Digital Reservoir Characterization Technology (DIRECT)

The DIRECT Industrial Affiliate Project aims to develop novel technologies, practical workflows, demonstrations and documentation to enable subsurface data analytics and machine learning.

Our goals are to combine best-practice and cutting-edge technology in  reservoir spatiotemporal characterization and modeling, real-time drilling control, production data integration and forecasting, reservoir petrophysical measures and geophysics with emerging technology in big data analytics and machine learning to optimize well trajectory and resource recovery.

Faculty Supervisors

Michael Pyrcz

John T. Foster

Carlos Torres-Verdin

Eric Van Oort

Researchers and Projects

wendi liu photo

Wendi Liu (2nd Year PhD)
Data analytics, geostatistics and machine learning for unconventional reservoirs including feature engineering, anomality detection and spatiotemporal geostatistical and machine learning accounting for sampling bias.
Supervisor: Dr. Pyrcz

jose salazar photo

Jose Salazar (2nd Year PhD)
Data analytics, geostatistics and machine learning for unconventional reservoirs including geostatistical significance and robust trend models.
Supervisor: Drs. Lake and Pyrcz

javier santos photo

Javier Santos (2nd Year PhD)
Multiscale spatial modeling with data analytics and machine learning, new methods to integrate physics and fast surrogate models.
Supervisors: Drs. Prodanovic and Pyrcz

honggeun jo photo

Honggeun Jo (3rd Year PhD)
Geostatistical and machine learning for heterogeneity modeling, including realistic training models and conditional machine learning-based models.
Supervisor: Dr. Pyrcz

wen pan photo

Wen Pan (4th Year PhD)
Machine learning for realistic subsurface models integrating petrophysical and geophysical data.
Supervisors: Drs. Torres-Verdin and Pyrcz

john d'angelo photo

John D’Angelo (4th Year PhD)
Multi-objective geosteering through integration of directional drilling and geophysical data".
Supervisor: Dr. van Oort

eduardo maldonado photo

Eduardo Maldonado (1st Year PhD)
Generalizable machine learning-based surrogate models for flow forecasting.
Supervisor: Dr. Pyrcz

akhil potla photo

Akhil Potla (1st Year MSc)
Integration of engineering physics into machine learning-based flow forecasts. Initial work with capacitance resistance modeling.
Supervisor: Dr. Foster

mide mabadeje photo

Midé Mabadeje (1st Year PhD)
Novel data analytics and machine learning workflows to address various sources of bias that impact decision making.
Supervisor: Dr. Pyrcz

Recent Highlights

The 1st Annual DIRECT Review Meeting took place on Zoom on June 24,2020. Overview Slides for this Meeting are available on the DIRECT website.

Application Areas

  • Conventional oil and gas
  • Tight and shale oil and gas
  • Heavy-oil/bitumen/tarmat
  • Geothermal
  • Carbon Storage
  • Subsurface Environmental Remediation
  • Ground Water (as a resource)
  • Subsurface reserves (SEC reporting)
  • Fundamental Processes

Technical Disciplines

  • Drilling Engineering
  • Production Engineering
  • Reservoir Engineering
  • Petrophysics/Formation Evaluation
  • Data Analytics and Machine Learning
  • Geomechanics/Rock Mechanics
  • Geology
  • Geophysics/Seismic Interpretation
  • at least not at the same level with IAPGeostatistics
  • Computational Sciences
  • Curriculum Development/Design

Engineering Tools

  • Large Scale Simulation (e.g. reservoir simulation)
  • Reduced-order models (e.g. capacitor resistance)
  • Software Development / Deployment (repositories, unit testing)
  • Design of Experiments / Uncertainty Quantification
  • Subsurface Visualization and Interpretation (e.g. Gocad, Petrel, Kingdom etc.)
  • Learning Management Systems

Member Companies

Saudi Aramco