LWD NMR for hydrocarbon typing and formation evaluation in a challenging offshore trajectory

Abstract

Copyright 2015, Society of Petroleum Engineers. The well within the context of this case study consists of reservoirs are sequences of permeable sands interbedded with variable proportions of silt and clay. Gas is the target hydrocarbon type, but light oil / condensate can be present unexpectedly. In these depleted reservoirs, hydrocarbons typing are complicated by their reduced volumes and corresponding diminished effect on conventional logs. Wells are highly deviated and targets don't align in the same direction leading to high trajectories tortuosity. This prevents to plan extensive wireline logging program. Formation evaluation is mainly based on LWD logs. For such challenging condition, fluids identification is traditionally made possible by stationary Nuclear Magnetic Resonance (NMR) from wireline conveyed logging devices, adopting Diffusion - Relaxation maps technique. Through Diffusion - Relaxation maps technique, the contrast on both diffu-sivity and relaxation time (longitudinal relaxation time T1 or transversal relaxation time T2) allow differentiation of gas, oil and water. Even though high gas diffusivity creates contrast on transversal relaxation time T2 to differ gas from the other fluids, the approach based on T2 domain only is long time neglected. This is because under wireline condition, formation gas is often flushed by mud filtrate and the formation oil can be mixed up with OBM signal. This study proves that LWD NMR, due to its logging while drilling features, enables the simple T2 spectra based method to differ formation fluids in an efficient way. Attentive BHA design and job planning ensure good data quality and reasonably fast logging speed. Because of short time after bit (TAB) while drilling, it probes directly the formation fluids without being affected by mud invasion. For the studied reservoirs, the measured T2 value for water, gas and light oil are well distinguished being approximately of 200 milliseconds, 450 milliseconds and 2000 milliseconds respectively. The intervals with the presence of light oil are revealed directly from T2 spectra and the gas-oil-contacts (GOC) are accurately determined by T2 distribution. The same result is hard to be achieved by triple-combo measurements only. A newly introduced statistical technique "factor analysis" is used to determine poro - fluid distributions and associated porosities. It automatically searches for the dominant T2 modes through T2 depth log and identify repeated T2 distribution patterns to provide a continuous fluid facies analysis. Density Magnetic Resonance Porosity (DMRP) method is used to estimate the total porosity and gas saturation. It provides a resistivity independent method to address the gas saturation. Considering the fresh formation water, the uncertainty on the petro - physical parameters is significantly reduced. This paper divulges the value of T2 based fluid typing method with LWD NMR tool. It provides a simple but efficient way to identify gas from light oil. The fluid information offered is essential for field completion decision making.