Viscosity Gradient Prediction in Heavy Oil Reservoirs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional reservoir simulators fail to accurately simulate heavy oil production processes due to inadequate understanding of asphaltene nanocolloidal structures and inability to handle viscosity gradients, which are critical for enhanced oil recovery in heavy oil reservoirs.
Innovation Solution
A downhole tool acquires reservoir fluid samples, analyzes asphaltene concentration, and uses a viscosity model based on corresponding state principles to predict viscosity gradients, allowing for accurate characterization of fluid properties and reservoir simulation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional reservoir simulators are used to simulate heavy oil production, then the simulation process can be performed with standard tools, but the accuracy of viscosity gradient prediction and asphaltene behavior characterization is insufficient
Solution Approach 1:
The simulation system is segmented into multiple functional modules: a downhole sampling tool for acquiring fluid samples at different depths, a laboratory analysis system for measuring asphaltene concentration and fluid properties, and a computational model that integrates the data. This segmentation allows each module to specialize in specific tasks, improving overall prediction accuracy while maintaining manageable complexity through modular design.
Solution Approach 2:
A corresponding states viscosity model serves as an intermediary between the measured asphaltene concentration data and the reservoir simulation. This model translates laboratory measurements into viscosity predictions that account for depth-dependent asphaltene behavior, enabling accurate viscosity gradient prediction without requiring direct measurement at every depth point.
2Measurement precision
If downhole fluid sampling and laboratory analysis are performed to characterize asphaltene concentration, then accurate fluid properties can be obtained, but the time and resources required increase
Solution Approach 1:
Fluid samples are collected downhole at multiple depth intervals before bringing them to the surface for analysis. This preliminary sampling action allows laboratory analysis to be performed on pre-collected samples rather than requiring immediate analysis, reducing the critical path time while maintaining measurement accuracy.
Solution Approach 2:
Instead of analyzing large volumes of reservoir fluid directly, the system uses small representative samples that copy the essential compositional characteristics of the bulk fluid. This allows accurate asphaltene concentration measurement with minimal fluid handling and reduced analysis time while preserving the representativeness of the data.
3Measurement precision
If a viscosity model based on corresponding state principles is used to predict viscosity gradients, then accurate viscosity prediction as a function of location can be achieved, but the computational complexity increases
Solution Approach 1:
The corresponding states viscosity model uses parameter changes based on measured asphaltene concentration to predict viscosity variations with depth. By transforming the complex asphaltene behavior into parameter adjustments within an established viscosity framework, the model achieves accurate predictions while leveraging existing theoretical foundations to manage computational complexity.
Solution Approach 2:
The model focuses computational effort on the most critical parameter - asphaltene concentration - and its direct impact on viscosity. Rather than attempting to model all possible fluid properties and interactions, the system applies partial action by concentrating resources on measuring and modeling asphaltene behavior, which dominates viscosity variations in heavy oil reservoirs.
Data Source
AI summary
A methodology that performs fluid sampling within a wellbore traversing a reservoir and fluid analysis on the fluid sample(s) to determine properties (including asphaltene concentration) of the fluid sample(s). At least one model is used to predict asphaltene concentration as a function of location in the reservoir. The predicted asphaltene concentrations are compared with corresponding concentrations measured by the fluid analysis to identify if the asphaltene of the fluid sample(s) corresponds to a particular asphaltene type (e.g., asphaltene clusters common in heavy oil). If so, a viscosity model is used to derive viscosity of the reservoir fluids as a function of location in the reservoir. The viscosity model allows for gradients in the viscosity of the reservoir fluids as a function of depth. The results of the viscosity model (and/or parts thereof) can be used in reservoir understanding workflows and in reservoir simulation.