Reservoir Crude-Oil-Viscosity Estimation From Wireline-NMR Measurements - Rajasthan, India

Abstract

This article, written by Technology Editor Dennis Denney, contains highlights of paper SPE 101689, "Reservoir Crude- Oil-Viscosity Estimation From Wireline- NMR Measurements-Rajasthan, India," by R.J. Zittel, SPE, D. Beliveau, SPE, T. O'Sullivan, SPE, and R. Mohanty, Cairn Energy India Ltd., and J. Miles, Oil Optimizers Ltd., prepared for the 2006 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 24- 27 September. In 2004, the Mangala, Aishwariya, and Bhagyam fields were discovered in Rajasthan, India. Oil viscosity is a main factor controlling performance from these high-permeability paraffinic reservoirs. Pressure/volume/temperature (PVT) data indicate areal and vertical variations in crude-oil properties. Meter-by-meter geochemical core data corroborate vertical variations in oil composition. Continuous wireline measurements of nuclear-magnetic-resonance (NMR) log properties and station NMR properties from wells drilled with water-based mud or synthetic-oil-based mud also were used to calculate a viscosity profile. This study correlated results from all techniques and showed how NMR measurements can provide oil-viscosity profiles in compositionally complex pools. Introduction Oil viscosity was estimated for the Mangala and Aishwariya fields by use of a variety of wireline data and methods. The objective was to investigate and use correlations between wireline and laboratory NMR measurements and in-situ PVT properties (especially oil viscosity) measured on oil samples collected from the fields. Further, correlations were made with geochemical analyses taken from field cores. The result was estimation of in-situ oil viscosity as a function of depth in the wells where appropriate data were available, along with fieldwide correlations of viscosity as a function of height above the oil/water contact (OWC). One desirable property of these viscosity estimates would be that they have the same depth resolution as the wireline-log data from which they were derived. The available data were from routine suites of wireline logs, NMR logs, station NMR measurements, PVT properties measured on reservoir-oil samples, and detailed geochemistry data. Estimating In-Situ Oil Viscosity The empirical relationship represented by µoil=4(T+273.16)/T2 between crude-oil viscosity, µoil (in centipoise), and the logarithmic mean of the measured T2 relaxation-time distribution of the oil, was derived from data for a variety of crude oils and has been demonstrated in industry literature. In this algorithm, T2 is the logarithmic mean of the NMR-T2 distribution of the bulk oil and T is the temperature in degrees Celsius. Laboratory measurements of the T2 distribution of bulk oil from drillstem tests were studied. Data acquired in Mangala, Bhagyam, and Aishwariya fields led to the speculation that crude-oil viscosity might be estimated from the continuous wireline NMR-T2 distribution data by use of an adaptation of the Kleinberg-Vinegar algorithm. A distinct peak in the distributions that occurs near 150 milliseconds apparently is caused by the oil in the reservoir having little or no influence from the water fraction. It was recognized that this peak was unlikely to represent the bulk-oil response, completely uncontaminated by diffusion and other effects, so it would be necessary to calibrate the results of the Kleinberg-Vinegar computation to the in-situ viscosities from PVT analyses on samples of the field crude oils.