petrophysical properties. Logging MR data at a low ROP was so far the recommended approach to ensure high MR data quality. Moreover, when high ROP could not be avoided, ROP effects were a reason to consider the MR data to be of very limited or no value. Recently, an ROP correction for MR logging data was introduced. The correction on porosity is particularly significant for large volumes of light hydrocarbon (HC) with a large T1 value. For best results, the MR logging data should comprise an extended data set, e.g., including echo trains with full and partial polarization of the light HC that enables the calculation of an in-situ T1 of the light HC. Extended data are, however, often not available in older data sets—a condition not discussed in detail before. This paper presents a case study based on MR data recorded offshore Nigeria many years ago. The well was drilled with an ROP exceeding 40 m/h, which was twice as fast as the typically recommended maximum ROP at the time. The logged reservoir has a porosity of 30 to 35 p.u., and it contains light oil and has a high oil saturation. The ROP polarization effect caused a severe porosity overestimation. At that time, no ROP correction was available, and the MR data were considered not suitable for petrophysical interpretation because of the porosity overestimation. The recently introduced ROP correction was successfully applied to the log data using a constant T1/T2 ratio value for estimating T1 of the light HC from the acquired T2 distribution. The correction results show that the MR total porosity was severely overestimated without correction. However, the porosity overestimation was only related to an increase of the light oil porosity. The volumetrics and the T2 distribution of the irreducible fluid fraction remain unaltered by the correction. Consequently, valuable information can still be extracted from the uncorrected T2 distribution. First, the uncorrected MR data enables a simple computation of saturation, which is in good agreement with the resistivity-based saturation. Second, the amount and position of the bound-water peak correlates to variations in rock type and quality. Third, the position of the light-oil T2 peak indicates variations in mineralogy, because it correlates well with variations in the gamma ray (GR) log curve.