Reservoir Characterization & Geological Modeling Study Notes

LECTURE OUTLINE

General Structure

  • Introduction to geological reservoir modeling workflows

  • Data for modeling

  • Geodata spatial variability:

    • Interpolation

    • Simulation

  • Gridding & Upscaling

  • Modeling workflow

WHY RESERVOIR MODELING?

Importance of Modeling

  • Purpose: to understand and predict a reservoir's key geological, geophysical, and engineering components.

  • Key applications include:

    • Identification of (new) prospects.

    • Estimation of Original Oil/Gas In Place (OOIP).

    • Quantification of reservoir potential.

    • Economic potential evaluation.

    • Development and field management.

WHY CELLULAR MODELS?

Advantages of Cellular Models

  • Develop a global understanding of integrated reservoir studies.

  • Integrate geological and dynamic constraints.

  • Understand the impact of reservoir heterogeneities on dynamic behavior.

  • Evaluate the relative impact of uncertainties in reservoir modelling.

QUESTIONS FOR MODEL BUILDING

Aspects to Consider

  1. Shape and Volume:

    • External envelope: iso contours, top-base.

  2. Structural Framework:

    • Faults, fractured areas, micro fractures.

  3. Internal Organization:

    • Correlations, layering, facies variation, petrophysics (drains, barriers).

  4. Fluids in Place:

    • Type of fluids, contacts, PVT (pressure, volume, temperature) composition, aquifer extension.

USES OF RESERVOIR GEOLOGICAL MODEL

Applications

  • Input for reservoir simulations.

  • Volumetric studies.

  • Quantifying uncertainties.

  • Well planning and geosteering.

  • Rock mechanics modeling.

DATA FOR MODELING

Input Data Types

  • Static Data:

    • Seismic structural data (horizons and faults).

    • Seismic characterization (attributes).

    • Well data (formation boundaries, logs, core data).

  • Dynamic Data:

  • Output:

    • Detailed geological model (lithofacies and petrophysics).

    • Coarse (upscaled) reservoir model (for dynamic simulation only).

SEISMIC DATA

Characteristics

  • Horizons:

    • Derived from seismic picking and time-to-depth conversion.

    • Good lateral definition but poor accuracy (12.5 to 25m in XY, ~3000m/s * 4ms = 12m in depth).

  • Faults:

    • Similarly derived, using seismic attribute maps (e.g., Net-to-Gross, average porosity).

    • Generate maps from seismic attributes (e.g., acoustic impedance) and calibrate them on well data.

Seismic Acquisition Methods

  • 2D Acquisition:

    • Overview of an area with low cost and sparse data coverage; interpretation can be difficult.

  • 3D Surveys:

    • Conducted in areas of special interest with data spacing of 12.5m or 25m; more expensive for detailed analysis.

WELL DATA

Well Path and Logs Analysis

  • Convert well path in MD azimuth to X, Y, Z coordinates; accurate only for vertical wells or those measured via GPS.

  • Logs utilized include:

    • PHI: derived from NPHI/RHOB or SONIC logs + cores.

    • Perm: deduced from PHI logs and cores.

    • Saturation: assessed from resistivity logs.

    • Facies classification: sedimentary/rock types recognized through logs.

DYNAMIC DATA

Types of Dynamic Data

  • Well Test:

    • Evaluate average permeability of drained areas.

    • Estimate distance to flow barriers.

  • Static Pressure Measurements:

    • Identify connected blocks within the reservoir.

  • Production History:

    • Estimate OOIP.

    • Evaluate capillary pressure curves for each facies.

GEOLOGICAL GRID BUILDING

Structural Modeling Components

  • Fault modeling (fault planes and intersections).

  • Horizon modeling.

  • Stratigraphic model related to sequence stratigraphy and correlations.

  • Construction of the 3D stratigraphic grid.

  • Lithological model via stochastic methods and fine-scale petrophysical model.

Summary of Components for "Container" Determination

  • Integrate various data including seismic interpretation, well log data, and geological structure to build effective models.

LITHOLOGICAL MODEL

Definition and Development

  • Not a mandatory step but serves as a powerful tool to guide petrophysical distribution.

  • Built by integrating:

    • Conceptual representation (sedimentological model).

    • Classification phase (facies definition).

    • Probabilistic approach to lithological distribution (stochastic model).

Key Elements of Lithological Model

  • Electrofacies Analysis characterized by:

    • Representation of log data.

    • Clustering of data to form electrofacies.

  • Two approaches:

    • Supervised: training samples guide clustering definition.

    • Non-supervised: a posteriori interpretation of electrofacies.

Gridding Processes

  • Involves associating facies with grid cells based on dominant characteristics and fine-scale analysis.

  • Grouping of well data for upscaling and effective reservoir characterization using categorical data.

UPSCALING AND GRIDDING

Overview

  • Upscaling refers to the coarsening of data to optimize computational efficiency while highlighting significant reservoir characteristics.

  • Different methods are applied depending on property types, including:

    • Arithmetic averages for N/G and porosity.

    • Weighted averages considering effective fluid flow in heterogeneous reservoirs.

Methods of Upscaling

  • Defined clearly for various properties such as permeability, water saturation, and lithotypes using qualitative and quantitative methods.

SIMULATION

Importance and Objectives

  • Simulation is key for representing real reservoir conditions under varied scenarios to manage uncertainties effectively.

  • Types of simulations include:

    • Deterministic model: clear predictions with no uncertainties.

    • Stochastic model: incorporates variability and uncertainty.

    • Static and dynamic models: differing from traditional interpolation methods by accurately reproducing the reservoir variability.

Simulation Techniques Overview

  • Uses probabilistic models and geostatistical tools for integrating various data points critically for producing risk assessments and management.

  • Kriging and Cokriging are both emphasized for handling spatial relationships and variabilities among data points.

RESERVOIR MODELING WORKFLOW

General Process

  • Integrating geological structures and data into a coherent model for practical application in decision making related to factors such as production forecasting and reservoir management.

  • Outputs include fluid flow models, volumetrics, and understanding spatial distributions to facilitate effective extraction and production strategies.

CONCLUSIONS

Key Takeaways

  • The processes of reservoir characterization, grid building, and modeling are interrelated and critical to evaluating potential reservoir performance. It incorporates vast amounts of data and various analytical and computational techniques, synthesizing them into actionable insights.