Abstract

The detailed knowledge of fluid composition is a key ingredient for successful management of oilfield reservoirs. The oil composition determines the PVT behavior and flow properties such as viscosity and can be used as a fingerprint for other reservoir properties, such as compartmentalization. The presence of large molecules can lead to asphaltene drop-out or waxing problems. It is highly desirable to measure fluid properties under downhole conditions since many properties depend critically on temperature and pressure. It is also well understood that samples can undergo irreversible changes as they are extracted from the formation and transferred to the laboratory.

The measurement of diffusion - relaxation distribution functions (D-T2) has proven to be a valuable tool for identifying and quantifying different fluids in the formation. Here we show that D-T2 measurements also contain detailed information about the composition of hydrocarbon fluids. We present measurements on a suite of stock tank oils to demonstrate that D-T2 distributions provide a fingerprint of the fluids that correlates with the chemical composition of the oils, i.e. the SARA analysis. For oils that have a high fraction of saturates or aromatics, the contributions in the D-T2 maps lie close to the standard oil line, but show systematically different slopes that depend on the chemical composition. Asphaltenes act as a relaxation contrast agent and shift all oil contributions to shorter relaxation times. In contrast, oils below the wax appearance temperature show restricted diffusion effects.

For oils without asphaltene or waxing effects, the diffusion or relaxation time distribution functions can be used in addition to extract quantitatively the chain length distributions of the oil molecules. These chain length distributions are in good agreement with gas chromatography measurements. We also note that the NMR based measurements extend to higher chain lengths than can be conventionally achieved with gas chromatography.

For samples with significant methane content, diffusion measurements are a useful probe for identifying the gas. Gas dissolved in the oil shows a characteristic diffusion coefficient that is significantly reduced from that of free gas by a degree that depends on the viscosity of the oil.

INTRODUCTION

Crude oils are complicated mixtures of fluids. For the understanding and prediction of fluid properties including viscosity and the onset temperatures for transitions such as gelling, waxing or asphaltene dropout, it is essential to have an accurate knowledge of the distribution of molecular sizes and the chemical composition. Standard laboratory techniques that give information on the distribution of molecular sizes include gas chromatography, mass spectroscopy, and NMR relaxation and diffusion measurements. Information on chemical composition can be obtained by standard chemical analysis and various spectroscopies, including optical, infrared and NMR spectroscopy.

NMR relaxation and diffusion measurements of crude oils have long been used to infer viscosity information of crude oils (Kleinberg 1996, Morriss 1997, Hirasaki 2003, Straley 2006). Here we take advantage of newly developed 2d-NMR techniques that allow the extraction of D-T2 and T1-T2 distribution functions (Hürlimann 2002b, Song 2002).

This content is only available via PDF.
You can access this article if you purchase or spend a download.