Datapages, Inc.Print this page

Determination and Statistical Estimation of Heavy Oil and Bitumen Viscosity

Adams, Jennifer J.1; Larter, Steve 1; Jiang, Chunqing 1; Bennett, Barry 1; Huang, Haiping 1; Oldenburg, Thomas 1; Noke, Kim 1; Snowdon, Lloyd 1; Gates, Ian 1
1 Gushor Inc., Calgary, AB, Canada.

Accurate mapping of in-reservoir oil mobility is crucial for effective exploration, design/optimization of production strategies especially in heavy oil and bitumen reservoirs which exhibit orders of magnitude fluid property variation laterally and vertically. Such maps require accurate estimation of subsurface reservoir and fluid properties; however, traditional bitumen exploration and production rely on maps of reservoir properties and fluid saturations, rather than the commonly heterogeneous viscosities in heavy oil fields. Also controlled inter-laboratory comparisons have revealed significant variability in viscosity determinations from stored core material.

Our lab experiments show that measured bitumen viscosity is a function of not only oil source and in reservoir alteration, but also storage conditions and time from core collection to oil analysis, and sample contamination. Loss of light end fractions of oil during sample storage, handling, and extraction most significantly affect measured heavy oil viscosity. Oil extraction from core by compaction preserves light ends in the oil and minimizes contamination of oil samples with water or solids better than centrifuge extracted oils, which show evaporation of hydrocarbons and higher viscosities. Small quantities of water and clay in extracted oils can increase viscosity significantly, adding to data variability.

Systematic variations in light end hydrocarbon (e.g. alkylmethylcyclohexanes, alkyltoluenes) loss during frozen core storage causes up to order of magnitude increases in viscosity over 1 year, but relative vertical viscosity trends are maintained. To minimize viscosity errors related to core storage, a storage time diffusive/evaporative viscosity correction algorithm (STVC) was developed using binary mixing of evaporative light and heavy oil end members to estimate viscosity at a constant epoch to normalize datasets in time. These storage time viscosity correction methods enable more effective field wide viscosity data comparison for development decision making. With time-corrected viscosity data linked to detailed total hydrocarbon concentrations, multivariate statistical models predicted dead oil viscosity across biodegraded oil fields where oil composition but not unaltered viscosity was determined. Sampling and lab protocols need revision to provide representative viscosities of reservoir fluids for recovery strategy design and management over the life time of a producing field.


AAPG Search and Discovery Article #90090©2009 AAPG Annual Convention and Exhibition, Denver, Colorado, June 7-10, 2009