Geostatistics is the mathematical engine of spatial data analysis and geomodeling. Dominant uses of geostatistics in the industry are multi-variate data analysis, mapping, integrating diverse variables, building geomodels, resource evaluation and decisions. Solid training in Geostatistics is an essential qualification of a proficient Geomodeler, data scientist and problem solver within subsurface teams. Deterministic and stochastic methods are combined. Uncertainty is a fundamental topic because many underlying applications are stochastic and data input are a sparse or imprecise sampling of reservoirs. Geostatistical basic theory and best practices are explained along with a variety of practical tips, tools of the trade. Uses of probabilistic results are discussed. Context for the subsurface team is given in context with workflow element descriptions and case studies. The course provides grounding in theory and thought process.
Afternoon exercises are designed to reinforce the theory and lecture through hands-on learning. The advanced exercises are scripted to allow flexibility to experience the impact of key parameter choices on model outcomes without getting bogged down in the software during a short course. The Isatis Geostatistics toolkit is used for exercises since it is a flexible software for learning where basic techniques are transferable to geomodeling packages.
Module 1 (computer exercises)
Introduction: Geomodeling and the Subsurface Team - What is Geostatistics?
Essential statistics and terminology
Purpose: Background for exploratory data analysis, preparing for mapping and model building
· Regionalized Variables: Data Types, definitions
· Univariate Statistics: Measures of position, spread, and shape; vertical facies proportions, stationarity, proportional effect
· Box plots, Q-Q plots, Normal score transform
· Bivariate Statistics: Covariance and correlation; principal components
· Quantifying Variability/Spatial Continuity: Variograms- experimental, anisotropy; hand calculations; variogram maps; Behaviour, impact of outliers and calculation tips
· Variogram Models: illustrations; nested, issues, fitting tips and tricks
Module 2 (computer exercises)
· General estimation techniques
· Kriging: simple and ordinary; Kriging by hand with a variogram model; Kriging weights, Cross validation, stationarity
· Multi-variate: Co-Kriging; collocated co-Kriging; Kriging with External Drift (universal Kriging)
· Trends in data: handling non-stationarity
· Case examples with mapping
· Geostatistical Depth conversion and QC, brief introduction, e.g., for model framework
· Multi-variate Special: Principal Components analysis; seismic attributes; statistical plays
Module 3 (computer exercises)
· Simulation versus Estimation concepts
· Conditional Simulation; random walk and search neighbourhood
· Sequential Gaussian Simulation processes
· Petrophysical Trends and secondary data
· Uses of models: probability mapping and uncertainty; risked volumetrics; avoiding bias in estimates
· Post-processing special topic: Checking results, direct forecasting without simulation
· Case History
Module 4 (computer exercises)
· Stochastic Methods summary
· Stratigraphic coordinate systems
· Deterministic facies trend modeling as a model constraint
· Object methods-summary
· Pixel methods: Illustrated description of algorithms for Truncated Gaussian (TGS), Truncated Pluri-Gaussian (PGS), Sequential Indicator (SIS), Multiple Point (MPS)
· Facies methods characteristics and choices
Module 5 & 6 (Lecture)
Generalized Subsurface Workflows and Workflow Elements
· Case Histories
· Compiling and checking the input databases, data types, model planning
· Defining the structure and stratigraphic framework; faults, grids and model sizing
· Facies inputs: Diverse Sources; Visual versus Electrofacies classification (machine learning); Issues, scale, rules for preparation of facies logs for modeling improvements
· Facies trend modeling: proper techniques and choices for building 1D, 2D and 3D facies proportions; integration of seismic attributes
· Topics on Petrophysical modeling for porous media and fluids: Porosity, water saturation methods, permeability, mechanical; scale and specific rules, oil in place
· Re-scaling for the simulator techniques, specific parameter choices and critical issues
· Post-processing: net pay, connectivity, summarizing uncertainty
· Linking static to dynamic behavior through direct forecasting (proxies and type curves) in resource plays, delineation, and developments
To provide grounding in subsurface data analysis, geomodeling and geostatistical thought process. Improved understanding of best practices, tools of the trade and important workflows. An introductory overview of necessary basic geostatistical theory for data analysis and subject knowledge to improve team communication. Improved understanding of the uses and limitations of geostatistics and geomodeling.
Who should attend:
Technical and decision makers working on subsurface hydrocarbon reservoirs in multi-disciplinary teams using or considering using geomodeling. This includes geomodelers, technicians, geologists, geophysicists, petrophysicists, reservoir engineers, technical managers, and new hires.
Windows based laptop. The Isatis (and Isatis.new) Geostatistics Toolkit Software from Geovariances will be provided for the course exercises and will be installed on attendees’ laptops at the start of the course.
Several on main topics with an integrated probabilistic volumetric study. Exercises are intended to reinforce concepts and practical application.
None, but general experience with integrated subsurface teams planning to use or using geomodeling would be helpful. Openness to seeing basic mathematical theory and new concepts.