Uncertainty Analysis of Visible and Near-Infrared Data of Hydrocarbons

Abstract

Measurement of physical and chemical properties of hydrocarbons plays an important role in the exploration and production of oil wells. In situ measurement of chemical properties of hydrocarbons makes use of visible and near-infrared (vis-NIR) absorption spectra of hydrocarbons. Uncertainty analysis of these fluid properties is central to developing a fundamental understanding of the distribution of hydrocarbons in the reservoir. In this manuscript, we describe an algorithm called the fluid comparison algorithm (FCA), which provides a statistical framework to quantify and compare hydrocarbon fluid properties and associated uncertainties derived from vis-NIR measurements. The inputs to FCA are the magnitude and uncertainty of vis-NIR spectroscopy data of two hydrocarbons. The output of FCA is a probability that two fluids are statistically different. FCA lays the foundations for subsequent optimization and capture of representative reservoir hydrocarbons. Furthermore, in some circumstances, it can also enable real-time decisions to identify reservoir compartmentalization and hydrocarbon composition gradients in natural oil reservoirs.