URTeC 2018 Julio

Program Schedule and Abstract Book
URTeC 2018
July 23 - 25, 2018
The Program Report was last updated July 21, 2018 at 01:17 AM EDT. To view the most recent meeting schedule
online, visit https://urtec2018.abstractcentral.com/planner.jsp
Monday, July 23, 2018
Time Session or Event Info
8:30 AM-10:00 AM, Grand Ballroom AB, Opening Plenary Session, Opening Plenary
8:30-8:35 AM Moderator Welcome
8:35-8:45 AM Opening Remarks S. Winberg
Re-Invent, Re-Tool, Re-Imagine: Finding Success in the Resources

8:45-9:00 AM
Arena V.A. Hollub
9:00-9:15 AM Enigmatic Shale S.W. Tinker
What the Investor Community Wants From the Unconventional

9:15-9:30 AM
Fracocene R. Brackett
9:30-10:00 AM Moderated Discussion
10:45 AM-12:05 PM, Room 310, Panel: Impact of Prior Depletion on Completion Efficiency
and Well Performance, Panel
10:45-10:50 AM Introductory Remarks
10:50-11:00 AM Reservoir Engineer, Shell R. Cao
11:00-11:10 AM Vice President of Development, Paramount Resources G. Stotts
11:10-11:20 AM Reservoir Development Manager, Alta Mesa S. Geetan
11:45-12:05 PM Audience Q&A
10:45 AM-12:05 PM, Room 320, Theme 04: Geomechanics and Pore Pressure, Oral,
Session Chair: Mehdi Mokhtari, mxm4487@louisiana.edu, University of Louisiana at
Lafayette; Session Chair: Yongcun Feng, yongcun.f@gmail.com, The University of Texas at
Austin
Novel Pore Pressure Prediction Technique for Unconventional

10:50-11:15 AM
Resources D. Yale; V. Swami; A. Perez
Using Traditional Methods to Predict Pore Pressure in Devonian

11:15-11:40 AM
Black Shale Basins of North East British Columbia S. Green
Impact of Pore Pressure Depletion on Stress Reorientation and its
11:40-12:05 PM Implications on the Growth of Child Well Fractures S. Agrawal; M.
Sharma
10:45 AM-12:05 PM, Room 322, Theme 03: Formation Evaluation – Integrated Workflows
and Interpretation Methods I, Oral, Session Chair: Brian Driskill, brian.driskill@shell.com,
Shell E&P; Session Chair: Livia Sivila, lsivila@enervest.net, EnerVest; Session Chair: Adam
McMullen, adam.mcmullen@shell.com, Shell
Montney Key Drivers: An Integration of Multidisciplinary Data

10:50-11:15 AM Analytics in a Low-Permeability Reservoir K. Hermanson; M. Kwan;
B. Papau
Porosity and Organic Content Analysis, Bone Spring and Wolfcamp

11:15-11:40 AM
Formations J.D. Walls; T. Rider; B. Driskill; M. Durand
A Water Saturation Interpretation Model for Organic-Rich Shale

11:40-12:05 PM Reservoir: A Case Study of North Sumatra Basin M.N. Akbar; B.
Milad
10:45 AM-12:05 PM, Room 330, Theme 07: Augmented Intelligence for Reservoir
Characterization and Performance Prediction , Oral, Session Chair: Lee Geiser,
lee.geiser@petrolink.com, Petrolink Services Inc; Session Chair: Sathish Sankaran,
Sathish.Sankaran@anadarko.com, Anadarko Petroleum
Spider Bots: Database Enhancing and Indexing Scripts to Efficiently

10:50-11:15 AM Convert Raw Well Data Into Valuable Knowledge G.S. Saini; H.
Chan; P. Ashok; E. van Oort; M. Behounek; T. Thetford; M. Shahri
Integrated Workflow for the Definition of a Type Well Using

11:15-11:40 AM
Probabilistic Methods D.S. Jones
Toward a Management Science for Unconventional Wells: A

11:40-12:05 PM
Methodological Approach R.R. Batsell; S. Paranji; J.S. Mintz
10:45 AM-12:05 PM, Room 332, Theme 12: International Emerging Plays, Oral, Session
Chair: Tomasz Ochmanski, t_ochmanski@yahoo.com, Geo-data Consulting LLC; Session
Chair: Luis Baez, Luis.Baez@shell.com, Shell
Emerging Shale Oil Plays in Hypersaline Lacustrine Qianjiang

10:50-11:15 AM Formation, Jianghan Basin, Central China M. Li; X. Ma; T. Cao; G.
Tao; Z. Li; Q. Jiang; S. Wu
Unconventional Reservoir Development in Egypt’s Western Desert:

11:15-11:40 AM Lessons Learned From the First Appraisal Wells M. Salah; M.H.
Ibrahim
11:40-12:05 PM TBD
10:45 AM-12:05 PM, Room 340, Theme 14: Drilling and Completions Optimization I, Oral,
Session Chair: Jennifer Miskimins, jmiskimi@mines.edu, Colorado School of Mines; Session
Chair: Geoff Gullickson, Geoffrey.Gullickson@halliburton.com, Halliburton
Mechanism Study of Casing Deformation in Multistage Hydraulic

10:50-11:15 AM
Fracturing Shale Reservoir F. Yin; S. Yang; Z. Xu; L. Han; X. Wu
Optimization and Drilling of Horizontal Wells using a Bayesian

11:15-11:40 AM
Network J.F. Fierstien; H. Winkler; P. Strauss; A. Klokov
TST3D: Automated Structural Interpretation in Horizontal Wellbores

11:40-12:05 PM
T. Zhang; D. McCormick; A. Nandlal; M. LeFranc
10:45 AM-12:05 PM, Room 342, Panel: National Labs – Leveraging Basic Science to
Advance Subsurface Understanding, Panel
10:50-10:58 AM Oak Ridge National Laboratory Y. Polsky
10:58-11:06 AM Los Alamos National Laboratory R. Pawar
11:06-11:14 AM Lawrence Berkley National Laboratory T. Kneafsey
11:14-11:22 AM Senior Fellow, Geologic and Environmental Systems G.S. Bromhal
10:45 AM-12:05 PM, Room 351, Theme 08: Geochemistry – Reservoir Characterization,
Oral, Session Chair: John Curtis, jcurtis@geomarkresearch.com, GeoMark Research Ltd.;
Session Chair: Christian Bradshaw, Christian.Bradshaw@encana.com, Encana
Reliable Solid Organic Matter Thermal Maturity Assessment Using

10:50-11:15 AM Surface Enhanced Raman Spectroscopy and Case Studies C. Jiang;
L. Gao; S. Wu; J. Shaw; A. Bishop; Y. Tang
Integrated Inorganic and Organic Geochemistry Approach in the

11:15-11:40 AM
Petroleum Systems Analysis of Permian Shale Plays C. Gong
Geochemical Characterization of the Eagle Ford Formation in

11:40-12:05 PM Northeast Mexico S. Ortega-Lucach; L. Gutierrez-Caminero; R.
Torres-Vargas; G. Murillo-Muñetón
10:30 AM-12:15 PM, Exhibit Hall Station A, Theme 13: Stakeholder Management and Social
Performance (HSSE), ePaper, Session Chair: Pedram Fanailoo,
pedram.fanailoo@dnvgl.com, DNV GL; Session Chair: Kent Perry,
kent.perry@gastechnology.org, Gas Technology Institute
Using Drone Magnetic and LiDAR Surveys to Locate Unmarked,

10:35-11:00 AM Abandoned Wells Prior to Unconventional Oil and Gas Development
R.W. Hammack; G. Veloski; J. Sams
Advanced Characterization and Novel Waste Management for Drill

11:00-11:25 AM Cuttings From Marcellus Shale Energy Development M.Y. Stuckman;
H.M. Edenborn; C.L. Lopano; A. Hakala
11:25-11:50 AM Estimating Carbon Intensity of Unconventional Plays D.N. Meehan
10:30 AM-12:40 PM, Exhibit Hall Station B, Theme 11: Reserves, Economics, and Field
Studies II, ePaper, Session Chair: Baosheng Liang, bliang@chevron.com, Chevron ; Session
Chair: Hosein Kalaei, kalaeimh@gmail.com, ConocoPhillips
Mapping the Barnett Shale Gas With Probabilistic Physics-Based

10:35-11:00 AM Decline Curve Models and the Development of a Localized Prior
Distribution R. Wanderley de Holanda; E. Gildin; P.P. Valko
The Use of the Bimodal Production Decline Curve for the Analysis of
11:00-11:25 AM Hydraulically Fractured Shale/Tight Gas Reservoirs C. Doughty; G.J.
Moridis; T. Blasingame
Hindcasting Production Forecasts in Four Shale Gas Basins Using a

11:25-11:50 AM
Physics-based Approach F. Male; M. Marder
Montney Versus North America – Completions Comparison K. Ogilvy;

11:50-12:15 PM
B. Papau; M. Kwan; S. Maliki
10:30 AM-12:15 PM, Exhibit Hall Station C, University Lands Special Session I, ePaper,
Session Chair: David Fulford, david.fulford@apachecorp.com, Apache Corporation; Session
Chair: Meilin Du, Meilin.Du@chevron.com, Chevron
10:35-11:00 AM Wolfcamp Geologic Reservoir Modeling Challenges B.J. Casey
Additional Applications of Optimal Artificial Lift Strategies in the

11:00-11:25 AM
Permian Basin Y.U. Pradhan; H. Xiong; J.K. Forrest; J. Zhu
Additional Applications on Determining Optimal Lateral Lengths and

11:25-11:50 AM Trajectories on University Lands’ Midland and Delaware Basins Y.U.
Pradhan; H. Xiong
1:45 PM-5:35 PM, Room 310, Theme 02: Integrated Characterization of Unconventional
Reservoirs – From Outcrops to Geomodels I, Oral, Session Chair: Samuel Simmons,
ssimmons@parsleyenergy.com, Parsley Energy; Session Chair: Kathryn Dianiska,
Kathryn.Dianiska@anadarko.com, Anadarko; Session Chair: Troy Beserra,
troy.beserra@anadarko.com, Anadarko Petroleum Corporation
1:45-1:50 PM Introductory Remarks
An Integrated View of the Petrology, Sedimentology, and Sequence

1:50-2:15 PM Stratigraphy of the Wolfcamp Formation, Delaware Basin, Texas M.
Thompson; P. Desjardins; J. Pickering; B. Driskill
Outcrop and Subsurface Geology Applied to Drilling, Sweet Spot and

2:15-2:40 PM Target Zone Detection of Resource Shales: The Woodford Shale
Example R.M. Slatt
Geological Facies Prediction Using Computed Tomography in a

2:40-3:05 PM
Machine Learning and Deep Learning Environment U. Odi; T. Nguyen
3:05-3:55 PM Refreshment Break
Recent–Pleistocene Immature Mudrocks of Anoxic Basins of

3:55-4:20 PM Venezuela and West Africa: Analogs for Unconventional Shale
Oil/Gas Reservoirs? U. Hammes; M. Grammer; J. Gregg
Integration of Core Fracture and Lithofacies Descriptions in the

Wolfcamp Shale: Implications for Mechanical Stratigraphy and
4:20-4:45 PM
Deformation History A. Salem; J. Solum; S. Naruk; D. Minisini; P.
Desjardins; J. Hnat
Integrated Geocellular Static Model for Geomechanical and Dynamic

4:45-5:10 PM Simulations in the Vaca Muerta Formation F.A. Lozano; J.P. Palacio;
V. Lazzari; C. Bernhardt; D.E. Hryb; F. González Tomassini
Advanced Downhole Measurements and 3-D Model-based

5:10-5:35 PM Geosteering Improves Wellbore Placement in the Permian Basin C.
Viens
1:45 PM-5:35 PM, Room 320, Theme 04: AAPG's Petroleum Structure and Geomechanics
Division (PSGD), Oral, Session Chair: David Haddad, david.e.haddad@gmail.com,
ConocoPhillips; Session Chair: Robert Hurt, robert.hurt@pxd.com, Pioneer Natural
Resources
3-D Anisotropic Damage Mechanics for Modeling Interaction Between

Hydraulic and Natural Fracture Planes in a Layered Rock –
1:50-2:15 PM
Application to Eagle Ford and Wolfcamp Y. Aimene; C. Hammerquist;
J. Nairn; A. Ouenes
Origin, Detection, Involvement in Hydraulic Stimulation, and
Consequences for Field Development of Large-scale Structural
2:15-2:40 PM
Lineaments in the Marcellus and Duvernay Plays B. Stephenson; E.
Galan; M. Fay; A. Savitski; T. Bai
A Method of Fracture Prediction Across Multiple Stratigraphic

2:40-3:05 PM Horizons in the Midland Basin, Texas, USA C. Pollock; C. Seiler; M.
Valcárcel; E. Macaulay
Can Seismic Inversion Be Used for Geomechanics? A Casing

3:55-4:20 PM Deformation Example J.J. Meyer; J. Gallop; A. Chen; S. Reynolds; S.
Mildren
Intensive Natural Fracture Study of Elk Hills Monterey Formation to

4:20-4:45 PM Better Understand Production Variability R. Gales; R. Sobczyk; N.
Harvey
Reservoir Geomechanic Heterogeneity Index (RGHI): Concept,

4:45-5:10 PM Methodology, and Application J. Zhou; S. Mandal; F. Chen; M. Quest;
D. Hume
Polygonal Fault System in the Paleogene of the Magallanes Foreland

5:10-5:35 PM
Basin, Southern Chile J.A. Pinto; D. Gonzalez; P. Mella; A. Gonzalez
1:45 PM-5:35 PM, Room 322, Theme 03: Formation Evaluation – Integrated Workflows and
Interpretation Methods II, Oral, Session Chair: Thaimar Ramirez,
Thaimar_Ramirez@oxy.com, OXY; Session Chair: Tyler Conner, Tyler.Conner@DVN.com,
Devon Energy
Developing Predictive Power in the Permian: Leveraging Advanced

1:50-2:15 PM Petrophysics to Deliver Cash to the Business A. Blount; T. Croft; M.
Durand; B. Driskill; A. McMullen
Integrated Reservoir Characterization Aids Target Selection,

Production Fluid Prediction, and Completions Optimization in the
2:15-2:40 PM
Southern Delaware Basin Resource Plays T. Tittlemier; J. Speight; S.
Satterfield; C. Hager; I. Easow; B. Chiniwala; A. Martocchia
Integrated Rock Characterization of a Shale Gas Field in the Horn

2:40-3:05 PM
River Basin, Canada T. Teklu; D. Park; H. Jung; J.L. Miskimins
Quantifying Nanoporosity: Insights Revealed by Parallel and

3:55-4:20 PM Multiscale Analyses A.C. Reynolds; S.A. Kelly; R.J. Bonnie; J.J.
Howard; R.L. Krumm
Quantifying the Mechanisms Contributing to Surface Relaxation of
4:20-4:45 PM Protons in Organic Pores of Organic-Rich Mudrocks S. Tandon; Z.
Heidari
Development of Raman Spectroscopy as a Thermal Maturity Proxy in

4:45-5:10 PM Unconventional Resource Assessment G. Myers; K. Kehoe; P.
Hackley
Spatial Characterization of Organic Matter Maturity by Raman

5:10-5:35 PM
Microscope Mapping E.G. Krukowski; J.J. Howard
1:45 PM-5:35 PM, Room 330, Operators' Forum – Completion Optimization, Oral, Session
Chair: Luis Baez, Luis.Baez@shell.com, Shell; Session Chair: Matías Fernandez-Badessich,
matias_badessich@hotmail.com, YPF SA; Session Chair: Andrey Bogdan,
andrey.bogdan@bjservices.com, BJ Services
New Mexico Delaware Basin Horizontal Well Heel Frac and Refrac
1:50-2:40 PM Program and Hydraulic Fracture Diagnostics M. Han; I. Tanakov; E.
Bunker; T. Vulgamore
Optimization of Completion and Well Spacing for Development of

Multi-stacked Reservoirs Using Integration of Data Analytics,
2:40-3:05 PM
Geomechanics, and Reservoir Flow Modeling K. Min; V. Sen; L. Ji; R.
Sullivan
Accelerated Stimulation Optimization via Permanent and Continuous

3:55-4:45 PM Production Monitoring Using Fiber Optic G.A. Ugueto; P. Huckabee;
M. Wojtaszek; A. Reynolds
Multivariate Study of Utica: Marrying the Rock to the Completion A.

4:45-5:35 PM
Trumbo; J. Bowman; L. Lasecki
1:45 PM-5:35 PM, Room 332, Theme 11: Reserves, Economics, and Field Studies I, Oral,
Session Chair: Usman Ahmed, theenergysource@yahoo.com, WellDog; Session Chair:
Susan Howes, showes@scacompanies.com, Subsurface Consultants & Associates, LLC;
Session Chair: Autumn Shannon, autumntriesch@gmail.com, Marathon Oil
Unconventional Field Development Optimization – Fit for Strategy

1:50-2:15 PM
Designs to Realize Your Corporate Goals R. Howrish; D. Anderson
Is GOR Truly Affecting Recovery? A Multi-variate Case Study in the

2:15-2:40 PM
Delaware Basin R. Dutta; S. Dawson; M. Maler
A New Look at Reserves Estimation of Unconventional Gas

2:40-3:05 PM
Reservoirs M.H. Ibrahim; O. Mahmoud; C. Pieprzica
3:55-4:20 PM Global Competitiveness of the US Tight Oil Cost Curve R.G. Clarke
4:20-4:45 PM Reserve Estimation With Unified Production Analysis M. Mehana
New Analysis of EUR Probability Plots Yields Better Uncertainty

4:45-5:10 PM
Assessment and Better Type Wells R. Freeborn
The Value of Building a Multiscale, Regional Geomodel for Reserves

5:10-5:35 PM Assessment of the Midland Basin R. Dommisse; L. Sivila; H. Hamlin;
F. Male
1:45 PM-5:35 PM, Room 340, Theme 05: Fracturing Fluid, Fracture, and Matrix Interactions,
Oral, Session Chair: Craig Cipolla, ccipolla@hess.com, Hess Corporation; Session Chair:
Randahl Roadifer, rdranman@yahoo.com, Amplify Energy, Inc.
Modeling and Experimental Investigation of Fluid-Related Damage to

1:50-2:15 PM Hydraulic Fractures P. Abivin; R. Prabhu; D. Khvostichenko; C.
Hilliard; C. Nelson; T. Kuo; Y. Li; P. Shukla; S. Makarychev-Mikhailov
Impact of Natural Fractures Beyond the Hydraulic Fracture

2:15-2:40 PM Complexity in Unconventional Reservoirs – A Permian Case Study P.
Pankaj; J. Li
Near Fracture Capillary End Effect on Shale Gas/Oil Production R.

2:40-3:05 PM
Elputranto; I. Akkutlu
Theoretical Investigation of Water Blocking in Unconventional

3:55-4:20 PM Reservoirs Due to Spontaneous Imbibition and Water Adsorption L.
Deng; M.J. King
Diagnosing Fracture-Wellbore Connectivity Using Chemical Tracer

4:20-4:45 PM
Flowback Data A. Kumar; M. Sharma
Impact of Authigenic Surface Roughness on Water Invasion and

4:45-5:10 PM Flowback in Fractured Media: A Micromodel Study A. Mehmani; S.A.
Kelly; C. Torres-Verdin; M. Balhoff
Effects of Hydraulic Fracturing Fluid Chemistry on Shale Matrix

5:10-5:35 PM Permeability A.A. Alalli; Q. Li; A. Jew; A. Kohli; J. Bargar; M. Zoback;
A. Kovscek
1:45 PM-3:05 PM, Room 342, Theme 10: Flowback and Artificial Lift for Unconventional
Reservoirs I, Oral, Session Chair: Jennifer Gujral, Jennifer.Gujral@shell.com, Shell; Session
Chair: Johannes Alvarez, johalvarez@chevron.com, Chevron

Flowback in Shale Wells: Proppant Transport and Distribution in the
1:50-2:15 PM
Wellbore K. Putri; H. Lu; C. Kwok; K. Moncada
Evaluating the Loss in Fracture Volume During Flowback and Its

2:15-2:40 PM Relationship to Choke-size: Fastback Versus Slowback H.
Dehghanpour; Y. Fu; S. Motealleh; C. Lopez
Defining the Optimal Drawdown Strategy in the Vaca Muerta

2:40-3:05 PM
Formation A.A. Lerza; B. Liang; D. Rojas
3:50 PM-5:35 PM, Room 342, Panel: Technologies That Will Make a Difference in
Unconventional Reservoir E&P, Panel
Executive Vice President and Chief Technology Officer – Pioneer

3:55-4:05 PM
Natural Resources C. Cheatwood
4:05-4:15 PM Vice President of Geoscience and Technology, Concho C. Spies
Vice President & Chief Information Officer, Occidental Petroleum

4:15-4:25 PM
Corporation Y. Charalambous
Corporate Vice President and Chief Technology Officer, National

4:25-4:35 PM
Oilwell Varco (NOV) H. Kverneland
4:35-5:05 PM Moderated Discussion
1:45 PM-3:05 PM, Room 351, Theme 08: Inorganic Geochemistry of Unconventional
Plays/Fluid Rock Interactions, Oral, Session Chair: Faye Liu, fayeliu0728@gmail.com,
ConocoPhillips; Session Chair: Jennifer Adams, Jennifer.Adams@Weatherford.com,
Weatherford
Bench-Top Experiments Evaluating Simulated Hydraulic Fracturing

Fluid Interactions With Marcellus Shale Core J. Moore; A. Hakala;
1:50-2:15 PM
C.L. Lopano; W. Xiong; T. Phan; A. Vankeuren; S. Sharma; J.
Pilewski; K. Jarvis; S. Brown; D. Crandall
Barium Sources in Hydraulic Fracturing Systems and Chemical

2:15-2:40 PM Controls on Its Release Into Solution A. Jew; Q. Li; D. Cercone; K.
Maher; G. Brown; J. Bargar
Imaging Pyrite Oxidation and Barite Precipitation in Gas and Oil

2:40-3:05 PM Shales Q. Li; A. Jew; A. Kiss; A. Kohli; A.A. Alalli; A. Kovscek; M.
Zoback; D. Cercone; K. Maher; G. Brown; J. Bargar
3:50 PM-5:35 PM, Room 351, University Lands Special Session II, Oral, Session Chair:
Yogashri Pradhan, ypradhan.pge@utexas.edu, Texas Oil and Gas Institute; Session Chair:
Jeff Spath, spath@tamu.edu
Lessons Learned From Existing Horizontal Fractured Wells in Midland

Basin of University Lands (UL): Rate Transient Analysis vs.
3:55-4:20 PM
Completion and Field Development Optimization J. Zhu; J.K. Forrest;
H. Xiong; Y.U. Pradhan
The Effect of Initial Conditions and Fluid PVT Properties on

4:20-4:45 PM Unconventional Oil and Gas Recoveries in the Wolfcamp Formation
in the Midland Basin J.K. Forrest; J. Zhu; H. Xiong; Y.U. Pradhan
A Practical Way to Prepare Physical-Based Type Well Performance

4:45-5:10 PM
Curves for Unconventional Reservoirs in the Permian Basin H. Xiong
Evaluating Underperforming Wells on Permian Basin University

5:10-5:35 PM
Lands Y.U. Pradhan; J. Spath; H. Xiong; J. Zhu; J.K. Forrest
1:45 PM-3:05 PM, Exhibit Hall Station A, Theme 09: EOR Applications for Unconventional
Reservoirs, ePaper, Session Chair: David Riestenberg, DRiestenberg@adv-res.com,
Advanced Resources International, Inc; Session Chair: Basak Kurtoglu,
bkurtoglu@quantumep.com, Quantum Energy Partners; Session Chair: Steven Carpenter,
steven.carpenter@uwyo.edu, EORI
Chemical Stimulation with Driving Process to Extract Oil from Tight

1:50-2:15 PM
Formation J. Zhang; D. Wang
Lithologic and Geomechanical Control on CO Huff-n-Puff Enhanced

2
2:15-2:40 PM Oil Recovery Processes Using Integrated Modeling Framework in
Wolfcamp T.N. Phan; Z.A. Reza
Scaling for Wettability Alteration Induced by Addition of Surfactants in

2:40-3:05 PM Completion Fluids: Surfactant Selection for Optimum Performance F.
Zhang; I. Saputra; I. Adel; D.S. Schechter
3:50 PM-4:45 PM, Exhibit Hall Station A, Theme 05: Unconventional Well Productivity,
ePaper, Session Chair: Baosheng Liang, bliang@chevron.com, Chevron ; Session Chair:
Bobby Poe, bobbypoe@yahoo.com, Texas A&M University
Alternative Production Mechanisms in Unconventional Reservoirs

3:55-4:20 PM
J.A. Acuna
Condensate Blocking and Mitigation in Liquid-Rich Shale Reservoirs:

4:20-4:45 PM An Integrated Evaluation Based on Systematic PVT Modeling and
Simulation Studies N. Nagarajan; A. Orangi
1:45 PM-4:45 PM, Exhibit Hall Station B, Theme 14: Drilling and Completions Optimization
IV, ePaper, Session Chair: Isaac Aviles, iaviles@slb.com, Schlumberger; Session Chair:
Jennifer Miskimins, jmiskimi@mines.edu, Colorado School of Mines; Session Chair: Geoff
Gullickson, Geoffrey.Gullickson@halliburton.com, Halliburton
A Unified Model for Predicting Flowing Pressure and Temperature

1:50-2:15 PM Distribution in the Horizontal Wellbore for Different Energized
Fracturing Fluids Z. Xu; K. Wu; X. Song; G. Li; W. Yu; Z. Zhu; Z. Pang
Development of A High-Performance Cement Slurry Antifoamer

2:15-2:40 PM Through Lab Evaluation and Field Trials L. Cabori; L. Jiang; B.
Abrams; J. Terracina
Fracture Initiation and Propagation Characteristics for Radial Drilling-

2:40-3:05 PM
Fracturing: An Experimental Study Q. Liu; K. Sepehrnoori; W. Yu
Rapid Evaluation of Diverter Effectiveness From Poroelastic Pressure

3:55-4:20 PM Response in Offset Wells C. Kahn; B. Cottingham; S. Kashikar; S.
Senften; E. Coenen
Geomechanical Modeling and Wellbore Stability Analysis Approach to

4:20-4:45 PM Plan Deep Horizontal Wells Across Problematic Shale Formation A.K.
Abbas; R. Flori; M. Alsaba
1:45 PM-3:05 PM, Exhibit Hall Station C, Theme 04: Beyond Brittleness: Geomechanical
Characterization II, ePaper, Session Chair: Bitao Lai, xilxlai@gmail.com, SPE member;
Session Chair: Abhijit Mitra, abhijit@metarocklab.com, MetaRock Laboratories
Identifying Volcanic Ash Beds and Lamina-scale Stratigraphy Using

1:50-2:15 PM
Rock Mechanical Properties A. Hildick; J. Havens
Compressibility, Porosity, and Permeability of Shales Involving Stress

2:15-2:40 PM
Shock and Loading/Unloading Hysteresis F. Civan
Rigorous Estimation of the Initial Conditions of Flowback Using a

2:40-3:05 PM Coupled Frac/Dynamic Drainage Area Model Constrained by
Laboratory Geomechanical Data Z. Zhang; B. Yuan; C.R. Clarkson
Tuesday, July 24, 2018

8:25 AM-12:15 PM, Room 310, Theme 02: Integrated Characterization of Unconventional
Reservoirs – From Outcrops to Geomodels II, Oral, Session Chair: Robin Pearson,
Robin.Pearson@chk.com, Chesapeake Energy; Session Chair: Todd Peters,
tpeters@parsleyenergy.com, Parsley Energy; Session Chair: Randy Lambert,
RLambert@oasispetroleum.com, Oasis Petroleum
Seismic to Simulation: Woodford Shale Case Study in Oklahoma,

8:30-8:55 AM USA E.J. Torres-Parada; S. Sinha; L.E. Infante-Paez; R.M. Slatt; K.
Marfurt
The Importance of Overburden and Pore Pressure on Horizontal

Stress Magnitude Determination: An Example From the Delaware
8:55-9:20 AM
Basin K. Kozlowski; M. Da Silva; D. Brown; J. Taylor; H. Willems; T.
Watson; D. Burch; T. Hutton; C. Christensen; M. Manohar
A Strategy for De-risking the Upper Vaca Muerta as a Dual-Layer

9:20-9:45 AM
Development K. Boyd; R. Notta; M. Thomas; D. Cakici
Geological Controls on Fluid Compositional Variations in

9:45-10:10 AM Unconventional Hybrid Plays: Insight From Gas Geochemistry
(Montney Play, Western Canada) T. Euzen; J. Chatellier; A. Mort
10:10-11:00 AM Refreshment Break
Outlier Analysis: A Systematic Method for Distinguishing Between

Sub-surface and Engineering Influence on Well Performance in the
11:00-11:25 AM
Montney F. Todea; B. Stephenson; A. Tomlinson; H. Pratt; W.
Williams; L. Acosta; B. Speidel
An Upscaled DFN Model to Understand the Effects of Natural

11:25-11:50 AM Fracture Properties on Fluid Flow in the Hunton Group Tight
Limestone B. Milad; S. Ghosh; R.M. Slatt
Sedimentological Characterization of the Pimienta Formation in the

Central Part of the Tampico-Misantla Basin, Veracruz, Mexico G.
11:50-12:15 PM Abascal-Hernández; J.M. León-Francisco; R. Torres-Vargas; D.E.
Garduño-Martínez; S. Franco-Navarrete; J. Méndez-Vázquez; S.
Ortega-Lucach; L. Gutierrez-Caminero; G. Murillo-Muñetón
8:25 AM-12:15 PM, Room 320, Theme 04: Geomechanics – Hydraulic Fracture Simulation I,
Oral, Session Chair: Gang Han, gang.han@aramcoservices.com, Aramco; Session Chair:
Andrew Munoz, amunoz615@gmail.com, Newfield
Increase NPV and Reduce Completion Cost by Using an In-House

Automated Fracture Design, Reservoir Simulator and Optimization
8:30-8:55 AM
Workflow to Design Well Spacing and Completions S. Bhattacharya;
E. Lake; R. Dombrowski
Numerical Study of Stress Shadowing Effect on Fracture Initiation and
8:55-9:20 AM Interaction Between Perforation Clusters B. Damjanac; S. Maxwell; A.
Pirayehgar
An Integrated Field and Numerical Study of the Impact of Formation

9:20-9:45 AM Anisotropy on Stage Spacing in Horizontal Wells V. Sesetty; A.
Ghassemi; I. Gil
Effect of Interbeds on Hydraulic Fracture Characteristics and

9:45-10:10 AM
Formation Pressure Response A. Bere; M. Profit; M. Dutko; U. Mutlu
A Coupled Three-Dimensional Hydraulic Fracture Propagation Model

11:00-11:25 AM Considering Multiple Bedding Layers J. Tang; K. Wu; L. Zuo; C. Ehlig
-Economides
The Impact of Natural Fracture Thickness on Hydraulic Fracture

11:25-11:50 AM
Interaction Mechanics W. Wang; J. Olson; M. Prodanović; R. Schultz
Field Scale Proppant Transport Simulation and Its Application to

11:50-12:15 PM
Optimize Stimulation Strategy R. Kou; G.J. Moridis; T. Blasingame
8:25 AM-12:15 PM, Room 322, Hydraulic Fracture Test Site (HFTS) Special Session I, Oral,
Session Chair: Kent Perry, kent.perry@gastechnology.org, Gas Technology Institute; Session
Chair: James Courtier, jcourtier@laredopetro.com, Laredo Petroleum
Hydraulic Fracture Test Site – Project Overview and Summary of

8:30-8:55 AM
Results J. Courtier; J. Ciezobka
Hydraulic Fractures in Core From Stimulated Reservoirs: Core

8:55-9:20 AM Fracture Description of the HFTS Slant Core, Reagan County,
Midland Basin, Texas J.F. Gale; S.J. Elliott; S.E. Laubach
Assessment of In Situ Proppant Placement in SRV Using Through-

9:20-9:45 AM
Fracture Core Sampling at HFTS D. Maity; J. Ciezobka; S. Eisenlord
Analysis and Distribution of Proppant Recovered From Fracture

9:45-10:10 AM Faces in the HFTS Slant Core Drilled Through a Stimulated Reservoir
S.J. Elliott; J.F. Gale
Natural and Hydraulic Fracture Density Prediction and Identification of

11:00-11:25 AM
Controllers J. Wicker; W. Campbell; J. Courtier
Inter-well Communication Study of UWC and MWC Wells in the HFTS

11:25-11:50 AM
T. Wood; R. Leonard; C. Senters; C. Squires
Well Interference Diagnosis Through Integrated Analysis of Chemical
11:50-12:15 PM Tracer and Pressure Interference Tests A. Kumar; P. Seth; K.
Shrivastava; R. Manchanda; M. Sharma
8:25 AM-12:15 PM, Room 330, Operators' Forum – Performance Prediction and Reservoir
Characterization, Oral, Session Chair: Craig Cipolla, ccipolla@hess.com, Hess Corporation;
Session Chair: Ali Sloan, asloan@parsleyenergy.com, Parsley Energy
Case Histories of Integrating DAS Fiber-Based Microseismic and

8:30-8:55 AM Strain Data for Monitoring Horizontal Hydraulic Stimulations R. Hull;
R. Meek; H. Bello
Integrated Statistical Workflow for Optimum Well Spacing With Data

Analytics, Pilots, Geomechanical-Reservoir Modeling, and Economic
8:55-9:45 AM
Analysis R. Cao; C. Chen; R. Li; T. Firincioglu; C. Ozgen; T. Croft; A.
Girardi; N. Chowdhury
Myths and Facts of Forecasting Horizontal Well Production in

11:00-11:25 AM Unconventional Reservoirs – Are We Complicating a Simple
Analysis? V. Muralidharan; K. Joshi
An Integrated Geomechanics-Reservoir Simulation Workflow for

Completion Design Optimization T. Yeh; J. Jennings; D. Cakici; J.
11:25-12:15 PM
Chavarria Guerra; M. Durand; B.L. Williams; T. Chen; R. Casillas; V.
Jain; R. Li; T. Bai
8:25 AM-10:10 AM, Room 332, Theme 07: Nanoparticles, Chemistry, and Machine Learning:
Tools for Enhancing Oil Recovery, Oral, Session Chair: David Langton,
David.Langton@dvn.com, Devon Energy; Session Chair: Leo Pirela,
leo.pirela@vplusenergy.com, VPlus Energy LLC
Nanoparticle-Encapsulated Acids for Stimulation of Calcite-Rich

8:30-8:55 AM
Shales R. Singh; S. Tong; K. Panthi; K. Mohanty
Post-Frac-Hit Mitigation and Well Remediation of Woodford

Horizontal Wells With Solvent/Surfactant Chemistry Blend C.
8:55-9:20 AM
Swanson; W.A. Hill; G. Nilson; C. Griman; R. Hill; P. Sullivan; C.
Aften; J.C. Jimenez; G. Pietrangeli; D.C. Shedd; J. Pursley
Ion Tuned Water Can Greatly Enhance Alteration of Carbonate

9:20-9:45 AM
Surface to Water-wet H. Ding; S. Rahman
Synthetic Well Log Generation Using Machine Learning Techniques

9:45-10:10 AM
O. Akinnikawe; S. Lyne; J. Roberts
10:55 AM-12:15 PM, Room 332, Theme 07: From Machine Learning to CT Scanning – Novel
Approaches to Old Problems, Oral, Session Chair: Isaac Aviles, iaviles@slb.com,
Schlumberger; Session Chair: Susan Nash, snash@aapg.org, University of Oklahoma
Novel Monitoring Technology Helps to Make Informed Decisions and

11:00-11:25 AM Maximize the Efficiency of Completion Strategy S. Parkhonyuk; R.
Korkin; A. Kabannik; A. Fedorov; M. Nikolaev
High-Quality 3-D MicroCT Imaging of Source Rocks – Novel

11:25-11:50 AM Methodology to Measure and Correct for X-Ray Scatter A. Katsevich;
M. Frenkel; Q. Sun; S. Eichmann; V. Prieto
Quantitative Analysis and Feature Detection for Scanning Electron

11:50-12:15 PM Microscopy Images Using Machine Learning and Image Processing
X. Tian; H. Daigle; F. Xu
8:25 AM-12:15 PM, Room 340, Theme 05: Transient Analysis, History Matching, and
Reservoir Modeling, Oral, Session Chair: Vincent Artus, artus@kappaeng.com, KAPPA
Engineering; Session Chair: Hosein Kalaei, kalaeimh@gmail.com, ConocoPhillips; Session
Chair: Mel Sorkin, mel.sorkin@apachecorp.com, Apache Corporation
A Methodology to Characterize Well Performance in the Delaware

8:30-8:55 AM
Basin: A Wolfcamp Case Study K. Mohan
Pressure and Rate Transient Behavior of a Horizontal Well

8:55-9:20 AM Intercepting Multiple Hydraulic Fractures Within a Fractal Reservoir
A.R. Valdes-Perez; T. Blasingame; L. Larsen
Improved Rate-Transient Analysis Coupled With Pore Mechanics and

9:20-9:45 AM Multi-Scale Fluid Dynamics in Heterogeneous Fractured Liquid-Rich
Shale B. Yuan; C.R. Clarkson
Transient Linear Flow Analysis of Multi-Fractured Horizontal Wells

9:45-10:10 AM Considering Three-Phase Flow and Pressure-Dependent Rock
Properties H. Behmanesh; H. Hamdi; S.A. Ghaderi; C.R. Clarkson
Understanding Production Drivers in the Vaca Muerta Shale Using an

11:00-11:25 AM Integrated Reservoir Simulation Approach R. Altman; D. Pederiva; R.
Mehranfar; M. Frydman
Reservoir Characterization to Understand Optimal Well Spacing – A

11:25-11:50 AM
Wolfcamp Case Study N. Bansal; T. Blasingame; J. Han; Y. Shin
Modeling Well Performance in Piceance Basin Niobrara Formation

11:50-12:15 PM Using Embedded Discrete Fracture Model Y. Xu; W. Yu; N. Li; E.
Lolon; K. Sepehrnoori
8:25 AM-11:50 AM, Room 342, Theme 06: Geophysics in the Permian Basin, Oral, Session
Chair: Martyn Hargrave, mmhargrave@ikonscience.com, Ikon Science ; Session Chair: Ulrich
Zimmer, Ulrich.Zimmer@shell.com, Shell
Predicting Pore-Pressure From On-shore Seismic Data in the

8:30-8:55 AM Delaware Basin M. Rauch-Davies; B. Schmicker; S.W. Smith; S.
Green; J.J. Meyer
8:55-9:20 AM Production Metric Analytics in the Wolfcamp Formation C.P. Ross
Geophysical Data Processing, Rock Property Inversion, and

9:20-9:45 AM Geomechanical Model Building in a Midland Basin Prospect,
Midland/Ector Counties, Texas S. Singleton
Revealing Geologic Features in the Midland Basin Through

9:45-10:10 AM
Frequency Decomposition R. Moore
Strategies for Improving the Performance of Child Wells in the

11:00-11:25 AM Permian Basin R. Manchanda; P. Bhardwaj; J. Hwang; M. Sharma;
M. Maguire; J. Greenwald
Estimation of Seismic Attenuation in the Delaware Basin Using Peak

11:25-11:50 AM Frequencies M. Drwila; L. Eisner; Z. Jechumtalova; D. Anikiev; R.
Keller
8:25 AM-12:15 PM, Room 351, Theme 11: Well Spacing Optimization, Oral, Session Chair:
Mohan Manohar, mohan.manohar@nblenergy.com, Noble Energy Inc.; Session Chair:
Alexsandra Martinez, adinorah.martinez@gmail.com, DeGoyler and MacNaughton; Session
Chair: Alejandro Lerza, Alerza@chevron.com, Chevron
Development of the Stacked Pay in the Delaware Basin F.

8:30-8:55 AM
Alimahomed; C. Defeu
Material Balance Approach to Determine Drainage Volume for Multi-

8:55-9:20 AM
Fracture Unconventional Oil Wells A. Gherabati
Characterizing Well Spacing, Well Stacking, and Well Completion

9:20-9:45 AM Optimization in the Permian Basin – An Improved and Efficient
Workflow Using Cloud Based Computing P. Pankaj
Advanced Modeling of Production-Induced Stress Change Impact on

9:45-10:10 AM Wellbore Stability of Infill Well Drilling in Unconventional Reservoirs
W. Zheng; L. Xu; P. Pankaj; F. Ajisafe; J. Li

Understand the Early Indicators for Long-term Performance of
11:00-11:25 AM
Unconventional Wells R. Cao; H. Liu
Depth of Investigation for Linear Flow: Theory and Practice A.

11:25-11:50 AM
Ravikumar; J. Lee
Rapid Field-Scale Well Spacing Optimization in Tight and Shale Oil

11:50-12:15 PM Reservoirs Using Fast Marching Method A. Iino; T. Onishi; F. Olalotiti-
Lawal; A. Datta-Gupta
9:40 AM-11:25 AM, Exhibit Hall Station A, Theme 04: Geomechanics: From Lab to Field,
ePaper, Session Chair: Ali Almomen, momiam0b@aramco.com, Saudi Aramco; Session
Chair: Yinghui Liu, Yinghui.Liu@shell.com, Shell; Session Chair: Yogashri Pradhan,
ypradhan.pge@utexas.edu, Texas Oil and Gas Institute
Physics-Driven Optimization of Drained Rock Volume for Multistage

9:45-10:10 AM Fracturing: Field Example From the Wolfcamp Formation, Midland
Basin S.G. Parsegov; K. Nandlal; D.S. Schechter; R. Weijermars
Experimental and Numerical Investigation of Fracture Toughness of

10:10-10:35 AM
Anisotropic Shale Y. Suo; Z. Chen; S. Rahman
The Elastic Moduli Change After Carbon Dioxide Flooding Into

10:35-11:00 AM Limestone: An Experimental Study Y. Zhang; M. Lebedev; M.
Sarmadivaleh; H. Yu; S. Iglauer
Pulsed Power Plasma Stimulation Technique – Experimental Study

11:00-11:25 AM on Single Pulse Test for Fractures Initiation Y. Xiao; W. House; E.
Unal; M. Soliman
9:40 AM-11:25 AM, Exhibit Hall Station B, Theme 05: Nanoscale PVT and IOR, ePaper,
Session Chair: Tuba Firincioglu, tfirincioglu@NitecLLC.com, NITEC LLC; Session Chair: Birol
Dindoruk, birol.dindoruk@shell.com, Shell
An Integrated Geological Modeling Approach to Assess Potential of

9:45-10:10 AM Field Wells for Application of a Surfactant Imbibition Process in an
Ultra-Tight Rock Formation S.L. Detwiler; A. Roth; D. Wang
Characterization of Thermal Evolution of Pores and Fluids in Shales

10:10-10:35 AM
Using NMR 2-D Measurement H. Jiang; H. Daigle; B. Zhang; X. Tian
Determination of Confined Fluid Phase Behavior Using Modified Peng

10:35-11:00 AM
-Robinson Equation of State G. Yang; Z. Fan; X. Li
Simulation of Gas Adsorption and Capillary Condensation in Shale

11:00-11:25 AM Nanopores Using Lattice Boltzmann Equation with Non-ideal
Equation of State R. Xu; M. Prodanović; C.J. Landry
9:40 AM-10:35 AM, Exhibit Hall Station C, Theme 10: Completion to Reservoir Optimization
and Diagnostics, ePaper, Session Chair: Pedram Fanailoo, pedram.fanailoo@dnvgl.com,
DNV GL; Session Chair: Autumn Shannon, autumntriesch@gmail.com, Marathon Oil
Fracture Closure (FC) Determination From Two Successive DFITs

9:45-10:10 AM (Diagnostic Fracture Injection Test) in One Formation Y. Rizwan; G.
Liu
A New Model for Pressure Transient Analysis of Fractured Horizontal

Wells in Shale Gas Reservoirs With Continuum and Discrete Fracture
10:10-10:35 AM
Networks Z. Chen; X.W. Liao; H. Zhang; W. Yu; X.D. Shen; X. Shang;
J.L. Zhang; H. Chu; Q. Liu
1:45 PM-5:35 PM, Room 310, Theme 09: Chemical EOR and Novel Techniques, Oral,
Session Chair: Johannes Alvarez, johalvarez@chevron.com, Chevron; Session Chair: Hosein
Kalaei, kalaeimh@gmail.com, ConocoPhillips
A Chemical Blend for Stimulating Production in Oil-Shale Formations

1:50-2:15 PM
K. Mohanty; C. Miller; S. Tong
Comprehensive Workflow for Lab to Field-Scale Numerical Simulation

to Improve Oil Recovery in the Eagle Ford Shale by Selective Testing
2:15-2:40 PM
and Modeling of Surfactants for Wettability Alteration I. Saputra; D.S.
Schechter
Comprehensive Study of Gas Cycling in the Bakken Shale A. Sanaei;

2:40-3:05 PM
A. Abouie; M. Tagavifar; K. Sepehrnoori
Simulation of Wettability Alteration Effect on Well Performance in

3:55-4:20 PM
Highly Fractured Reservoirs J. Li; K. Wu
An Efficient Method to Determine Wormhole Propagation During

4:20-4:45 PM
Matrix Acidizing Z. Fan; X. Li; R.D. Ostermann; J. Jiang
The Impact of the Fracturing Additives on the Near Fracture Face

4:45-5:10 PM Matrix Permeability for Shale and Low Permeability Sand Formations
A.A. Al-Ameri; T.D. Gamadi; L. Heinze; I. Ispas; S. Gorell
Surfactant Based EOR for Tight Oil Unconventional Reservoirs

Through Wettability Alteration: Novel Surfactant Formulations and
5:10-5:35 PM
Their Efficacy to Induce Spontaneous Imbibition P.D. Patil; N. Rohilla;
W. Yu; A. Katiyar; S. Falcone; B. Brown; S. Duckworth; P. Rozowski
1:45 PM-3:05 PM, Room 320, Theme 04: Fracture Monitoring and Diagnostics, Oral,
Session Chair: James Hnat, J.Hnat@Shell.com, Shell Exploration & Production; Session
Chair: Caleb Pollock, caleb.pollock@pxd.com, Pioneer Natural Resources
Geomechanical Simulation of Different Conceptual Models for

1:50-2:15 PM
Microseismic Interpretation S. Maxwell; A. Pirayehgar
Modeling Distributed Fiber Optic Sensor Signals Using Computational

2:15-2:40 PM Rock Mechanics C.S. Sherman; R.J. Mellors; J.P. Morris; F.J.
Ryerson
Estimation of Fracture Geometries From a Network of Poroelastic

2:40-3:05 PM
Pressure Responses in Offset Wells S. Spicer; E. Coenen
3:50 PM-5:35 PM, Room 320, Theme 04: From Perforation to Performance: Geomechanical
Applications, Oral, Session Chair: Katerina Yared, katerina.yared@gmail.com, RWTH
AACHEN University; Session Chair: Ahmed Mubarak,
AHMED.MUBARAK.8@ARAMCO.COM, Saudi Aramco
Paradigm Shift in Wolfcamp Shale Improves Well Performance by

3:55-4:20 PM 70% E. Ejofodomi; R. Sethi; E. Aktas; J. Padgett; B. Mackay; A.
Mirakyan; B. McCrackin; C. Douglas
Production Optimization Using Machine Learning in Bakken Shale G.

4:20-4:45 PM
Luo; Y. Tian; M. Bychina; C. Ehlig-Economides
An Industrialized Approach to Stage-by-Stage Completion Designs

4:45-5:10 PM
W.D. Logan; P. Dalamarinis; B. Williams; K. Urbis; S. Lipari
A Simple and Cost-Effective Workflow for Engineered Perforations B.

5:10-5:35 PM
Cottingham; J. Truax; M. Padgham; X. An; S. Denney
1:45 PM-5:35 PM, Room 322, Theme 03: NMR and Electrical Measurements, Oral, Session
Chair: Stephanie Perry, Stephanie.Perry@anadarko.com, Anadarko; Session Chair: Mohan
Manohar, mohan.manohar@nblenergy.com, Noble Energy Inc.
Investigation of the Shale Electrical Resistivity Reversal Commonly

1:50-2:15 PM Observed at the Wet- to Dry-Gas Transition S.P. Cumella; A.P.
Byrnes; M.D. Sonnenfeld
Observations of Induction Dielectric Measurements and Their Role in

2:15-2:40 PM Determining Thermal Maturity of Organic Mudrocks J.C. Rasmus; D.
Homan; G. Wang; N. Uschner
Quantifying the Influence of Rock Fabric, Composition, and Electrical

2:40-3:05 PM Double Layer on the Broadband Dielectric Dispersion of Organic-rich
Mudrocks A. Posenato Garcia; Z. Heidari

Multiscale Pore Systems in Shales and Their Effect on Laboratory
3:55-4:20 PM Measurements and Transport Properties H. Daigle; C. Jiang; X. Tian;
H. Jiang
NMR Considerations in Shales at Elevated Temperature A.

4:20-4:45 PM
Chakravarty; C. Rai; C. Sondergeld
Measurement of Natural Gas Isotherms and Imaging Gas in Shale

4:45-5:10 PM
Using NMR M. Dick; D. Veselinovic; D. Green
High-Field (400 MHz) T2 Measurements Using a Custom-Built NMR

5:10-5:35 PM Probe, Eagle Ford Shale, Gonzales and La Salle Counties, Texas B.
McDowell; A.N. Tutuncu; Y. Yang
1:45 PM-5:35 PM, Room 330, Operators' Forum – Well Spacing and Field Development,
Oral, Session Chair: Robert Hull, Robert.Hull@pxd.com, Pioneer; Session Chair: Hao Sun,
war.1982@gmail.com, Chevron ; Session Chair: Patrick Boyle,
Patrick.R.Boyle@conocophillips.com, ConocoPhillips
Integrating Microseismic, Geomechanics, Hydraulic Fracture

Modeling, and Reservoir Simulation to Characterize Parent Well
1:50-2:40 PM
Depletion and Infill Well Performance in the Bakken C.L. Cipolla; M.
Motiee; A. Kechemir
An Integrated Approach to Optimizing Completions and Protecting

3:55-4:45 PM Parent Wells in the Montney Formation J. Nieto; G. Janega; B. Batlai;
H. Martinez
Tank Development in the Midland Basin, Texas: A Case Study of

Super-charging a Reservoir to Optimize Production and Increase
4:45-5:35 PM
Horizontal Well Densities J. Thompson; N. Franciose; M. Schutt; K.
Hartig; J. McKenna
1:45 PM-5:10 PM, Room 332, Theme 13: Water Management, Oral, Session Chair: Bryce
Levett, bryce.levett@dnvgl.com, DNV GL; Session Chair: Sarah Eisenlord,
sarah.eisenlord@gastechnology.org, Gas Technology Institute
“Fit-for-Purpose” Treatment of Produced Water for Hydraulic

1:50-2:15 PM
Fracturing R.R. Sharma
Produced Water Treatment R&D: Developing Advanced, Cost-

2:15-2:40 PM Effective Treatment Technologies E. Folio; O. Ogunsola; E. Melchert;
E. Frye
Tailoring Treated Brines for Reuse Scenarios M. Wenzlick; N. Siefert;

2:40-3:05 PM
A. Hakala
The Water Challenge Program – Permian Basin Pilot Results A.

3:55-4:20 PM
Wilcox
Optimal Planning for Wastewater Disposal Facilities: Application of

4:20-4:45 PM
Geographic Information System and Data Analytics A. Jamali
Replacing Freshwater With Seawater: Problems, Solutions, and

4:45-5:10 PM
Applications T. Almubarak; J. Ng; H. Nasr-El-Din
1:45 PM-5:35 PM, Room 340, Theme 05: Fluid Flow – Nanoscale Compositional and
Diffusion Processes, Oral, Session Chair: David Fulford, david.fulford@apachecorp.com,
Apache Corporation; Session Chair: John Thompson, jthompson@atreservoir.com, Anderson
Thompson Reservoir Strategies
In Situ Fractionation in Liquids-Rich Shales and Its Implications for

1:50-2:15 PM EOR: Experimental Verification and Modeling Study A. Tinni; F.
Perez; D. Devegowda; T. Trong; S.T. Dang; C. Sondergeld; C. Rai
Minimum Miscibility Pressure Calculation for Oil Shale and Tight

2:15-2:40 PM Reservoirs With Large Gas-Oil Capillary Pressure K. Zhang; B.
Nojabaei; K. Ahmadi; R. Johns
Solution Gas Drive in Tight Oil Reservoirs: New Insights From

2:40-3:05 PM Capillary Evaporation Experiments E. Barsotti; M. Piri; J. Chen; S.
Althaus
Flow Behavior From Organic- and Mineral-Hosted Porosity

3:55-4:20 PM Systems—From Pores to Production R.M. MacDonald; S. Geetan; D.
Klemin
Experimental and Theoretical Investigation of the Confinement Effect

4:20-4:45 PM on Gas Properties in Nano-scale Porous Media S. Salahshoor; M.
Fahs
The Effect of Mineral Composition on Shale Oil Recovery A. Fakhry;

4:45-5:10 PM
T. Hoffman
Determining the Impact of Mineralogy Composition for Multiphase

5:10-5:35 PM Flow Through Hydraulically Induced Fractures J. E. Santos; C.J.
Landry; M. Prodanović
1:45 PM-3:05 PM, Room 342, Panel: Induced Seismicity – Perspectives and Challenges,
Panel

1:50-2:00 PM Manager of Geophysics, Cimarex D. Klepacki
2:00-2:10 PM Texas State Seismologist A. Velasco
2:10-2:20 PM Science Writer and Reporter, Dallas Morning News A. Kuchment
2:20-2:45 PM Moderated Discussion
3:50 PM-5:35 PM, Room 342, Induced Seismicity Special Session, Oral, Session Chair: Hal
Macartney, hal.macartney@pxd.com, Pioneer Natural Resources; Session Chair: Cody
Comiskey, Cody.Comiskey@anadarko.com, Anadarko
Integrating Poroelastic Effects of Wastewater Injection and Rupture

3:55-4:20 PM
Dynamics to Understand Induced Seismicity D. Szafranski; B. Duan
Source Mechanisms of Hydraulic-Fracturing Induced Event

4:20-4:45 PM
Sequences in the Fox Creek Area H. Zhang; D.W. Eaton
Case Study: Fault Slip and Casing Deformation Induced by Hydraulic

4:45-5:10 PM
Fracturing in Sichuan Basin Z. Chen; L. Zhou; R. Walsh; M. Zoback
Statistical Controls on Induced Seismicity: A Physics-based Data

5:10-5:35 PM Mining Approach Integrating Seismic and Well Data S. Sinha; Y.
Wen; R.A. Pires de Lima; K. Marfurt
1:45 PM-3:05 PM, Room 351, Theme 14: Drilling and Completions Optimization II, Oral,
Session Chair: Samuel Noynaert, noynaert@tamu.edu, Texas A&M University; Session
Chair: Mel Sorrell, MSorrell@coveypark.com
Modeling of Azimuthal Gamma Ray Tools for Use in Geosteering

1:50-2:15 PM Unconventional Reservoirs H. Wang; E. Stockhausen; D. Wyatt; D.
Gulick
Monitoring Wellbore Quality in Real-Time Using a Geometrically

2:15-2:40 PM Derived Tortuosity Metric J. D'Angelo; P. Ashok; E. van Oort; M.
Shahri; T. Thetford; B. Nelson; M. Behounek
Cement Placement in Severe Doglegs and Its Impact on Well

2:40-3:05 PM
Integrity: A Numerical Assessment H. Yu; A. Dahi Taleghani
3:50 PM-5:35 PM, Room 351, Theme 14: Drilling and Completions Optimization III, Oral,
Session Chair: Samuel Noynaert, noynaert@tamu.edu, Texas A&M University; Session
Chair: Mel Sorrell, MSorrell@coveypark.com

Impact of Cyclic Pressure Loading on Well Integrity in Multi-Stage
3:55-4:20 PM
Hydraulic Fracturing M. Shahri; D. Barreda; R. Wagner; G. King
Diagnostic Applications of Borehole Hydraulic Signal Processing C.J.

4:20-4:45 PM
Clark; J.L. Miskimins; D.L. Gallegos
Enhancing Placement of Microproppant in Microfractures for

4:45-5:10 PM Increasing Stimulated Reservoir Volume in Shale Reservoirs L. Xu; K.
He; P. Lord; P. Nguyen
Is Conductivity Still Important in Unconventional Reservoirs? A Field

5:10-5:35 PM
Data Review R. Shelley; B. Davidson; K. Shah; T. Palish
1:45 PM-3:05 PM, Exhibit Hall Station A, Theme 06: The Use of Geophysical Technologies in
Unconventional Plays, ePaper, Session Chair: Erkan Ay, Erkan.Ay@shell.com, Shell;
Session Chair: Hemali Patel, Hemali.Patel@bp.com, BP
Novel 3-D Field-scale Characterization of Reservoir Fractures Using

1:50-2:15 PM Surface Seismic Data by the Double-Beam Method and Field
Applications Y. Zheng; H. Hu
Improving Marcellus Imaging Through the Use of FWI and Joint

2:15-2:40 PM
Tomographic Inversion for Velocity and Epsilon D. McCann
Quantitative Interpretation Efforts in Seismic Reservoir

2:40-3:05 PM Characterization of Utica-Point Pleasant Shale – A Case Study S.
Chopra; R.K. Sharma; H. Nemati; J. Keay
1:45 PM-3:30 PM, Exhibit Hall Station B, Theme 05: Well-Scale Modeling and Simulation,
ePaper, Session Chair: Andrew Yarotsky, Andrew.Yarotsky@bhpbilliton.com, BHP; Session
Chair: Vincent Artus, artus@kappaeng.com, KAPPA Engineering
Applications of a Novel Hybrid Model for Unconventional Reservoirs

1:50-2:15 PM
J.A. Ayoub; B. Blakey; S. Krishnamurthy; M. Thambynayagam
Multiscale Modeling to Evaluate the Mechanisms Controlling CO -

2
Based Enhanced Oil Recovery and CO Storage in the Bakken
2:15-2:40 PM 2
Formation J. Torres Rivero; L. Jin; N. Bosshart; L.J. Pekot; J.A.
Sorensen; K. Peterson; P. Anderson; S.B. Hawthorne
Impact of Cluster Spacing on Infill Completions in the Eagle Ford S.

2:40-3:05 PM
Evans; J. Magness; S. Siddiqui
Numerical Modeling and Optimization of Condensate Banking

3:05-3:30 PM Treatment in the Hydraulic-fractured Shale Gas Condensate
Reservoir W. Liu; R. Ganjdanesh; A. Varavei; W. Yu; K. Sepehrnoori
3:50 PM-4:45 PM, Exhibit Hall Station A, Theme 11: Decline Curve Analysis and Reservoir
Models II, ePaper, Session Chair: Wei Yu, yuwei127@gmail.com, Texas A&M University;
Session Chair: Srimoyee Bhattacharya, s.bhattacharya3@shell.com, Shell Oil Company
RTA-Assisted Production Forecasting in Shale Reservoir

3:55-4:20 PM
Development C. Aniemena
Integrating Model Uncertainties in Probabilistic Decline Curve

4:20-4:45 PM Analysis for Unconventional Oil Production Forecasting A. Hong; R.B.
Bratvold; L.W. Lake
3:50 PM-4:45 PM, Exhibit Hall Station B, Theme 02: Integrated Characterization of
Unconventional Reservoirs – From Outcrops to Geomodels III, ePaper, Session Chair:
Martyn Hargrave, mmhargrave@ikonscience.com, Ikon Science ; Session Chair: Brian
Driskill, brian.driskill@shell.com, Shell E&P; Session Chair: Travis Loseke,
TLoseke@parsleyenergy.com, Parsley Energy
SCOOP/STACK Periphery Plays: A Multidisciplinary Approach D.

3:55-4:20 PM
Yee; G. Johnston; S. Ahmed; J. Wakter; D. Howard
Using Data Analytics to Maximize Value Within the Denver-Julesburg

4:20-4:45 PM
Basin K. Repchuk; A.P. Reimchen; D. Gregoris
3:50 PM-4:45 PM, Exhibit Hall Station C, Theme 08: Geochemistry – Applications to
Unconventionals, ePaper, Session Chair: Luis Baez, Luis.Baez@shell.com, Shell; Session
Chair: Irene Arango, irene.arango@chevron.com, Chevron; Session Chair: Steve Macalello,
smacalello@gmail.com, ConocoPhillips
Are Redox-Sensitive Geochemical Proxies Valid in Mature Shales?

3:55-4:20 PM
D.R. Lindsey; S.M. Rimmer; K.B. Anderson
Interpretation of High Resolution XRF Data From the Bone Spring

4:20-4:45 PM and Upper Wolfcamp, Delaware Basin, US B. Driskill; J. Pickering; H.
Rowe
Wednesday, July 25, 2018
8:25 AM-12:15 PM, Room 310, Theme 09: Gas Injection Projects, Oral, Session Chair: Tuba
Firincioglu, tfirincioglu@NitecLLC.com, NITEC LLC; Session Chair: Theo Mallinson,
Theodore.Mallinson@Aramcoservices.com, Aramco Services Company

Field Test of CO Injection in a Vertical Middle Bakken Well to
2
Evaluate the Potential for Enhanced Oil Recovery and CO Storage
8:30-8:55 AM 2
J.A. Sorensen; L.J. Pekot; J. Torres Rivero; L. Jin; S.B. Hawthorne; L.
Jacobson; T. Doll; S. Smith; M. Flynn
Miscible EOR Process Assessment for Unconventional Reservoirs:

8:55-9:20 AM Understanding Key Mechanisms for Optimal Field Test Design V.
Sahni; S. Liu
Huff-n-Puff Gas Injection Performance in Shale Reservoirs: A Case

Study From Duvernay Shale in Alberta, Canada H. Hamdi; C.R.
9:20-9:45 AM
Clarkson; A. Ghanizadeh; S.A. Ghaderi; A. Vahedian; N. Riazi; A.N.
Esmail
Laboratory Investigation of EOR Techniques for Organic Rich Shales

9:45-10:10 AM
in the Permian Basin S. Liu; V. Sahni; J. Tan; D. Beckett; T. Vo
Recovery Mechanisms During Gas Injection for EOR in Organic Rich

11:00-11:25 AM
Shale Reservoirs F. Tovar; M. Barrufet; D.S. Schechter
The Influence of Organic Matter on Supercritical CO Migration in

2
Organic-Rich Shales B.A. Kurz; J.A. Sorensen; S.B. Hawthorne; S.
11:25-11:50 AM
Smith; H. Sanei; O. Ardakani; J.D. Walls; L. Jin; S. Butler; C. Beddoe;
B. Mibeck
Ethane Flooding as an Alternative to CO Injection in Tight

2
11:50-12:15 PM Formation: A Bakken Case Study B.N. Yolo; H. Jabbari; W. Yue; Y.
Agbor
8:25 AM-12:15 PM, Room 320, Theme 04: Beyond Brittleness: Geomechanical
Characterization I, Oral, Session Chair: Rob Fulks, rob.fulks@weatherford.com,
Weatherford; Session Chair: Katelyn Huffman, Katelyn.Huffman@chevron.com, Chevron
Full-Field Strain Measurement on Naturally-Fractured Rocks M.

8:30-8:55 AM
Mokhtari; F. Nath; S. Jiang
Rock Dilation and Its Effect on Fracture Transmissivity H. Zhou; Q.

8:55-9:20 AM
Zhao; G. Grasselli
Effects of Porous Properties of Rock on Near-Wellbore Hydraulic

9:20-9:45 AM
Fracture Complexity Y. Feng; E.E. Podnos; K. Gray
A Novel Method for Experimental Characterization of the Poroelastic

9:45-10:10 AM Constants in Unconventional Formations D. Gokaraju; M. Aldin; S.
Govindarajan; A. Thombare; O. Abdulbaki; R. Patterson

Layered Modulus Effect on Fracture Modeling and Height
11:00-11:25 AM
Containment K. Yue; J. Olson; R. Schultz
Stimulation Mechanisms in Unconventional Reservoirs Z. Ye; A.

11:25-11:50 AM
Ghassemi; S. Riley
Hydraulic Fracture Propagation in a Vertically and Laterally

11:50-12:15 PM Heterogeneous Stress Media in the Permian Basin L. Cruz; G. Izadi;
C. Barton; T. Hoeink; B. Elliott
8:25 AM-12:15 PM, Room 322, Theme 03: Physical Properties of Low-Permeability Rocks,
Oral, Session Chair: Robert Hurt, robert.hurt@pxd.com, Pioneer Natural Resources ; Session
Chair: Katerina Yared, katerina.yared@gmail.com, RWTH AACHEN University
A Novel Methodology for Mercury Intrusion Porosimetry Analysis,

8:30-8:55 AM Data Reduction, Blank Correction, and Interpretation for Shales K.E.
Gorynski; T. Dewane; T. Smagala; M.H. Tobey; J. Mansoori
Application of Integrated Core and Multiscale 3-D Image Rock

Physics to Characterize Porosity, Permeability, Capillary Pressure,
8:55-9:20 AM and Two- and Three-Phase Relative Permeability in the Codell
Sandstone, Denver Basin, Colorado A.P. Byrnes; S. Zhang; L.
Canter; M.D. Sonnenfeld
A New Method for Quantifying Cation Exchange Capacity: Application

9:20-9:45 AM
to Organic-Rich Mudrock Formations K. Cheng; Z. Heidari
Impacts of Thermal Maturity and Geochemical Properties on

9:45-10:10 AM Wettability of Kerogen and Organic-rich Mudrocks A. Jagadisan; Z.
Heidari
Characterization of Transport Properties of Shale Using Novel

11:00-11:25 AM Theoretical Pore-scale Dendroidal Model D. Zheng; Y. Zapata; Z.A.
Reza
Probing the Wettability of Mudrocks at the Pore-scale Using

11:25-11:50 AM
Nanoparticle Tracers C.J. Landry; M. Prodanović; K. Mohanty
Measurement and Analysis of Wellbore Micro-Losses and Rock

Properties While Drilling: A Novel Approach to Identification of
11:50-12:15 PM
Fractures in the Osage and Meramec Formations of Anadarko Basin
B. Chiniwala; A. Palakurthy; I. Easow; E. Russo
8:25 AM-12:15 PM, Room 330, Operators' Forum – Operating in the Permian, Oral, Session
Chair: Brendan Elliott, brendan.elliott@dvn.com, Devon Energy; Session Chair: Jonah Amini,
jamini@parsleyenergy.com, Parsley Energy; Session Chair: James McCulla,
james.mcculla@pxd.com, Pioneer Natural Resources
Life Isn’t Perfect: The Petrophysical Difficulties of Modeling the

8:30-8:55 AM Permian T. Croft; A. Blount; M. Durand; S. Warneke; A. McMullen; B.
Driskill
What Can Good Data Do For You? Machine Learning Applied to

8:55-9:20 AM Completion Design Optimization A. Sommer; R. Flumerfelt; J.
Parkhurst
Completion Design and Optimization Program Overview in the

9:20-10:10 AM
Permian Basin O. Jaripatke; G. Barzola; R. Flumerfelt
8:25 AM-12:15 PM, Room 332, Theme 07: Big Data and Artificial Intelligence Are Rapidly
Changing the Oilfield – Don't Be Left Behind, Oral, Session Chair: Lee Geiser,
lee.geiser@petrolink.com, Petrolink Services Inc; Session Chair: Sathish Sankaran,
Sathish.Sankaran@anadarko.com, Anadarko Petroleum; Session Chair: Jichao Han,
jichao.han@anadarko.com, Anadarko
Rate of Penetration (ROP) Modeling Using Hybrid Models:

8:30-8:55 AM
Deterministic and Machine Learning C. Hegde; C. Soares; K. Gray
Identification and Evaluation of Viscoelastic Surfactants Including

Smart Viscoelastic Systems for Generation and Stabilization of Ultra-
8:55-9:20 AM Dry N and CO Foam for Fracturing Fluids and Proppant Transport
2 2
S. Alzobaidi; M. Lotfollahi; C. Lu; M. Bloom; X. Zhang; M. Prodanovic;
K. Johnston; D. DiCarlo
Degradation Study on Materials for Dissolvable Frac Plugs S.

9:20-9:45 AM
Takahashi; A. Shitsukawa; M. Okura
Accelerating Well Construction Using a Digital Twin Demonstrated on

9:45-10:10 AM Unconventional Well Data in North America G.S. Saini; P. Ashok; E.
van Oort; M. Isbell
Oilfield Data Analytics: Linking Fracturing Chemistry and Well

11:00-11:25 AM
Productivity D. Khvostichenko; S. Makarychev-Mikhailov
A Fiber Optic-Assisted Multilayer Perceptron Reservoir Modeling: A

Machine Learning Approach in Prediction of Gas Production From the
11:25-11:50 AM Unconventional Reservoirs, a Case Study From the Marcellus Shale
P. Kavousi Ghahfarokhi; S. Bhattacharya; T. Carr; A. Shahkarami; J.
Elliott
Extracted Pore-Network Model for Shales Characterizing Geometry of

11:50-12:15 PM
Void Space D. Zheng; Z.A. Reza
8:25 AM-12:15 PM, Room 340, Theme 05: Permeability Measurement and Modeling, Oral,
Session Chair: Baosheng Liang, bliang@chevron.com, Chevron ; Session Chair: Wei Yu,
yuwei127@gmail.com, Texas A&M University; Session Chair: Ryan Hassen,
ryan.hassen@apachecorp.com, Apache Corporation
Geological Controls on Liquid Hydrocarbon Permeability of Tight Oil

8:30-8:55 AM and Liquid-Rich Gas Reservoirs A. Ghanizadeh; C. Song; A.
Vahedian; C.R. Clarkson
Impacts of Kerogen and Clay on Stress-Dependent Permeability

8:55-9:20 AM
Measurements of Shale Reservoirs C. An; X. Guo; J. Killough
A Finite-difference Based, Multi-scale Approach for Electromagnetic

9:20-9:45 AM
Digital Rock Modeling M. Yu; Y. Wang; J. Chen; J. Chen
Determination of Shale Matrix Permeability Through Methane

9:45-10:10 AM Dynamic Production Experiments Using Variable Pressure Gradients
K. Fan; Y. Li; D. Elsworth; M. Dong; H. Yu; C. Yin; Y. Li
Use of Rate-Transient Analysis Techniques for Evaluating

Experimental Core Permeability Tests for Unconventional Reservoirs
11:00-11:25 AM
A. Vahedian; C.R. Clarkson; A. Ghanizadeh; B. Zanganeh; C. Song;
H. Hamdi
Integrated Effects of Pore Volume Compaction and Connectivity Loss

11:25-11:50 AM on Intrinsic Permeability of Shale Samples D. Davudov; R.G.
Moghanloo
A Validated Digital Rock Workflow to Accurately Predict Apparent

11:50-12:15 PM Permeability in Tight Rocks J.F. Bautista; D. Freed; B. Crouse; G.
Balasubramanian; H. Cheng; R. Zhang; C. Ghodke
8:25 AM-12:15 PM, Room 342, Theme 06: Geophysical Reservoir Characterization in
Unconventional Plays, Oral, Session Chair: Andrew Munoz, amunoz615@gmail.com,
Newfield; Session Chair: Scott Singleton, scott.singleton@comcast.net, Independence
Resources Management
Considerations in Azimuthal Processing and Velocity Inversion for

8:30-8:55 AM
Unconventional Plays M.J. Perz; W. Keller; V. Kriechbaum
Unconventional Play Fracture Characterization Through

8:55-9:20 AM Orthorhombic Depth Model Building G. Hilburn; A. Pendharkar; W.
Keller; R. Mott; J. Peinado; A. Jumper; V. Kriechbaum
Steps for Improving the Utility of Land Seismic Data for

9:20-9:45 AM
Unconventional Reservoirs C. Stork
Coherence Attribute Applications on Seismic Data in Various Guises
9:45-10:10 AM
S. Chopra; K. Marfurt
Fracture Productivity Prediction Considering Natural Fracture

11:00-11:25 AM Formation Proximal to Fault Damage Zone T. Ramsay; L. Hernandez;
J. Li; M. Erdogan
Time-Lapse Petro-Elastic and Seismic Modeling to Evaluate

11:25-11:50 AM Fracturing Efficiency in Low-Permeability Reservoirs M. Alfi; Z. Chai;
A. Pradhan; T. Ramsay; M. Barrufet; J. Killough
First Unconventional Play From Peruvian Northeast: Muerto

11:50-12:15 PM Formation W. Morales Paetan; A. Arguedas; J. Rodríguez; H. Taipe;
J. Porlles
8:25 AM-12:15 PM, Room 351, Theme 11: Decline Curve Analysis and Reservoir Models I,
Oral, Session Chair: Rick Walker, rick.walker@bhpbilliton.com, BHP; Session Chair: Leo
Pirela, leo.pirela@vplusenergy.com, VPlus Energy LLC; Session Chair: Frank Male,
frank.r.male@gmail.com, University of Texas at Austin
A Physical Decline Curve for Fractured Horizontal Wells V. Artus; O.

8:30-8:55 AM
Houzé
Variation of Hyperbolic-b-parameter for Unconventional Reservoirs,

8:55-9:20 AM and 3-Segment Hyperbolic Decline Curve Model S. Varma; H.
Tabatabaie; J.R. Ewert; L. Mattar
Criteria for Proper Production Decline Models and Algorithm for

9:20-9:45 AM
Decline Curve Parameter Inference P. Zhou; Y. Pan; H. Sang; J. Lee
Variable Exponential Decline – Modified Arps to Characterize

9:45-10:10 AM Unconventional Shale Production Performance I. Gupta; C. Rai; C.
Sondergeld; D. Devegowda
A Model-Based Diagnostic Workflow for Time-Rate Performance of

11:00-11:25 AM
Unconventional Wells D.S. Fulford
Straightforward Representative Fluid Flow Models for Complex

11:25-11:50 AM
Fracture Networks in Unconventional Reservoirs J.A. Acuna
A Novel Production Forecast Model for Hydraulic Fractured Wells

11:50-12:15 PM Based on Anomalous Transport Phenomenon due to the Fractal
Geometry of the Fracture Networks S. Liu; H. Li; P.P. Valko
9:40 AM-11:25 AM, Exhibit Hall Station A, Theme 03: Quantification and Evaluation of
Reservoir Quality in Unconventional Reservoirs, ePaper, Session Chair: Usman Ahmed,
theenergysource@yahoo.com, WellDog; Session Chair: Kanay Jerath,
Kanay.Jerath@anadarko.com, Anadarko; Session Chair: Baosheng Liang,
bliang@chevron.com, Chevron
Advances in Borehole Imaging in Unconventional Reservoirs M.

9:45-10:10 AM
Morys; S. Knizhnik; A.R. Duncan; B.E. Tingey
Laboratory Evaluation of Apparent Conductivity of UFP D. Hu; J.C.

10:10-10:35 AM
Montalvo; U. Inyang; R. Dusterhoft; M. Apostolopoulou
Causes of Resistivity Reversal in the Vaca Muerta Formation,

10:35-11:00 AM
Argentina A.C. Ortiz; C. Bernhardt; F. Tomassini; S.P. Cumella
Natural Fractures, Fracture Facies, and Their Applications in the Well

Completion – Case Studies From the Permian Wolfcamp Formation,
11:00-11:25 AM
Midland Basin, West Texas, USA B. Li; J. Wan; P. Lascelles; A.
Coker
9:40 AM-11:00 AM, Exhibit Hall Station B, Theme 04: Geomechanics – Hydraulic Fracture
Simulation II, ePaper, Session Chair: Ahmad Ghassemi, ahmad.ghassemi@ou.edu; Session
Chair: Xiaowei Weng, xweng@slb.com, Schlumberger
Microscale Laboratory Studies for Determining Fracture Directionality

9:45-10:10 AM in Tight Sandstone and Shale During Hydraulic Fracturing M.A. Ante;
G. Manjunath; B. Jha; F. Aminzadeh
Geomechanical Investigation of Fracture Hits and Its Implications on

10:10-10:35 AM
Well Integrity and Productivity P. Pankaj
Micromechanical Modeling of Hydraulic Fracturing in Kerogen Rich

10:35-11:00 AM
Shales Y. Fang; Y. Han
9:40 AM-12:15 PM, Exhibit Hall Station C, Hydraulic Fracture Test Site (HFTS) Special
Session II, ePaper, Session Chair: Kent Perry, kent.perry@gastechnology.org, Gas
Technology Institute; Session Chair: James Courtier, jcourtier@laredopetro.com, Laredo
Petroleum
Downhole Microseismic Mapping of More Than 400 Fracturing Stages

on a Multiwell Pad at the Hydraulic Fracturing Test Site (HFTS):
9:45-10:10 AM
Discussion of Operational Challenges and Analytic Results N.A.
Stegent; C. Candler
Using Stage Level Microseismic Analysis to Gain Insight Into Fracture
10:10-10:35 AM Efficiency and Completion Effectiveness R. Fairfield; I.H. Wang; D.
Gray; J. Courtier; S. Lee
Using Stage Level Microseismic Analysis to Correlate and Ground

10:35-11:00 AM Truth Cored Hydraulic Fractures J. Wicker; J. Courtier; T. Campbell;
S. Lee; R. Fairfield; S. Trowbridge
Surface Seismic Monitoring of Hydraulic Fracturing Test Site (HFTS)

11:00-11:25 AM in the Midland Basin, Texas A. Kumar; K. Chao; R.W. Hammack; W.
Harbert
Microseismicity Analysis for HFTS Pad and Correlation With

11:25-11:50 AM
Completion Parameters D. Maity
Environmental Monitoring of the Hydraulic Fracture Test Site (HFTS)

11:50-12:15 PM
S. Eisenlord; T. Hayes
1:45 PM-3:30 PM, Room 310, Theme 12: Emerging Plays in North America, Oral, Session
Chair: David Hume, david.hume@corelab.com, Core Laboaratories; Session Chair: Matthew
Poole, matthew.poole@shell.com, Shell Exploration and Production
Review of the Bone Spring Hybrid Play in the Delaware Basin K.

1:50-2:15 PM
Schwartz; A. Starr; H. Meier; N. Stolte
Predicting Success in the Haynesville Shale: A Geologic, Completion,

2:15-2:40 PM
and Production Analysis B. Johnston
The Niobrara Formation in the Southern Powder River Basin,

2:40-3:05 PM Wyoming: An Emerging Giant Continuous Petroleum Accumulation S.
Sonnenberg
Regional Appraisal of Shale Resource Potential Within the Permian,

3:05-3:30 PM Anadarko, and Arkoma Basins: How Does the Alpine High Stack Up?
A. Bromhead; T. Butt
1:45 PM-3:30 PM, Room 320, Theme 10: Flowback and Artificial Lift for Unconventional
Reservoirs II, Oral, Session Chair: Johannes Alvarez, johalvarez@chevron.com, Chevron;
Session Chair: Jennifer Gujral, Jennifer.Gujral@shell.com, Shell
Diagnosing the Health of Your Well With Rate Transient Analysis D.

1:50-2:15 PM
Anderson; J.M. Thompson; H. Behmanesh
Artificial Lift Selection and Its Applications for Deep Horizontal Wells
2:15-2:40 PM
in the Unconventional Reservoirs P. Pankaj
Use of Chemical Tracers Reveals Details of Cleaning of a Non-

2:40-3:05 PM Conventional Gas Well in Vaca Muerta Shale J.V. Ramirez; D.
Garcia; D. Ceccon; C.D. Ferlaza
Multifunctional Surfactant Provides Superior Post Frac Production by
3:05-3:30 PM Enhancing Polymer and Load Fluid Recovery R.C. Plasier; J.
Delorey; K. Cooney; J. Leguizamon; C. Thomson
1:45 PM-3:30 PM, Room 322, Theme 08: Produced Fluid Geochemical Surveillance –
Drained Rock Volume, Oral, Session Chair: Olivia Woodruff,
olivia.woodruff@kimmeridgeenergy.com, Kimmeridge Energy; Session Chair: Matt
Laughland, Matt.Laughland@pxd.com, Pioneer Natural Resources
Fluid Heterogeneity for Tight Unconventionals on a Well-Box Scale

1:50-2:15 PM
C.H. Whitson; F. Alqahtani; E. Chuparova
Production of Migrated Oil From Horizontal Wells Drilled Into the

2:15-2:40 PM
Eagle Ford Formation on the San Marcos Arch A.S. Kornacki
Permian Basin Petroleum Systems—Geochemical Insight Into

2:40-3:05 PM Hydrocarbon Generation, Migration, and Well Performance J.B.
Curtis; J.E. Zumberge
Insights From Stable Isotope Geochemistry Surveillance in the

3:05-3:30 PM
Unconventional Horn River Basin Play G. Norville
1:45 PM-3:30 PM, Room 332, Theme 07: Surveillance of Unconventional Production and
Rock Physics, Oral, Session Chair: Sathish Sankaran, Sathish.Sankaran@anadarko.com,
Anadarko Petroleum; Session Chair: John Thompson, jthompson@atreservoir.com,
Anderson Thompson Reservoir Strategies
Rapid Reservoir Modeling With Automated Tops Correlation C. Grant;

1:50-2:15 PM
W.M. Bashore; S. Compton
Stimulation Performance Indicators and Machine-Learning-based

2:15-2:40 PM Analytics in the Utica shale: Case Study and Lessons Learnt S.
Perrier; A. Delpeint; A. Shrestha; Z. Shawuti
Unlocking Reservoir Potential With a Multilayer Inversion Technique

2:40-3:05 PM From a Directional Resistivity Tool T. Rathmann; P. Lemay; A.
Nandlal; J. Gremillion; M. Flowers
Novel Diversion Case Study for Improved Near-Wellbore Connection

3:05-3:30 PM Between Wellbore and Hydraulic Fracture W.P. Scanlan; K.
Pierskalla; D.W. Sobernheim; R. Boehringer
1:45 PM-3:30 PM, Room 340, Theme 05: Fluid Flow – Fracture Simulation and
Geomechanics, Oral, Session Chair: Craig Cipolla, ccipolla@hess.com, Hess Corporation;
Session Chair: Wei Yu, yuwei127@gmail.com, Texas A&M University

Simulation of Proppant Transport in Foam Fracturing Fluid Based on
1:50-2:15 PM
Experimental Results K. Mohanty; S. Tong; M. Gu; R. Singh
Multi-Physics Pore-Scale Modeling of Particle Plugging Due to Fluid

2:15-2:40 PM Invasion During Hydraulic Fracturing Y. Zapata; X. Dong; T.N. Phan;
Z.A. Reza
Understanding the Mechanism of Interwell Fracturing Interference

2:40-3:05 PM Based on Reservoir-Geomechanics-Fracturing Modeling in Eagle
Ford Shale X. Guo; K. Wu; J. Killough; J. Tang
Modeling of Fluid Injection in Depleted Parent Wells to Minimize

3:05-3:30 PM
Damage Due to Frac-Hits D.P. Gala; R. Manchanda; M. Sharma
1:45 PM-3:30 PM, Room 342, ARMA (American Rock Mechanics Association): Principles,
Simulation, and Practice, Oral, Session Chair: John McLennan, jmclennan@egi.utah.edu,
University of Utah
The EGS Collab Project: A Field Stimulation Study in Crystalline Rock

1:50-2:15 PM
to Validate Models D.A. Blankenship; T. Kneafsey
Completion Engineer for a Day: How Geology and Geomechanics

2:15-2:40 PM
Can Influence Completion Designs in Unconventionals N. Nagel
Modeling of Hydraulic Fracture Height Growth Through Weak

2:40-3:05 PM
Interfaces X. Weng
The Formation and Properties of Complex Fracture Networks in

3:05-3:30 PM
Shales M. Sharma
1:45 PM-3:30 PM, Room 351, Theme 07: Advanced Materials and Chemistry, Oral, Session
Chair: Susan Nash, snash@aapg.org, University of Oklahoma; Session Chair: David Livasy,
dlivasy@gmail.com, Parsley Energy
Quantitative Mineralogy of Vaca Muerta and Alum Shales From Core

Chips and Drill Cuttings by Calibrated SEM-EDS Mineralogical
1:50-2:15 PM Mapping C.I. Fialips; B. Labeyrie; V. Burg; V. Maziere; Y. Muneral; H.
Haurie; I. Jolivet; R. Lasnel; J. Laurent; L. Lambert; L. Jacquelin-
Vallee
Development of a Mixed Polymer Fracturing Fluid for High

2:15-2:40 PM Temperature Applications T. Almubarak; J. Ng; K. Sokhanvarian; H.
Nasr-El-Din; M. Khaldi
Enhancing Friction Reducer Performance in High Salt Conditions B.

2:40-3:05 PM
Seymour; A. Sanders; D. Friesen
Can Friction Reducers Transport Sand During Fracturing Treatment?

3:05-3:30 PM
L. Shen; L. Vigderman; D. Heller; D. Fu
Opening Remarks
1
S. Winberg; ;
1. Office of Fossil Energy, U. S. Department of Energy, Washington, DC, United States.
Objectives/Scope: N/A
Methods/Procedures/Process: N/A
Results/Observations/Conclusions: N/A
Applications/Significance/Novelty: N/A
Re-Invent, Re-Tool, Re-Imagine: Finding Success in the Resources Arena
1
V. A. Hollub; ;
1. Occidental Petroleum , Houston, TX, United States.
Enigmatic Shale
1
S. W. Tinker; ;
1. BEG, Austin, TX, United States.
What the Investor Community Wants From the Unconventional Fracocene
1
R. Brackett; ;
1. Bernstein Research, Washington, DC, United States.
Reservoir Engineer, Shell
1
R. Cao; ;
1. Shell, Houston, TX, United States.
Vice President of Development, Paramount Resources
1
G. Stotts; ;
1. Paramount Resources, Houston, TX, United States.
Reservoir Development Manager, Alta Mesa
1
S. Geetan; ;
1. Alta Mesa, Houston, TX, United States.
Novel Pore Pressure Prediction Technique for Unconventional Resources
1 2 3
D. Yale; ; V. Swami; ; A. Perez; ;
1. Yale Geomechanics Consulting, Estes Park, CO, United States.
2. CGG Services, Calgary, AB, Canada.
3. CGG, Houston, TX, United States.
Objectives/Scope: Standard seismic/ acoustic log Pp prediction techniques developed for young sediments in offshore
basins are not very effective in unconventional shale gas/oil reservoirs. The age and lithification of shale reservoirs,
the variability in lithology, and different overpressure generation mechanisms and basin histories all lead to poor
quality predictions using standard Eaton or Bowers methods. But Pp prediction remains important in unconventional
reservoirs due to the correlation between overpressured areas and productivity, and the correlations between thermal
maturity and pore pressure.
Methods/Procedures/Process: We have developed a method that extends the theoretical basis of the Eaton and
Bowers methods to the geologic and basin history conditions of unconventional reservoirs. The method has been
developed using standard log suite along with dipole acoustic logs. Key components of the method are:
- use of a normal pressure trend section of wells and extension of that trend to deeper, overpressured zones,
- correction of acoustic velocities for lithologic and porosity variations through a rock physics model and lithology logs
- use of lith logs and the rock physics model to determine Biot alphas and thus accurate effective stresses and thus
pore pressures,
- calibration of the predicted pressure profile with measured pore pressures from DFIT and/or high quality DST data
and/or stress measurements.
Results/Observations/Conclusions:
Tests of the method in a series of wells in Wyoming’s Powder River Basin have shown it to be very effective for a
range of productive formations across the basin. The calibration factors needed to match measured data highlight how
different these older, lithified sediments, with overpressure controlled by thermal maturity and long term leakage, are
from younger sediments in offshore basins. But the effectiveness of the method over a range of wells and formations
gives credence to the theoretical underpinnings of the method. Integration of results from multiple wells is important,
especially when measured Pp comes from different wells than the predictions.
Applications/Significance/Novelty: Further correlations between the predicted pore pressures and thermal maturity &
basin history give insights into basin development and hydrocarbon generation. The method can also be extended to
seismic PPP.
Comparison of this new method to standard Eaton or Bowers predictions, to measured Pp, and to other
Unconventionals PPP techniques shows how important key aspects of the technique are for accurate PPP in
unconventional reservoirs.
Using Traditional Methods to Predict Pore Pressure in Devonian Black Shale Basins of North East British Columbia
1
S. Green; ;
1. Ikon Science, London, United Kingdom.
Objectives/Scope: In unconventional resource plays, pore pressure plays a critical role in the ability to predict fracture
behaviour for the exploitation of these plays. Traditional pore pressure prediction assumes that all shales have a
porosity/effective stress relationship. Shales in unconventional plays are uplifted, affected by chemical processes and
mechanical compaction such that porosity is not relatable to effective stress. This paper shows how the traditional
methods can be used to predict pressure using a Pressure Reference Trend (PRT) in-lieu of a Normal Compaction
Trend (NCT) as used offshore.
Methods/Procedures/Process: Building a velocity-pore pressure cross-plot (Bowers) is the best way to build a pore
pressure model in unconventionals but this technique requires a large amount of well data. The advantages of using
the Eaton or Equivalent Depth Method techniques for pressure prediction is that they are more suited to datasets with
less wells and fewer direct pressure measurements, often found in unconventional plays. The PRT is developed in the
same way as an NCT; data from several wells are used to build and calibrate the models while a few wells are
excluded to be used as a blind test. The PRT is not linked to the expected compaction behaviour of the rock (as
inferred from an NCT) but it is simply an empirical depth trend from which the pore pressure can be predicted using
industry standard formulae.
Results/Observations/Conclusions: The PRT model was calibrated to several wells and applied to a blind test well.
The Blind test well has low mudweight but a high pore pressure measurement. The PRT model matches the direct
pressure measurement and not the mudweight. This shows that wells can be drilled underbalanced without
experiencing overly detrimental wellbore stability problems. More importantly, mudweight cannot be used as a proxy
for pore pressure as it can severely underpredict the magnitude of pore pressure, which will have a significant impact
on the development of hydraulic fractures. The accuracy of the PRT is critical as it allows the magnitude of the pore
pressure to be predicted in a 3D model, which can be input in to a geomechanical model leading to more efficient
fracturing and more cost effective wells.
Applications/Significance/Novelty: This paper presents a case study showing that traditional pore pressure prediction
techniques (Eaton, Equivalent Depth) can be adapted to predict pore pressure in an unconventional play. The
advantages of using these techniques is that they are standard to most industry software, they work better than other
techniques (Bowers) in areas with less data, they work with industry standard logs (Vp, Rho), and they can be easily
translated into the 2D/3D domain for predicting pressure in inter-well regions. Accurate pore pressure is important not
only to safer, faster drilling of cleaner boreholes but also as an important input into mechanical earth models used to
design hydraulic fractures and as an input for rock physics model such as those used to predict lithology and fluid fill.
Impact of Pore Pressure Depletion on Stress Reorientation and its Implications on the Growth of Child Well Fractures
1 1
S. Agrawal; ; M. Sharma; ;
1. Petroleum Engineering, The University of Texas at Austin, AUSTIN, TX, United States.
Objectives/Scope: It is well known that hydraulic fractures in the field propagate towards regions of low pore pressure.
Fracture hits are often observed from child wells into parent wells that are depleted by production. The objective of this
study is to reconcile the different effects of pore pressure observed in the laboratory under low far-field stresses with
that observed in the field.
Methods/Procedures/Process: Experiments were conducted to show that under low far-field stresses in the laboratory,
hydraulic fractures propagate towards regions of high pore pressure. However, under high in-situ stresses in the field,
they have been observed to propagate towards depleted regions. In this paper, preferential hydraulic fracture
propagation towards a neighboring injector or producer well is investigated numerically.
We utilize our peridynamics-based hydraulic fracturing simulator to study the effects of poroelasticity on fracture
propagation. Different simulations are run to represent lab and field conditions by changing the stress state, and the
injection/production scheme. Fracture trajectory and breakdown pressures are compared with the experiments.
Results/Observations/Conclusions: Under laboratory conditions, the tensile strength of the rock is greater than the
applied stress. However, under field conditions, in-situ stresses are much greater than the tensile strength of the rock.
We explain the differences in fracture growth behavior based on two criteria – magnitude of effective normal stress
exceeding the tensile strength to cause failure in tension, and fluid pressure inside the fracture exceeding total stress
to result in a positive fracture width. The pore pressure generated in the field can be quite complex because of the
interaction between different pressure sources and sinks. It is shown that this pressure field affects the effective and
total stresses around the fracture, which in turn control the preferred direction of fracture propagation.
Applications/Significance/Novelty: We provide a clear mechanistic explanation of the role pore pressure and pore
pressure gradients play in controlling the preferred direction of fracture growth. By combining experiments with
poroelastic theory, we show that when in-situ stresses are larger than the tensile strength of the rock, fractures will
propagate towards regions of low pore pressure. However, when the far-field stresses are lower than the tensile
strength, they propagate towards regions of high pore pressure. This for the first time helps to reconcile low-stress
laboratory experiments with field observations.
Montney Key Drivers: An Integration of Multidisciplinary Data Analytics in a Low-Permeability Reservoir
1 1 1
K. Hermanson; ; M. Kwan; ; B. Papau; ;
1. RS Energy Group, Calgary, AB, Canada.
Objectives/Scope: Low-permeability reservoirs, including the Haynesville, Utica and Montney, have become the focus
of many major oil and gas producers. The Montney is Canada’s largest oil and gas producing field due to its vast size,
over 10,000 square miles, and multiple producing intervals, which combined are up to 1,000-feet thick. The Montney is
overpressured, self-sourced and is productive across all fluid windows. This case study analyzes a multidisciplinary
data set including geological and engineering data such as digital wireline logs, core and completion data to determine
the key drivers of the play.
Methods/Procedures/Process: The Montney is a quartz-rich, siltstone-dominated, tight Triassic reservoir that was
deposited in an arid environment along a passive margin slope setting; it is found at depths of 4,000 to 10,000 feet.
Areas of interest are the newer fringe Montney developments in northeast British Columbia, north of Fort Saint John,
and southeast of Grande Prairie (Gold Creek) in Alberta. To perform the analysis, a regional stratigraphic and
petrophysical model of the Montney divided the play into three zones into which all horizontal wells were allocated.
The multidisciplinary data set was then put through a multivariate analysis to determine the most influential factors in
low-permeability conditions.
Results/Observations/Conclusions: The study analyzed 2,000 wells with digital logs to produce the geological model.
In addition, completion and production data from ~10,000 wells were used to create type curves based on operator,
year and region to account for geological differences. Results of this study identify the what makes the Montney
successful, especially in the newer fringe developments, and highlight the multidisciplinary analytical approach to
characterize low permeability rocks. The ideal geological and engineering parameters that are being targeted in the
basin for each stratigraphic horizon will be discussed.
Applications/Significance/Novelty: In 2016, the Montney’s daily output of 5.2 Bcf/d accounted for 30% of Canada’s gas
production. In comparison, the Haynesville and Utica produced 5.8 and 3.6 Bcf/d, respectively. The variability in the
Montney’s fringe areas must be understood to determine what factors can drive these areas to further success. The
Gold Creek development area has less than 100 horizontal wells primarily targeting the liquids-rich Middle Montney
zone (4,000- to 6,000-feet deep). The North Montney fringe development is targeting a relatively shallow (~4,600 feet),
overpressured Upper Montney zone, with the Middle and Lower Montney as additional targets. Liquids percentages in
the Northern Fringe area are ~ 11 %, lower than Gold Creek at 20%. These factors make these fringe areas
prospective areas of interest.
Porosity and Organic Content Analysis, Bone Spring and Wolfcamp Formations
1 1 2 2
J. D. Walls; ; T. Rider; ; B. Driskill; ; M. Durand; ;
1. Ingrain - a Halliburton Service, Houston, TX, United States.
2. Shell Exploration and Production, Houston, TX, United States.
nd
Objectives/Scope: A core analysis program was conducted on a Delaware basin well in west Texas. The 2 Bone
Spring and Wolfcamp formations were selected for a core testing program. The main objective was to understand
reservoir quality, including porosity, fluid saturation, mineralogy, pore size, and pore type. An additional objective was
to compare scanning electron microscope- (SEM) based digital rock analysis (DRA) results to physical laboratory
tests, such as GRI crushed rock porosity and total organic carbon (TOC).
Methods/Procedures/Process: Plug samples were obtained for X-ray diffraction (XRD) mineralogy, pyrolysis, TOC,
DRA, and retort analysis. DRA included Fourier transform infrared (FTIR) mineralogy, argon-ion milling, SEM imaging,
image processing, and segmentation. The pore size and volume fraction of organic porosity, inter- and intra-granular
porosity, and solid organic matter were computed. Rock material from near each plug was used for Dean-Stark
extraction (Sw, So) and helium porosity. Thus, the relationships between key rock properties, such as clay content vs.
TOC, bulk volume water vs. clay content, and bulk volume oil vs. porosity associated with organic matter (PAOM),
were determined.
Results/Observations/Conclusions: In these samples, TOC was observed to be related to clay content, but not
linearly. Rather, TOC increases with clay content up to approximately 30% by weight clay, then begins to decrease.
Total clay, silica, and carbonate from XRD and FTIR were in good agreement, although XRD indicated slightly higher
clay content than FTIR. As reported elsewhere, SEM-derived porosity is generally lower than helium porosity. In
addition, the difference between the two is strongly related to total clay, suggesting that clay-bound water may be a
key factor. This is supported by GRI and XRD data that show bulk volume water is directly correlated to total clay with
a near-zero intercept.
Applications/Significance/Novelty: Understanding that a sweet spot for TOC exists at approximately 30% clay content
in these formations may help to target the best reservoir.
Although image resolution and damage to organic porosity from ion-milling are commonly provided explanations for
why SEM porosity on ion-milled samples is routinely lower than helium porosity, this work suggests a different reason:
capillary-bound water on clay mineral surfaces. This water is driven off in Dean-Stark and retort methods, thus
counted as part of total porosity. However, the SEM images do not resolve this adsorbed layer of water on clay
surfaces.
When clay-bound water volume is added to SEM porosity and plotted vs. GRI porosity, a linear fit with a slope of near
2
1 and a correlation coefficient r of approximately 0.8 is obtained.
A Water Saturation Interpretation Model for Organic-Rich Shale Reservoir: A Case Study of North Sumatra Basin
3 2, 1
B. Milad; ; M. N. Akbar; ;
1. LEMIGAS-Indonesia Research and Development Centre for Oil and Gas Technology, Jakarta Barat, DKI Jakarta,
Indonesia.
2. Petroleum Geoengineering, University of Miskolc, Miskolc, Hungary.
3. University of Oklahoma, Norman, OK, United States.
Objectives/Scope: Water saturation in the organic-rich mudstone is a crucial parameter in characterizing and
evaluating reservoir hydrocarbon in-place. The presence of organic matter and the over-complex of shale pore system
have remained a lack of understanding of how to interpret the water saturation in this reservoir by using wireline log
data. Many practitioners only used conventional methods while they are inappropriately applied due to the presence of
organic matter. This paper offers a method for evaluating organic-rich shale water saturation and applied it to a case
study of North Sumatra Basin.
Methods/Procedures/Process: In general, all interpretations are conducted by integrating the conventional and NMR
log, special and routine core analysis, petrography analysis, and geochemical analysis. We started to determine the
total and effective porosities and also the total porosity considering the presence of kerogen. These parameters were
estimated by using the conventional and NMR log and validated by core data. Then, the water saturation equation of
Simandoux model was modified by involving the weight percent of total organic carbon (TOC) while the TOC was
predicted in advance by averaging between Carbolog and Passey’s ΔlogR methods. Equally important, we used
saturation exponent (n), cementation factor (m), and the tortuosity factor (a) which obtained from formation factor and
resistivity index analysis.
Results/Observations/Conclusions: Finally, the interpretation results of water saturation under this developed method
have been made. In order to compare this technique’s result, we also applied several approaches those usually used
in many case studies (the revised water saturation-TOC, water saturation separation, and some conventional water
saturation methods). In conclusion, the water saturation from the offered method gives a reasonable fluid saturation
distribution result for both free gas and absorbed gas while the conventional methods only consider to the free gas
filled in the fracture and/or matrix pore spaces. In addition, this interpreted water saturation was used for determining
the bright-spot interval of unconventional shale by combining with geochemical and rock mechanic properties.
Applications/Significance/Novelty: The novelty of this study offers the prominent ways and effectively applied method
on how to integrate and maximize the petrophysical, geological and geochemical data in order to solve the problems
in estimating the water saturation of unconventional organic mudstone. This analysis presents the further
understanding of water saturation for better interpretation in the organic-rich shale reservoir, especially in the
application of this case study.
Spider Bots: Database Enhancing and Indexing Scripts to Efficiently Convert Raw Well Data Into Valuable Knowledge
1 1 1 1 2 2 2
P. Ashok; ; G. S. Saini; ; H. Chan; ; E. van Oort; ; M. Behounek; ; T. Thetford; ; M. Shahri; ;
2. Apache Corporation, Houston, TX, United States.
Objectives/Scope: Oil and gas industry generates large datasets that often remain unused by operators and service
companies. Key to improving data usage is converting data into information to be able to easily answer pivotal
questions on well performance enhancement, and other improvement opportunities. This is quite difficult today, as raw
data is typically not converted into information when it is stored. “Spider bot” scripts, however, can effectively pull
valuable data from sources that may be structured/unstructured, static/dynamic, etc. at predefined time intervals,
indexing and updating it in a database.
Methods/Procedures/Process: Scripts were developed for a dataset consisting of a large number of unconventional
wells. A noSQL database was used to store all the data. Three classes of spider bots were developed. First, scripts
were created to automatically curate and clean the data, a vital step. Second, scripts were written to convert the
cleansed data into information. Example of such scripts are characterizing metrics such as wellbore quality (tortuosity
index, dog leg severity etc.), drilling dysfunction type and severity, and invisible lost time. Finally, scripts were devised
to create indices and visuals based on the generated information to rank well performance using various KPIs. These
bots enable the end user to meaningfully compare well performance metrics, derive insight and knowledge quickly.
Results/Observations/Conclusions: Spider bots greatly enhance the value of original raw well datasets. To test the
spider bots after deployment, we queried the database with relevant questions. Sample questions included “What was
the best well drilled in Region A?”, “Which wellbore showed the best quality”, etc. Previously, such questions would
have taken an excessive number of analyst man hours to answer. On our system, however, these queries were
answered in mere seconds, chiefly due to indexing performed by the spider bots. When we added new well data to the
original dataset and posed the same questions again, the answers changed to reflect the new data. Another key
advantage of our approach was the ability to easily update and improve the spider bots by modifying the existing
scripts or by creating entirely new ones.
Applications/Significance/Novelty: Spider bots are programs that run in the background and constantly work towards
deriving valuable information. This is the first time such bots have been applied to drilling data. Three main categories
of bots were developed and are considered essential: data cleansing bots, data processing bots and data indexing
bots. These bots were primarily applied on drilling data, but can be applied universally on all kinds of oil and gas data,
including e.g. completions and production. They significantly facilitate and accelerate the analysis of big data. Such
bots will become a necessity as the industry moves towards trying to generate more value from big data.
Integrated Workflow for the Definition of a Type Well Using Probabilistic Methods
1
D. S. Jones; ;
1. Reservoir Engineering, Chesapeake Energy, OKLAHOMA CITY, OK, United States.
Objectives/Scope: Current methodologies for describing type wells are often biased to a producing well set from which
the analogy is described. For an evaluation team, this can be problematic, because the well set is biased to the
current producing conditions of the wells, the completion techniques tested in the dataset, the current effective well
spacing, and geologic environment that the dataset has tested.
The goal of this paper is to estabilsh a workflow methodology that utilizes ranges avaialable data to generate a
probabalistic type well that captures the range of possible outcomes from a system
Methods/Procedures/Process: This paper will set forth a methodology and workflow that will attempt to reconcile
production, physics, and uncertainties that are inherent in the production of hydrocarbons from multi-fractured
horizontal wells. This will be methodology will make use of Monte Carlo techniques to build probability distribution
functions of common inputs and reservoir engineering software that combines rate transient analysis and analytical,
hybrid, or numerical models, to better describe a type well and the range of possible outcomes in a type well areas.
Results/Observations/Conclusions: The result of this workflow is a robust procedure that is scaleble from exploration
to development programs and captures the full range of outcomes given the availability and confidence of the given
data.
Applications/Significance/Novelty: This bias created by current type well methodologies creates the over or under
prediction of flowing rates and the estimate ultimate recovery of reserves (EUR). Furthermore, the reliance on purely
empirical (production) data divorces the evaluation from the elements of physics that control that production and
ultimate recovery. This bias will then ultimately lead the destruction of value and loss of reserves due to inefficient
recovery of hydrocarbons. It is imperative that techniques be developed that can combine the physics of a petroleum
system, empirical production data, and control for the bias of that empirical well set.
Toward a Management Science for Unconventional Wells: A Methodological Approach
3 1 2
J. S. Mintz; ; R. R. Batsell; ; S. Paranji; ;
1. Jones Graduate School, Rice University, Houston, TX, United States.
2. Anadarko, The Woodlands, TX, United States.
3. Apache, Houston, TX, United States.
Objectives/Scope: Accurate multivariate modeling of production from oil and gas wells enables the testing of important
managerial questions. 1) Does target area B, which costs over $200,000 more to complete, result in enough increased
production to warrant the extra cost over target area A? 2) Does increasing lateral length result in a linear increase in
production? 3) Is there a difference in production associated with plug and perf versus sliding sleeve? By controlling
for geology and completion factors and achieving a high goodness of fit in a multivariate model, the above questions
are reliably addressed.
Methods/Procedures/Process: Based on data from 450 gas wells in the Marcellus region in Pennsylvania, multivariate
modeling was used to address several of these questions. Principal components factor analysis was used to extract
the 9 underlying dimensions from the 40 original geologic variables. The same factor analysis approach was used to
extract the 7 underlying dimensions of 20 completion variables. Based on these 16 derived dimensions, multivariate
regression was then used to model the production from the wells. Because the models were very successful, it was
possible to then conduct statistical tests of the managerial questions.
Results/Observations/Conclusions: The dependent variables in the regression modeling were 30 day, 60 day, 90 day,
180 day and 365 day cumulative production. Remarkable fits were achieved for the 180 day and 365 day cumulative
production numbers. The models fit production so well that they allowed the tests of serious managerial questions.
The target area that cost over $200,000 a well more to complete did not result in a significant difference in production.
Eliminating pursuit off the more expensive target saved the company over $40,000,000 for the next 200 wells.
Increasing lateral length was deemed to result in linear increases in production. Plug and perf resulted in significantly
greater production than sliding sleeve. Premium fracking approaches from major suppliers produced significantly
greater production.
Applications/Significance/Novelty: This paper is the first to apply principal component factor analysis to extract
orthogonal underlying dimensions from a large group of highly co-linear geologic and completion variables and then
use those resulting dimensions to model production in a multivariate framework. The orthogonal independent
variables allow the reliable designation of the importance of each contributing variable. A second novel result was
empirical in nature: over 70% of the variance in production for 180 day cum and 365 day cum was explained by the
models. This very strong goodness of fit allowed reliable testing of important managerial and cost questions. The
accumulated experience from the work lead to a series of suggestions on completion specifications that can lead to
well optimization of production.
Emerging Shale Oil Plays in Hypersaline Lacustrine Qianjiang Formation, Jianghan Basin, Central China
1 1, 2
M. Li; ; X. Ma; ; T. Cao; 1; G. Tao; 1; Z. Li; 1; Q. Jiang; 1; S. Wu; 3;
1. China State Key Laboratory of Shale Oil and Shale Gas Resources and Effective Development, Sinopec Petroleum
Exploration and Production Research Institute, Beijing, China.
2. China University of Petroleum (Beijing), Beijing, China.
3. Sinopec Jianghan Oilfield Company, Wuhan, Hubei, China.
Objectives/Scope: Organic-rich lacustrine shale may be the holy grail for petroleum industry in China. One remarkable
feature of lacustrine shales is the high spatial heterogeneity compared to the marine shales. Recognizing and
characterizing the spatial heterogeneity of the lacustrine shales are therefore essential to successful exploration for
and exploitation of hydrocarbon resources in lacustrine basins. This presentation reports the key geological elements
of an emerging hypersaline lacustrine shale play in Eocene Qianjiang Formation, Jianghan Basin, China.
Methods/Procedures/Process: To help understand such a shale oil play, we studied seven production test oil samples,
and three continuous cores of Qianjiang Formation organic rich shales. We employed an integrated suite of high-
resolution analytical techniques to determine quantitative mineralogy, measure quantity, quality and thermal maturity
of organic matter, acquire detailed molecular marker data from saturate and aromatic hydrocarbon fractions of solvent
extracts, describe and classify porosity, and develop a diagenetic and oil accumulation history of the Qianjiang
Formation in the region.
Results/Observations/Conclusions: The Qianjiang Formation shales are minerologically complex. The inter-salt rocks
are dominated by calcite or dolomite, whereas the intra-salt shales are often associated with sporadic glauberite. All of
the inter-salt organic-rich intervals meet the requirements for commercial shale oil reservoirs. The generally high
porosity (7-20%) and Hydrogen Index (300-950 mgHC/gTOC) indicate the rocks are currently at low thermal maturity.
Based on the large difference in the measured rock porosity before and after sequential solvent extraction, most of the
porosity in the Qianjiang shales appears to be associated with conventional rock matrix.
Applications/Significance/Novelty: Results of this investigation provide useful constraints for the likely models of inter-
salt shale oil accumulation in the Jianghan Basin. In regions where structural traps were formed as a result of salt
tectonic development, up-dip lateral oil migration along variably deformed shallow shale laminates may be conducive
for hybrid shale oil/conventional oil systems. In thermally mature source kitchens and the deeper Eq4 section, in
contrast, continuous oil accumulation in structurally undisturbed areas may produce even more attractive
unconventional drilling targets. In both cases, the presence of halite, organic rich and relatively brittle laminated shale,
suitable maturity and early oil emplacement are among the key elements for the self-contained, inter-salt, petroleum
system.
Unconventional Reservoir Development in Egypt’s Western Desert: Lessons Learned From the First Appraisal Wells
2 1
M. H. Ibrahim; ; M. Salah; ;
1. Shell, Cairo, Egypt.
2. Apache, Houston, TX, United States.
Objectives/Scope: The paper will discuss the journey of developing first unconventional gas in Egypt. The
transformation of the development concept through successful appraisal and pilot phases will be discussed,
emphasizing the key learnings and outcomes from the first pilot project. The integration of early results from the pilot
into some stochastic techniques to evaluate a range of EUR/well and range of production forecasts will also be
discussed. This paper will review of results and share lessons learned related to the recent appraising and
development activities of unconventional plays in Egypt.
Methods/Procedures/Process: The country has launched studies to appraise several prospects for unconventional gas
in Shoushan and AbuGharadig basins. The search is characterized by shale gas at Khatatba source rock in Shoushan
basing and tight Chalk gas in AbuGharadig basin. Exploratory 6 vertical and 3 horizontal pilot data wells were drilled
and completed via multistage fracturing in the appraisal program for collecting the required data to evaluate reservoirs
qualities, demonstrate the reserves availability, and identify optimal technology to maximize productivity and set the
foundation for future development of these unconventional plays. These understandings helped reduce uncertainty
and the lessons learned and findings out from this work helped define optimum completion designs for optimal
development strategy.
The results from this study will shed light on the results of recent unconventional gas appraisal and development
activities, which indicate that the western desert of Egypt holds substantial resources of unconventional gas. This
unconventional gas can help to change the slope of production rates in the country positively and set the foundation
for future development of these plays.
Applications/Significance/Novelty:
Most of Egypt’s Western Desert reservoirs are characterized to be low permeable and reservoir quality. In the early
development stages, only layers with the high permeability were produced while the unconventional, tight and
challenging layers were not considered economic. As these high permeability layers became more mature and decline
in production, tight and unconventional reservoirs became operator’s alternative to unlock the enormous amount of
hydrocarbons retained in these rock and achieve economical production targets from these marginal fields.
Mechanism Study of Casing Deformation in Multistage Hydraulic Fracturing Shale Reservoir
1 1, 2
Z. Xu; ; F. Yin; ; S. Yang; 3; L. Han; 3; X. Wu; 1;
1. Mewbourne School of Petroleum and Geological Engineering, University of Oklahoma, Norman, OK, United States.
2. College of Energy, Chengdu University of Technology, Chengdu, Sichuan, China.
3. Tubular Goods Research Institute of CNPC, Xi'an, Shanxi, China.
Objectives/Scope: Hydraulic fracturing is the key technology to develop unconventional resources. Field practices
have shown that the large scale hydraulic fracturing could compromise the well integrity through casing deformation or
failure, which results in the premature abandonment of well completion and low well productivity and ultimate
recovery. Many factors can cause the casing deformation, and they will be firstly reviewed. this study focuses on
existing geological features such as natural fractures.
Methods/Procedures/Process: Starting from fundamental rock mechanics and fluid flow in porous media, we set up a
coupled model of casing-cement-formation system under hydraulic fracturing in Abaqus. The porous elastoplastic
constitutive relation is employed to simulate the changes of stress and flow fields during hydraulic injection. The
casing and cement sheath in injection and production wells are simplified as the pipe element embedded in formation.
The effect of formation response on casing integrity is revealed with parametric sensitivity analysis. Sensitivity factors
including fracture geometry, perforation position, the mechanical properties of casing and cement and injecting
pressure are simulated. For given casing grade and thickness, injection flow rate and pressure are recommended.
Results/Observations/Conclusions: Results indicate that formation, cement and casing stress fields redistribute after
the injection of fracture fluid. The formation deforms in both horizontal and vertical directions, which give additional
shear and stretch stress. The shear stress accumulates along the lithologic interfaces. Casing stress concentration
intensifies with the increase of injection rate and pressure. Fracturing parameters, cementing and casing strength are
optimally designed. The research findings provide a reference for understanding formation response and sustaining
well integrity. Based on the above analysis, new casing design is proposed.
Applications/Significance/Novelty: Even though casing failure during hydraulic fracturing is a relative low risk scenario,
its consequence could be significant on both economics and HSE. This work investigates the mechanism of casing
failure in hydraulic fractured wells caused by shear and stretch stress for the purpose of optimal casing design
accounting for the wellbore long term integrity and constraints of fracturing operation.
Optimization and Drilling of Horizontal Wells using a Bayesian Network
1 1 1 1
J. F. Fierstien; ; H. Winkler; ; P. Strauss; ; A. Klokov; ;
1. Factor Technology, Dripping Springs, TX, United States.
Objectives/Scope: Today’s horizontal wells have moved from technology marvels to everyday occurrence. Like all
maturing technology we are moving past accomplishment to making it as efficient as possible. Efficiency is essential,
as horizontal drilling costs remain high and oil prices remain well below their 2015 high (as of this writing). New
technologies will drastically change how wells are steered and possibly drilled are now available that that will make
drilling horizontal wells more accurate and more efficient.
Methods/Procedures/Process: Recent advancements in technology (such as described by Winkler 2017 Geophysics)
use a sophisticated Bayesian Network to better predict exactly where the formations are and also better predict where
the wellbore is in 3D space. The system uses MWD and LWD data and calculates the most likely position given the
errors inherent in today’s tools. Knowing the best likely position of the tool, the system can then make a better
prediction where it is in relation to the formation it is drilling through. The system knows where the bit is, not just where
the tool sensors are, and appropriately updates the wellbore plan and directs future drilling on the corrected
information.
Results/Observations/Conclusions: A better wellbore means a straighter wellbore with less tortuosity and elimination
of all the problems that go along with a twisted or porpoising wellbore.
Increasing the accuracy of drilling means placing the wellbore in exactly the best producing hydrocarbon zone
possible.
The system can also detect faults and determine if the desired formations are above or below the current drillbit
location. This means the wellplan can be updated on the fly and wouldn’t require human intervention. Directional
changes to the drill bit can be made in a matter of minutes. Changes outside of a given tolerance would trigger a
notification but anything inside the tolerance would adjust and continue as normal.
Applications/Significance/Novelty: Autonomous drilling might be closer than we think, with this new system able to
determine where the bit is in relation to formations, detect faults and readjust the well plan automatically without any
human intervention. Additionally the technology monitors drilling and alerts staff when formations are not where they
are anticipated or something is outside a predetermined tolerance. This allows staff to do more and intercede only
when it is critical for a human to make an interpretation call.
This new technology does a much better job determining where the drill bit is and where it is going so corrections can
be made sooner, creating a wellbore that is less prone to porpoising and other over corrections.
TST3D: Automated Structural Interpretation in Horizontal Wellbores
1 1 2 1
T. Zhang; ; D. McCormick; ; A. Nandlal; ; M. LeFranc; ;
1. Reservoir Geosciences, Schlumberger-Doll Research, Cambridge, MA, United States.
2. Schlumberger, Denver, CO, United States.
Objectives/Scope: 1. Objective
Horizontal wells dominate the development of unconventional shale reservoirs. Using real time drilling data to steer in
a target zone is the key to economic success. Today structural interpretation in horizontal wells is a manual process
that is time consuming, tedious, and error-prone, especially since gamma-ray (GR) logs are commonly the only
available logging-while-drilling (LWD) data in unconventional horizontal wells. For the first time, we have developed a
method to automate interpretation of subsurface structure.
Methods/Procedures/Process: 2. Methodology and Novelty
True Stratigraphic Thickness in 3D (TST3D) automates structural interpretation using pattern recognition. Given an
initial structural model, TST3D automatically computes TST as the shortest distance from each wellbore survey
location to the initial surface and then matches GR patterns in the horizontal well to those seen in a pilot well in TST
space. TST3D inserts fold hinges, bends the structure, then recomputes the model GR response, progressively
matching the pilot well log signature, from heel to toe in the horizontal well. We make two assumptions in the current
TST3D: constant layer thickness across the drilled interval and GR variation follows structural layers, which are
reasonable in most shale plays.
Results/Observations/Conclusions: 3. Applications
TST3D can be applied in either a post-drill mode for structural interpretation or real-time mode for automated
geosteering. Field tests in different shale plays and complex well trajectories demonstrate that TST3D runs quickly: a
structural model of a 10,000-foot horizontal section can be computed in minutes, and a real-time update of 100 feet of
new data takes less than a minute. TST3D eliminates manual insertion of fold hinges and adjustment of formation
dips. It automatically builds 3D near-wellbore structural models using true formation dips and azimuths, creating an
optimal structure that matches observed well measurements with minimal user interaction. The final computed surface
can be used to update existing 3D models.
Applications/Significance/Novelty: 4. Significance and Conclusions
Automating the geosteering correlation process would allow well placement engineers to cover multiple wells
simultaneously, increasing the efficiency of the team while maintaining service quality. Beyond structural
interpretation, we believe that TST3D has great potential to contribute to the digital transformation of formation
evaluation and drilling automation.
Oak Ridge National Laboratory
1
Y. Polsky; ;
1. Oak Ridge National Laboratory, Houston, TX, United States.
Los Alamos National Laboratory
1
R. Pawar; ;
1. Los Alamos National Laboratory, Los Alamos, NM, United States.
Lawrence Berkley National Laboratory
1
T. Kneafsey; ;
1. Lawrence Berkely National Laboratory, Houston, TX, United States.
Senior Fellow, Geologic and Environmental Systems
1
G. S. Bromhal; ;
1. National Energy Technology Laboratory, Washington, DC, United States.
Reliable Solid Organic Matter Thermal Maturity Assessment Using Surface Enhanced Raman Spectroscopy and Case
Studies
1 1 1 1 1 1
Y. Tang; ; C. Jiang; ; L. Gao; ; S. Wu; ; J. Shaw; ; A. Bishop; ;
1. Power Environmental Energy Research Institute, Covina, CA, United States.
Objectives/Scope:
In this study, we present a Surface Enhanced Raman Spectroscopy (SERS) technique to measure maturity of solid
organic matter and report the application to the Anadarko Basin.
Methods/Procedures/Process:
Coal and kerogen samples were first pulverized to powder, and an aqueous solution of Ag nanoparticles was
prepared and applied to the powders. The sample mix were then analyzed under the Raman Spectroscopy with
optimal laser light intensity and duration. Comparison of SERS results and known maturity values of different samples
(coal vs. kerogen) was performed to derive the calibration curves by kinetics modeling.
Our results of coal and kerogen calibration sets show that organofacies impacts Raman responses with increasing
maturity for the two different types of organic matter. Application to the Woodford Formation in the Anadarko Basin
shows that SERS results are larger than vitrinite reflectance equivalent values based on the reflectance of solid
hydrocarbons (%VRE) to different extents. The SERS data also showed a more consistent trend with increase in well
depth and well fluid data such as American Petroleum Institute (API) gravity and/or gas to oil ratio (GOR).
Source rock maturity is an essential parameter for petroleum system understanding. Direct assessment is often
problematic with conventional methods such as vitrinite reflectance, due to lack of ‘true’ vitrinitic material,
heterogeneity and/or lack of effective calibration for available proxies, etc. Our results have demonstrated the ‘proof of
concept’ of the SERS method for maturity determination, and provided major constraints on the interpretation of
generation, cracking and migration of hydrocarbons.
Integrated Inorganic and Organic Geochemistry Approach in the Petroleum Systems Analysis of Permian Shale Plays
1
C. Gong; ;
Objectives/Scope: This study is to use an integrated inorganic and organic geochemistry method to characterize
Permian black shale rocks in Midland Basin, to understand controls on organic carbon richness, such as primary
productivity, depositional environments, sediment supply, and bottom water preservation conditions, and its
implications on petroleum generation/charge and on shale oil development.
Methods/Procedures/Process: A high frequency sampling of core samples from a thick sequence (~1000 ft) of
argillaceous mudstones, organic-rich mudstones, siliceous mudstones, and carbonate mudstones in Midland basin
were characterized for elemental concentrations and source rock geochemistry, and for high resolution gas
chromatography and biomarkers on extracts. The small maturity differences (<0.05 VRo% ) from the top to the bottom
of this 1000 ft section and limited migration of hydrocarbons into the rock pore space due to low permeability and high
capillary entry pressure allows us to interpret the extract geochemistry fingerprints to compositions of the source rock
kerogen and its in-situ generated bitumen.
Results/Observations/Conclusions: Geologic, petrophysical, and elemental analysis have divided the section into
many depositional packages (chemozones) with distinct signatures. This has suggested cycles of parasequences with
varied sediment (clastic vs carbonate) supply and preservation conditions. Source rock characters respond to these
depositional environmental variations with changes in total organic carbon contents and HI/OI values. Depositional
environment dependent biomarker parameters (Pr/Ph, DBT/Phen, etc.) from core extracts also show systematic
changes reflecting variations in bottom water oxygen conditions and source rock facies.
Applications/Significance/Novelty: This integrated approach greatly enhances our description of Permian age resource
play petroleum systems, which in turn helps with sweetspot mapping of the most prolific play area and the selection of
lateral landing zones.
Geochemical Characterization of the Eagle Ford Formation in Northeast Mexico
1 1 1 1
S. Ortega-Lucach; ; L. Gutierrez-Caminero; ; R. Torres-Vargas; ; G. Murillo-Muñetón; ;
1. Instituto Mexicano del Petróleo, Mexico City, Mexico.
Objectives/Scope: The Eagle Ford Formation is a well-known shale gas/oil unconventional reservoir in the United
States. The purpose of this study was to evaluate the hydrocarbon potential and thermal maturity of Eagle Ford
Formation in the Galaxia Area, northeast Mexico (Coahuila State).
Methods/Procedures/Process: The geochemical characterization was performed using Rock Eval pyrolysis results,
vitrinite reflectance data and optical studies of organic matter, from well cores and cutting samples from twelve wells
within the Galaxia Area. Pemex provided the set of information of these wells. Source rock potential and thermal
maturity were assessed in order to identify the sweet spots in the area.
A detailed description of two continuous analogous outcrop coring, constrained the variation of organic content along
the Eagle Ford Formation. The two cores comprise a complete record of the Eagle Ford Formation with thickness of
47.5 m and 86.8 m, respectively.
Results/Observations/Conclusions: Lower Eagle Ford Formation (LEF) presents higher TOC values (0.12-7.68%) than
the Upper Eagle Ford Formation (0.22-3.53% TOC). Furthermore, the higher TOC values (up to 7%) in the LEF are
found in the southwest and northwest portions of the Galaxia Area, which is related to the deep marine facies
identified in those regions. In contrast, minor TOC values (ca 1.7%) are recognized in the central portion of the study
area, where the outer platform facies have been interpreted for the LEF. Remnant kerogen correspond to Type II/III
while Tmax and %Ro values indicate that the Eagle Ford Formation exhibits a thermal maturity ranging from the oil
window to the dry gas window (0.6-2.0 %Ro eq.) showing a maturity trend increasing towards the southeast portion of
the Galaxia Area.
Applications/Significance/Novelty: This study confirmed the relevance of the Eagle Ford Formation as an
unconventional reservoir in northeastern Mexico, and allowed to identify the sweet spots in the Galaxia Area. In
addition, the detailed geochemical characterization performed, provided essential data that allowed to calibrate the
unconventional petroleum system modelling of the Eagle Ford Formation in the Galaxia Area, Mexico.
Using Drone Magnetic and LiDAR Surveys to Locate Unmarked, Abandoned Wells Prior to Unconventional Oil and
Gas Development
1 1 1
R. W. Hammack; ; G. Veloski; ; J. Sams; ;
1. National Energy Technology Laboratory, U.S. Dept. of Energy, Pittsburgh, PA, United States.
Objectives/Scope: Unrecorded wells pose HSSE risk to unconventional development. The development of effective,
rapid, low-cost methods to locate existing wells provides operators with a tool to identify and mitigate this risk.
Methods/Procedures/Process: Wells with steel casing can be quickly located using aeromagnetic surveys, but the
high performance helicopter and skilled pilot needed for these terrain compliant, low-altitude surveys is expensive.
Wells where steel casing has been removed or where the casing is non-magnetic (ie. wood) cannot be detected by
magnetic surveys. High-resolution, airborne LiDAR surveys can look down through the tree canopy and image subtle
topographic features such as well pads that can help limit the ground seach for unrecorded wells, including wells with
no steel casing. This study uses new, light-weight magnetic sensors and high-resolution LiDAR scanners deployed on
small unmanned aircraft systems (sUAS) to quickly and inexpensively survey large areas for the presence of
unrecorded wells.
Results/Observations/Conclusions: In this study, a sUAS magnetic survey over an area containing buried wells was
compared with a manned helicopter survey of the same area flown using the same flightlines and altitude. The buried
wells were detected by both surveys, but the sUAS magnetic survey exhibited higher resolution (because of increased
sampling rate) and the cost was much less. High-resolution LiDAR surveys were flown using sUAS over an 1860s-era
oilfield known to contain wells with wood casing. The last return imagery from LiDAR surveys clearly showed
topographic evidence of where well pads, storage tanks, and flow lines had been. This knowledge limited the ground
search needed to locate unrecorded wells, saving time and associated costs.
Applications/Significance/Novelty: The Commonweath of Pennsylvania requires that operators identify existing wells
within a 1000-ft-wide buffer zone surrounding unconventional wells that are to be hydraulically fractured. This is to
mitigate for the concern that fluids and gas from the treatment well might communicate with unrecorded and possibly
unplugged wells that are the legacy of Pennsylvania's early oil and gas history
Advanced Characterization and Novel Waste Management for Drill Cuttings From Marcellus Shale Energy
Development
1 1, 2
A. Hakala; ; M. Y. Stuckman; ; H. M. Edenborn; 1; C. L. Lopano; 1;
1. Research and Innovation Center, DOE-National Energy Technology Lab, Pittsburgh, PA, United States.
2. AECOM, Pittsburgh, PA, United States.
Objectives/Scope: Millions of tons of drill cuttings generated from shale gas development are currently disposed of in
landfills, buried in-situ, or reused as road fill. However, current waste management practices only use one regulatory
test for waste evaluation and has not explored the potential for land application. This study conducted advanced solid
characterization and leaching tests on Marcellus shale drill cuttings to explore more diverse disposal environments.
The potential practical application of drill cuttings as a supplemental plant growth medium on building green roofs was
demonstrated.
Methods/Procedures/Process: Drill cutting samples with different mineralogical traits (e.g., high in pyrite and carbon,
or high in calcite) from the lateral portion of a Marcellus shale well in WV were studied. Solid characterization (e.g.,
elemental distribution and mineral composition) and geochemical leaching experiments (e.g., rainfall vs. landfill, wet
vs. dry conditions, short-term batches vs. long-term columns, and pH) were conducted to decide whether the materials
meet the USDA and USEPA standards for waste disposal and land application. The effects of added drill cuttings on
soil quality, plant germination and long-term plant growth were determined in the laboratory and on a building green
roof.
Results/Observations/Conclusions: Results show that, although drill cuttings passed the regulatory short-term
leaching tests for landfill disposal, long-term release of Ba occurred in all samples and release of As, Ni, and Sb were
possible in high pyrite and low calcite samples under simulated rainfall conditions. Selected trace metal release (e.g.,
Ba, V, Mo, Sr and Sb) increased 2-10 fold when the waste was first air-dried and subsequently underwent simulated
rainfall leaching. Residual salt content in drill cuttings was inhibitory to plant seed germination, but plant growth on
green roofs in cutting-amended soil was not inhibited once soluble salts were leached by rainfall.
Applications/Significance/Novelty: This improved characterization informs waste managers of the different
geochemical factors that control trace metal release from drill cuttings and improves the ability to assess proposed
methods of solid waste characterization and handling. There may be an environmental benefit if drill cuttings are kept
wet (as opposed to being allowed to air dry) prior to disposal in landfills as a low cost option for disposal pretreatment.
Novel commercial uses, such as the addition of drill cuttings to green roof soils to offset costs, may be feasible if the
long-term safety of the material can be demonstrated.
Estimating Carbon Intensity of Unconventional Plays
1, 2
D. N. Meehan; ;
1. Gaffney, Cline & Associates, Houston, TX, United States.
2. Baker Hughes, A GE Company, Houston, TX, United States.
Objectives/Scope: Carbon intensity of oil and gas production varies widely across global oil plays. Life cycle extraction
from some unconventional plays (e.g. tar sands) have some of the highest CIs but even many North American shale
plays have relative high carbon intensity. Flaring and venting of associated or non-associated natural gas and
methane dramatically increase CI.
This paper applies peer-reviewed processes across broad averages of shale activity in North America and compares
them with sample global activity. Ways to lower the carbon intensity of North American shale activity are discussed.
Methods/Procedures/Process: We perform well-to-refinery calculations of CI for major unconventional oil plays in
North America This approach accounts for emissions from exploration, drilling &completions, production, processing,
and transportation.
The analysis tool is an open-source engineering-based model called Oil Production Greenhouse Gas Emissions
Estimator (OPGEE). OPGEE makes estimates of emissions accounting using up to 50 parameters for each modeled
field. This model was developed at Stanford University. Data sources include government sources, technical papers,
satellite observations, and commercial databases.
Results/Observations/Conclusions: Applied globally, OPGEE estimates show highest values in areas with extensive
flaring of natural gas and very heavy crude oils. Heavy oils requiring large energy inputs (e.g. steam flooding) and/or
the use of light hydrocarbon diluents for transportation offset. Other major areas included for reference are Saudi
Arabia, Russia and Iran.Carbon Intensity values are estimated for all North American unconventional plays. A
sensitivity analysis to Bakken flaring volumes illustrates these impacts.
Further sensitivity analyses to pad drilling, and improving well performance are shown CI impacts associated with
hydraulic fracturing. Examples illustrate how OPGEE can be used to evaluate the CI of public policy actions are
provided.
Unconventional production, especially from light tight oil is the most significant new source of fossil fuels in the last
decade. Under a wide variety of carbon constraints, oil usage will continue for many decades and increase in the near
term. Operators, governments and regulators need to be able to avoid “locking in” development of suboptimal
resources and providing incentives and for shale operators to manage resources sustainably. Oil producers must
prepare by eliminating development of marginal projects, optimizing hydraulic fracture treatments, using improved
recovery methods (e.g. enhanced oil recovery using anthropogenic CO ), reducing energy use, and eliminating
2
unnecessary gas waste.Most actions that lower CI are shown to increase operator flexibility.
Mapping the Barnett Shale Gas With Probabilistic Physics-Based Decline Curve Models and the Development of a
Localized Prior Distribution
1 1 1
R. Wanderley de Holanda; ; E. Gildin; ; P. P. Valko; ;
1. Texas A&M University, Bryan, TX, United States.
Objectives/Scope:
A new methodology is introduced to couple publicly available geospatial and production data from shale plays aiming
to reliably reduce uncertainty in the performance of new wells drilled and reveal geological trends. Uncertainty in the
production forecasts depends on model selection, observed data, and prior experience. In this work, a physics-based
model is probabilistically history-matched to 869 gas wells from the Barnett shale. A prior distribution is automatically
generated for each well based on the type of fluid produced and previous production of surrounding wells.
This work is a continuation of the paper SPE-185589-PA, mapping decline curve parameters, and reserves of 869 gas
wells in the Barnett shale. Initially, the production data is automatically filtered with adjustable heuristic rules that
probabilistically calibrate the hindcasts, and wells are clustered by type of reservoir fluid. A three parameter physics-
based decline model that accounts for material balance and linear flow is implemented.
The P-50 values of the prior distribution’s parameters are mapped and variograms are matched to model the spatial
continuity of these properties. At specific dates, local simple kriging is applied to the maps of the 10-quantiles to
estimate the respective quantiles of newly drilled wells, generating a localized prior distribution.
- Maps of predicted EUR, windows of reservoir fluid types, total depth, and decline model parameters are created
using solely publicly available geospatial and production data.
- Regions where the wells present a slower production decline are identified.
- In general, wells in the dry gas window have higher predicted EUR per lateral length.
- The definition of a prior distribution is quite subjective. The criteria established for the localized prior reduces this
subjectivity while taking advantage of the observed performance of surrounding wells to reduce the uncertainty locally.
- When the neighboring wells are beyond the range of the variograms, the localized prior algorithm is consistent,
reconstituting a general prior per reservoir fluid type that is initially defined.
- Publicly available geospatial and production data can be used to quickly and cheaply unlock valuable information
(e.g. spatial distribution of EUR) and understand the development of major resource plays.
- The proposed framework can be strategic for land acquisition and uncertainty reduction in well performance
evaluation.
- In unconventional reservoirs with good spatial continuity and sufficient production data, a comparison of localized
priors can be a useful tool when deciding among potential infill drilling locations (i.e. estimating and comparing
performance before any production data is available).
- While operators may have more detailed data from their wells, they still rely on public data to estimate the
performance of their competitors, so the need to maximize learning from it.
The Use of the Bimodal Production Decline Curve for the Analysis of Hydraulically Fractured Shale/Tight Gas
Reservoirs
2 1 2, 1
T. Blasingame; ; C. Doughty; ; G. J. Moridis; ;
1. Energy Geosciences Division, LBNL, Berkeley, CA, United States.
2. Petroleum Engineering Dept., Texas A&M University, College Station, TX, United States.
Objectives/Scope: The capability to conduct a rapid, near real-time model-based analysis of production data from
tight/shale (TS) gas fields is important in determining fracture and matrix properties. Model-based analysis of
production can range from simple analytical solutions to complex numerical models. The objective of this study is to
develop a simple, Excel-based tool for the analysis of the complex problem of gas production from a fractured TS gas
reservoir that is based on a robust model that is faithful to the underlying physics and can provide rapid estimates of
the important system parameters.
Methods/Procedures/Process: The scientifically robust model used as the basis for the Excel tool is a significant
modification and expansion of the bimodal production decline curve of Silin and Kneafsey (2012). The production
period is divided into two regimes: an early-time regime before the extent of the SRV is felt, where an analytical
similarity solution for gas production rate is obtained, and a late-time regime where the rate can be approximated with
an exponential decline describing a semi-analytical solution obtained by numerical integration. The expanded model
was programmed into an Excel spreadsheet to develop an interactive, user-friendly application for curve matching of
well production data to the bimodal curve, from which matrix and fracture properties can be extracted.
Results/Observations/Conclusions: This tool allows significant insight into the model parameters that control the
reservoir behavior and production: the geometry of the hydraulically-induced fracture network, its flow and transport
properties, and the optimal operational parameters. This information enables informed choices about future
operations, and is valuable in several different ways: (a) to estimate reserves and to predict future production,
including expected ultimate recovery and the useful lifetime of the stage or the well; (b) if curve-matching is
unsuccessful, to indicate the inadequacy of the mathematical model and the need for more complex numerical model
to analyze the system; (c) to verify/validate numerical models, and to identify anomalous behavior or measurement
errors in the data.
Applications/Significance/Novelty: This study provides a spreadsheet-based, user-friendly tool for near real-time
decline-curve analysis that is founded on a novel, scientifically robust mathematical model that involves a bimodal
semi-analytical solution and incorporates key features of hydraulically fractured gas reservoirs. Curve-fitting can
estimate fracture attributes, SRV flow properties and operational parameters, thus providing guidance for a more
efficient hydraulic fracturing process. Additional benefits include reserve estimation, prediction of future production
decline and verification complex numerical models. The present approach can be adapted to gas-flow problems in
dual-permeability media (hydraulically or naturally fractured) or highly heterogeneous sedimentary rock, as well as to
retrograde condensation.
Hindcasting Production Forecasts in Four Shale Gas Basins Using a Physics-based Approach
1 2
F. Male; ; M. Marder; ;
1. Bureau of Economic Geology, University of Texas at Austin, Austin, TX, United States.
2. Department of Physics, University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: In this work we attempt to characterize uncertainties on production forecast. The analysis focuses
on how decline analysis results that depend on the production profile of older wells change as the population ages.
Also we investigate systematic and random errors that affect EURs. This work is a retrospective on production
forecasting performed in the Marcellus, Fayetteville, Haynesville and Barnett shale gas resource plays.
Methods/Procedures/Process: Publicly available and subscription-based production data on 34,000 shale gas wells
were archived and forecasted multiple times over a six year period. We perform hindcasting analysis on these wells,
where production has been forecast using a physics-based decline curve analysis (DCA). This allows us to test the
accuracy of the decline methods and assess reasons for errors in this analysis. We also review the quality of the data
and its effect on production forecasting.
Results/Observations/Conclusions: Using this analysis, we find that there are a number of systematic and random
errors that affect ultimate recovery estimates. Production forecasts for wells in which the production has been back
allocated from lease level to well level show regression to the mean over time. This has an effect on the errors
associated with individual well forecasts in these fields. Also, fields where few wells have entered boundary-dominated
flow (BDF) show higher uncertainty for field-wide production forecasts.
Applications/Significance/Novelty: A proper accounting for individual well and population uncertainties is necessary for
sampling data and risk assessment. Understanding the sources and magnitudes of errors and uncertainties in EUR
values for unconventional gas wells allows operators to account for these in determining economic outcomes and
financial planning of wells.
Montney Versus North America – Completions Comparison
1 1 1 1
K. Ogilvy; ; B. Papau; ; M. Kwan; ; S. Maliki; ;
1. Research, RS Energy Group, Calgary, AB, Canada.
Objectives/Scope: The upstream industry has become very competitive with the recent rise in activity, making it
imperative to understand where the Montney sits amongst its peers. This study looks at completion trends across
North America, focusing on recent completion designs such as proppant intensities and lateral lengths. This analysis
provides a statistical evaluation of Montney and Deep Basin completions and looks at the risk associated with frac’ing
offsetting child wells. The study highlights the upside potential in Canadian plays, particularly the Montney.
Methods/Procedures/Process: This study utilized over 13,000 wells from 2015 to compare proppant intensities and
lateral lengths between the major resource plays across North America and to draw broader conclusions on trends in
the industry. The statistical analysis focused on the Deep Basin and Montney, comparing well performance with
changes in lateral lengths, proppant intensity, and fluid intensity for over 2,000 wells drilled since 2013. The impact of
completion parameters in the Montney was reviewed by identifying over 400 downspaced parent wells with three or
more child wells.
Results/Observations/Conclusions: The study shows Canadian operators have yet to test the levels we see in the
other basins across North America, implying the upper bound on proppant intensity has not been found. Results
indicate larger completion designs improve well performance, providing operators with upside potential. The offsetting
well failure analysis in the Montney showed the time between the parent and the child well being on production had
the largest influence on parent well failure.
Applications/Significance/Novelty: It is extremely relevant and important for both operators and investor to understand
where the Montney sits in comparison to the top plays in North America. This presentation offers an overview and
comparison of seven major resource plays in North America. It provides statistical evidence that more intensive fracs
improve EURs and highlights the optimal cases for completions and how operators can mitigate well failure and
maximize resource recovery in the Montney.
Wolfcamp Geologic Reservoir Modeling Challenges
1, 2
B. J. Casey; ;
1. Geosciences, Texas Oil and Gas Institute, Bellville, TX, United States.
2. Geosciences, University Lands, Houston, TX, United States.
Objectives/Scope: As input for reservoir simulation studies used to optimize Wolfcamp development, geologic models
have been generated across extensive University Lands acreage in the southern Midland Basin. This acreage covers
about 32 blocks in Upton, Reagan, Crockett, and Irion counties, and crosses the southern end of the Midland Basin
from the Central Basin Platform to the Eastern Shelf. Geologic models that contain enough information to characterize
this region are challenging to construct, but can be justified if they reduce uncertainty and allow greater attention to be
placed on simulation studies.
Methods/Procedures/Process: The distribution of individual geologic facies may extend over large distances, the
distribution of petrophysical data (e.g., porosity, permeability, clay volume, etc.) is often irregular, and the distribution
of geomechanical reservoir data (e.g., minimum horizontal stress, Young’s modulus, Poisson’s ratio, etc.) is often very
limited. Consequently, large geologic models must be constructed to distribute the geologic facies and to capture
enough reservoir data to generate a suitable population size. Vertical layers must be thin enough to capture key
geologic facies and generate sufficient data. Lateral grids require sufficient cells between wells to capture lateral
variations.
Results/Observations/Conclusions: With 8 horizontal wells typically drilled across a square mile, the ideal lateral cell
size would be about 200 feet square. To capture the thinnest geologic facies, vertical cells would need to be about 6
inches thick. However, if the area of interest is hundreds of square miles and the Wolfcamp is over 1000 feet thick, the
resultant model could exceed several billion cells. Computation time for such a huge model will be very slow. Thus,
the evaluation interval is reduced, the lateral grid size is increased, and the layer thicknesses are increased to reduce
cellular volume, while maintaining a statistically useful reservoir data population. Although hardware limitations of
stand-alone workstations are being challenged, a working model of 800 million to 900 million cells is possible.
Applications/Significance/Novelty: The reservoir simulation sector model to be used for production history matching
may only cover a single well path, typically about two square miles, but the geologic model used to derive the sector
model may cover several hundred square miles. Derivation of the geologic model assumes a good understanding of
the geologic facies that are being distributed, and sufficient well data to populate the model with a statistically
meaningful population size. If the geologic model can be sufficiently tied to regional outcrop studies, core, well data
and seismic data, our confidence in the geologic model should be high. This would then allow more attention to be
given to such variables as effective drainage volume and the hydraulic fracture model.
Additional Applications of Optimal Artificial Lift Strategies in the Permian Basin
1 1 1 1
Y. U. Pradhan; ; H. Xiong; ; J. K. Forrest; ; J. Zhu; ;
1. Texas Oil and Gas Institute, Rosenberg, TX, United States.
Objectives/Scope: Maximizing well performance strongly depends on artificial lift installation (AL), operational
maintenance, and reservoir performance over time. Operators in the Permian Basin often follow neighboring
operators’ AL methods without further optimization. While implementing one form of AL could maximize initial
production and yield quicker payout, these installations could hinder long-term hydrocarbon recovery and economics.
Therefore, an AL strategy using a holistic reservoir performance analysis could optimize production and operators’
economics.
Methods/Procedures/Process: This study evaluates additional applications various AL strategies in southern Midland
and Delaware Basins based on production histories, general operation practices, and reservoir modeling. The study
reviews the regional reservoir characteristics and general AL practices in Wolfcamp and Lower Spraberry formations,
identified more candidate wells, and built and calibrated numerical simulation models with production histories. The
study then used those calibrated models to run various cases with different AL methods. Finally, the study performed
economic analyses for different AL methods.
Results/Observations/Conclusions: This paper reviews the workflow to select an optimal AL strategy, compares
strategies in the Southern Midland Basin Wolfcamp and in the Northern Midland Basin Lower Spraberry shale, and
reviews additional opportunities on University Lands. This paper will also discuss various strategies to optimize
production on University Lands’ Delaware Basin unconventional horizontal wells. The study suggests best practices
that can be used by other Permian Basin operators. Similar previous findings in other parts of the Midland Basin,
these new case studies confirm that operators who implement AL strategies on a larger scale have a larger probability
of success in their economics.
Applications/Significance/Novelty: The productivity index (PI) of wells decline with time in unconventional reservoirs,
and characterizing well PI over time should benefit these long term AL strategies decision. The scale of development
has a large influence on the AL strategy. In addition to that costs and reservoir properties are known parameters of
influence, this paper will also discuss how run time and equipment failure rates proved to be additional parameters of
consideration as well.
Additional Applications on Determining Optimal Lateral Lengths and Trajectories on University Lands’ Midland and
Delaware Basins
1 1
Y. U. Pradhan; ; H. Xiong; ;
Objectives/Scope: Operators in the Permian Basin continue to advance their knowledge on unconventional horizontal
wellbore constructions. Motivations to drill and complete the longest possible laterals could come from either
maintaining lease agreements held by production or reducing costs. Therefore, analyzing production effects over time
due to trajectories is necessary to advance our understanding of production optimization. This paper aims to
determine optimal lateral lengths and trajectories in the Permian Basin by studying the impact of the lateral length and
trajectory on well performance.
Methods/Procedures/Process: This study covers additional applications on the impact of lateral lengths and
trajectories on Midland Basin Lower Spraberry , Wolfcamp, and Delaware Basin Wolfcamp wells. We couple reservoir
simulation models with transient multiphase wellbore models. The reservoir simulation model is calibrated with
production histories. The calibrated model is then used to forecast the long-term well productivity. This paper details
the following multiphase flow simulation cases: (1) horizontal wellbores with vertical deviations, and (2) horizontal
wellbores with vertical deviations and tortuosity, combining with different lateral lengths, and reservoir and production
conditions.
Results/Observations/Conclusions: These new case studies confirm that uniform, toe up wells yield higher productivity
and production potential. However, the study also reveals that long lateral lengths do not completely ensure
proportionately more production, and wells with high tortuosities could hinder production potential. Fluid properties
could influence recoveries per lateral length when drilling longer laterals. For instance, wells producing fluid with high
gas-to-liquid ratios could produce less than twice the EUR at double the average lateral length compared to the EUR
at an average lateral length. However, wells producing fluid with low gas-to-liquid ratios could produce roughly twice
the EUR at double the average lateral length. The paper also discusses how liquid holdup contributes to these results.
Applications/Significance/Novelty: Wellbore lateral lengths and uniformity are important for production optimization to
minimize liquid holdup and maximize productivity throughout the wells’ production lives. However, wellbore
construction designed for holding leases and completed under time constraints could negatively impact production
and limit operators’ potential for their acreage. To determine an optimal lateral wellbore length, it is necessary to study
wellbore multiphase flow behavior and resulting production for unconventional horizontal wells.
An Integrated View of the Petrology, Sedimentology, and Sequence Stratigraphy of the Wolfcamp Formation,
Delaware Basin, Texas
1 2 1 2
M. Thompson; ; P. Desjardins; ; J. Pickering; ; B. Driskill; ;
1. Shell International Exploration and Production Company, Houston, TX, United States.
2. Shell Exploration and Production Company, Houston, TX, United States.
Objectives/Scope: This study was conducted in 2016-2017 to advance the understanding of how small-scale
elements, such as texture, composition, pore-types, and diagenesis affect the rock properties of the Wolfcamp
Formation. The objectives of this study were to define key rock types from thin section (petrofacies) and core
(lithofacies) to describe predictable, cyclic packages in a sequence stratigraphic framework to better understand what
is controlling reservoir quality and its distribution in the Wolfcamp.
Methods/Procedures/Process: The Wolfcamp A and B have been examined in detail in two proprietary cores from the
central part of the Delaware Basin. Detailed sedimentological and petrographic techniques were employed to
document the different facies types and bed boundaries, their characteristics, and associated rock properties to
characterize the vertical changes in facies and reservoir properties. The robustness of the sequence stratigraphic
framework was enhanced with seismic, biostratigraphic, geochemical, and sedimentological data from additional
available cores from Reeves, Loving, and Ward Counties in Texas.
Results/Observations/Conclusions: Facies are stacked in predictive, repetitive packages that are linked to their
position within a sequence stratigraphic framework. Four distinct cycles were identified: debrite, calciturbidite, fine-
grained turbidite, and distal turbidite. The debrites and calciturbidites represent episodic events during low stands in
sea-level where carbonate sediments from the platform are deposited in the basin. Fine-grained turbidites are thinly
bedded and often interbedded with replacive microcrystalline dolomite are also deposited during low stands. Distal
turbidites represent the most distal, muddy tails of turbidites where silty mudstones fine-upward into organic-rich,
siliceous mudstones. Distal turbidites have the best reservoir quality and are thickest and more abundant during
maximum flooding.
Applications/Significance/Novelty: The sequence stratigraphic approach improves the lateral and vertical predictability
of sweet spots and the zones/areas to stay away from, which ultimately drives appraisal and development decisions.
Integrating our understanding of the cyclic nature of the Wolfcamp with the calibrated e-facies derived from logs in a
stratigraphic framework allows for quick high-grading of acreage away from current control points support of
development planning and new acreage evaluations.
Outcrop and Subsurface Geology Applied to Drilling, Sweet Spot and Target Zone Detection of Resource Shales: The
Woodford Shale Example
1, 2
R. M. Slatt; ;
1. Geology and Geophysis, University of Oklahoma, Norman, OK, United States.
2. Institute of Reservoir Characterization, Norman, OK, United States.
Objectives/Scope: “Over the past three years, more than 100 oil and gas companies in North America ....... have filed
for bankruptcy…” (Jenkins, 2017). Among the technical reasons for this, Jenkins includes a lack of understanding of
the shale reservoir and lack of quality and quantity of properly-sampled rock and fluid data as contributors to less-
than-desirable drilling and completion results. As discussed below, Geology (and Geophysics) play an important role
in mitigating these shortcomings in order to improve hydrocarbon discovery and production efficiencies.
Methods/Procedures/Process: The Woodford Shale provides an example of some key geologic factors for reservoir
characterization . Outcrops of strata equivalent to the subsurface reservoir of interest provide a good starting point for
describing the reservoir. Outcrop stratification (layering) and composition are first documented, then a outcrop gamma
ray log is obtained for comparision with a non-cored well in the reservoir. The gamma log provides the link to transfer
outcrop stratigraphy to the subsurface well. Closely-spaced outcrop samples are taken for analysis of chemical (XRF)
and mineralogical (XRF, XRD, petrography) composition and stratification; Leeb hardness; microfabric, porosity/
permeability (SEM); organic content (TOC, RockEval, biomarkers); fractures/folds (visual observations); and
biostratigraphy.
Results/Observations/Conclusions: The resulting outcrop-well data set is used to determine depositional environment,
determine lateral continuity of strata beyond the wellbore, document fracturability, characterize reservoir quality, and
predict the potential for strata to hold proppant. 3D seismic provides a broad image of reservoir geometry and
structure, but internal reservoir intervals often are too thin to be seismically resolved. The Woodford 3D seismic
volumes sometimes do detect internal variability that may image a horizontal landing or drilling hazard zone. Also, by
examining 3D proportional slices, potential sweet spots of extra thickness and organic richness can be imaged. In the
Woodford, an erosional surface atop the underlying limestone has helped define "sweet spots" stratigraphically and
geographically.
.
Applications/Significance/Novelty: Numerous Woodford studies which follow the above recommended series of
analyses have provided insights into the importance of geologic (and geophysical) input to reservoir characterization
of shales, and have been successfully employed by companies in our STACK-MERGE-SCOOP consortium. There is
an informal consensus by operators that exploration and exploitation of shales in Oklahoma has bottomed out, placing
Oklahoma oilfield activity in the No. 2 ranking in the U.S. (Wilmoth, 2017). Once again quoting Jenkins (2017) “This
process requires maintaining consistent drilling and completion practices so variations in reservoir quality can be
understood. The time to do this is NOW, before prices rise again ……”
Geological Facies Prediction Using Computed Tomography in a Machine Learning and Deep Learning Environment
1 1
U. Odi; ; T. Nguyen; ;
1. Advanced Analytics, Devon Energy Corporation, Oklahoma City, OK, United States.
Objectives/Scope: Machine learning and deep learning, a subset of machine learning, have become successful
predictive modeling tools. Machine learning applications include classifying DNA sequences and fraud detection.
Deep learning applications include language processing and facial recognition. Both methods have shown to be useful
in understanding unconventional reservoirs when traditional techniques used to understand conventional reservoirs do
not readily transfer. In this work, machine learning and deep learning are used to predict geological facie
classifications using computed tomography (CT).
Methods/Procedures/Process: A combined approach is developed to predicted geological facies. The first aspect of
this approach utilizes molecular weight, density, and porosity from CT scans on core samples to learn from existing
user defined geological facie classifications. Once the machine understands the relationship between geological
facies and CT scan physics, the resulting model is used to predict subsequent geological facies using the CT scan
parameters. The second aspect of this approach utilizes CT scan thin section images of core samples to learn from
corresponding geological facie classifications. Once a deep learning model learns from these CT scan images, the
resulting model is used to predict subsequent geological facies using CT scan images.
Results/Observations/Conclusions: Results show misclassification rates at approximately less than 30% which are
more favorable to error rates associated with a single person classifying facies. In addition, unsupervised machine
learning results show robust clustering of geological facies. Machine learning and deep learning show the value of
utilizing CT scan parameters and images in unconventional oil and gas reservoir workflows. While typically used in
other areas such as cancer detection, CT scans in conjunction with machine learning and deep learning can
drastically reduce the error rate and time it takes to fully categorize an unconventional play.
Applications/Significance/Novelty: Current methodologies involve having a team of petrophysicist characterize core
samples at every specified CT scan resolution depth. These workflows can be cumbersome and time consuming for a
team of petrophysicist and at times result in facies classification governed by human bias. The proposed methodology
in this work drastically reduces human error and the time to understand geological facies. As a result, business
decisions can be made quicker. These models are expected to provide substantially more accurate and consistent
geological facie predictions.
Recent–Pleistocene Immature Mudrocks of Anoxic Basins of Venezuela and West Africa: Analogs for Unconventional
Shale Oil/Gas Reservoirs?
1 2 2
U. Hammes; ; M. Grammer; ; J. Gregg; ;
1. Department of Geosciences, Texas A&M University, Austin, TX, United States.
2. Boone Pickens School of Geology , Oklahoma State University, Stillwater, OK, United States.
Objectives/Scope: Geochemical and sedimentological studies of modern anoxic basins offer analogues for ancient
basins and have the potential to elucidate answers to long-standing questions. However, a direct comparison of
Pleistocene anoxic basins (e.g., Cariaco Basin, offshore Venezuela, the Walvis and Mid-Cape Basins, offshore
southwest Africa. and Porcupine Basin, offshore Ireland) to Mesozoic source rocks, such as the Upper Jurassic
Haynesville and Upper Cretaceous Eagle Ford Formations of Texas, and the Mississippian Lime/Chert has not been
conducted to date and are investigated in this study,
Methods/Procedures/Process: Cores from three ocean cruise legs, ODP 165 and 175 and IODP 307, have been
selected for this study because of the similarity of the basins and their associated deposits. These Neogene
sediments share remarkable lithological similarities to calcareous shales such as the Eagle Ford Group (Cretaceous),
the Haynesville (Jurassic) Shale or Mississippian Lime/Chert. Our study of Leg 165 and 175 sediments provide the
opportunity to examine lithologies deposited under anoxic conditions in a sub-tropical region with high TOC, at similar
levels as found in the Eagle Ford and Haynesville shales.
Results/Observations/Conclusions: The Cariaco Basin and the Eagle Ford and Haynesville cores exhibit similar
features, such as alternating layers of organic-rich, laminated marls and planktonic-foram-rich limestones. Alternating
deposits exhibit distinct geochemical and sedimentologic signatures probably related to sea-level fluctuations,
upwelling, and varying detrital sediment supply. Core sediments obtained in Leg 307 were deposited largely under
more oxygenated conditions and have relatively low TOC content. Identifying controlling factors of organic-rich versus
frackable intervals are key aspects of unconventional plays.
Applications/Significance/Novelty: Characterizing recent mudrocks with detailed insight into controls of water and
sediment geochemistry, oceanographic changes, age, and nutrient and sediment flux provide in-depth information for
depositional controls on ancient source rocks. These samples provide the opportunity to compare and contrast the
diagenetic effects of organic composition on the Neogene sediments as well as the development of micro-nanoscale
pores within the organic material. Studying early diagenesis of potential unconventional shale reservoirs provide
insights in origin and evolution of unconventional reservoirs.
Integration of Core Fracture and Lithofacies Descriptions in the Wolfcamp Shale: Implications for Mechanical
Stratigraphy and Deformation History
1 1 1 1 2 2
A. Salem; ; J. Solum; ; S. Naruk; ; D. Minisini; ; P. Desjardins; ; J. Hnat; ;
1. Shell International Exploration & Production, Inc., Houston, TX, United States.
2. Shell Exploration & Production, Co, Houston, TX, United States.
Objectives/Scope: Here we present findings from an extensive survey of natural fractures from Wolfcamp Shale cores
from the Delaware Basin in west Texas, one of the most active unconventional plays in North America. Understanding
natural fractures in the Wolfcamp is important for reservoir characterization, with potential implications for optimizing
reservoir stimulation and improving well performance. The objective of this study was to characterize fractures in
whole cores and integrate with lithofacies to establish mechanical stratigraphy.
Methods/Procedures/Process: Fractures were described and measured over more than 6500 ft. of core from 16 wells.
Detailed descriptions were made of every fracture including texture, mineral fill, length, width, cross-cutting
relationships and lithofacies observed in. Fracture types include subvertical veins, horizontal veins, joints and faults.
Fracture intensity was quantified by measuring frequency and areal density. Fracture intensity logs were compared to
lithofacies logs to determine if fracture intensity was associated with particular lithofacies. Fracture intensity per well
was compared with mapped faults to see if fracture intensity correlated with proximity to faulting. Constraints on the
kinematics and timing of fracture formation are discussed.
Results/Observations/Conclusions: Inclined veins, the most common fracture type, are typically less than 1 ft long with
apertures of less than 1 mm and are typically filled with calcite cement, some dolomite and quartz. Fractures terminate
at lithology changes or against horizontal veins. Horizontal veins typically have apertures of 1-2 mm. Fibrous texture
suggests “crack and seal” growth of the veins. Slickenlines were observed on some vein walls. Joints and faults are
observed less frequently Joints have similar dimensions to inclined veins. Faults are mostly syndepositional with
secondary faulting observed only in a few wells. Carbonates and mudstones contain far more inclined veins than
sandstones, siltstones and mass transport deposits. Horizontal veins are observed exclusively in organic-rich
mudstones.
Applications/Significance/Novelty: Lithofacies are the primary control on fracture intensity and height. Considerable
overlap in the range of fracture intensity values was observed for all Wolfcamp members, except for the X and Y
sands of the A interval due to the propensity of sandstone and siltstone facies.Horizontal veins are interpreted to pre-
date vertical veins and likely formed as the result of pore fluid pressure exceeding lithostatic pressure during peak
hydrocarbon generation. Inclined veins are primarily oriented NW-SE with a secondary fracture set trending NE-SW
and are interpreted to have formed as tensile cracks with σ oriented NE-SW, perpendicular to the primary fracture
3
set. This orientation suggests these fractures formed during Basin and Range uplift and extension.
Integrated Geocellular Static Model for Geomechanical and Dynamic Simulations in the Vaca Muerta Formation
1 1 1 1 1 1
F. A. Lozano; ; J. P. Palacio; ; V. Lazzari; ; C. Bernhardt; ; D. E. Hryb; ; F. González Tomassini; ;
1. YPF, Buenos Aires, Argentina.
Objectives/Scope: This project had two objectives: 1) create a semi-regional geocelular model that preserved the
geological trends, in a sequence stratigraphic framework, through the integration of interpreted geological,
petrophysical, geophysical and geomechanical data. 2) Crop a single pad with three horizontal wells completed in
different landing zones. Integrate the microseismic to define the SRV and build a DFN. Matrix and fracture properties
were delivered for geomechanical and dynamic simulation to predict well performance, define optimal scenarios for
the plan development and reserve estimations.
Methods/Procedures/Process: The regional geocelular model integrated 3D seismic, 22 vertical wells with
petrophysical and geomechanical data, 5 vertical cores and 10 horizontal image logs with facies, and fracture
analysis. Porosity and geomechanical property model distributions were based on correlations with seismic acoustic
impedance. Permeability and saturation used porosity correlations for their distribution. The cropped pad model
included filtered microseismic properties (magnitude, S/N) to define the stimulated reservoir volume (SRV). The
Discrete Fracture Network (DFN) model was built based on mechanical stratigraphy, fracture analysis, and the
estimated SRV. DFN fracture planes were used as input for the geomechanical simulator. Using the ODA method we
obtained the fracture properties for dynamic simulation
Results/Observations/Conclusions: Regional geocelular modeling was indispensable to preserve geological and
property trends required in the cropped model for simulation. Rock-physics models showed good correlations between
seismic acoustic impedance, porosity and young modulus allowing robust regional distributions. Filtered microseismic
properties prove to be valuable to define the SRV and showed a good correlation with the geomechanical properties.
Azimuthal microseismic growth showed minor changes between the same well stages but large variations between
different well stages. We interpreted it to be the product of the interaction between fracture families and changes in the
local stress during well completions. Geological constrained DFN were necessary to deliver robust models for
simulation.
Applications/Significance/Novelty: YPF has acquired and interpreted numerous regional multidisciplinary data that we
integrated building a unique regional geocelular used as database. This model helped us to better understand the
interaction between the different variables and their correlations. Integrated regional models did not only allow us to
preserve the geological trends but also will permit us to crop multiple single pads for systematic simulation of different
pilot zones. Regional models produced more accurate inputs for geomechanical and dynamic simulations.
Microseismic property modeling techniques presented in this study are being implemented systematically in different
areas of the basin in the Vaca Muerta Fm. This work lead to obtain better scenarios for well spacing optimization and
performance prediction.
Advanced Downhole Measurements and 3-D Model-based Geosteering Improves Wellbore Placement in the Permian
Basin
1
C. Viens; ;
1. Nabors, Houston, TX, United States.
Objectives/Scope: Proper wellbore placement is critical to the economic success of the well. Given the nature of the
bare minimum total gamma and sparse 90’ survey data used in most horizontal drilling applications, even the best
geosteering geologists often find themselves running multiple simultaneous interpretations in hopes that one will start
to make sense as new data comes in. Utilizing advanced downhole measurements and a 3D model based
geosteering solution has led to significant improvements is wellbore placement and protocol for handling common
pitfalls associated with the geosteering process.
Methods/Procedures/Process: Employment of azimuthal gamma imaging in conjunction with continuous inclination
brings clarity to stratigraphic position, bed dip, faults, and zones where bedding is laterally discontinuous or completely
absent. In some of these situations geosteering in its traditional sense cannot be executed. Through enhanced
visibility to the subsurface through azimuthal gamma imaging, geological anomalies can be quickly identified and
handled in the best possible manner. Furthermore, in many cases, the status quo of 2D geosteering software
packages creates further confusion and uncertainty as they give little to no indication of what may lay ahead.
Results/Observations/Conclusions: Wolfcamp examples of azimuthal gamma image data showing faults, laterally
discontinuous bedding, and high energy sedimentary deposition such as debris flows, channels, and mass transit
complexes will be presented. To improve geological interpretation confidence, protocol has been developed to identify
and handle these situations using a model based 3D geosteering solution with multiple well and seismic surface inputs
to provide options beyond traditional correlation alone.
Applications/Significance/Novelty: The data examples and developed protocal shown have not been previously seen
or acknowledged by the industry.
3-D Anisotropic Damage Mechanics for Modeling Interaction Between Hydraulic and Natural Fracture Planes in a
Layered Rock – Application to Eagle Ford and Wolfcamp
1 1 2 1
Y. Aimene; ; C. Hammerquist; ; J. Nairn; ; A. Ouenes; ;
1. FracGeo LLC, The Woodlands, TX, United States.
2. Oregon State University, Corvallis, OR, United States.
Objectives/Scope: A full 3D anisotropic geomechanical model is developed and aims to model 3D hydraulic fracturing
in naturally fractured and layered anisotropic reservoirs. The geomechanical tool provides the ability to model the
propagation of 3D hydraulic fracture planes and their complex intersection with a 3D fracture network. The natural
fractures are represented using explicit discontinuities or initial damaged zones, both estimated from seismic data.
New approaches for modeling 3D intersections between hydraulic and natural fractures are introduced for the first
time.
Methods/Procedures/Process: The modeling of layered and fractured reservoirs was achieved using an anisotropic
damage mechanics model implemented in the material point method. The approach is based on the fourth-rank
anisotropic damage tensor D, which augments the material constitutive model. Damage initiation is predicted by a
failure surface in principle stress space and damage propagation direction is defined by the normal to the failure
surface. Finally, the damage evolution is dictated by softening laws to the fracture plane, one for normal and two for
shear. Frictional contact using imperfect interface model was used to model the effects of interface properties between
layers.
Results/Observations/Conclusions: The validation of this 3D anisotropic model is demonstrated on elementary tests
and on an actual Wolfcamp well that showed significant vertical interferences between Lower Sprayberry and
Wolfcamp A. First, laboratory experiments were performed, and numerical results closely matched the experimental
ones. Two interfaces properties, soft and stiff, were used. Numerical results showed that the damage propagation was
highly affected by the presence and nature of the interfaces. Soft interfaces played a barrier role, which ended
damage progression. In addition, cases of damage reflection and bifurcations were observed. Layered media shows
different pattern of fracture initiation, where high density is observed in thin brittle layers, while less damaged zone is
developed in thick soft layers.
Applications/Significance/Novelty: The full 3D model was applied to a Wolfcamp A hydraulic fracturing problem where
the results explained the interference with the Lower Sprayberry. The input for the 3D models were derived mainly
from 3D seismic and the results validated with microseismic. The resulting damaged zones is a key indicator of the
stimulated permeability in the reservoir and it is a critical input to reservoir simulation. The 3D anisotropic damage
mechanics model demonstrated its ability to reproduce field and laboratory observations in complex layered and
naturally fractured reservoirs, thus it represents a good tool to analyze and optimize hydraulic fracturing. The new
technology can help oil and gas operators adjust their development strategies to vertical and lateral well spacing to
avoid interferences.
Origin, Detection, Involvement in Hydraulic Stimulation, and Consequences for Field Development of Large-scale
Structural Lineaments in the Marcellus and Duvernay Plays
1 1 1 2 2
B. Stephenson; ; E. Galan; ; M. Fay; ; A. Savitski; ; T. Bai; ;
1. Shell Canada, Calgary, AB, Canada.
2. SEPCO, Houston, TX, United States.
Objectives/Scope: The evidence for large-scale structural features (lineaments/faults) affecting a hydraulic stimulation
is much more compelling than for small-scale features (natural fractures). Large-scale features are weaker and have
similar dimensions to a typical hydraulic fracture. But is it beneficial to stimulate these features and what are the
potential consequences? An analysis of structural features from the Marcellus and Duvernay formations has been
undertaken to understand whether, in the presence of structural features, any development decisions might get
impacted.
Methods/Procedures/Process: Maps of structural features supported by seismic attributes are commonly challenged
as to what they physically represent. An outcrop analogue is used to demonstrate how strain is distributed in
intrinsically layered media, such as a shale. Chemical water tracers have been used to calibrate the conductive
length-scale of these features in the sub-surface. A novel interpretation method of microseismic events in time reveals
how lineaments are involved during a hydraulic fracture treatment and geomechanical modeling reveals the impact on
overall stimulation effectiveness. More compelling evidence of lineament involvement is derived from observations of
external fluids brought into the well-bore from underlying formations.
Results/Observations/Conclusions: Shales will preferentially fold above a fault due to the strong layering anisotropy,
resulting in strain partitioning, with bedding-parallel slip (shear) limiting the vertical extent and opening (dilation) of
discrete planes, making vertical offsets more difficult to detect from seismic. Lineaments in the Marcellus are broad,
planar zones of fracturing, forming axial kink-band deformation and in the Duvernay they are some places associated
with a subtle flexure above faults. Geomechanical modelling shows that when a lineament/fault at a high angle to
sHmax is stimulated, it can dilate and the opposite tail of the bi-wing frac does not propagate. Diagnostics from plays
with lineaments oriented close to sHmax, show that the length-scale of hydraulic communication is increased.
Applications/Significance/Novelty: Whether a lineament helps well productivity depends partly whether it is ‘contained’
or ‘uncontained’ within the over-pressured formation. In the ‘contained’ case, stimulation of a lineament may enhance
productivity of a stand-alone well, but will potentially exacerbate ‘parent-child’ relationships on a pad. A potential
mitigation measure may be to use small-scale proppant to plug-off these high conductivity (or leak-off) pathways.
Moreover, this may also improve the stimulation efficiency in a multi-cluster plug and perf completion design by
encouraging diversion. Paradoxically, the best way to handle large-scale structural lineaments may be to stimulate
them in order to shut them off.
A Method of Fracture Prediction Across Multiple Stratigraphic Horizons in the Midland Basin, Texas, USA
1 2, 3
C. Pollock; ; C. Seiler; ; M. Valcárcel; 2; E. Macaulay; 2;
1. Pioneer Natural Resources, Irving, TX, United States.
2. Midland Valley Exploration Ltd., Glasgow, United Kingdom.
3. Geoscience Australia, Symonston, ACT, Australia.
Objectives/Scope: Natural fractures influence the development of unconventional reservoirs in many ways. Production
results indicate enhanced fluid rates associated with natrual fractures. Microseismic and well interference during
hydraulic stimulation suggest that they influence completions. Numerical models suggest that natural fractures may
influence length and height dimensions of hydraulic fractures. Methods commonly used to quantify them can be costly,
possibly risky or impractical. Here we present a method of fracture prediction using commonly available data in
conjunction with commercial software.
Methods/Procedures/Process: Two distinct fracture trends (NE-SW and NW-SE) are observed in the Wolfcamp and
Spraberry unconventional reservoirs of the Midland Basin. A third, less prevalent E-W trending set, is observed in
some intervals. These sets are consistent in orientation, spatially and stratigraphically, across the basin with mean
trends varying by only a few degrees. Fracture intensity logs, calculated from a basin-wide set of horizontal image log
interpretations, demonstrate an increasing intensity with proximity to faults in the middle Wolfcamp. In areas where
multiple intervals are sampled, a similar relationship of fracture intensity to faults is observed in each. This suggests
that the fractures are tectonic in origin and are coeval with faults or are the result of later fault reactivation.
Results/Observations/Conclusions: Fracture formation mechanisms were evaluated in a study area with 3D seismic
data (for horizon and fault surfaces) and image logs in multiple stratigraphic units. Evaluated mechanisms included:
fractures formed by folding above fault tips (curvature and geomechanical modeling) and fault-related fractures formed
during fault reactivation under various stress regimes (elastic dislocation modelling). Plausible paleostress regimes
were determined by an inversion, varying the orientation of the principal stress axes and the differential stress ratio to
maximize the slip and dilation tendencies on the observed fractures. Of the considered mechanisms, only fault
reactivation under a strike-slip sress regime (S1 ~80°) predicted fracture orientations and failure modes consistent
with observations.
Applications/Significance/Novelty: Relative fracture intensities were evaluated using the principal strains magnitutes
predicted from elastic disloction modeling as a proxy. In the strike-slip fault reactivation model, horizontal and vertical
trends of maximum extensional strain (e1) correspond well with observed variaions in fracture intensity, decreasing
exponentially away from faults and towards shallower depths. This proxy relationship provides a promising means to
estimate fracture orientations and intensities in areas where little image log data is available, whereas the more
ubiquitous 3D seismic is available.
Can Seismic Inversion Be Used for Geomechanics? A Casing Deformation Example
1 1 1 1 1
J. J. Meyer; ; J. Gallop; ; A. Chen; ; S. Reynolds; ; S. Mildren; ;
1. Ikon Science, Adelaide, SA, Australia.
Objectives/Scope: An Alberta operator has experienced casing deformation in horizontal wells after hydraulic fracture
stimulation. The occurrence of the casing deformation varies with location and formation and is primarily observed
close to faults. A comprehensive dataset is available for the field that includes 3D seismic, multi-finger caliper data of
the casing deformation, dipole sonic logs, image logs and mini-fracture data, making it an ideal candidate for
investigating both the causation of the casing failure and the potential to use the 3D seismic to predict the risk of
casing failure pre-drill.
Methods/Procedures/Process: A 1D geomechanical model was constructed for a vertical pilot on a pad with two
laterals at different depths. The laterals are drilled in opposite directions, sub-parallel to a large fault and have been
hydraulically stimulated. Casing deformation is only observed in the lower lateral. The 3D seismic was inverted
creating a 3D property volume, which was then calibrated to create a 3D geomechanical model.
Slippage planes and slip-sense were interpreted from the multi-finger caliper data and the risk of reactivation was
calculated using the 1D geomechanical model from the pilot well, with a good match observed between the failure
orientations, and predicted and observed slip sense, suggesting that shear failure of the pre-existing faults is the likely
cause of the casing deformation.
Results/Observations/Conclusions: The geomechanical model in the pilot well shows lower stress anisotropy in the
upper zone than the lower zone. The magnitude of the shear stress on planes of the same orientation is dependent on
the stress anisotropy. Consequently, the magnitude of the shear stress is elevated in the lower zone, increasing the
risk of shear failure and associated casing deformation.
Analytic 1D geomechanical models built in the region are consistent and can explain the observed wellbore
deformation, but do not quantitatively predict the shear fracture failure and casing deformation. However, the
calibrated peak shear stress magnitude appears to correlate with to observed casing deformation in the subsurface.
Volumes derived in this way may potentially be used as a ranking tool.
Applications/Significance/Novelty: Variation in casing deformation can be explained by understanding the risk of
reactivation, maximum shear stress and stress anisotropy, as it is the shear stress that directly correlates with the
magnitude of the failure event and the energy released. The greater the magnitude, the greater the risk of casing
deformation. By building 3D geomechanical models and understanding the variation in both risk of reactivation and
maximum shear stress it is possible to lower the risk of casing deformation pre-drill.
Intensive Natural Fracture Study of Elk Hills Monterey Formation to Better Understand Production Variability
1 1 2
R. Sobczyk; ; R. Gales; ; N. Harvey; ;
1. California Resources Corporation, Los Angeles, CA, United States.
2. Harvey Rock Physics, Perth, WA, Australia.
Objectives/Scope: Elk Hills, initially designated as Naval Petroleum Reserve No. 1, is a large anticlinal field discovered
in 1911 in the southern San Joaquin Valley of California. Elk Hills is a complex structure with multiple pays and highly
fractured. Focus will be on the Antelope Shale (Antelope) member of the Monterey Formation which has variability in
production due to variations in net pay and degree of natural fracturing. The goal of this project was to understand the
natural fracture complexity of the Antelope interval and influence on production variability and provide insight for
development.
Methods/Procedures/Process: We focus on the scope of work and workflows used to provide information obtained in a
study involving 188 image logs over the Elk Hills structure. The study was designed to identify variability in fracture
type, orientation, aperture and relationship to bedding over the field. We will also outline the systematic approach from
Scope of Work, project Quality Control and final results including the benefit of a field wide approach compared to
prior small area studies. The approach involved selection of high quality data sets that provided geographic coverage
vertically, laterally and by tool type. Data sets came from different vendors showed that each tool type gave consistent
response that could be used for comparable evaluation of fracture type, orientation and aperture.
Results/Observations/Conclusions: The fractures in the target section were picked in considerable detail focussing not
so much on one particular fracture type but providing a spatially well populated and distributed data set allowed for the
mapping of orientation, fracture type, aperture and frequency by zone, which provided insights into the nature and
extent of the fracture distribution by zone within the Antelope Interval. A positive comparison to production was
achieved by integrating weighted fracture intensity and reservoir quality from the petrophysical model.
Applications/Significance/Novelty: We present an approach for handling large borehole image data sets and
presenting the information in a way that is beneficial to selection of production intervals and understanding the
character of different fracture sets across a field or an exploration area. The techniques presented are applicable
across a wide range of environments provided there is the data available.
Reservoir Geomechanic Heterogeneity Index (RGHI): Concept, Methodology, and Application
1 1 1 1 1
J. Zhou; ; S. Mandal; ; F. Chen; ; M. Quest; ; D. Hume; ;
1. Integrated Reservoir Solutions Division, Core Laboratories, Houston, TX, United States.
Objectives/Scope: In the last decade, the widespread application of lateral drilling and hydraulic fracturing
technologies has led to an unprecedented boom in unconventional oil and gas production in North America. However,
a key challenge remains; accurately predicting propped fracture height growth. Rock mechanical properties calculated
from wireline logs generally underestimate the vertical heterogeneity of reservoirs due to log resolution issues. To
address this, this study proposes a new method that uses high-resolution Leeb (rebound-hammer) hardness data to
fully capture reservoir heterogeneity.
Methods/Procedures/Process: Geomechanical rock properties (e.g. Young’s Modulus and Poisson’s Ratio) have long
been used as key parameters to design stimulation treatments in unconventional reservoirs. However, conventional
rock mechanics measurements are commonly widely spaced. In contrast, Leeb Hardness measurements made on
core slabs are fast and cost effective, with a flexible sampling frequency. More importantly, Leeb Hardness varies
significantly with bulk mineralogy and responds to rock types (e.g. ‘soft’ mudrock vs. ‘hard’ limestone). A mean value
of numerous measurements around a small sample area has been shown to be representative of individual lithofacies.
For vertically heterolithic reservoirs, a six-inch or higher resolution is recommended to capture the full spectrum of
reservoir heterogeneity.
Results/Observations/Conclusions: We have established a series of empirical relationships between lab-measured
geomechanical attributes (e.g. Young’s Modulus) and Leeb Hardness using Core Lab’s North America unconventional
core database. This has allowed Core Lab to build high-resolution, continuous, geomechanical profiles through many
of the most prolific reservoirs. The high-resolution geomechanical profiles tend to reveal higher degrees of vertical
heterogeneity than those derived from conventional logging data (e.g. sonic logs), especially in heterolithic, thin-
bedded reservoirs (e.g. Wolfcamp in the Midland and Delaware basins). The frequency and magnitude of
geomechanical changes with depth can be more accurately captured by the high-resolution Leeb Hardness data,
allowing more accurate geomechanical modeling.
Applications/Significance/Novelty: A new quantitative parameter, Reservoir Geomechanic Heterogeneity Index
(RGHI), has been used as an additional input in reservoir modeling to more accurately predict hydraulic fracture height
growth. RGHI is generated by statistically processing closely spaced Leeb Hardness data, and can be used to
quantify vertical changes of geomechanical rock properties. A broad range of averaged RGHI values have been
acquired based on the studies of various formations in North America, and this quantitative method is consistent with
core observations. A continuous RGHI surrogate can be obtained if a core is measured with high-resolution hardness
data. This RGHI surrogate has been successfully applied in GOHFER modeling of fracture height growth to better
account for ‘layered’ producing intervals.
Polygonal Fault System in the Paleogene of the Magallanes Foreland Basin, Southern Chile
1 1 1 1
J. A. Pinto; ; D. Gonzalez; ; P. Mella; ; A. Gonzalez; ;
1. Exploration, ENAP Magallanes, Punta Arenas, Chile.
Objectives/Scope: The purpose of the present work is twofold: (1) to describe the widespread and well-developed
polygonal fault systems (PFS), and (2) to highlight its potential effects on the Paleogene petroleum system of the
Basin. The data belong to three areas located in the central part of the basin, from west to east: the Brunswick
Peninsula (foredeep domain) and the Coiron-Dorado-Riquelme (CDR) and Arenal (AR) blocks (both slope domain).
The two blocks delimit important unconventional gas accumulations in Paleogene strata.
Methods/Procedures/Process: This work is based on the integration of 3D-seismic attribute analysis, mini-frac and
fracking data and drill core evaluation. Key reflectors that represent the base and top of the faulted interval were
interpreted in the seismic reflection dataset. Amplitude and variance maps were used to delineate the PFS and
construct fault trace maps, which were used to analize fault distribuition and orientations. Well data, including logs and
drill cores, provided the base for stratigraphic correlation, lithology and fault plane characteristics. Microfrac and
fracking data were used to evaluate the posible influence of the PFS in the size and orientation of induced fractures.
Results/Observations/Conclusions: The Paleogene depocenters of the Magallanes foreland basin contain thick marine
mudstones locally interbedded or flanked by sandstones. Within the Paleocene-Eocene succession, the deformed
interval sandwiched between relatively undisturbed and continuous reflections, shows extensional faults with small
throws (<40 m) that form polygonal patterns in map view. The development of this distinctive interval seems to be
diachronic: Paleocene in the foredeep and lower to upper Eocene in the AR and CDR blocks, respectively. The
polygons are rectangular to pentagonal in shape and range from 1.5 to more than 3 km across. Faults are normal and
have no preferred direction. In drill cores, single or conjugate set of faults occur and they display striations with no
filling material between fault planes.
2
Applications/Significance/Novelty: Because the PFS extend for more than 13000 km and their formation is
dependent, among other factors, on grain-size and lithology, they are useful as a basin-wide correlation tool. Although
mini-frac and fracking data suggest that the PFS exert no control over the size and orientation of induced hydraulic
fractures in the study area, they probably do affect the geometry and extension of reservoirs. As a result, PFS might
control production characteristics and fluid flow on a regional scale, so an understanding of the timing and distribution
of the PFS in the Magallanes basin may provide important clues not only about the petroleum migration and
compartmentalization history but also about future well and reservoir management for unconventional resource plays.
Developing Predictive Power in the Permian: Leveraging Advanced Petrophysics to Deliver Cash to the Business
1 1 1 1 1
A. Blount; ; T. Croft; ; M. Durand; ; B. Driskill; ; A. McMullen; ;
Objectives/Scope: The advent of Unconventional Resources has created a breadth of challenges for all subsurface
disciplines, and technical understanding has generally fallen behind the pace of development. Petrophysicists have
been repeatedly challenged and pressured to do more with less. In this paper, the authors detail and demonstrate
successful applications of petrophysical modeling that have generated significant predictive power and the ability to
quickly assess the rock quality and pay zones within the Permian Basin.
Methods/Procedures/Process: A multi-tiered petrophysical model is presented which shows a practical way to
evaluate wells and stitch together a more accurate subsurface picture of the field. The Permian Basin -- and many
other unconventional plays -- benefit from a significant legacy dataset that can be leveraged to infill data gaps;
however, this is regularly underutilized due to quality and vintage concerns. Through targeted pilots of advanced log
and core acquisition (including FMI, wellsite NMR, MDT, and pressurized sidewall coring), a calibrated petrophysical
model was created for the lowest common denominator -- the vintage triple combo -- allowing the practicing
petrophysicist to gain more granularity in subsurface evaluation and optimize field development.
Results/Observations/Conclusions: Advanced research and application of retort analysis shows a clear correlation
between the mobile/immobile core fluids and the water cut of producing wells, allowing operators to minimize the
detriment of water production. NMR -- both via wireline logging and on core samples -- can both predict the optimal
landing zone within a reservoir and provide a direct hydrocarbon indicator downhole. Pressurized sidewall coring
greatly increases the quantification of in-situ saturations. Log acquisition still has significant value, and the ability to
predict well performance increases greatly with proximity to quality data anchor points -- this will be fully
demonstrated. Data processing can be practically applied to logs of any vintage to infill holes in subsurface maps.
Applications/Significance/Novelty: The application of petrophysics is growing increasingly challenging as data
acquisition shrinks and lateral length grows; today's petrophysicist is tasked with predicting the performance of several
miles worth of rock in a single wellbore. However, the workflow presented shows how targeted, high quality data
acquisition in combination with a significant legacy dataset can be utilized to regularly predict the most productive
rock. This tiered petrophysical model allows operators to get useful subsurface information throughout the basin, while
still anchoring the model to the highest quality data and increasing the risk profile as distance from data increases.
Integrated Reservoir Characterization Aids Target Selection, Production Fluid Prediction, and Completions
Optimization in the Southern Delaware Basin Resource Plays
1 1 1 1 2 2 2
T. Tittlemier; ; J. Speight; ; S. Satterfield; ; C. Hager; ; I. Easow; ; B. Chiniwala; ; A. Martocchia; ;
1. Trey Resources, Midland, TX, United States.
2. Geolog Americas, Houston, TX, United States.
Objectives/Scope: The Southern Delaware basin has proven to be prolific for production of liquid rich hydrocarbons
with operators targeting sweet spots in the various layers of the Wolfcamp and deeper Pennsylvanian to Devonian. In
these deeper formations, the thermal maturity of the source rocks increases and the reservoir fluid expected is
predominantly Gas. However, there have been indications of potential liquid rich production in specific geological,
structural and thermal maturity settings, but data points are limited and not many laterals have been drilled, evaluated
and completed yet.
Methods/Procedures/Process: Recent log and core evaluations have shown presence of natural micro and large
aperture fractures in the Wolfcamp and other formations in the basin. These fractures can potentially enhance
production and sometimes cause challenges to drilling processes. Operators can utilize superior technology from
advanced surface logging to characterize many important aspects of formation evaluation which can be performed
while drilling. Cost-effective technologies utilized in this study include Advanced mud gas geochemistry, Fracture
detection from mud flow, Chemostratigraphy and source rock evaluation. These data were then combined with open
hole wireline logs to determine the best approach to identify targets and plan completions.
Results/Observations/Conclusions: This paper will demonstrate how an integrated approach to reservoir
characterization was utilized by an operator in the southern Delaware basin to evaluate their prospect in a vertical pilot
and 2 laterals targeting the Paleozoic formations. By utilizing the integrated reservoir characterization approach the
operator selected the best points for the landing the laterals in both the Wolfcamp and the Woodford by evaluating the
reservoir quality, production potential reservoir compartmentalization, source rock character, identifying micro-
fractures (compared to and interpreted Image log), and estimating reservoir fluid type and production GOR. While the
techniques were used to aid geo-steering the laterals and provide significant inputs for effective completions.
Applications/Significance/Novelty: This integrated approach is a non-intrusive and cost-effective alternative for
superior quality formation evaluation which can aid operators to map regional hydrocarbon distribution, location of
fracture prone reservoirs, predict porosity and permeability related to fractures, identify faults, estimate GOR or CGR,
identify potentially higher water production and plan future drilling and completions strategies for targeting primary and
secondary targets within the Southern Delaware basin and other areas in the greater Permian basin.
Integrated Rock Characterization of a Shale Gas Field in the Horn River Basin, Canada
1 1 2, 1
J. L. Miskimins; ; T. Teklu; ; D. Park; ; H. Jung; 2, 1;
1. Colorado School of Mines, Golden, CO, United States.
2. KOGAS, Daegu, Korea (the Republic of).
Objectives/Scope: Integrated rock characterization of a shale gas field in the Horn River basin, Canada, is studied.
Three main gas bearing shale lithofacies, the Muskwa, Otter Park, and Evie, are productive. The study objectives are
twofold: (1) compare and correlate porosities and permeabilities of crushed cores vs. intact cores measurements to
assess information on the natural micro and nano fractures; and (2) to correlate well log data, mineralogy, organic
matter, etc. with matrix and fracture porosity and permeability, fracability, and to identify sweet-spots and optimize
production of the field.
Methods/Procedures/Process: Permeability and porosity measurements of more than forty intact shale cores were
performed at net stress conditions between 1000 to 5000 psia. The stress dependent permeabilities of each cores
were determined and correlated with mineralogy, gamma ray, organic matter, and crushed core (matrix) permeability
and porosity. SEM image of representative core samples were also performed to assess micro and nano fractures.
Since the intact and crushed core measurements were performed on samples taken from the same depths, the
contrast provide excellent tools to identify the presence of nano/micro fractures and multiple-scale permeabilities that
will be very useful when designing hydraulic fracturing treatments in these shale formations.
Results/Observations/Conclusions: Stress dependent permeability and porosity were measured and correlated to
micro/nano fractures, geomechanics, and fracability.
Petrophysical data, SEM images, and crushed vs. intact core samples porosity and permeability information were
used to characterize matrix and natural fractures; which are useful in reservoir simulation and hydraulic fracturing
design optimization.
Crushed core porosities were about five times higher than corresponding intact core porosities. The reason is that the
crushed-core porosity measurments include nano and micro pores (pore diameters of 50 nm or less) that are not
easily accessible when testing intact-core samples. On the other hand, intact-core permeabilities exhibited two or
more orders of magnitude higher than corresponding crushed-core permeabilities.
Applications/Significance/Novelty: Crushed core vs. intact core porosity and permeability measurements in
conjunction with SEM image, geomechanical and other petrophysical data provided a new tool to identify the presence
of sweet-spots (nano and micro natural fractures).
Integrated mineralogy, organic matter, and petrophysical data were used to determine the fracability of the shale
formations to better design hydraulic fracturing treatments in the main three shale formations of the Horn River basin,
Canada.
Stress dependent permeability and porosity measurements are correlated to SEM micro/nano fractures,
geomechanics, and fracability.
Quantifying Nanoporosity: Insights Revealed by Parallel and Multiscale Analyses
1 1 1 2 2
A. C. Reynolds; ; S. A. Kelly; ; R. J. Bonnie; ; J. J. Howard; ; R. L. Krumm; ;
1. ConocoPhillips, Houston, TX, United States.
2. Premier Oil Field Laboratories, Houston, TX, United States.
Objectives/Scope: Over the last decade, multiple studies have outlined the challenges of adapting core analyses to
unconventional rocks, from challenges in measuring and modeling very small, high surface area pores to round robin
discrepancies among laboratories. Multiple authors have demonstrated the challenges of characterizing mobility in
nanoporous systems: from the scale and variability of the pore system to the change in fluid composition as it
reequilibrates to surface conditions. This study used a number of different parallel and multiscale analyses to
characterize a broad variety of West Texas samples.
Methods/Procedures/Process: This study is designed to compare current core analyses for nanoporous rocks, identify
strengths and limitations of each analysis, and to design appropriate upscaling workflows to compare analyses. We
examined 50+ samples from West Texas with a range in mineralogy and organic matter. The integration of SEM, N
2
adsorption, NMR, GRI, HPMI, and thin section data on each of these samples enabled an understanding of pore size
distributions, pore types and connectedness, and inferred wettability.
Results/Observations/Conclusions: Results from the analytical comparisons – both where they agree and disagree –
reveal insights into unconventional pore systems. We find that total porosity is well constrained by a tight agreement
(± 1 p.u.) between crushed rock helium porosimetry and plug NMR, confirming the methodologies for total pore
volume. Low-field NMR relaxation measurements at several laboratory-controlled liquid saturation states assist in
understanding potential liquid volumes at reservoir conditions and the nature of the liquid-pore wall interactions that
reflect wettability behavior. High resolution imaging is used to calibrate the distribution of pores and organic matter at
multiple scales.
Applications/Significance/Novelty: The wettability state of organic-rich low-permeability unconventional reservoir
samples is unclear since many samples spontaneously imbibe both water and light oil. While this may signify a neutral
wetting state, another interpretation is that two pore systems, one water-wet and the other oil-wet, reside adjacent to
each other in these rocks. This study focuses on whole rock samples rather than mineral/organic isolates and employs
parallel and multiscale analyses to pinpoint how each method informs aspects of reservoir quality. Take-aways include
a stronger understanding of analytical capabilities and upscaling in organic-rich unconventional reservoirs.
Quantifying the Mechanisms Contributing to Surface Relaxation of Protons in Organic Pores of Organic-Rich
Mudrocks
1 1
S. Tandon; ; Z. Heidari; ;
1. Department of Petroleum and Geosystems Engineering, The University of Texas at Austin, Austin, TX, United
States.
Objectives/Scope: Two dimensional (2D) Nuclear Magnetic Resonance (NMR) measurements can enhance the
quantification of petrophysical properties in organic-rich mudrocks. However, the interpretation of these
measurements is still challenging as the mechanisms of proton surface relaxation in organic pores are not well
understood. The objectives of this paper are to quantify the effects of homonuclear dipolar coupling on proton surface
relaxation and 2D NMR responses, to develop a new model for surface relaxivity in organic pores, and to apply it for
simulation of NMR responses in organic-rich mudrocks.
Methods/Procedures/Process: We start by extracting pure kerogen from different samples of organic-rich mudrocks.
The specific surface areas of kerogen packs were measured using Brunauer–Emmett–Teller (BET) analysis. We then
conducted transverse relaxation (T ), longitudinal and transverse relaxation (T -T ), and T -T measurements on the
2 1 2 2 2
extracted kerogen samples using both polar (water) and non-polar (decane) solvents. We used the results of NMR
measurements to develop a new model for proton surface relaxation considering the effect of homonuclear dipolar
coupling in kerogen pores. The new model was then used to simulate NMR T and T -T measurements in pore-
2 1 2
scale images of organic-rich mudrocks using our previously developed finite-volume numerical simulator.
Results/Observations/Conclusions: We performed NMR measurements on 6 pure kerogen samples at different
thermal maturities. Using BET surface areas and NMR measurements, we obtained kerogen surface relaxivities of 2-
15 µm/s and 1-6 µm/s for decane and water, respectively. T -T measurements show T /T values of 1.1-2 and 2-6
1 2 1 2
for decane and water, respectively, demonstrating the impact of homonuclear dipolar coupling on surface relaxation in
kerogen pores. The T -T measurements yielded 2-3 off-diagonal peaks for both water and decane indicating the
2 2
effects of inter-molecular dipolar coupling during surface relaxation in organic pores. The results of numerical
simulations show that misidentifying the coupling mechanism can lead to an error of up to 25% in the assessment of
hydrocarbon pore-volume in organic-rich mudrocks.
Applications/Significance/Novelty: In this paper, we developed a new surface relaxivity model using an experimental
approach that improves the previous models by including the effects of both inter-and intra-molecular dipolar coupling
on surface relaxation in organic pores. This new model quantifies the impacts of inter-and intra-molecular dipolar
coupling on surface relaxation in organic-rich mudrocks and can be used for quantifying surface relaxivity at higher
temperatures and fluid viscosities, which enables interpretation of NMR logs in situ condition. Enhanced quantification
of surface relaxivity also improves interpretation of both lab- and log-scale NMR measurements, enhances NMR-
based reservoir characterization, and helps to improve estimates of hydrocarbon reserves in organic-rich mudrocks.
Development of Raman Spectroscopy as a Thermal Maturity Proxy in Unconventional Resource Assessment
1 1 2
G. Myers; ; K. Kehoe; ; P. Hackley; ;
1. WellDog, Laramie, WY, United States.
2. USGS, Reston, VA, United States.
Objectives/Scope: The objective of this study was to correlate shale hydrous pyrolysis with measurements of solid
bitumen reflectance (BRo) at the U. S. Geological Survey (USGS) and Raman microscopy (RM) at WellDog. In
semiblind Phase I, BRo values of the initial set of 8 samples were withheld prior to RM analysis. As reported
previously, a strong correlation was observed between BRo and Raman parameters. For Phase-II, BRo values for the
second set of 8 samples were shared before RM. Observations from Phase-II are reported here as well as the ability
of RM to correctly order the semi-blind Phase I samples.
Methods/Procedures/Process: Immature shale samples from the Bakken (Phase-I) and Duvernay (Phase-II)
formations were subjected to hydrous pyrolysis for 72 hours at temperatures from 280°C to 360°C. Rock residues
from both series were mounted and polished (ASTM D2797) for analysis of BRo (ASTM D7708) and confocal
laserscanning Raman microscopy. For RM, multiple hyperspectral maps were collected from each sample, resulting in
tens of thousands of spectra per sample. Map areas were ~5,000 μm2, with a spectrum collected from every square
micrometer. The organic carbon G- (Graphitic-) and D- (Disordered) bands in each Raman spectrum were fit
algorithmically to a multi-peak model, yielding a number of diagnostic parameters that correlate with changes
occurring in samples as a result of thermal maturation and pyrolysis.
Results/Observations/Conclusions: Parameters extracted from analysis of Raman spectra were plotted against the
previously determined BRo values to determine which Raman parameters best correlate with thermal maturity.
Plotting two of the sample-averaged anonymized spectral parameters versus BRo in the Bakken series indicated an
exponential trend with strong correlations (R2>0.8) as reported at URTeC in 2017 (MS-2671253). Similar strong
relationships occurred in the Duvernay samples with respect to increasing maturity. Significantly, the magnitude of
differences among measured Raman parameters in low-maturity samples is greater than the differences in measured
reflectance for the same samples, suggesting RM may be a better discriminant of shale thermal maturity in the pre- to
mid-oil window.
Applications/Significance/Novelty: Quantifying thermal maturity is crucial for properly evaluating an unconventional
resource. However, the most widely used methods–vitrinite reflectance and programmed pyrolysis–may be inadequate
for shales. Reflectance of solid bitumen has wider applicability in shales, because of greater abundance than vitrinite.
But additional techniques that can accurately and objectively characterize the thermal maturity of shale samples are
still needed. The promising results from this study will act as the foundation for a future Phase-III study, in which the
same methods and analyses will be applied to a series of samples from the Lower Huron Member of the Ohio Shale
that have been naturally matured at varying depths across the northern Appalachian Basin.
Spatial Characterization of Organic Matter Maturity by Raman Microscope Mapping
1 1
E. G. Krukowski; ; J. J. Howard; ;
1. Premier Oilfield Laboratories, Houston, TX, United States.
Objectives/Scope: There is a strong correlation between the ordered and disordered C-C peaks observed in Raman
spectra and conventional vitrinite reflectance values that encourage the use of Raman microscopes to characterize
micron-sized particles that are not detected with optical methods. The spatial distribution of minerals with respect to
pores and organic matter is examined as an explanation to the range of maturity index values that are observed within
an individual sample.
Methods/Procedures/Process: A Raman-enabled microscope was used to collect large field-of-view images of
organic-rich shale at micron resolution with the use of an automated stage. The acquired spectra were processed with
standard spectral analysis software (LabSpec6) that resolved the ordered and disordered carbon bands used to
estimate maturity. The non-organic phases were identified with peak-match algorithms that used an in-house mineral
database. Mineral maps were compiled with a Classical Least Squares (CLS) fitting or a Principal Component
Analysis (PCA). These maps were often compared with SEM-based mineral maps acquired over the same region.
Results/Observations/Conclusions: The Raman mapping approach generated maturity values that compared
favorably with vitrinite reflectance and Rock-Eval-based indices. The range in maturity values within a single field-of-
view is linked to whether the dominant mineralogy that enclosed the particle was carbonate or silicate-rich. The local
environment in a fine-grain rock influences the extent of organic-matter diagenesis.
Applications/Significance/Novelty: Raman microscope methods to characterize maturity of organic matter adds the
capability of measuring smaller particles in an automated process that limits subjective sampling strategies. The ability
to characterize the surrounding matrix and pore space for each organic matter particle opens up new ways to study
diagenesis in these low-permeability fine-grain reservoir rocks.
New Mexico Delaware Basin Horizontal Well Heel Frac and Refrac Program and Hydraulic Fracture Diagnostics
1 1 1 1
M. Han; ; I. Tanakov; ; E. Bunker; ; T. Vulgamore; ;
1. Occidental Oil and Gas, Houston, TX, United States.
Objectives/Scope: Previous completion strategies of 2nd Bone Spring horizontal wells in New Mexico Delaware Basin
have left pay behind pipe by only perforating and stimulating up to 80 degrees inclination at the heel. Completion of
the heel pay (heel frac) has been implemented and has been shown to provide economic productivity. In the latest
heel frac programs, re-stimulation of the last existing stages were added. Proppant and chemical tracer were used in
order to understand the stimulation and diversion effectiveness and evaluate the incremental production gained from
the new heel and refrac stages.
Methods/Procedures/Process: The original heel frac candidate list was selected by modeling hard line intersections
with perf data and the final list was filtered down based on IPs, WI, proppant loading, lift constraints, etc. In this study,
comprehensive post-job hydraulic fracture diagnostics were done on three 2nd Bone Spring horizontal wells by a
multi-discipline group. Proppant tracer results and stimulation treatment data were integrated to diagnose
abnormalities such as having unstimulated clusters and creating propped fracture growth out of the intended zones.
Geology and Petrophysics data were evaluated in the sections with effective and ineffective clusters while post-
production and chemical tracer samples were monitored to provide insights on hydrocarbons gained from each stage.
Results/Observations/Conclusions: Early chemical tracer results showed that re-fracturing the last existing stage
added economic production. Proppant tracer results showed that stimulation diversion was not very effective in the
refrac stage. Unstimulated clusters and propped fracture growth out of the intended zones were also observed in
some stages. Alongside this, both fracture treatment behavior and proppant tracers suggested communication behind
pipe between stages where it is suspected that bad cement and changes in rock properties along the wellbore may
have contributed to the stimulation ineffectiveness. These case studies have shown that the integrated fracture
diagnostics provide an understanding on stimulation effectiveness, causes of ineffective stimulation, and the impact on
production/economics.
Applications/Significance/Novelty: Operators normally perforate and stimulate up to 80 degrees inclination at the heel
due to execution challenges. To our best knowledge, this is the first study sharing challenges, results and learnings on
stimulation above 80 degrees inclination and its combination with re-stimulation stages. The techniques used in this
study enabled both qualitative assessment and quantitative assessment on zonal contribution, which was crucial for
future business decisions. Lessons learned from this study will be applied to optimize future heel frac/refrac designs
and assist with candidate selection and data gathering. Alongside this, this study will be one of the few published
nd
integrated analyses on a modern frac design applied in the New Mexico Delaware Basin 2 Bone Spring Sand.
Optimization of Completion and Well Spacing for Development of Multi-stacked Reservoirs Using Integration of Data
Analytics, Geomechanics, and Reservoir Flow Modeling
1 1 1 1
K. Min; ; V. Sen; ; L. Ji; ; R. Sullivan; ;
1. Anadarko Petroleum Corp, The Woodlands, TX, United States.
Objectives/Scope: In multi-stacked unconventional development, optimization of completion and well spacing is a

major challenge. Because of complex interactions between tightly placed wells in different target zones, it is necessary
to model the interference between the wells together with depletion and flow across zones. This integrated scenario
modeling is a pre-requisite for full field development and we have developed a new workflow combining statistical
methods geomechanics and reservoir modeling. Using this hybrid method, we can efficiently identify key drivers for
completion and well spacing optimization
Methods/Procedures/Process: Our method involves statistically selecting multiple single wells from our portfolio, using
them in Dynamic SRV Modeling (using geomechanically influenced permeabilities during fracturing and production) to
build prediction functions and then validating them against actual multi-well field data. For real life development
scenarios for multi-stacked reservoirs, we applied Design of Experiment (DoE) method and response surface
methodology (RSM) to screen and rank scenarios. For completion optimization, multiple production profiles predicted
from simulated DoE cases were used to generate response surfaces which were used to predict the outcome from
various permutations and combinations of completion parameters and well spacings in a multi-stacked configuration
and fed into economic analyses.
Results/Observations/Conclusions: Through this study, we have identified that the relevance of key completion drivers
actually change over production periods as well as well spacing. Thus, some completion parameters have a high
impact at only early times, while other completion parameters dominate for later production. Furthermore, when well
spacing and targeting are changed, the key completion derivers may also change. This demonstrates to us that
completion design, targeting and well spacing should be optimized based on the field-wide development plan,
honoring interaction of wells during growth as well as draining of the SRV’s, with appropriate representation of the
geomechanical nature of the reservoir. Our integrated workflow demonstrated that data analytics makes us much
faster calibration and prediction.
Applications/Significance/Novelty: Optimization based on multi-well error functions safeguards against the fatal flaw of
building rigorously tuned singe well functions, which are very limited and sometimes misleading in their ability to
predict beyond and extrapolate. Acknowledging that all inverse problems are beset by the issue of know unknowns vs.
unknown unknowns, we also show that it is necessary to avoid certain types of over-parameterization. We therefore
built a workflow comprising appropriate physical proxies, multiple scenario modeling and advanced analytics.
Following the same approach, users will be able to build effective workflows for directional guidance in completion
design and full field development in tight rock reservoirs.
Accelerated Stimulation Optimization via Permanent and Continuous Production Monitoring Using Fiber Optic
1 1 1 1
G. A. Ugueto; ; P. Huckabee; ; M. Wojtaszek; ; A. Reynolds; ;
1. Shell, Houston, CO, United States.
Objectives/Scope: Since the start of the “Shale Revolution” more than ten years ago, hydraulic fracturing stimulation
providers have offered operators an increasing number of completion technologies and stimulation designs. The
evolution of completion practices reveals a trend towards increasing stimulation intensity in all the unconventional
plays. Although the result of this trend has generally been more productive wells, the optimization of completion and
stimulation practices have been slow and in many cases resulted in significant over-capitalization.
Methods/Procedures/Process: The question we need to answer as operators is: Which combination of completion and
stimulation options is the most cost effective and maximizes economic value? The traditional approach of industry
towards stimulation de-risking has been through the implementation of “trials” comparing the well performance of
several wells, completed with the new technology, against the production of a group of “reference” wells. While this
empirical approach appears to be relatively straight forward, there are some obvious challenges associated with
sample size and the cycle-time duration of these tests. Typically, this approach also requires a relatively large number
of wells to account for uncertainties associated with the well-to-well subsurface variability.
Results/Observations/Conclusions: This paper outlines an alternative approach that can performed in few wells, that
allows to accelerate the optimization process. It includes experimental design considerations, acquisition of stimulation
distribution effectiveness information, as well as, production data obtained by continuously monitoring a well using
Fiber Optics (FO). In the example, we used Distributed Acoustic Sensing (DAS) to derive discrete production profiling
results over a period of more than two years. This has enabled us to track the gas production of two competing
completion designs in a single well with minimum production deferral and HSE exposure.
Applications/Significance/Novelty: Continuously monitoring production of each stage, perforation cluster on sleeve-
entry opens the possibility of testing different completion technologies or designs by comparing the performance of
each completion type in different segments of the same well. This not only enables to accelerate the completion
optimization process and reduce overcapitalization risk but also the FO can be interrogated to investigate other
aspects of the stimulation that can impact stimulation quality and production results. This alternative approach towards
completion de-risking using few wells and continuous FO monitoring provides critical information that could make the
evaluation of stimulation technology faster, with higher confidence, and more cost effective than the traditional
optimization methods.
Multivariate Study of Utica: Marrying the Rock to the Completion
1 1 1
A. Trumbo; ; J. Bowman; ; L. Lasecki; ;
1. Chesapeake Energy, Oklahoma City, OK, United States.
Objectives/Scope: Multi-variate statistics was employed in the Utica to provide guidance for optimal completion
techniques related to varying rock type. In the past, this method has been used for sweet spot mapping, but in
unconventionals today, all acreage will be drilled and completed regardless of whether it is in the sweet spot or not.
The more powerful use of this approach is marrying the rock to the completion. Engineering, geologic, and
geophysical parameters were used to create a model with which we were able to test different completion methods in
areas of varying geology.
Methods/Procedures/Process: A large portion of the project was spent on data collection, loading, and QC before the
multi-variate workflow began and the predictive model was created. The workflow allows analysis of how each
variable interacts with other input and response variables as well as any data biases or trends. An unbiased and
robust predictive model requires sufficient variability for each attribute; including competitor data. Geologic clustering
was used to identify “like” geologic areas to help prioritize key subsurface play driver relationships. Extractions of
geologic maps, seismic attributes, and engineering data were then made to the wellbores to be compared with a well
performance indicator. Ultimately, a subset of wells within three engineering type curve areas was chosen to model
AsqrtK/ft.
Results/Observations/Conclusions: The main outputs from the project were: confirmation of key play driver
relationships, well optimization tool marrying geology with engineering, and predicted AsqrtK/ft. While the most
appealing output of the model are the predicted AsqrtK/ft maps, the model can be used to optimize each well on the
drill schedule. The geologic variables in the model become fixed for the pad based on well log measurements and
Monte Carlo simulation is used to show the probability distribution of different completion types. This led us to change
the completion on a pad and resulted in 50% uplift in AsqrtK/ft which is the highest in the field.
Applications/Significance/Novelty: The study was a multi-discipline look at how a reservoir produces and what we
could do to best exploit the asset. In unconventional reservoirs, production is driven by a complex interplay of geology
and engineering. The multi-variate model gives us our best understanding of interplay between rock and completion.
This study has highlighted areas of the play that have unrealized upside potential.
Interdisciplinarity (Operators’ Forum only): The model was able to integrate data from all disciplines to explain
productivity based on rock AND completion.
Unconventional Field Development Optimization – Fit for Strategy Designs to Realize Your Corporate Goals
1 1
R. Howrish; ; D. Anderson; ;
1. Anderson Thompson Reservoir Strategies, Houston, TX, United States.
Objectives/Scope: Shareholder or stakeholder value is created in the oil and gas industry in multiple ways. This paper
will demonstrate that no single evaluation metric can be used to adequately compare companies with dramatically
different corporate strategies. Strategies range from proving acreage to sell with a minimum number of wells and
capital investment to maximizing long term resource recovery. This concept is also of importance for non-operated
partners to ensure their strategy and organizational capability are aligned with the operator.
Methods/Procedures/Process: We selected one operator with a unique execution strategy relative to its offset peers
and evaluated their capital program, completion design, and well performance results using a combination of public
and proprietary data. The evaluation includes regression and cluster modeling to identify performance drivers and high
graded areas, hydraulic fracture modeling to define fracture geometries associated with different stimulation designs,
probabilistic production forecasting, and economic modeling single wells. Based on the resulting model, a wide range
of full field development scenarios are generated to investigate the impact on the chosen business and economic
metrics based on varying inputs to the full field scenarios
Results/Observations/Conclusions: The value metrics utilized to evaluate the various strategies include net present
value, IRR, discounted profitability index, payback period, free cash flow generation, reserve additions, annual
production growth, and resource recovery factor. The results show that no single field development strategy can
maximize all of these metrics. Thus, an operator’s objective value metric must be clearly identified prior to evaluating a
potential strategy, and it must be clearly known by all stakeholders in the process. Strategies that favor the best single
well metrics ultimately have a negative impact on some or all metrics that are maximized in a full field development
scenario while a strategy that maximizes resource recovery could have a negative impact on capital efficiency related
metrics.
Applications/Significance/Novelty: To our knowledge, there is no existing paper that links the requirement of
understanding corporate strategy to unconventional field development optimization. This work highlights the
importance of having the interests all stakeholders clearly aligned in a field development design or optimization
project. It also demonstrates the potential folly of apples-to-oranges comparisons among operators and their peers, if
they do not share a common metric for defining value creation.
Is GOR Truly Affecting Recovery? A Multi-variate Case Study in the Delaware Basin
1 1 1
R. Dutta; ; S. Dawson; ; M. Maler; ;
1. Drillinginfo, Houston, TX, United States.
Objectives/Scope: The Permian basin has been a hotbed of activity, with huge potential in previously untapped
reservoirs. As this continues to grow, questions on the accuracy of EUR predictions continue to linger. Citing an
increase in initial GORs, recent articles have concluded that oil wells will reach their economic limit sooner than
expected. This has led to further scrutiny of empirical forecasting methods for Permian oil wells. To understand the
effect of GOR on recovery, we focus on the Wolfcamp A of the Delaware basin and examine both geologic conditions
and production history which influence GOR.
Methods/Procedures/Process: GOR variations, both temporal and spatial, can be due to processes that occur on both
production and geologic scales: reservoir depletion, uplift and unroofing, source rock maturity, etc. In the Delaware
basin, the geology and the drilling history combine to yield an overall increasing GOR trend for wells completed in the
Wolfcamp A. We start by identifying the geological and petrophysical properties that can be used to predict GOR. This
nonlinear multi-variate model includes attributes such as temperature, TOC, and deep resistivity and is calibrated
using produced 6-month GOR as the independent response variable. We then leverage this predicted GOR model,
particularly spatial variations, to better quantify locally recoverable volumes and consequently EURs as a function of
GOR variability.
Results/Observations/Conclusions: The model predicts higher GOR in the western part of the basin, agreeing with
existing studies citing the higher geothermal gradient of the Diablo Platform. As Wolfcamp A drilling has been
marching westward, the wells have been increasingly exposed to reservoirs with higher proportions of solution gas.
Thus, little can be inferred about reservoir management with respect to GOR trends until the larger geological
overprint is removed. We focus on a core area with similar predicted GOR and the earliest drilling history and show a
correlation between rising GORs and decreasing recovery. As gas continues to come out of solution, we highlight the
uncertainty in conventional empirical production forecasting methods and the importance of using a predictive GOR
model in forecasting volumes.
Applications/Significance/Novelty: As operators continue to plan future activity in the Permian basin, it is important to
understand GOR in the context of both the basin history and production history. Understanding pressure depletion in
the area will help plan for better reservoir management, consequently affecting oil recovery. This study presents a
novel multi-variate approach to quickly understand GOR spatial variations purely from geological and petrophysical
observations rather than an elaborate forward model. This predicted GOR model can then be used to better delineate
multiple type curves and thus better quantify resource estimates. Reducing uncertainty in empirical forecasting
methods is important for avoiding resource over- and under-estimation.
A New Look at Reserves Estimation of Unconventional Gas Reservoirs
1 1 1
O. Mahmoud; ; M. H. Ibrahim; ; C. Pieprzica; ;
1. Apache Corporation, San Antonio, TX, United States.
2. Texas A&M University, College Stataion, TX, United States.
Objectives/Scope: Forecasting future production and estimating ultimate recovery (EUR) in supertight reservoirs and
shale plays have long been problematic. Developing a reliable and more accurate production forecast have always
been a main goal of any petroleum operation. Effectively assessing the reservoir volume and well producing life and
deciding development scenarios and strategies are directly depending on the accuracy of the forecast. This study is
intended to review and compare the methods and models developed and used for such a reason.
Methods/Procedures/Process: Reserve estimation is a process that is thoroughly renewed during the life of a
reservoir. Its accuracy depends on the amount of data available and the method of forecast. Analytical models or Rate
Transient Analysis (RTA) methods are widely used for production forecast of unconventional reservoirs. Numerical
simulation is also used for predicting EUR. Different relations have been introduced to model the rate/time behavior in
unconventional plays as alternatives for traditional Arp’s decline curve analysis. Modified Hyperbolic decline, Power
Law Exponential Decline (PLE), Stretched Exponential Decline (SEPD), Duong method, and Logistic Growth Model
(LGM) are developed for forecasting the production in shale reservoirs, but all are based on empirical observations of
a particular scenario.
Results/Observations/Conclusions: In this study, different methods of history matching the production of hydraulically
fractured unconventional reservoirs were investigated by forecasting future production and predicting EUR. Different
synthetic scenarios and actual field data were used. The traditional Arp’s decline over-forecasting reserves. PLE,
SEPD, LGM, and Duong method were intended to represent the character of rate/time production data for the
standard well completion in a multiple-fractured horizontal. These methods provide different forecasts as they have
different equation forms and can't forecast production for all unconventional reservoirs. The RTA analytical models
required different modifications of the reservoir and fracture parameters to provide optimistic EUR when compared to
the numerical simulation.
Applications/Significance/Novelty: Different methods for forecasting unconventional well data have been reviewed and
compared in this work based on the production forecast and EUR prediction. Both synthetic and field case production
data have been used to reveal the goodness of fit, the similarity of reserve estimation, and relationship to the reservoir
theory.
Global Competitiveness of the US Tight Oil Cost Curve
1
R. G. Clarke; ;
1. Upstream Research, Wood Mackenzie, Houston, TX, United States.
Objectives/Scope: Lower 48 tight oil barrels are some of the lowest cost, new drill supply in the world. Many of the
best operators have become virtual Permian tight oil pure-plays, while others industry stalwarts have completely
avoided the resource theme. What project-specific nuances could be contributing to the division?
Methods/Procedures/Process: Wood Mackenzie models specific tight oil projects of over 150 US Lower 48 operators.
The team has constructed - and continually updates - discounted cash flow asset models for over 500 tight oil assets
domestically. The post-taxbreakevens figures are output from those models and are calculated for future, yet-to-drill
wells.
Results/Observations/Conclusions: Permian breakevens – particularly in the Delaware – broke below the US$40/bbl
breakeven threshold in 2017. Wood Mackenzie models indicate that tighter cluster spacing was a critical variable in
that reduction. Even though the longer-term impact on well EURs have is yet to be fully seen, next generation
completions resulted in much higher cumulative 180-day volumes and shorter well payback periods.
Large breakeven reductions were also seen in the STACK Mid-con play where average well performance improved as
the number of failed completions fell rapidly. We attribute this to better basin models, more robust multivariate analysis
for appraisal projects, and higher pump rates used during completions.
Applications/Significance/Novelty: Low-cost barrels today may not stay that way . Well designs and completion
formulas are no doubt continually evolving, but cost structures moving forward will be dynamic too with varying
degrees of inflation seen across the supply chain. Additional productivity gains can offset some of this, but technology
advancements will be massively critical as sweet spots become exhausted and less productive reservoir zones
become reclassified as core areas.
Reserve Estimation With Unified Production Analysis
1
M. Mehana; ;
1. MPGE, University of Oklahoma, Norman, OK, United States.
Objectives/Scope: Reserve estimation is a topic of continuous importance to the petroleum industry. The
establishment of a reserve estimation method with probabilistic guidelines acceptable to the Petroleum Reserve
Management System (PRMS) for the brunt of US actual current production – multi-fractured horizontals is our
objective.The challenge of developing estimation method is largely due to the completion inhomogeneity. For
homogeneous completions, only two regimes are required to quantify and predict performance: infinite acting linear
and boundary-dominated flow. However, this is not valid for shale plays.
Methods/Procedures/Process: Our method involves determination of flow regime features that can be statistically
characterized. Unification is obtained by utilizing this characterization from mature wells. The flow regime features can
be stochastically input in order to generate proved, probable and possible reserves once convergence is reached.
Monte Carlo simulation is utilized to determine reserves for all producing wells. A field study was performed in the
Bone Springs formation of the Delaware Basin. Flow regime characteristics determined from monthly production data
are presented.
Our results include reserve reductions when young wells are allowed to be trained by other wells in the formation of
greater longevity rather than relying on the individual well’s production alone. We also show that for inhomogeneous
stimulations a correlation can be made between the terminal Arps’ hyperbolic exponents and the number of flow
regimes present.
Applications/Significance/Novelty: The unified process provides a path for the determination of an individual well’s
production forecast to be aided with offset formation wells with more production experiences. Capturing flow regime
features and stochastic predictions fulfills our objective of reserve determination for multi-fractured horizontals.
New Analysis of EUR Probability Plots Yields Better Uncertainty Assessment and Better Type Wells
1
R. Freeborn; ;
1. 3 esi-Enersight, Calgary, AB, Canada.
Objectives/Scope: Evaluators use a lognormal probability distribution of estimated ultimate recovery (EUR) to predict
the EUR that may be expected when drilling more wells. The expected range is captured by the difference or ratio of
P10 and P90 EUR. Even with many analog wells, the analog and population distribution may differ, creating
meaningful risk usually absent when using distributions. A statistical method is used to quantify the possible deviation
between analog and population distributions (risk) and to construct risk adjusted type wells for use in drilling and
reserves evaluations.
Methods/Procedures/Process: The process uses Monte Carlo simulation to select EURs from known population
distributions and create trial distributions of the selections. The population P50 EUR is the P50 from analog wells, the
P10/P90 range is random and the selection count is the number of analog wells.
We select trials with near equal sample and analog P10/P90 ratio. A population scatter plot of P50 versus P10/P90
ratio for this subset shows the range and frequency of possible population distributions. The EUR distribution from all
wells selected for this subset will be a close estimate to the risked true population.
Analog samples are properly positioned on the risked population distribution to determine the probability for each
analog sample. Reserves and type well construction follow from this distribution.
Results/Observations/Conclusions: We test our observation that risk is understated using SPEE Monograph 3. For
P10/P90 = 4, 60 wells will obtain reliable results. Our analysis obtains a range of +/- 25% for the mean and +/- 10% for
the ratio P10/P90 at 80% confidence, values we do not consider reliable.
Probability distributions are used for building aggregation type wells (SPE 175967). With many wells, the P10/P90
ratio is small (no risk), but this is not observed in practice. Employing the methods described herein, other risk is
captured and persists for large numbers of wells.
We find that analog distributions are unlikely to represent the population due to non-uniform sampling. The proposed
method works to identify the uncertainty, correct distributions to align with the population and obtain better
evaluations.
Applications/Significance/Novelty: Our application is to obtain more reliable type wells that do not reflect an absence
of risk for drilling many wells in the future. Because the process is Monte Carlo based, errors in calculating EUR are
also incorporated in the analysis, providing those errors are not systemic.
When statisticians construct probability or frequency distributions there is only one uncertain variable. The methods
described herein identify the equivalent of a frequency distribution for two inter-related variables that are not easily
analysed using multivariate analysis.
The Value of Building a Multiscale, Regional Geomodel for Reserves Assessment of the Midland Basin
1 1 1 1
R. Dommisse; ; L. Sivila; ; H. Hamlin; ; F. Male; ;
Objectives/Scope: Modern resource assessment approaches can be greatly aided by incorporating all available data
and interpretations in a three dimensional geomodel. Our goal is to offer a regional perspective to augment the
interpretations from local, field-scale 3D models developed by the industry. In this work we highlight the benefits of the
continuous development of the geomodel for reserves assessment in an unconventional play.
Methods/Procedures/Process: We generated a three dimensional, faulted Midland Basin geomodel – containing over
1 billion cells, including stratigraphic, petrophysical, core description, and production data for the Spraberry and
Wolfcamp intervals. The model is based on over 1500 correlated wells, 700 wells with petrophysical and facies
interpretations and approximately 6000 horizontal production wells with decline curve and completion data analyses.
Results/Observations/Conclusions: There are three areas of focus for which we present results: 1) Testing of the
facies model derived from log and core analyses using different deterministic and stochastic attribute distribution
techniques 2) Characterizing the relationship between data density requirements and model predictability 3) Exploring
the influence of geological trends on productivity.
Applications/Significance/Novelty: We show the value of providing a multiscale, regional perspective to the practice of
3D reserves assessment geomodeling in the Midland Basin.
Modeling and Experimental Investigation of Fluid-Related Damage to Hydraulic Fractures
1 1 1 2 3 3 3 1
P. Abivin; ; R. Prabhu; ; D. Khvostichenko; ; C. Hilliard; ; C. Nelson; ; T. Kuo; ; Y. Li; ; P. Shukla; ; S.
1
Makarychev-Mikhailov; ;
1. Schlumberger, Sugar Land, TX, United States.
2. The Dow Chemical Company, Freeport, TX, United States.
3. The Dow Chemical Company, Midland, MI, United States.
Objectives/Scope: The efficiency of hydraulic fracturing operations can be significantly impaired by multiple damage
mechanisms, induced by the fracturing fluid, that affect both the formation and the proppant pack. The two key
damage factors are a) unbound immobile water trapped in pores preventing the flow of hydrocarbons and b) insoluble
polymer residue originating from degraded fracturing fluid gelling agents. This study aims at elaborating on the
mechanisms of the fluid-related damage and on new solutions to mitigate their detrimental effect.
Methods/Procedures/Process: On the modeling side, a commercial reservoir-centric stimulation-to-production
software was used to study the impacts of fracture network conductivity and effective fracture length/area on
simulated well production. On the experimental side, a comprehensive study of a gelling agent degradation
mechanism and the formation/proppant pack dewatering process was performed. A new experimental setup was
developed, focused on relative oil/water permeability measurements in mixed packs of proppant and rock.
Results/Observations/Conclusions: An extended sensitivity analysis of production for different fracturing job types
(slickwater and crosslinked) was conducted for a major US play, the Eagle Ford shale. The effects of fracture
geometry and conductivity on initial and cumulative production were quantified. These reveal general trends and
enable further job design optimization for a variety of parameters. Several concepts were tested in the laboratory
targeting improvement of effective fracture geometry through minimization of polymer residue and reduction of water
saturation. The selected concepts enabled development of a new highly effective multifunctional chemical additive.
The performance of this additive was further validated in application-oriented testing.
Applications/Significance/Novelty: This paper investigates the impact of prominent damage mechanisms on well
production and proposes solutions for damage mitigation. The results demonstrate how the combined approach
relying on production simulations and laboratory development can be used to optimize well performance.
Impact of Natural Fractures Beyond the Hydraulic Fracture Complexity in Unconventional Reservoirs – A Permian
Case Study
1 1
P. Pankaj; ; J. Li; ;
1. Schlumberger, Richmond, NE, United States.
Objectives/Scope: The study outlined in this paper is aiming to uncover the impact of natural fracture network (NFN)
properties and its interaction on productivity of a multi-well pad system in an unconventional reservoir setting. The
investigation utilizes a theoretical model to characterize the dynamic fluid leak-off behavior in the naturally fractured
reservoirs. Effect of well spacing, sequence of stimulation and timing on well’s productivity is studied within the
naturally fractured settings.
Methods/Procedures/Process: A field case study from the Permian basin with characterized and calibrated geo-model
is used for the investigation. A complex fracture model to incorporate the impact of variable fracturing fluid leak-off as
it propagates is utilized to evaluate the hydraulic fractures geometry. History matched production on the calibrated
model serves to elucidate the productive fracture geometry. Further, a boundary element computation technique to
predict the stress shadow in zipper and non-zipper scenarios is utilized to predict the multi-well pad production
performance with time.
Results/Observations/Conclusions: Most NFNs present in the rock fabric of unconventional reservoirs are mineralized.
However, being a plane of weakness in the system, when they come under the direct or indirect pressure perturbation
from the hydraulic fractures, they are easily sheared. It evident that the production performance for wells are severely
impacted by the overall geometry and footprint of the hydraulic fractures in these naturally fractured reservoirs. The
dynamic fluid leak-off to the NFN and its dilation while propagating the hydraulic fracture has a consequential impact
on the subsequent frac stages (same well and the offset). The loss of fracturing fluid to the NFN affects the
productivity of the multi-well system. Well spacing decisions are greatly impacted with the density of NFN and fluid
leak-off.
Applications/Significance/Novelty: A new workflow to calibrate complex fracture model to incorporate dynamic NFN
fluid loss is created. The unique approach not only serves a methodology to predict well performance in a multi-well
stimulation scenario and history match production, but also provides a consistent approach to optimize well spacing,
stimulation sequencing, and timing of stimulation considering the impact of the NFN.
Near Fracture Capillary End Effect on Shale Gas/Oil Production
1 1
R. Elputranto; ; I. Akkutlu; ;
1. Petroleum Engineering, Texas A&M University , College Station, TX, United States.
Objectives/Scope: Capillary end effect develops in tight gas and shale formations near hydraulic fractures during flow
back of the fracturing treatment water and extends into the natural gas production period. Previous studies showed
the potential flow impairment mechanisms in tight gas and shale formations and discussed to a certain extent their
influence on a well’s performance during production. However these studies did not consider the existence of capillary
end effect near the fractures. The objective is to quantify the capillary end effect and compare to permeability change
due to changes in total stress
Methods/Procedures/Process: A fully-implicit integral finite difference method is used for modelling and simulation of
multi-phase flow in the formation. The model has a multi-scale pore structure with secondary-pores consisting of
micro-fractures and cracks, and small primary-pores that belong to the inorganic particles, interlayer clays and
kerogen. The model simulates water-gas multi-phase flow in the secondary pore network with a capillary discontinuity
at the hydraulic fracture-matrix interface. In addition, the model considers changes in total stress due to changes in
the primary pores pressure, driven by osmosis and clay-swelling, and by gas desorption and diffusion from kerogen
pores.
Results/Observations/Conclusions: The simulation results show that the capillary end effect cause significant
formation damage during the flow back and production periods by holding the water saturation near the fracture at a
higher level than that based on spontaneous imbibition of water. The effect makes water less mobile in the formation
during the flow-back, and tends to block gas flow severely. The stress-related effects of the primary pores are
relatively less important. We showed that the capillary end effect cannot be removed completely but can be reduced
significantly by controlling the wellbore flowing pressure and by altering the formation wettability.
Applications/Significance/Novelty: A new multi-phase reservoir flow simulation model including capillary end effect is
developed to understand the impact of capillarity near the fracture on the removal of the injected water from the
formation and on the gas production. In the presence of capillary end effect, the values of water saturation, relative
permeability and capillary pressure are computed dynamically during the simulation. This model can be used to
characterize the formation damage during the fracturing.
Theoretical Investigation of Water Blocking in Unconventional Reservoirs Due to Spontaneous Imbibition and Water
Adsorption
1 1
L. Deng; ; M. J. King; ;
1. Texas A&M University, College Station, TX, United States.
Objectives/Scope: After fracturing, only a small fraction of the fracturing fluid flows back into the well. The water
invasion may induce a loss in hydrocarbon mobility, called water blocking. This phenomenon is expected to hinder
hydrocarbon productivity, but is contradicted by field observations where increased water invasion during soaking may
lead to better hydrocarbon production. This paper presents a novel semi-analytic approach to quantitatively evaluate
the trade-off between the potential hydrocarbon mobility reduction, water adsorption and counter-current flow
production due to spontaneous imbibition.
Methods/Procedures/Process: An analytic solution for spontaneous imbibition has been coupled with a fracture pore
volume constraint to calculate the water invasion during soaking. In cases where the initial water saturation is below
the mobile saturation, water adsorption due to the presence of clays controls the water invasion. The developed
solution generates saturation profiles from which the apparent hydrocarbon mobility can be evaluated. The imbibition
flux is compared with achievable viscous hydrocarbon flux based on the mobility calculation, and the cross-over of the
two phenomena represents an optimal soaking time after which further imbibition would only damage the hydrocarbon
productivity. The total backflow volume is translated into a compensation time for the production loss due to water
blocking.
Results/Observations/Conclusions: When the initial water saturation is below the mobile saturation, water invasion
occurs due to a combination of imbibition and adsorption while the relative permeability of hydrocarbon stays fairly
constant. The blocking effect does not impact flow until the mobile water saturation is reached. Then spontaneous
imbibition dominates and the hydrocarbon mobility decreases. Gravity segregation and a resulting flowing area
reduction in the fracture will cause higher decline of the imbibition rate compared with a 1D self-similar solution. The
mobility reduction, soaking time and compensation time are studied along with sensitivities on parameters like
permeability and wettability. The result shows the invasion water may even benefit production with proper control over
the soaking period.
Applications/Significance/Novelty: The present analysis examines the matrix/fracture interaction after hydraulic
fracturing in the context of water adsorption and spontaneous imbibition, and presents a theoretical explanation for
increased production after soaking a well. The mechanisms studied reconcile the apparent contradiction between
current operational observations and the conventional multiphase flow theory on mobility reduction. It demonstrates
that the water blocking effect is less severe compared with traditional imbibition models, and the results may also
provide the basis for the application of EOR techniques in tight formations. In future applications, these relationships
can be used for well soaking management and water blocking evaluation based on intrinsic properties and fracturing
design.
Diagnosing Fracture-Wellbore Connectivity Using Chemical Tracer Flowback Data
1 1
A. Kumar; ; M. Sharma; ;
1. Petroleum Engineering, University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: Existing fracture diagnostic methods such as micro-seismic monitoring and tiltmeters do not
provide information about fracture connectivity to the wellbore. In this work, we present a chemical tracer flowback
based fracture diagnostic method to (a) estimate the length of the fractures which are open and connected to the
wellbore, and (b) understand the effect of Induced un-propped (IU) fracture closure on the tracer response.
Methods/Procedures/Process: We present a fully coupled geomechanical, reservoir simulation model to study tracer
flowback in complex fracture networks. The model captures the effect of fracture opening and closure due to changes
in in-situ effective stress during flowback. As the pore pressure is reduced, fractures close over time. This directly
affects the tracer response during flowback. The impact of the density, geometry and closure rate of induced
unpropped (IU) fractures, on tracer response was demonstrated through simulation results and field data. Fracture
permeability and lengths were lumped to define effective connected fracture length, a parameter which correlates with
production. Neural network based inverse modeling was performed to estimate effective connected fracture length
using tracer data.
Results/Observations/Conclusions: Simulation results indicate that the tracer response is dominated by the fractures
which are open and connected to the wellbore. Multiple peaks in the tracer response can be explained by the closure
of IU fractures. The number of peaks is found to be proportional to the density of IU fractures and their closure rate.
Fracture closure also explains the low tracer recovery typically observed in field tests. Tracer recovery is found to be
proportional to the production. Based on these observations, tracer peaks and recovery parameters were selected for
training the neural network for inverse modeling. The trained neural network was used to estimate the effective
connected fracture length. We observed a good match between neural network prediction and the fracture parameters
in the simulation.
Applications/Significance/Novelty: We present a new method to analyze chemical tracer data which includes the effect
of flow and geomechanics on tracer flowback. The proposed approach, for the first time, allows operators to get a
relatively inexpensive, quantitative estimate of the degree of connectivity between the wellbore and open connected
fractures as well as the connectivity between propped and unpropped portions of the fracture network.
Impact of Authigenic Surface Roughness on Water Invasion and Flowback in Fractured Media: A Micromodel Study
1 2 1 1
A. Mehmani; ; S. A. Kelly; ; C. Torres-Verdin; ; M. Balhoff; ;
1. Petroleum and Geosystems Engineering, The University of Texas at Austin, Austin, TX, United States.
2. Applied Geoscience, ConocoPhillips, Houston, TX, United States.
Objectives/Scope: The objective of this work is to investigate the impact of physical surface roughness on fluid
trapping during imbibition and drainage. Matrix pores and fractures exhibit varying degrees of surface roughness, due
to authigenic cement coatings, clays, etc. This roughness increases the surface area contacted by fluids during
imbibition/drainage, two critical immiscible flow phenomena that determine the efficacy of hydraulic fracturing
operations. The physical and chemical properties of surfaces are especially important in tight rocks due to their high
pore surface area to volume ratios.
Methods/Procedures/Process: This study uses glass micromodels (microfluidic chips) with patterns that are proxies for
fracture-matrix networks with varying connectivity. Surface roughness is incorporated into the micromodel
fractures/matrices with a controlled fabrication method and surface roughness profiles are quantified. Imbibition is
conducted by pipetting water droplets on the inlet port of the micromodels whereupon the liquid spontaneously enters
the chips. Drainage is replicated by injecting air at a constant flow rate into the chips subsequent to imbibition. In some
experiments, injection flow rates and viscosity ratios are also controlled variables. Light microscopy and image
analysis are utilized to quantify the saturations and spatial distributions of trapped phases after each experimental flow
cycle.
Results/Observations/Conclusions: Surface roughness is found to increase trapped nonwetting phase saturation
during spontaneous imbibition and reduce viscous fingering up to an order of magnitude, thereby increasing the
surface area contacted by fluids. These results and other findings provided by the micromodel work aid in upscaling
by establishing quantitative relationships between trapped phase saturation and spatial distribution as a function of
fracture network properties and surface roughness, variables that can be ascertained from reservoir quality
(petrographic) studies. These relationships can be calibrated to the invasion radius of fracture fluid, its degree of
chemical impact on the formation, and its flowback, all directly related to the recovery factor of hydraulically fractured
formations.
Applications/Significance/Novelty: This study’s micromodel imbibition experiments are proxies for the invasion of
fracture fluid into fractured rock matrix while the drainage experiments are proxies for hydrocarbon production from
fractured rock matrix. Connecting reservoir quality (pore scale) to fracture performance is not always straightforward;
micromodels can help to bridge this gap. Imbibed fracture fluids react with rock matrix; trapped nonwetting saturations,
such as are quantified in this work, affect the amount of surface area available for reaction. Conversely, the pore
volumes of trapped wetting phases after drainage (also referred to as flowback), such as are also quantified in this
work, has a first order impact on well skin factor and, ultimately, recovery factor.
Effects of Hydraulic Fracturing Fluid Chemistry on Shale Matrix Permeability
1 1, 2
A. A. Alalli; ; Q. Li; ; A. Jew; 2; A. Kohli; 2; J. Bargar; 2; M. Zoback; 1; A. Kovscek; 1;
1. Stanford University, Stanford, CA, United States.
2. SLAC National Accelerator Laboratory, Stanford, CA, United States.
Objectives/Scope: Our objective is to understand how shale mineralogy in unconventional shale reservoirs responds
to hydraulic fracturing fluids and measure how chemical reactions occurring during completions affect the intrinsic
permeability of the shale matrix. There are two competing effects occurring in these systems: 1) enhanced
permeability due to secondary porosity formation (mineral dissolution) and 2) reduced permeability due to mineral
precipitation and fluid imbibition. Deconvolving the factors that influence permeability is crucial in order to optimize
production and promote better recovery factors.
Methods/Procedures/Process: Initial characterization of each unreacted microcore (9 mm diameter) was acquired for
both a carbonate-rich Eagle Ford sample and a clay-rich Marcellus sample using synchrotron X-ray microtomography
(28μm resolution). Permeability was measured afterwards using pressure pulse decay method at incrementally
increasing pore pressures and at a constant effective pressure. The matrix permeability was calculated from the
relative contributions of advection and diffusion flows through Klinkenberg analysis. In ongoing work, the cores will be
reacted with synthetic fracturing fluid (pH 2) in a batch reactor for 3 days and then vacuum oven dried till mass is
stable. Subsequently, X-ray imaging will be acquired and permeability will be measured on reacted microcores using
the same procedure.
Results/Observations/Conclusions: After completing baseline permeability measurements of unreacted shale
samples, we calculated the Eagle Ford shale (high carbonate/low clay) had an intrinsic permeability of ~30
nanodarcies. The high clay Marcellus shale sample showed that it was impermeable after 7 days of measurement.
This observation is inferred to be due to mineral composition effects on shale permeability. Our ongoing hypothesis is
that following reaction, the permeability will increase due to dissolution of carbonate minerals from the acidic synthetic
fracture fluid. However, a competing chemical reaction (mineral precipitation) will reduce permeability resulting in
shale permeability measurements below the predicted theoretical maximum values using reactive transport modeling.
Applications/Significance/Novelty: Our results will demonstrate the effects of hydraulic fracturing fluid on the shale
rock matrix measured from both chemical and physical changes on the sub-micron scale and correlating these
changes to permeability evolution. Consequently, we can tailor specific fracturing fluid recipes that account for shale
mineralogical composition in order to enhance permeability of stimulated shale volumes. By optimizing the injected
fracturing fluids, increases in permeability can be achieved while the effect of secondary mineral precipitation can be
reduced which will ultimately enhance both permeability and overall recovery factors.
Flowback in Shale Wells: Proppant Transport and Distribution in the Wellbore
1 2 2 2
K. Putri; ; H. Lu; ; C. Kwok; ; K. Moncada; ;
1. Colorado School of Mines, Golden, CO, United States.
Objectives/Scope: Flowback of hydraulically fractured shale wells is a key step to deliver optimum early return rates
and long term economic gain of the field. This study aims to provide insights of wellbore conditions during flowback
from two aspects: transient bottomhole pressure profile and solid particle distribution in the horizontal wellbore. Case
study is conducted to investigate the proppant flowback of a slick water treated well from Delaware basin.
Methods/Procedures/Process: High resolution flowback data including water, oil, gas and sand flowrates were
collected from a multiphase flowmeter, along with wellhead pressure and choke opening data through a
commercialized flowback service. A transient multiphase flow simulator incorporating dynamic choke opening
schedule was utilized to history match the field measurements. It also serves to accurately capture the rheology in
downhole condition. Furthermore, a solid particle tracking model embedded in the simulator is applied to simulate the
proppant movement controlled by bean-up procedures. The parametric study is followed to examine proppant
properties impact on flowback duration and residual proppant distribution in the wellbore lateral section.
Results/Observations/Conclusions: The data showed multiphase flow behavior presented in the wellbore during early
stage of flowback due to formation fluid inflow from reservoir. History matching of flowback showed good agreement
with the measurements. The simulation analysis suggested a higher bottomhole pressure observed with proppant
sand included in the model. Furthermore, the study results indicated that a sudden increase in choke opening led to
higher sand production at the surface. Sand accumulation in the lateral section was observed even at high flowrate
condition. In addition, proppant density, proppant diameter and wellbore undulation were found to be the key factors
affecting proppant settling in the wellbore.
Applications/Significance/Novelty: The work in this paper extends the understanding of proppant behavior in the
dynamic multiphase flow environment. It demonstrates the integration of field practice and dynamic simulation
approach to be effective to ensure optimized flowback operation, which has profound impact on maximizing estimated
ultimate recovery (EUR) of the field.
Evaluating the Loss in Fracture Volume During Flowback and Its Relationship to Choke-size: Fastback Versus
Slowback
1 1 2 2
Y. Fu; ; H. Dehghanpour; ; S. Motealleh; ; C. Lopez; ;
1. Civil and Environmental Engineering, University of Alberta, Edmonton, AB, Canada.
2. BP America, Houston, TX, United States.
Objectives/Scope: This paper presents a workflow for flowback data analysis 1) to estimate the initial effective fracture
volume (Vi), 2) to evaluate the loss in effective fracture volume (dV) during flowback, and 3) to relate the normalized
change in fracture volume (dV/Vi) to choke-size changes for 18 gas wells completed in Eagleford Formation.
Methods/Procedures/Process: The workflow mainly consists of three key steps: First, we apply rate-decline analysis
on water flowback data to estimate the ultimate water recovery volume, approximated as initial effective fracture
volume; Second, we estimate dV by applying a material balance model on the flowback data; Third, we investigate the
effect of choke-size variations on dV/Vi for the 18 wells with fastback and slowback strategies. In particular, we
investigate the effect of water flowback rate on the loss in fracture volume.
Results/Observations/Conclusions: Semi-log plots of water flowback rate versus cumulative water volume are linear,
representing a harmonic-decline behavior. The results show a relatively good match between the measured water rate
during post-flowback and that predicted by the harmonic-decline behavior. The initial fracture volume (Vi) estimated by
rate-decline analysis accounts for about 60% of the total injected water volume. Up to 30% of Vi can be lost during
early flowback, while fracture volume generally remains constant during late flowback. Comparing dV/Vi with choke-
size indicates that slowback leads to lower loss in fracture volume compared with fastback. Comparative analysis also
suggests that the loss in fracture volume is more sensitive to choke-size changes during early flowback than that
during late flowback.
Applications/Significance/Novelty: The paper proposes a novel and efficient method to estimate initial fracture volume,
and a workflow to investigate the change in fracture volume during flowback processes. Analysis of the field data
leads to an improved understanding of the factors controlling water flowback and effective fracture volume. The results
can help the operators to reduce the loss in fracture volume through optimizing choke-size variations during flowback
operations.
Defining the Optimal Drawdown Strategy in the Vaca Muerta Formation
1 2 3
A. A. Lerza; ; B. Liang; ; D. Rojas; ;
1. Latin America Business Unit, Chevron, Houston, TX, United States.
2. Mid Continent Business Unit, Chevron, Houston, TX, United States.
3. YPF, Buenos Aires, Argentina.
Objectives/Scope: Due to geomechanical effects during a quick pressure depletion of a hydraulically fractured
reservoir, well operational practices in shale formations have a significant impact over final well economics. Defining
“operational drawdown” as the Flowing Bottom Hole Pressure (FBHP) drop over time, this paper aims to determine
what drawdown strategy generates the highest economic returns in Vaca Muerta formation, located in Neuquen basin,
South of Argentina.
Methods/Procedures/Process: This paper presents a novel approach to improve the current drawdown management
based on empirical observations of 100+ wells. Analyzed well data was used to bound the results of an in-house
geomechanical simulator fully coupled with a reservoir simulator, to completely capture the geomechanical effects of
different operational practices, while considering uncertainties such as reservoir and injected proppant properties.
Design of Experiment technique was used to cover the uncertainty range of each scenarios, and to understand if there
is a single optimal drawdown strategy that holds throughout the expected variations of the considered uncertainties
and completion designs.
A field pilot was conducted and conclusions are going to be presented.
Results/Observations/Conclusions: A proxy function capturing a massive number of DOE results was built to test
thousands of new alternatives through Monte Carlo simulation. An optimal operational drawdown was consequently
found and further proved to be the same for the whole range of uncertainty. It showed an average increase of 20% in
discounted production from the previous practice which was slightly more aggressive and 8% increase compared to
curent less aggressive strategy, respectively. These increments are roughly translated to the same increase in NPV,
representing a significant economic uplift from the current status. Finally, the obtained optimal drawdown in Vaca
Muerta formation is potentially applicable to any other formation if physical properties and completion designs are
within the analyzed range.
Applications/Significance/Novelty: Previous workflows for conducting this analysis consisted on including a “proxy”
stress calculation into reservoir simulators, generally due to unavailability of data, geomechanical simulator or
extensive CPU time required. The caveat of that approach, is the incapability of the “proxy” to fully capture the real
stress variations, which shows up to 25% discrepancies between lab measured values and the proxy calculation,
leading to higher uncertainty conclusions and riskier decisions. The current work overcame the “proxy” limitations
using a geomechanical simulator fully coupled with a reservoir simulator. An optimal drawdown strategy was further
identified from the combination of a large number of DOE runs and rigorous uncertainty analysis which covered the
whole operational range.
Executive Vice President and Chief Technology Officer – Pioneer Natural Resources
1
C. Cheatwood; ;
1. Pioneer Natural Resources, Dallas, TX, United States.
Vice President of Geoscience and Technology, Concho
1
C. Spies; ;
1. Concho, Denver, CO, United States.
Vice President & Chief Information Officer, Occidental Petroleum Corporation
1
Y. Charalambous; ;
1. Occidental Petroleum , Bakerfield, CA, United States.
Corporate Vice President and Chief Technology Officer, National Oilwell Varco (NOV)
1
H. Kverneland; ;
1. National Oilwell Varco (NOV), Houston, TX, United States.
Bench-Top Experiments Evaluating Simulated Hydraulic Fracturing Fluid Interactions With Marcellus Shale Core
2 1, 3
A. Hakala; ; J. Moore; ; C. L. Lopano; 2; W. Xiong; 2, 4; T. Phan; 2, 4; A. Vankeuren; 5; S. Sharma; 6; J.
6 1, 3
Pilewski; ; K. Jarvis; ; S. Brown; 1, 3; D. Crandall; 1;
1. National Energy Technology Laboratory, Department of Energy, Morgantown, WV, United States.
2. National Energy Technology Laboratory, Department of Energy, Pittsburgh, PA, United States.
3. AECOM, Morgantown, WV, United States.
4. Oak Ridge Institute for Science and Education, Oak Ridge, TN, United States.
5. Geology, Sacramento State University, Sacramento, CA, United States.
6. Geology and Geography, West Virginia University, Morgantown, WV, United States.
Objectives/Scope: The injection of hydraulic fracturing fluids (HFF’s) into shales can result in a variety of chemical
reactions that affect hydrocarbon recovery during production. Experiments were conducted on Marcellus Shale core
taken from the Marcellus Shale Energy and Environmental Laboratory (MSEEL) site in Morgantown, WV and outcrop
samples using a simulated HFF’s formulated from regional averages. The experiments were designed to evaluate how
mixing-to-injection times, HFF components, and rock composition affected geochemical transport out of the system
and fracture to matrix reactions.
Methods/Procedures/Process: Marcellus shale cores were fractured in the laboratory and loaded with 40/70 mesh
proppant. A Hassler style core holder was used to apply overburden and pore pressures of 3000PSI and 2800PSI,
respectively. Core temperature was held at 160F while simulated HFF’s were injected at 0.03 ml/min to simulate
diffusive flow during a shut-in for a period of 96 hours. Samples were taken every 24 hours to evaluate bulk
anion/cation changes, carbon and strontium isotopes, and pH. The samples were scanned prior to reaction utilizing
Computed Tomography (CT), and post-reaction with CT, Scanning Electron Microscopy (SEM), and Raman
Spectroscopy (RS). Pre and post-reaction samples were also analyzed for carbon content.
Results/Observations/Conclusions: Earlier studies using spring water as the base fluid showed significant precipitation
of barite, especially around proppant packs demonstrated by CT and SEM images. In later experiments using
deionized water as the base, less precipitation was observed at the CT spatial scale, however precipitation was found
with SEM and RS. Matrix dissolution occurred in the carbonate rich Marcellus samples, but non-carbonate rich
MSEEL samples appeared unreacted. Increased Ca was observed in all experiments indicating carbonate dissolution.
Barium and sulfate were lower in the effluents relative to the injected fluid, indicating precipitation within the core.
Effluent pH was higher than the injected fluid, indicating that acidic reactions that reduced pH were buffered by
carbonate dissolution.
Applications/Significance/Novelty: Knowledge of in-situ reactions between HFF’s and the reservoir rock is critical to
long-term production planning and operations. These experiments provide meaningful insights into inorganic
water/rock interactions at the micro-scale and illustrate component based effects on the host rock. These experiments
also provide a basis for conducting more detailed experiments evaluating organic mediated reactions within the
reservoir.
Barium Sources in Hydraulic Fracturing Systems and Chemical Controls on Its Release Into Solution
1, 2
A. Jew; ; Q. Li; 2, 1; D. Cercone; 3; K. Maher; 2; G. Brown; 2, 1; J. Bargar; 1;
1. SLAC National Accelerator Laboratory, Stanford, CA, United States.
2. Stanford University, Stanford, CA, United States.
3. National Energy Technology Laboratory, Pittsburgh, PA, United States.
Objectives/Scope: Precipitation of barite scale on fracture surfaces and microcracks is a large concern in
unconventional systems because it reduces rock permeability and production efficiency. Though the main source of
Ba in these systems is poorly understood, source rocks can contain Ba up to 1 g/kg while drilling muds contain
upwards 30 wt.%. This work not only determines the dominant Ba source in these systems, but also identifies the
important chemical parameters responsible for the release of Ba into the hydraulic fracturing system, that will
subsequently precipitate resulting reduced rock permeability.
Methods/Procedures/Process: Three shale samples (Marcellus, Eagle Ford, and Barnett) and drilling mud (DM) used
at the MSEEL site in Morgantown, WV were selected for this work. Shale samples were ground to 150-250 μm and
o
reacted with simulated fracture fluid, water, or HCl (pH=2) for 3 weeks at 80 C. Because Ba in drilling mud (DM) is
comprised solely of barite, both pure barite and DM were reacted with a variety of pH's, Ionic Strengths (I.S.),
o
organics, and simulated fracture fluid. Samples were mixed on an end-over-end tumbler for 72 hrs at 80 C. All filtered
samples (20 nm) were monitored for dissolved Ba concentrations using ICP-MS throughout the experimental time
period (shales) or at the end of the experiments in the case of barite/drilling mud because steady state for these
reactions was reached at ~60 hrs.
Results/Observations/Conclusions: The percentage of Ba released from Marcellus and Barnett was quite low (3.3%
and 4.1%, respectively) while Eagle Ford was 44.1%. Dissolution experiments of pure barite and DM show that Ba
release into solution is dominated by acidic conditions followed by high I.S. In the presence of an injected acid slug,
15% HCl, 6% of the total Ba was released from pure barite while 23% was leached from the DM. At pH 7, organic
additives had little to no effect on Ba leaching from barite or DM. Surprisingly however, at pH2, the opposite was
observed; more Ba was released in the presence of organic additives. In nearly all cases more Ba was leached from
DM than pure barite suggesting that either organics in the DM accelerate dissolution of the barite present or that there
is another Ba phase in DM.
Applications/Significance/Novelty: In unconventional systems the total amount of Ba leached from the shale in these
experiments is insufficient to explain either the scaling issues or high Ba concentrations in produced waters. This work
suggests that the primary issues with Ba scale are due to the drilling process. Without removal of the DM prior to fluid
injection, a significant amount of Ba is being introduced into these systems. Unlike pure barite, the Ba in DM is not
nearly as stable as expected indicating that Ba release from the DM and by chemicals in both the DM and fracture
fluid are most likely creating scale induced reductions in oil/gas recovery. By understanding the source and major
chemical constraints on Ba release better formulations of DMs and completion fluids can be made to mitigate Ba scale
precipitation.
Imaging Pyrite Oxidation and Barite Precipitation in Gas and Oil Shales
1 1, 2
A. Jew; ; Q. Li; ; A. Kiss; 1; A. Kohli; 1, 2; A. A. Alalli; 2; A. Kovscek; 2; M. Zoback; 2; D. Cercone; 3; K.
2 1, 2
Maher; ; G. Brown; ; J. Bargar; 1;
1. Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA, United
States.
2. School of Earth, Energy & Environmental Sciences, Stanford University, Menlo Park, CA, United States.
3. Strategic Center for Natural Gas and Oil, National Energy Technology Laboratory, Pittsburgh, PA, United States.
Objectives/Scope: The complexity of fracture fluid causes numerous chemical interactions with shale. Acid, one of the
most common additives, dissolves carbonates and barite-laden drilling mud, leading to barite precipitation, and
oxygen entrained in injected fluids drives pyrite oxidation. Though Fe/Ba scale can reduce oil/gas transport, these
precipitation reactions have rarely been studied inside the shale matrix. In this research we aim to image these
reactions along a cross section extending from the fracture surfaces into the matrix, as a function of reaction type,
shale composition, and fluid chemistry.
Methods/Procedures/Process: Experiments were conducted to examine the spatial extent of pyrite oxidation and
barite scale formation in the shale matrices. Marcellus (low carbonate/high clay) and Eagle Ford (high carbonate/low
clay) cores of 1 cm diameter were reacted with simulated fracture fluid. The fracture fluid contained dissolved oxygen
and was acidified with HCl (pH 2) before reaction. A subset of experiments contained additional BaCl and NaSO
2 4
o
(log (Q/K) = 1.3) to promote barite precipitation. After 3 weeks of reaction at 80 C and 77 bars, the cores were
10
scanned using micro CT, and then cross sectioned for Fe and S mapping using synchrotron microprobe.
Results/Observations/Conclusions: Solution pH increased from 2 to 3 (Marcellus) and 2 to 7 (Eagle Ford) due to the
dissolution of carbonates. High pH is predicted to favor barite precipitation, and CT images showed barite precipitation
on Eagle Ford shale surfaces and within near-surface microcracks, whereas for Marcellus, barite accumulated only on
shale surfaces. Synchrotron x-ray chemical imaging showed that pyrite oxidation occurred throughout the entire 1 cm
Marcellus core, suggesting a minimum of 5 mm penetration within the 3 week reaction time. Our results also suggest
that dissolved oxygen penetration was hindered due to occlusion of porosity by barite precipitation on Marcellus
surface. In contrast, Eagle Ford cores experienced pyrite oxidation only on the surfaces of the core.
Applications/Significance/Novelty: Dissolution of primary minerals such as carbonates and pyrite will enhance porosity
for gas/oil flow. However, released ions will subsequently precipitate in pore spaces, clogging pores throats and
reducing permeability. As shown in this study, chemical reactions do not occur solely on fracture surfaces, but also
extend deep (up to cms) into shale matrices. Understanding the types and depths of chemical reactions can help
refine reactive transport models to predict their impact on gas/oil production, leading to optimization of fracture fluid
chemistry for unconventional resource stimulation.
Lessons Learned From Existing Horizontal Fractured Wells in Midland Basin of University Lands (UL): Rate Transient
Analysis vs. Completion and Field Development Optimization
1 1 1 1
J. Zhu; ; J. K. Forrest; ; H. Xiong; ; Y. U. Pradhan; ;
1. Texas Oil & Gas Institute, Houston, TX, United States.
Objectives/Scope: A traditional way to evaluate a well performance is to look at the initial production rate, different
time cumulative production, and EUR based upon DCA etc., which does not consider the geology, fluid properties,
completion design, and production operation. On the other hand, it may take tremendous efforts and long time to
finalize an optimal field development plan with corresponding optimal completion designs by building geology models,
petrophysical models, DFN models and geo-mechanical models, then the hydraulic fracture models, and reservoir
models to perform reservoir simulations.
Methods/Procedures/Process: In this paper we will demonstrate a quick way to analyze the well performance by
performing the multiphase rate transient analysis (RTA) on individual wells. Wells with production volume, downhole
gauge pressure data or well head pressure will be chosen as candidate wells. After preforming RTA analysis, we will
combine the results with completion parameters, such as cluster spacing, stage length, proppant and fluid volume,
landing zone, well spacing, well lateral length, and well operation method etc., to identify critical parameters that have
significant impact on completion design and field development strategy in different area of UL in the Midland Basin.
Results/Observations/Conclusions: The paper provides the cluster spacing and well spacing trends based on multiple
operators’ field trial and error approach through different geology and fluid areas on UL. The data show that tighter
cluster spacing with large completion jobs, which will maximize stimulated Reservoir volume, yields higher well
productivity and enhances well economics, and may improve ultimate recoveries.
Although well spacing vs EUR data is provided, the well spacing may not be optimized. The optimal well spacing
essentially depends on fracture half-length (Xf). In general, tighter cluster tends to generate short Xf and more efficient
completion. Accurately estimate the Xf is the key of optimal well spacing. Optimal cluster spacing and well spacing are
also driven by drilling, completion and operation cost.
Applications/Significance/Novelty: Unconventional reservoir production heavily depends on a successful hydraulic
fracturing treatment. More than 2000 horizontal wells were drilled since 2010 on UL in the Midland Basin. Finding the
optimal completion and field development plan on the UL acreage is always the top priorities for Texas Oil and Gas
Institute.
This paper provides a way to speed up the analysis of well performance thus to reduce the cycle of trial and error
approach. The identified optimal cluster spacings and well spacings throughout UL will provide a guideline to the
operators when they design completion field development plan strategy, which will also provide a reference outside
the UL acreages where there are similar geology and fluid properties since UL is major production contributor of
Permian Basin.
The Effect of Initial Conditions and Fluid PVT Properties on Unconventional Oil and Gas Recoveries in the Wolfcamp
Formation in the Midland Basin
1 1 1 1
J. K. Forrest; ; J. Zhu; ; H. Xiong; ; Y. U. Pradhan; ;
1. UT System Administration, Texas Oil & Gas Institute, Southlake, TX, United States.
Objectives/Scope: The objective of this paper was to determine the variation in estimated ultimate gas and oil
recoveries (EUR) from the Wolfcamp B formation (WC-B) due to variations in the initial conditions such as pressure,
temperature, and fluid PVT properties. The study focused on the University Lands within the Midland Basin, which
have significant acreage in the southern part and the northwestern edge of the basin. These two areas are separated
by roughly 90 miles, and the WC-B depths vary by thousands of feet across this expanse.
Methods/Procedures/Process: This study included the following steps:
Created a dual-porosity, reservoir simulation model and history matched a WC-B horizontal well located in the
Northern Midland Basin.
Examined PVT data and initial conditions for other WC-B wells located in the North and South parts of the basin.
Examined industry fluid correlations that match the reported PVT data.
Used the history-matched model as representative of WC-B reservoir properties, and performed sensitivity analyses of
EUR forecasts by varying the initial conditions of depth, pressure, temperature, and fluid PVT properties.
Summarized the effects of variations in initial conditions on EUR.
Results/Observations/Conclusions: A dual-porosity simulation model was successfully history-matched to a WC-B
horizontal well located in the Northwest area of the Midland Basin (Andrews County). This model was then used as a
proxy representation of the formation properties. This proxy model was then used to examine variations in estimated
ultimate recoveries using the observed ranges of reservoir depths, initial pressures and temperatures, and fluid PVT
properties. The predicted simulation results show that the initial conditions and fluid PVT properties can account for a
30% difference in recovery between a WC-B well located in the north (University Block 7 in Andrews County) and the
south (University Block 3 in Upton County).
Applications/Significance/Novelty: The significance of these results are as follows:
These results are useful to operators and investors in partially explaining why Wolfcamp wells located in the southern
part of the Midland Basin generally underperform the wells located in the central and northern part of the Midland
Basin. Because of the structural dip of the basin, wells in the north are deeper and have higher reservoir pressures
and initial gas-oil ratios.
These results show that it is important to understand the initial conditions of each unconventional horizontal well.
A Practical Way to Prepare Physical-Based Type Well Performance Curves for Unconventional Reservoirs in the
Permian Basin
1
H. Xiong; ;
1. Texas Oil and Gas Institute, Houston, TX, United States.
Objectives/Scope: It has been a huge challenge to generate type well curves for UR plays due to the reservoir
heterogeneity, complicated fracturing mechanisms, different completion designs and operations. The industry has
been trying to use analytical, statistic, and numerical modeling methods. Complexity of fracture networks limits the
usage of analytical method, and lack of physics mechanisms and long-term prod data constrains the application of
statistic method. Numerical method often requires significant resources and expertise.
This paper illustrates a practical approach to generate robust type well curves.
Methods/Procedures/Process: 1. For a given geologically similar area (GSA), we build a sector numerical model, and
then calibrate the model with available production history with assisting of automatically-history matching method.
2. Identify the key uncertainties and investigate their possible ranges.
3. Study the effectiveness of past completion designs and investigate the possible completion designs.
4. Perform multiple runs with Monte-Carlo simulation method based upon the ranges and the distribution types of
those uncertainties and completion design parameters.
5.Compare the modeling results with the real production data from the same GSA, and perform multiple runs again by
adjusting the ranges and/or distribution shapes of the uncertainties if needed.
6.Build type well curves based upon the calibrated multiple runs.
Results/Observations/Conclusions: We have applied the workflow and built multiple type well curves for several GSAs
of those URs in the Permian Basin. The automatic process of history matching and forecasting significantly reduces
the manpower requirement. Since the approach enables us to incorporate the different flow mechanisms, formation
heterogeneity, and completion designs, we observed that the outcomes of multiple runs from those calibrated models
are very consistent with well production historical behaviors. We witnessed that the workflow was very easy to follow,
and less experience engineers were able to generate those type well curves very efficiently.
Applications/Significance/Novelty: Our practical workflow is very easy to follow and can be applied into various UR
plays in different basins. The effective approach enables us to easily integrate various physics flow mechanisms in the
unconventional reservoirs with various completion designs. Therefore, the resulting type well curves should be more
reliable to forecast the overall performance of multiple wells in a given geologically similar area. It is very convenient to
develop type well curves for a subset of the GSA or completion designs, which enable us to get into the critical and
insightful info about the subsurface and/or completion effectiveness.
Evaluating Underperforming Wells on Permian Basin University Lands
1 1 1 1 1
Y. U. Pradhan; ; J. Spath; ; H. Xiong; ; J. Zhu; ; J. K. Forrest; ;
Objectives/Scope: University Lands’ production and revenue are driven by over 270 operators in the Midland and
Delaware Basins, warranting a Permian-wide study of underperforming wells. Texas Oil and Gas Institute has access
to all well data from University Land operators, upholding production optimization objectives by identifying
underperforming wells and determining improvements either by refracturing or artificial lift optimization. This study
develops a thorough workflow to identify 100 underperforming wells out of the ~700 horizontal Wolfcamp and Lower
Spraberry unconventional wells on University Lands.
Methods/Procedures/Process: We first constructed two sets of type curves were constructed for each assigned
geologically similar area—P50 based on production and P10 based on recent completion intensities. We then
developed and calibrated numerical models for various Wolfcamp benches and the Lower Spraberry formations to
forecast what each well could produce at a given lateral length. We then compared each well’s normalized cumulative
production and EUR to these type curves over time to determine low performance. Then, we performed analysis,
including RTA and/or AL analysis to diagnose the underperforming reasons. We then narrowed down potential
candidates to enhance production by selecting wells based on high quality reservoir, low completion intensity, and
suitable wellbore quality (if known).
Results/Observations/Conclusions: We determined that these shortlisted wells are potential refract or AL optimization
candidates that University Land operators could improve despite critical commodity prices. Different completion
methods over time significantly contributed to varying performance on University Land wells. For instance, larger
ratios of hybrid fluid or cross-linked fluids compared to the total fluid pumped for the frac jobs yielded poorer
performing wells. Wells considered for artificial lift optimization have marginal, daily operational costs and minimal
workover times as recommendations, ensuring feasible implementation and consistent production across University
Lands.
Applications/Significance/Novelty: Our workflow is very flexible depending on the available data, man power and
expertise. Completion, production, reservoir, and geologic data from operators are used to evaluate operators’
performance on 2.1 million acres, four formations, and eight intervals. The evaluations gave a better understanding to
TOGI staff engineers and geologists of best operational and field development practices, and revealed insights on the
key reservoir characters, recoverable potentials, and corresponding optimal well completion design and AL strategy.
Chemical Stimulation with Driving Process to Extract Oil from Tight Formation
1 1
J. Zhang; ; D. Wang; ;
1. Geology & Geological Engineering Department, University of North Dakota, Grand Forks, ND, United States.
Objectives/Scope: Oil recovery factor from the Bakken and other shale formations is still a small portion of the original
oil in place (typically less than 10%) with existing techniques. Using surfactant imbibition to stimulate oil recovery has
been studied in our group since 2009, and in the process of field application. However, the contact area between
stimulating fluid and the tight rock matrix is not enough to speed oil extraction rate up. This paper aims the approach
of increasing contact area by driving process in tight rocks, and to determine if penetration rate can be improved by
the combined method.
Methods/Procedures/Process: Semi-preserved Bakken cores with visible natural fractures were used for aqueous
formulation imbibition combined with driving process at reservoir condition. Also, various comparisons among the
conditions of remaining oil saturation, initial water saturation, surfactant concentration, and surfactant solution salinity
were discussed using paired cores.
Results/Observations/Conclusions: Based on the laboratory studies, we conclude: (1) using driving process by forcible
flooding, the surfactant solutions penetration rate was improved, and the contact area was expanded through the
fractures. (2) Favorable oil extraction rate was observed compared to the spontaneous imbibition. (3) Surfactant
-3
flooding by driving process for tight rocks with low permeability (10 md) has good potential for oil recovery
enhancement at various conditions. (4) Wettability alteration is the dominant mechanism for oil recovery enhancement
during the driving process from tight rocks. (5) Remaining oil saturation, initial water saturation, surfactant
concentration, and salinity effect on oil recovery may be masked by permeability, heterogeneity or lithology
differences.
Our study demonstrates, for the tight rocks, using surfactant driving process to pull off the oil out from matrix through
natural fractures is also a feasible approach for oil recovery enhancement compared with spontaneous imbibition
under the good fracture connections.
Lithologic and Geomechanical Control on CO Huff-n-Puff Enhanced Oil Recovery Processes Using Integrated
2
Modeling Framework in Wolfcamp
1 1
T. N. Phan; ; Z. A. Reza; ;
Objectives/Scope: Enhanced Oil Recovery (EOR) is the future for unconventional reservoirs suffering from rapid
depletion and low recovery factor. CO EOR method, especially huff-n-puff CO scheme has been evolving in recent
2 2
years from lab experiments, simulation modeling and field-pilots. However, existing studies use simplified
geomechanical properties and neglect the effect of fracture complexity in assessing the EOR efficacy in these
reservoirs. This study aims to assess the lithological and geomechanical controls on incremental recovery in CO
2
huff-n-puff and the impact of gas interference to nearby wells.
Methods/Procedures/Process: We construct 3-D structural models capturing Midland Basin stratigraphy (Spraberry,
Dean and Wolfcamp formations) having multiple horizontal wells. Pressure gradient ranges within 0.46-0.52 psi/ft.
Matrix porosity ranges from 4% - 12%, while permeability is 0.01 - 10 microDarcy. 3-D mechanical earth models are
derived using isotropic shale materials and available logs. We employ a fracture simulator to create longitudinal and
tranverse fractures using these geological models. Resulting fracture geometry and conductivity profiles are fed back
to 3-D numerical reservoir simulation models. Several scenarios of CO huff-n-puff injection are evaluated under
2
various fracturing operational and design constraints using several lithological and geomechanical realizations.
Results/Observations/Conclusions: Base case primary depletion forecasting run is matched with observed production
data from available resources. Adequate matching is acquired by changing basic petrophysical and rock mechanical
parameters. CO huff-n-puff results positive oil increment over the primary depletion in some certain huff-n-puff
2
operating conditions and hydraulic fracturing design. Low permeability reservoirs require sufficient soaking period for
miscible condition occurs before bringing back on production. A combination of reasonable injection time and soaking
time is recommended in this study for 5-year period of huff-n-puff. Different behaviors of hydraulic fracturing response
to highly heterogenous reservoir conditions and complexity of fracture networks complicates the impact of CO huff-n-
2
puff efficiency.
Applications/Significance/Novelty: Studying the effect of operational hydraulic fracturing design in CO huff-n-puff in
2
unconventional helps address the existing issue of undesirable early breakthrough during flood process. It provides
some insights to avoid the limitation of CO huff-n-puff in the future. Operators might consider multiple parameters in
2
fracturing process when designing CO huff-n-puff for their unconventional reservoirs, especially in Wolfcamp
2
formation. Adsorption and diffusion effect are investigated in these models. An integrated approach which covers full
range of data from geology to geomechanical and dynamical reservoir parameters provides more robust and accurate
modeling and prediction of the efficiency of the enhanced recovery in unconventional reservoirs.
Scaling for Wettability Alteration Induced by Addition of Surfactants in Completion Fluids: Surfactant Selection for
Optimum Performance
1 1 1 1
F. Zhang; ; I. Saputra; ; I. Adel; ; D. S. Schechter; ;
1. Petroleum Engineering, Texas A&M University, College Station, TX, United States.
Objectives/Scope: Enhanced spontaneous imbibition by adding surfactants into the completion fluid has recently been
targeted as a technique to increase EUR from unconventional liquid reservoirs well. In this work, a new approach in
the capillary pressure-based scaling analysis of surfactant-assisted spontaneous imbibition is presented coupled with
numerical simulation, used to determine the relative permeability curves. In addition, a new surfactant selection
method is provided to obtain the best performing surfactant in spontaneous imbibition based on contact angle,
interfacial tension and zeta potential.
Methods/Procedures/Process: The improved scaling model for the spontaneous imbibition are based on available
scaling model with modification based on the surfactant-related properties measured in the lab. Contact angle (CA),
interfacial tension (IFT), zeta potential, and ultimately the spontaneous imbibition experiments were performed on
samples retrieved from a reservoir. Based on the capillary pressure curve obtained from the scaling model, relative
permeability is approximated through history matching process on core-scale numerical models. CT-Scan technology
is used to visualize the process that water imbibed into core plugs to investigate flow path for these core samples. CA,
IFT, and zeta potential are correlated to the ultimate oil recovery to provide a surfactant selection method.
Results/Observations/Conclusions: The change in capillary force as the wettability is altered by the surfactant leads to
oil being expelled as water imbibes into the pore space. Capillary pressure profiles are generated from the novel
scaling model with the addition of relative permeability curves which are obtained from numerical simulation history
match. A set of normalized oil recovery vs dimensionless time curves are plotted to investigate the best scaling model
for spontaneous imbibition in shale reservoir. CT images are used to indicate the flow path that water imbibed into the
core plugs. By considering the contact angle, interfacial tension, zeta potential and ultimate oil recovery from
laboratory measurements, the surfactant performance in the field can be predicted from the novel surfactant selection
method.
Applications/Significance/Novelty: The novelty of this paper is the generation of improved scaling model to be used in
surfactant-assisted spontaneous imbibition on unconventional liquid-rich reservoir. In addition, a novel surfactant
selection method is provided to evaluate surfactant performance by contact angle, interfacial tension and zeta
potential. This methodology allows selection of the appropriate surfactant for optimal completion design. Our
correlated set of experiments is explained and the complete governing equations are provided in detail to permit
readers to replicate all results.
Alternative Production Mechanisms in Unconventional Reservoirs
1
J. A. Acuna; ;
1. ETC, Chevron, Houston, TX, United States.
Objectives/Scope: Reservoir simulation in unconventional reservoirs utilizes stress-dependent properties (i.e.,

permeability or compaction tables) as well as desorption processes (in gas reservoirs) for production history
calibration. We propose that the reliance on these calibration approaches is due to the lack of other physical
mechanisms able to explain well productivity behavior. This paper offers two alternative production mechanisms to
explain well performance in unconventional reservoirs that reproduce the effect of stress-dependent properties and
gas desorption respectively.
Methods/Procedures/Process: In the absence of proppant crushing or embedment, propped hydraulic fractures should
have little sensitivity to reservoir pressure changes. A fraction of natural fractures are reactivated during hydraulic
fracturingby shear dilation and propped open by unmatched asperities along the fracture wall. This could form a
complex network with connected permeable fractures in varying orientations and rock fragments of many different
sizes. Changes in effective stress due to pressure depletion affects fractures differently depending on their respective
orientation with respect to the stress field and should not cause a significant permeability degradation. In gas reservoir
many fractures with very small conductivity will play a role in reservoir depletion due to the high mobility of gas.
Results/Observations/Conclusions: This paper complements another one submitted for this conference (Acuna,
Straightforward Representative Fluid Flow Models for Complex Fracture Networks in Unconventional Reservoirs) that
addresses the reservoir simulation aspects. This paper lays the fundation to state that altough some degree on
fracture compaction and gas desorption should ocuur, they should not be dominant mehcanism to explain reservoir
performance. The combined response of many rock fragments of different sizes in a relatively high conductivity
fracture network produce the same effect of compaction tables without any changes in properties with time or
pressure. The simultaneous presence of large and very small conductivity fractures fractures creates an effect similar
to desorption in gas reservoirs.
Applications/Significance/Novelty: These two new mechanisms provide an alternative view of unconventional reservoir
depletion that do not require stress-dependent properties or desorption to reproduce unconventional reservoirs
behavior. These mechanisms also explain many apparently inconsistent observations made in cores such as the
existence of a large number of unpropped fractures. More importantly it shows a new way to understand performance
in unconventional reservoirs that may lead to improved reservoir characterization and better resource recovery.
Condensate Blocking and Mitigation in Liquid-Rich Shale Reservoirs: An Integrated Evaluation Based on Systematic
PVT Modeling and Simulation Studies
1 2
N. Nagarajan; ; A. Orangi; ;
1. Hess Corporation, Houston, TX, United States.
2. Apache Corporation, Tulsa, OK, United States.
Objectives/Scope: Economic production from liquid-rich shale reservoirs is viable only through sustainable oil and gas
rates over the reservoir life. One of the critical challenges to economic production from these reservoirs is the
condensate dropout below the fluid dew point pressure (P ) leading to condensate blocking. It not only degrades
dew
the well productivity but also leads to significant liquid losses because of high critical condensate saturations
experienced by these tight formations. This article focusses on the impacts of condensate dropout in tight shale
formations with fracture networks and attempts to identify potential mitigations based on both compositional PVT
process simulations and sector-model reservoir simulations.
Methods/Procedures/Process: Mitigation of condensate blocking through solvent-injection (e.g., propane, CO , or
2
lean/rich gases) is possible. Injected solvent can suppress P , delaying liquid dropout or enhance ‘dropped-out’
dew
liquid recovery through partial vaporization and miscibility. We demonstrate these effects through compositional PVT
modeling of a rich-gas-condensate and subsequent simulation of various recovery mechanisms by solvent injection in
a tank-model approach.
We have also conducted detailed reservoir simulation studies on a sector model to confirm the gas-injection results
obtained through PVT simulation. The reservoir model uses a pair of horizontal wells and a network of propped and
un-propped hydraulic and natural fractures. The simulation uses fluid models employed in PVT simulation to quantify
incremental liquid recovery. Sensitivity studies capture the effects of altered dewpoint and liquid dropout
characteristics (pore-confinement), interfacial-tension dependent relative permeability curves, and critical condensate
saturations (S ).
cc
Results/Observations/Conclusions: Above P , C injection will significantly lower the original dew point pressure
dew 3
delaying or sometimes avoiding liquid-dropout regions altogether during the reservoir life. At lower temperatures
(<200° F), C3-mixed fluid may behave as an oil with significantly lowered bubble point, enhancing liquid
recovery. Below Pdew, C3 mixed with the equilibrium leaner gas can develop first or
multiple-contact miscibility with dropped-out liquids.
- CO2 or rich gas injection yields similar results. When injected above Pdew, CO2
alters liquid dropout to a lesser extent than propane. However, below Pdew, injected
CO2 may potentially lead to multi-contact miscibility to enhance liquid production. In case of rich gas,
single-phase region extends only over a small pressure range with limited enhancement to liquid recovery.
- Lean gas injection has the potential to vaporize large amounts of dropped out liquids, particularly if C7+
fraction composition is more skewed towards lower carbon numbers thus increasing liquid recovery.
Applications/Significance/Novelty: Overall, this study indicates with proper solvent injection will mitigate condensate
blocking effects and lead to improved liquid recovery with minimal productivity loss.
A Unified Model for Predicting Flowing Pressure and Temperature Distribution in the Horizontal Wellbore for Different
Energized Fracturing Fluids
1 1 2 2 2 2 2
Z. Xu; ; K. Wu; ; X. Song; ; G. Li; ; W. Yu; ; Z. Zhu; ; Z. Pang; ;
1. Harold Vance of Petroleum Engineering, Texas A&M University, College Station, TX, United States.
2. College of Petroleum Engineering, China University of Petroleum, Beijing, China.
Objectives/Scope: Energized fluids, including foam, CO , and N , are widely used for multi-stage fracturing in
2 2
horizontal wells. However, since density, rheology, thermal properties, and phase state are sensitive to temperature
and pressure, it is important to understand flow and thermal behaviors of energized fluids along the long horizontal
lateral. This study is aim to develop a unified model to accurately predict flowing pressure and temperature distribution
for different energized fracturing fluids and investigate the changes of fluid properties from the wellhead to the toe of
the horizontal wellbore.
Methods/Procedures/Process: We develop a unified model to simulate flow and thermal behaviors of different
energized fluids. The velocity and pressure are calculated based on continuity and momentum equations. By
simultaneously solving energy equations for different thermal regions in wellbore and formation, energized fluid
temperature and cooling-formation distance could be obtained. Besides, the effects of flow work, heat source/sink,
energized fluid property variation, and phase transition are incorporated into the model. The state-of-the-art equation
of state and transport models for different energized fluids are employed to calculate in-situ thermodynamic and
transport properties. Temperature, pressure, and properties of energized fluids are coupled in both depth and radial
directions using an iteration scheme.
Results/Observations/Conclusions: We verify this model against field data from energized fluid injection operations.
The relative average errors on injection pressure and maximum temperature are 3.41% and 2.55%, respectively.
Neglecting flow work and heat source/sink effect results in underestimating energized fluid temperature. At a large
injection rate, heat transfer is dominated by energized fluid properties, while at a low injection rate, formation
properties dominate the heat transfer process. As the well depth increases, fluid velocity and foam quality decrease.
To maintain a desired foam quality along the whole wellbore, optimum injection foam quality for CO foam is larger
2
than that for N foam. High-quality foam injection requires a higher injection pressure and achieves a higher
2
temperature at a deeper depth.
Applications/Significance/Novelty: This work, for the first time, provides a unified numerical model to simulate flowing
pressure and temperature distribution in the horizontal wellbore that is suitable for different energized fracturing fluids.
Our study offers an improved understanding of the impacts of key energized fluid and formation properties on injection
pressure, optimum injection foam quality, and required energized fluid temperature tolerance. Operators may use this
model to select the most suitable energized fluid and further optimize energized fluid fracturing treatments.
Development of A High-Performance Cement Slurry Antifoamer Through Lab Evaluation and Field Trials
1 1 1 1
L. Cabori; ; L. Jiang; ; B. Abrams; ; J. Terracina; ;
1. Hexion, Inc., Houston, TX, United States.
Objectives/Scope: Control of air entrainment in the cement slurry blending process is critical to cement slurry density,
mixing adequacy, gas permeability and pump cavitation avoidance. These factors have a direct impact on cement job
quality. This paper will document the development and field trials of an antifoamer, which significantly improves
cement quality and performance. Cement quality and performance are important to ensure proper well integrity and
prevent contamination of fresh water aquifers. Field case studies will be presented to support laboratory test results
using various antifoamers.
Methods/Procedures/Process: Previous research on cement antifoamers will be reviewed in detail. This paper will also
analyze laboratory tests that were performed to measure the performance of various commonly used cement
antifoamers. Laboratory tests included air entrainment tests, compressive strength tests, cement density tests, and
storage stability tests. A vital part of the paper will review field trials performed in wells utilizing an Advanced Cement
Antifoamer (ACA). Cement quality and performance results using ACA were compared to conventional cement
antifoamers.
Results/Observations/Conclusions: Laboratory cement foaming tests showed over 50% less air entrainment in cement
systems compared to conventional antifoamers. Less air entrainment results in cement with higher compressive
strength and reduces the risk of having to do remedial cement treatments for the life of the well. The field case studies
prove that ACA provided over 50% less air entrainment and improved cement density control compared to
conventional antifoamers used at twice the concentration. This antifoamer allows operators to reduce well completion
costs by reducing the antifoamer concentration by 50% and no liquid antifoamer added during the cement treatment.
The use of ACA resulted in reduced risk and liability of low quality cement from improper cement density control using
conventional antifoamers.
Applications/Significance/Novelty: This paper will introduce an innovative cement antifoamer designed with unique
chemistry to significantly reduce air entrainment, while using a smaller dosage compared to leading alternative
commercial products. This performance has been validated in various cement systems and a wide range of surface
and downhole well conditions.
Fracture Initiation and Propagation Characteristics for Radial Drilling-Fracturing: An Experimental Study
1 1 2
Q. Liu; ; K. Sepehrnoori; ; W. Yu; ;
1. University of Texas at Austin, Austin, TX, United States.
2. Texas A & M University, College Station, TX, United States.
Objectives/Scope: Field experience of hydraulic fracturing indicates two main problems: fracture initiating pressure is
high; and fracture geometry is complex. Radial drilling-fracturing, integrating radial drilling and hydraulic fracturing, is
an innovative approach to settle above problems. The radial borehole is drilled from the main wellbore. Then,
fracturing fluid is injected through the radial borehole to the formation, and fracture initiates and propagates along the
radial borehole. Investigating fracture initiation and propagation characteristics are essential for the successful design
and operation.
Methods/Procedures/Process: In this study, we conduct a series of tri-axial fracturing experiments. The maximum
confining pressure of the equipment is 20MPa, and maximum fluid injection rate is 60ml/min. The cement cubes are
used, with size of 300mm*300mm*300mm.The main well and radial boreholes are prefabricated uniformly in the
cement cubes. Six crucial factors are considered, including: in-situ stress regime (normal fault, reverse fault and strike
slip), lateral orientation, lateral number (1~6), lateral length (2cm~7cm), lateral diameter (5mm~10mm) and natural
fracture. We also make a comparison between with conventional perforation fracturing. The fracture initiation pressure
is recorded and the finial fracture geometry is analyzed.
Results/Observations/Conclusions: Results show that the fracture initiation pressure from radial borehole is much
lower than perforation fracturing. With the lateral number, length and diameter increasing, the fracture initiation
pressure reduces. Besides, we find that when the lateral number is less than 5, multiple longitude fractures propagate
along the radial boreholes, forming planar fractures, and the fracture reorientation by in-situ stress is reduced.
Therefore, the radial borehole guide the fracture propagating in the desired direction. Furthermore, we find that when
the radial lateral intersects with the high permeable faults or joints, fracture firstly initiates from these weakness
planes. Hence, radial drilling-fracturing technology could achieve good effects in fractured reservoirs, such as coalbed
methane.
Applications/Significance/Novelty: The experiment study proves that radial drilling-fracturing technology could reduce
fracture initiation pressure and guide the fracture propagation. Besides, we also found that this technology could
achieve good effects in fractured reservoirs.
Rapid Evaluation of Diverter Effectiveness From Poroelastic Pressure Response in Offset Wells
1 2 1 1 1
C. Kahn; ; B. Cottingham; ; S. Kashikar; ; S. Senften; ; E. Coenen; ;
1. Reveal Energy Services, Houston, TX, United States.
2. Linn Energy, Houston, TX, United States.
Objectives/Scope: Multiple diagnostic studies have shown that diverters used in hydraulic fracturing result in 40% to
50% of the perforation clusters to remain un-stimulated. This paper describes four case studies from the Eagle Ford,
Permian and Stack plays, showing the application of pressure data acquired from isolated “monitor” stages on offset
wells during treatment of adjacent wells, to make rapid evaluation of diverter effectiveness and adjust the diversion
techniques. In real-time optimization, the results of a diverter stage are used to adjust the next diverter stage to
optimize placement.
Methods/Procedures/Process: Fracture geometries are estimated from a network of poroelastic pressure responses
collected from isolated “monitor” stages on offset well during treatment of adjacent well using a fully-coupled, three-
dimensional, finite element effective stress model. This method enables computation of fracture growth over the
treatment interval, providing fracture growth patterns before and after diversion placement. Diversion effectiveness is
quantified using these fracture growth patterns. Diversion effectiveness is classified into four categories from
successful in stopping dominant fracture growth, to diversion having no impact as well as adverse effect of diversion in
accelerating dominant fracture growth.
Results/Observations/Conclusions: Offset pressure monitoring was implemented on wells in several unconventional
basins. Fracture geometry was computed to quantify the fracture growth patterns before and after individual diverter
drops. Both far-field and near-field diverters were evaluated, including single and multiple diverter drops per stage.
Results showed wide variations in success of a given diversion technique – ranging from 65% of diverter drops being
successful to 24% of the drops being counter-productive. Near real-time analysis permitted rapid modifications of the
diversion techniques enabling operators to adjust and identify diversion technique that provided consistent diversion.
Applications/Significance/Novelty: A unique approach that represents a significant step change in evaluating diversion
effectiveness has been developed using offset pressure measurements, enabling a process to fully stimulate each
stage, enhance cluster efficiency and improve fracture distribution across the stage. Knowledge of diverter
effectiveness will lead to better understanding of the diversion process and enable real-time optimization, where the
results of a diverter stage are used to adjust treatment design for subsequent stages. These diagnostic measurements
will augment and enhance other diagnostic methods currently in practice.
Geomechanical Modeling and Wellbore Stability Analysis Approach to Plan Deep Horizontal Wells Across Problematic
Shale Formation
1, 2
A. K. Abbas; ; R. Flori; 2; M. Alsaba; 3;
1. Drilling and Completion , Iraq Drilling Company , Rolla, MO, United States.
2. Geosciences, Geological and Petroleum Engineering , Missouri University of Science and Technology, Rolla, MO,
United States.
3. Australian College of Kuwait, Kuwait City, Kuwait.
Objectives/Scope: Zubair oil reservoir has a significant potential to contribute to the petroleum supply in Iraq. Drilling
long horizontal wells is one of many technologies commonly used to enhance the production in mature fields. Drilling
these wells in Zubair shale formation has been a challenge due to severe wellbore instability issues. This study aims
to find the optimum azimuthal direction to place long horizontal wells that will minimize the risk of wellbore breakouts
and ensure sustainable production under depletion mode.
Methods/Procedures/Process: In this study, an integrated workflow was implemented mainly to build a geomechanical
model using offset well data. The in-situ principal stresses and their orientations were obtained from wireline logging
measurements and mini-frac tests. The repeat formation test was used to constrain pore pressure. Rock mechanical
properties were calculated using empirical correlations that have been derived from laboratory tests on core samples.
Three different failure criteria including Mohr–Coulomb, Modified Lade, and Mogi–Coulomb were used to address safe
mud weight for drilling horizontal wells successfully. The model was verified with wellbore failure and also with
wellbore instability events while drilling the horizontal wells.
Results/Observations/Conclusions: The heterogeneity of Zubair shale formation and the inappropriate drilling practice
have signified the severity of the drilling problems. The results obtained from the geomechanical model shows that
both intermediate and production sections of planned wells would have narrow mud weight windows. The analyses
also emphasized that the shale interbeds in the intermediate section require a minimum mud weight of 10-10.5 ppg to
limit hole breakouts to a manageable level considering hole cleaning challenges in high-angled wells. The drilling
programs for future horizontal wells through the problematic Zubair shale formation was improved by designing an
optimal mud weights, safe drilling margins, and casing setting points for different well trajectories focusing on the
azimuthal and inclination.
Applications/Significance/Novelty: The presented study highlights that geomechanical model can be applied as cost-
effective tools to assess and address existing wellbore instability problems and to guide future well plans for better
drilling efficiency with a reduced non-productive time (NPT) by using proper mud weight with respect to the inclination
angle and the azimuthal direction.
Identifying Volcanic Ash Beds and Lamina-scale Stratigraphy Using Rock Mechanical Properties
1 1
A. Hildick; ; J. Havens; ;
1. Fracture ID, Denver, CO, United States.
Objectives/Scope: The ability to differentiate clay-rich volcanic ash beds from other fine-grained beds is important for
optimizing production in the Austin Chalk and Eagle Ford reservoirs. Core acquisition or expensive logging tools are
typically needed to identify these ash layers. They can also be identified using mechanical properties, however
potential tools for calculating mechanical properties including LWD or wireline sonic are also rarely acquired due to
high costs and risk. This study uses drilling vibration observations recorded while drilling to differentiate ash beds from
other fine-grained beds.
Methods/Procedures/Process: Continuous, high-resolution measurements of drilling induced vibrations were obtained
in a cored interval from the Lower Austin Chalk through the Eagle Ford Shale. The recording tool was run behind a
coring assembly for approximately 250’ to record the rock mechanical data at in situ pressure and temperature.
The collection of this data was used to provide stiffness coefficients and determine relative values of mechanical
properties including Young's Modulus and Poisson’s Ratio. These elastic parameters were determined by populating
stress-strain relations with forces and displacements measured near the drill bit. The resulting mechanical curves
were compared directly with core images in order to test the validity of in situ geomechanical measurements.
Results/Observations/Conclusions: Volcanic ash beds were successfully differentiated from other fine-grained beds in
the cored interval based on their distinct mechanical properties. Differences in mineralogy, in-situ water content,
temperature and pressure determine the mechanical properties of each layer. Delineation of volcanic ash beds in the
wellbore can be of tremendous value when designing a completion strategy, since they are known to inhibit fracture
height growth.
The mechanical curves also captured small-scale fining- or coarsening-up packages and thinly-bedded intervals not
identified by logging tools. This high-resolution mechanical and stratigraphic data gives operators the opportunity to
map core measurements back to reservoir characteristics in the lateral wellbore.
Applications/Significance/Novelty: Determining lithologic and elastic property changes is key to optimizing production.
Volcanic ash layers can be important stratigraphic markers, but present challenges for completion and production of
horizontal wells. Identifying ash beds and other mechanical changes without expensive acquisition of data alleviates
much of the typical economic and operational burden associated with these types of analyses. Therefore, mechanical
rock property data gathered using drilling vibrations can strongly influence economic efficiency of completion
operations.
Compressibility, Porosity, and Permeability of Shales Involving Stress Shock and Loading/Unloading Hysteresis
1
F. Civan; ;
1. Mewbourne School of Petroleum and Geological Engineering, University of Oklahoma, Norman, OK, United States.
Objectives/Scope: This paper presents the theory and formulation of the relationships of the compressibility, porosity,
and permeability of shale reservoirs by considering the effects of stress shock and loading/unloading hysteresis.
The slope discontinuities observed during loading and unloading in the porosity and permeability data of various
sources are properly represented by the present formulation. This slope discontinuity happens because the
contribution of the matrix and fracture porosity and permeability changes significantly at a critical effective stress at
which condition the fractures close or open depending on whether the process is loading or unloading. This paper
shows that the hysteresis phenomenon occurs as a result of partially reversible and partially irreversible deformations
of the various shale rock constituents by various complicated processes that the formation rock undergoes during the
loading and unloading processes.
Accurate correlations of experimental data are obtained by honoring the effect of slope discontinuity observed in the
relationship of the compressibility, porosity and permeability to effective stress during the loading/unloading
hysteresis. It is shown that the stress shock phenomenon occurs at a critical effective stress which causes the natural
fractures to close or open during loading and unloading.
Applications/Significance/Novelty: The present rigorous approach enables the correlations of the below and upper
critical stress data separately with different relationships. In contrast, the previous efforts published in various
literatures attempted to correlate the whole data by one equations without addressing the slope discontinuity observed
at a critical yield stress
Rigorous Estimation of the Initial Conditions of Flowback Using a Coupled Frac/Dynamic Drainage Area Model
Constrained by Laboratory Geomechanical Data
1 1 1
B. Yuan; ; Z. Zhang; ; C. R. Clarkson; ;
1. Geoscience, The University of Calgary, Calgary, AB, Canada.
Objectives/Scope: The application of rate-transient analysis (RTA) concepts to flowback data gathered from multi-
fractured horizontal wells (MFHWs) completed in tight/shale reservoirs has recently been proposed as an independent
method for quantitatively evaluating hydraulic fracture properties. However, the initial fluid pressures and saturations
in the fracture network, and adjacent reservoir matrix, are generally unknown at the start of flowback, creating
significant uncertainty in the quantitative analysis of flowback data. An approach to recreating these initial conditions is
to simulate fluid leakoff during hydraulic fracture propagation (during the stimulation treatment) and subsequent shut-
in period prior to flowback. In this study, we present a semi-analytical flow model, coupled with a frac model and
constrained with laboratory-based geomechanical data, for evaluating the initial conditions of flowback.
Methods/Procedures/Process: In previous work, a semi-analytical model based on the dynamic drainage area (DDA)
concept was used to simulate water-based fluid leakoff from a MFHW into an oil saturated tight oil reservoir in western
Canada during and after fracturing operations. The model assumed that each fracturing stage could be represented
by a primary hydraulic fracture (PHF, containing the majority of the proppant), and adjacent non-stimulated reservoir
(NSR) or enhanced fracture region (EFR, area of elevated permeability in reservoir caused by the stimulation
treatment). Each region was represented by a single-porosity system. The DDA propagation speed within the PHF
during the stimulation treatment was constrained through the use of a simple analytical frac model. While this
approach was considered novel, several improvements to the modeling approach, and additional laboratory
constraints, were considered necessary to yield more accurate predictions of flowback initial conditions.
Results/Observations/Conclusions: In the current work, the modeling approach described above was improved by 1)
representing the EFR with a dual-porosity system and 2) fully coupling the frac model (used for PHF creation and
propagation) with the DDA model for fluid leakoff simulation into the EFR. Improvement 1) was considered necessary
to more realistically represent the spatial distribution of fluids in the EFR and associated saturations and pressures.
Improvement 2) was considered necessary to more realistically control PHF propagation speed. Finally, to ensure that
model geomechanics were properly constrained, a comprehensive suite of laboratory data was gathered. Laboratory-
derived propped (PHF) and unpropped (EFR) fracture permeability data as a function of pore pressure, as well as
fracture compressibility data, were used as constraints to the model. The improved model was re-applied to the tight
oil field case and yielded more realistic estimates of flowback initial conditions, enabling more confident history-
matching of flowback data.
Applications/Significance/Novelty: The results of this study will be of importance to those petroleum engineers
interested in quantitative analysis of flowback data to obtain fracture properties, but who are concerned about correct
initialization of models for flowback simulation leading to more realistic results.
Seismic to Simulation: Woodford Shale Case Study in Oklahoma, USA
1 1 1 1 1
E. J. Torres-Parada; ; S. Sinha; ; L. E. Infante-Paez; ; R. M. Slatt; ; K. Marfurt; ;
1. Conoco Phillips School of Geology and Geophysics, The University of Oklahoma, Norman, OK, United States.
Objectives/Scope: In this paper, we perform a reservoir characterization of Woodford shale integrating seismic
attributes, seismic inversion study, well logs and cores. We then use the results of the reservoir characterization to
identify sweet spots in our area of interest. We divide our area in six regions based on our sweet spot rankings. We
then use a fully coupled geomechanical simulation on our model to generate reference model for every area and
propose a methodology to calibrate the type curves (TC) in Woodford shale for resource estimation.
Methods/Procedures/Process: We first use well logs to determine the TOC, porosity, permeability etc. and calibrate it
to the core derived properties. We then use seismic multi-attribute analysis to derive facies in the region and use
inversion guided property model to populate petrophysical properties away from wells.
We use the inversion derived geomechanical properties to populate our grid. We divide our area into six different
zones based on an in-house developed machine learning algorithm constrained by geological boundaries. We extract
sub-models representative of these regions and use them to perform a fully coupled geomechanical simulation.
Finally, we base our TC on the reference production curves from this model.
Results/Observations/Conclusions: Our results show multi-attribute analysis provide a better way to populate the
geocellular model instead of individual seismic attributes and hence we illustrate multivariate way of property
modeling. Geomechanical simulation provides a robust way to model gradual closing of fractures and hence a time
variant “shrinking” simulated rock volume instead of conventional history matching with multiple permeability zones.
Our natural fracture model along with hydraulic fractures provide a better way to observe the pressure distribution in
the area and can be used to optimize well spacing or future enhanced oil recovery (EOR) projects.
Applications/Significance/Novelty: Our study identifies geological sweet spots and type curves in the Woodford shale
along with major facies. We provide a sound physical basis to generate robust models as a reference solution to base
the type curves. As our workflow is integrated from seismic to simulation it provides a general guideline for operators
to scale their models from well logs to seismic and use good of both worlds. We include tight boundary conditions to
base our methodology and hence we provide more confidence level on type curves to estimate the reserves for the
operators.
The Importance of Overburden and Pore Pressure on Horizontal Stress Magnitude Determination: An Example From
the Delaware Basin
1 1 1 1 1 1 1 1
K. Kozlowski; ; M. Da Silva; ; D. Brown; ; J. Taylor; ; H. Willems; ; T. Watson; ; D. Burch; ; T. Hutton;
; C. Christensen; 1; M. Manohar; 1;
1. Noble Energy, Inc, Houston, TX, United States.
Objectives/Scope: Accurate estimation of minimum horizontal stress is often challenged in unconventional plays due
to uncertainties in pore pressure and overburden stress. Lack of direct pressure measurements, and sparse data
collection, can result in stress models that are poorly constrained. The influence of calcite and kerogen on log
response further complicates pressure interpretation.
The sensitivity of horizontal stress magnitude to variations in overburden stress and pore pressure is investigated
using a 3D mechanical earth model to characterize mechanical facies, pore pressure and insitu stresses.
Methods/Procedures/Process: The Permian Delaware Basin, located in west Texas, has a complex depositional and
burial history. To capture these complexities, a 3-D earth model was built to integrate data from both geoscience and
engineering. The foundations of the model are core calibrated petrophysical analyses, detailed geological
interpretations, drilling parameters and events, and pressure data from both drilling and completions operations. For
pore pressure model calibration at the reservoir level, we utilize both flowback and mini-frac data from completions
operations. In the shallow overburden interval we use flow events tabulated from daily drilling reports. Lastly, for
minimum horizontal stress calibration we use a combination of pressure data from casing shoes tests as well as mini-
fracs.
Results/Observations/Conclusions: Overburden stress estimation in the Delaware Basin is strongly influenced by
lithology and thickness variations of shallow, mixed layer evaporites of Ochoan Age. Using novel integrated workflows,
we construct 3-D density models to assess the impact of overburden stress on pore pressure and minimum horizontal
stress at the reservoir level. The resultant 3-D models show abrupt vertical variations in pore pressure as well as
notable lateral variability. Furthermore, shallow evaporite lithology and thickness changes produce significant
overburden stress variations at the reservoir target.
Applications/Significance/Novelty: In unconventional reservoir systems, accurate minimum horizontal stress
determination is critical for well design, drilling operations efficiency and completions optimization.
As both pore pressure and overburden stress have a first order effect on minimum horizontal stress, it is critical to
characterize this information accurately. A 3-D earth model that captures these variations, facilitates more accurate
insitu stress determination for well planning, drilling operations and completions optimization.
A Strategy for De-risking the Upper Vaca Muerta as a Dual-Layer Development
1 1 1 1
K. Boyd; ; R. Notta; ; M. Thomas; ; D. Cakici; ;
1. Shell Oil Company, Houston, TX, United States.
Objectives/Scope: Shell Argentina has been exploring the Vaca Muerta Play in Neuquén Basin, Argentina since 2012.
During this time, two landing targets were identified in the Lower and Upper Vaca Muerta, respectively. Initially, drilling
focused on the Lower target, a reservoir that is well understood and predictable within Shell’s acreage. Now, Shell’s
technical team is broadening its focus to further de-risk the more heterogenous and challenging Upper Vaca Muerta
with the objective to optimize a two-layer development.
Methods/Procedures/Process: The existing Sierras Blancas multi-well pad (SB-1004, 1005 & 1006) forms the
foundational data set for current subsurface understanding, with one horizontal well landed on each target and one
slant well crossing both targets.
To further de-risk the play, Shell refined the stratigraphic interpretation of the upper target, finetuning
paleoenvironmental interpretation and its link to sweet spots, containment, heterogeneity and pay thickness. The
resulting subsurface maps are good indicators of the most likely areas for a two-layer development.
Next, we will execute an instrumented multi-well pad, designed to assess production interference between both
targets. This pad will be equipped with fiber optics, chemical tracers, microseismic and geochemical fingerprinting of
produced oils.
Results/Observations/Conclusions: Within Shell’s acreage, the Lower VM target is comprised of toe of slope to basin
floor deposits resulting from the initial transgression and early progradation phases. This reservoir is wide spread,
homogeneous and its containment is well defined between Catriel Member, at the base, and a regressive sequence,
to the top.
The Upper VM target is a more complicated, distal slope depositional setting. This proximity to the shelf increases the
impact of relative sea level changes and results in more heterogeneity, and variability in rock properties. Also, fracture
containment is an issue, as the hard limestone of the Quintuco Fm. is a strong barrier to upward growth where it
overlies the reservoir, but, where a transition zone with poor marls overlies it, the fractures are not well contained.
Applications/Significance/Novelty: The ideal unconventional development plan is one where all the target reservoirs
are drilled and completed in one pass; thus, lowering the drilling and completion risks and maximizing pad efficiency.
By judicious and timely multidisciplinary data gathering, we aim to optimize the development plan as early as possible.
Geological Controls on Fluid Compositional Variations in Unconventional Hybrid Plays: Insight From Gas
Geochemistry (Montney Play, Western Canada)
1 2 3
T. Euzen; ; J. Chatellier; ; A. Mort; ;
1. IFP Technologies (Canada) Inc., Calgary, AB, Canada.
2. Tecto Sedi Integrated, Calgary, AB, Canada.
3. Land & Minerals Sector, Geological Survey of Canada, Calgary, AB, Canada.
Objectives/Scope: Condensate-gas ratio in liquid-rich unconventional plays is primarily controlled by thermal maturity
and hydrocarbon migration pathways. While regional maturity trends have long been recognized, predicting fluid
compositional variations at sub-regional to local scales remains challenging, due to stratigraphic and structural
complexity. The present study aims at demonstrating geological controls on hydrocarbon compositional variations in
the Montney Play of Western Canada through the integration a large gas analysis dataset within a well-defined
stratigraphic framework.
Methods/Procedures/Process: Understanding the mechanisms and geological heterogeneity that control the present-
day distribution of in situ fluids in unconventional plays, is key to better predict spatial variations of well productivity
and liquid/gas ratio. Recombined gas wetness, iC4/nC4 ratio and other compositional attributes give indirect
indications of hydrocarbon thermal maturity, migration and mixing. Based on production data from over 1,500
producing wells, compositional variations were mapped within individual stratigraphic units over an area of about
1,500 square kilometers. Recombined gas composition trends reveal discontinuities highlighting local stratigraphic and
structural controls on fluid distribution.
Results/Observations/Conclusions: Compositional trends mapped in this study suggest that stratigraphic and
structural discontinuities influence fluids distribution in the Montney Formation. Third-order sequence boundaries and
internal clinoform architecture, provide the framework controlling facies belts and hydraulic units. Fluids composition
produced from different stratigraphic intervals show distinct spatial trends, stressing the need for analyzing
compositional variations within a well-defined stratigraphic framework. Within individual stratigraphic units, both gas
wetness and iC4/nC4 ratios show local departure from the regional maturity trend and the influence of linear
discontinuities interpreted as faults. Furthermore, gas-enriched elongated areas consistent with regional gas migration
pathways are also observed.
Applications/Significance/Novelty: The huge amount of well data released by operators in recent years, combined with
an improved understanding of the high-resolution stratigraphic architecture of the Montney Formation, allows for an
unprecedented review of the geological controls on fluids distribution in an unconventional hybrid play. This
knowledge can be used by Montney operators to better predict and target liquid-rich sweet spots at sub-regional and
local scales and as a conceptual analogue for other unconventional hybrid plays.
Outlier Analysis: A Systematic Method for Distinguishing Between Sub-surface and Engineering Influence on Well
Performance in the Montney
1 1 1 1 1 1 1
F. Todea; ; B. Stephenson; ; A. Tomlinson; ; H. Pratt; ; W. Williams; ; L. Acosta; ; B. Speidel; ;
1. Shell Canada Ltd., Calgary, AB, Canada.
Objectives/Scope: A fundamental conundrum common to hydraulically stimulated resource plays is how to understand
the relative contribution of the sub-surface versus the drilling and completion parameters to well performance. A
systematic methodology has been applied on a study of over 400 dry-gas wells in the Montney siltstone of British
Columbia, with a variety of completion intensities, to extract the primary drivers of this multi-variate problem. The goal
is to know how to tailor the drilling and completion design across the acreage as the sub-surface changes, to provide
the optimum economic outcome.
Methods/Procedures/Process: A systematic workflow, Hypothesis-Based Development, provides a rigorous
framework to guard against biases, by forcing normalization of completions, well-spacing and the primary geological
variability; in this play, depositional facies controlling effective permeability. The key analytical step, Outlier Analysis, is
a methodology to prove, or disprove, hypotheses for well-performance by looking for opposite signatures in the very
best and very worst wells; the ‘outliers’. A data analysis tool was created to interactively screen hypotheses against all
the wells, explore trends and search for data-driven correlations. Being interactive, asset teams can arrive at solutions
together, fostering alignment quickly, allowing focus to shift on how to affect development decisions.
Results/Observations/Conclusions: A long-standing conundrum in the Montney study area was why there were no
strong correlations between individual subsurface parameters, like porosity, and well performance. After testing key
hypotheses for well-spacing and completion intensity, these parameters were found to have a dominant influence on
outlier well productivity. The sub-surface signature was detected by choosing a high-graded, sub-set of ‘stand-alone’
wells, completed with single-point cemented sleeves at the same tonnage and spacing (50T, 56m), in which the
primary engineering parameters had effectively been normalized out. A bulk-volume gas term, normalized for clay
2
content and depth has a robust correlation (R > 0.7) with an inverse productivity metric, normalized for 1000T
proppant.
Applications/Significance/Novelty: These results from Outlier Analysis have many applications. Using the metric for
reservoir quality as a proxy for effective permeability, decisions around stage length can be optimized for frac to frac
production interference. A map of potential deliverability allows easy recognition of under-performing wells, which can
be flagged as candidates for intervention. Competitor data can also be quickly assessed for completion optimization
learnings, saving capital. In summary, Outlier Analysis is a powerful technique applying the scientific method by
learning from negative, as well as positive outcomes, which can be applied on any number of wells; with confidence
increasing as more wells are drilled.
An Upscaled DFN Model to Understand the Effects of Natural Fracture Properties on Fluid Flow in the Hunton Group
Tight Limestone
1 1 1
B. Milad; ; S. Ghosh; ; R. M. Slatt; ;
1. Geology , University of Oklahoma, Norman, OK, United States Minor Outlying Islands.
Objectives/Scope: Natural fractures likely have a significant role in the Hunton Group tight carbonate effective
permeability due to its low matrix porosity/permeability. This paper presents a realistic, multidisciplinary approach to
fracture characterization in the Hunton Group. The fracture parameters were used to build a DFN model within a
seismic volume. The volume was then upscaled to obtain reservoir grid properties. The objective is to use this model
and subsequent models to understand the effect of fracture parameters (mainly fracture permeability and length) on
fluid flow in the Hunton Group.
Methods/Procedures/Process: Fracture data were obtained from various sources. Fracture sets were identified from
flat outcrops and borehole images. Besides, fracture lengths were studied at the outcrops. Matrix porosity and
permeability were obtained from the core. Fracture intensities, on the other hand, were obtained from image logs.
Also, the fracture apertures from image logs (high uncertainty) were adjusted to fracture apertures from the core and
outcrops. In addition, fracture intensities in the seismic volume were assigned based on the facies identified by
prestack inversion. The cubic law was used to obtain fracture permeabilities.
Results/Observations/Conclusions: Fracture aperture and permeability of the Hunton Group is heterogeneous due to
fracturing at different scales. Core observations indicate that fractures originated due to karsting. Outcrop
observations, on the other hand, indicate that fractures (joints) are related to overpressure or minor structural bending.
Outcrop fractures are mostly cemented. Interpreted fractures from borehole images are open, partially open, and
o o o o
cemented. Four fracture sets (E-W [81-110 ], N-S [340-10 ], NW-SE [111-160 ], and NE-SW [11-80 ]) were
observed using both borehole images and outcrops indicating outcrops can be good analogs for the subsurface.
Fracture length and apertures mostly follow characteristic-size distributions. Also, wackestones have higher fracture
intensity compared to the mudstones.
Applications/Significance/Novelty: The aforementioned observations and measurements were used to build a
stochastic discrete fracture network (DFN) in Petrel. Subsequently, we upscaled the DFN (calibrated to seismic) using
the Oda method to assign grid porosity/permeability values to the seismic volume. Most of the upscaled grids contain
permeability values in the 0-100 md range. Simulations will be performed in dual porosity/dual permeability modes to
understand fluid flow. This model and subsequent models can be utilized by the industry as the models come from the
integration of varied and comprehensive datasets.
Sedimentological Characterization of the Pimienta Formation in the Central Part of the Tampico-Misantla Basin,
Veracruz, Mexico
1 1 1 1
G. Abascal-Hernández; ; J. M. León-Francisco; ; R. Torres-Vargas; ; D. E. Garduño-Martínez; ; S. Franco-
1 1 1 1 1
Navarrete; ; J. Méndez-Vázquez; ; S. Ortega-Lucach; ; L. Gutierrez-Caminero; ; G. Murillo-Muñetón; ;
1. Instituto Mexicano del Petróleo, Mexico City, Mexico.
Objectives/Scope: The Tithonian Pimienta Formation represents the main hydrocarbon source rock in Mexico,
recently being targeted as an unconventional reservoir. This work shows the results of a detailed sedimentological
characterization of the Pimienta Formation in the Tampico-Misantla Basin, Mexico.
Methods/Procedures/Process: This study integrates the well information, gamma ray log (GR), seismic interpretation
and a high-resolution sedimentological analysis of an analogue outcrop core which involved facies identification,
biostratigraphy and chemostratigraphy. The continuous core covers the whole Pimienta Formation and its lower and
upper stratigraphic boundaries (San Andres and Tamaulipas Inferior formations respectively).
Results/Observations/Conclusions: Three stratigraphic members were recognized in the Pimienta Fm, the lower
member consists mainly of saccocoma wackestone and calcispheres wackestone, with increasing GR values towards
the top, suggesting a deposition on an outer platform to deep marine environment during early Tithonian time. The
middle member comprises predominantly argillaceous carbonate mudstone and radiolarian wackestone, both with
high organic matter, V, Si and Al content, high GR values, indicating a deep marine environment and a middle
Tithonian age. The upper member contains mainly bioclastic wackestone and mollusk-fragments wackestone with
decreasing GR values towards the top and low organic matter content, implying an outer platform environment and a
late Tithonian age.
Applications/Significance/Novelty: According to their sedimentological features, both lower and middle members of the
Pimienta Formation exhibit the best characteristics as an unconventional reservoir. This analysis contributed to
identify sweet spots within the Pimienta Formation in the study area.
Increase NPV and Reduce Completion Cost by Using an In-House Automated Fracture Design, Reservoir Simulator
and Optimization Workflow to Design Well Spacing and Completions
1 1 1
S. Bhattacharya; ; E. Lake; ; R. Dombrowski; ;
1. Shell Oil Company, Houston, TX, United States.
Objectives/Scope: Unconventional reservoirs are intended to provide cash flow in the longer term. They rely on
horizontal drilling and hydraulic fracturing to produce from them. Longer fractures in terms of well completion means
higher pumping rates, higher treating pressure and surface horse power and more materials such as fracture fluid and
proppant. “More” does not necessarily mean “optimal” in this case. In this work, engineering optimization is performed
that uses subsurface based models along with the treatment costs for different completion efficiency (uniform, non-
uniform and dominant completions).
Methods/Procedures/Process: In this research, we have developed an in-house automated approach to import
spatially variable fracture properties with non-rectangular fracture geometry generated by a 3D fracture simulator into
a reservoir simulator. Coupling this model to an optimizer, allows optimizing several well and completion parameters.
The proposed workflow would allow optimization on several parameters, for example, well spacing, well orientation,
landing depth, lstage spacing, perforation cluster spacing, etc. Three different completion efficiencies were considered
in this study: uniform fractures, non-uniform fractures and dominant fractures. Total 1800 multiwell fracture realizations
from different completion parameter’s combinations for the three completion efficiency cases were modeled and
compared.
Results/Observations/Conclusions: The optimal pareto front obtained for the three different completion types show the
optimal cases where objective is to maximize NPV and reduce cost with VIR used as a constraint. Comparing the
three completions, it is observed that the optimal cases for uniform fractures case are easier to achieve than the non-
uniform and dominant fractures. The non-uniform and dominant fractures require complementary long fractures to be
in the optimal category which can be difficult to control operationally. Also, uniform fracture cases provide significant
value as compared to non-uniform and dominant cases, which conforms with the observed best practice completions
from the field application.
Applications/Significance/Novelty: The workflow integrating fracture simulator, reservoir simulator and optimizer is
applied on the Shell's Uncoventional Asset.
Numerical Study of Stress Shadowing Effect on Fracture Initiation and Interaction Between Perforation Clusters
1, 2
B. Damjanac; ; S. Maxwell; 2; A. Pirayehgar; 2;
1. Itasca Consulting Group, Inc., Minneapolis, MN, United States.
2. IMaGE, Calgary, AB, Canada.
Objectives/Scope: During stimulation of stages with multiple perforation clusters, complex interactions between
fractures take place, resulting in different numbers of propagated fractures (usually greater than one but less than the
number of the perforation clusters), non-planar fracture surfaces, asymmetric propagation and variable fracture
heights along the fracture lengths. A numerical model is used to investigate the effects of fracture interactions as a
function of completion designs and operational parameters.
Methods/Procedures/Process: A coupled hydro-mechanical model based on the lattice approach is used to simulate
fracture initiation and propagation during stimulation of one stage. The lattice method is used to represent deformation
and fracturing of the reservoir rock subject to changing fluid pressures. Stress state, a function of the initial state and
fracture-induced changes, is simulated evolving throughout the reservoir. The stress-shadow effect (of the fractures
that initiate first, on suppression of initiation and propagation of the fractures from the adjacent perforation clusters) is
automatically accounted for. The fracture geometry is part of the model solution, potentially resulting in asymmetrical,
non-planar fractures of variable height.
Results/Observations/Conclusions: The model predicts that the number of fractures propagated within one stage with
multiple perforation clusters is greater than one but typically less than the number of perforation clusters. The actual
number depends on the number of parameters, including the cluster spacing, injection rate, viscosity of fluid and well
orientation. Typically, fractures are non-planar. Depending on the spacing of the perforation clusters, the fractures can
attract (eventually merging) or repulse each other. In some cases, as a result of stress shadow, the perforation
clusters propagate asymmetric fractures with a single wing, alternating the sides of the propagated wing between the
adjacent perforation clusters. Another similar effect is variable height along the fracture length.
Applications/Significance/Novelty: The numerical model, and to some extent the presented results, can be used to
better understand the effects of completion and different operational parameters on the number and geometry of
hydraulic fractures within a stage. Ultimately, the model can be used in the completion design and optimization of
pumping schedules based on criteria of uniform fracturing of the rock mass and maximization of stimulated rock
volume. The modeling approach is physics based, fully-coupled hydro-mechanically and allows general interaction of
the hydraulic fractures.
An Integrated Field and Numerical Study of the Impact of Formation Anisotropy on Stage Spacing in Horizontal Wells
1 1 2
V. Sesetty; ; A. Ghassemi; ; I. Gil; ;
1. Petroleum Engineering, University of Oklahoma, Norman, OK, United States.
2. BP, Houston, TX, United States.
Objectives/Scope: Despite the evidence suggesting that shales are strongly anisotropic, current hydraulic fracture
models do not take rock anisotropy into account. This paper presents improved traditional plane strain hydraulic
fracture propagation models such as pseudo three-dimensional height growth model (P3D) and two-dimensional fixed
fracture height model for the case of anisotropic rock.
Methods/Procedures/Process: Singular integral equation that relates fracture opening with fluid pressure inside the
fracture in an elastic anisotropic body is used to extend P3D model to anisotropic case. On the other hand,
displacement discontinuity method with height correction factor is used to model 2D fracture propagation in anisotropic
rock. Fracture deformation is fully coupled with fluid flow inside the fracture.
Results/Observations/Conclusions: Numerical analysis of a Utica shale field-scale fracturing job is presented using
anisotropic rock properties. The variation of fracture height, length and pressure with time and stress shadow around
the fractures are calculated and the results are compared with available fracturing field data and micro-seismicity to
interpret the stimulation. The numerical results along with field observations indicate the effect of rock anisotropy on
fracture apertures and stress shadow could be substantial even for a formation with a low degree of anisotropy (e.g.,
1.5). The results are used to optimize completions design for the next target zone.
Applications/Significance/Novelty: As rock anisotropy affects interaction between the clusters, the optimum spacing
between them to achieve uniform growth among multiple parallel fractures in anisotropic rock turns out to be different
from that for isotropic rock. The insights from these results could be used to optimize completion strategies to achieve
maximum cluster efficiency.
Effect of Interbeds on Hydraulic Fracture Characteristics and Formation Pressure Response
2 1 1 1
U. Mutlu; ; A. Bere; ; M. Profit; ; M. Dutko; ;
1. Rockfield Software, Swansea, Wales, United Kingdom.
2. Rockfield Global Technologies America LLC, Houston, TX, United States.
Objectives/Scope: 2D and 3D geomechanical models are used to simulate hydraulic fracture propagation in high
frequency interbedded/laminated rocks. Models integrate routinely obtained reservoir & geomechanical log/core data
and mechanical stratigraphy with interbed properties to quantify hydraulic fracture branching, complexity and
conductivity. The primary objective of this study is to gain an understanding of the physical and operational controls
that result in the observed hydraulic fracture patterns. The models attempt to provide guidelines for well placement,
cluster spacing and stimulation parameters.
Methods/Procedures/Process: The approach is based on finite discrete element methodology with flow and
geomechanical coupling. This coupling allows the realization of the complex interactions between fracture, mechanical
stratigraphy and interbeds/laminations. Rock matrix is represented by a poro-elasto-plastic model that can honor
mixed mode rock fracturing. Interbeds are modeled either as plane of weaknesses or as discrete contacts where flow
and strength properties are assigned. Pore fluid coupling is single phase and flow is controlled by conductivity
changes due to applied mechanical and hydraulic loads. Simultaneous fracture propagation from multi-clusters with
stress shadows can also be simulated. Methodology is described in detail and governing equations are documented
for the readers.
Results/Observations/Conclusions: Simulations quantify the impact of key parameters on fracture characteristics and
formation pressure response while taking into account the interaction between hydraulic fractures and interbeds.
Formation pressure response and fracture networks emerge from the models as dictated by the interbeds,
stratigraphy, stress state and stimulation parameters. Sensitivity analysis show that interbed properties, e.g. hydraulic
conductivity & frictional characteristics, can dictate the mechanical & flow behavior of the resulting network.
Stimulation parameters can be adjusted to control the degree of interaction to optimize fracture conductivity. Results
also highlight the need for new experimental techniques to quantify interbed properties. These properties include
strength and flow properties.
Applications/Significance/Novelty: Historically the effects of laminations and high frequency interbeds are lumped
within stress and material profiles in conventional numerical models. Therefore key parameters that control the
physics of the interaction between hydraulic fractures and pre-existing interbeds are ignored. The geomechanical
models presented in this paper allow the inclusion of high frequency interbeds with key parameters assigned to them.
These models simulate a more realistic interplay between hydraulic fractures, pre-existing interfaces/laminations and
the resulting fracture network.
A Coupled Three-Dimensional Hydraulic Fracture Propagation Model Considering Multiple Bedding Layers
1 1 1 2
J. Tang; ; K. Wu; ; L. Zuo; ; C. Ehlig-Economides; ;
1. Petroleum Engineering, Texas A&M University, COLLEGE STATION, TX, United States.
2. Petroleum Engineering, University of Houston, Houston, TX, United States.
Objectives/Scope: Weak bedding planes that exist in many tight oil and shale gas formations may strongly affect
fracture height growth during hydraulic fracturing treatment. Few of the hydraulic fracture propagation models
developed for unconventional reservoirs are able to quantitatively estimate the fracture height containment or predict
fracture geometry under the influence of multiple bedding planes. This paper introduces a coupled three-dimensional
hydraulic fracture propagation model considering the effects of bedding planes.
Methods/Procedures/Process: In this model, a fully three-dimensional displacement discontinuity method (3D DDM)
governs the rock deformation. The advantage of this approach is that it addresses both the mechanical interaction
between hydraulic fractures and weak bedding planes in three-dimensional space and the physical mechanism of
slippage along weak bedding planes. Fluid flow governed by finite difference methodology considers the flow in both
vertical fractures and opening bedding planes. An iterative algorithm couples fluid flow and rock deformation
mechanisms. Comparison between the new model and the PKN model showed good agreement.
Results/Observations/Conclusions: Sensitivity studies investigate the impact on fracture propagation of parameters
such as pumping rate, fluid viscosity, Young’s modulus and Poisson’s ratio. We investigated the fracture width at the
injection point and near the junction between a fracture tip and a bedding plane, the latter including tensile opening of
bedding planes and shear displacements along them. Shear displacements along the planes drop sharply with
distance from the junction. Pumping rate is proportional to the width near the junctions, opening of bedding planes,
and shear displacements along the planes. As viscosity increases, the junction width increases as do the shear
displacements. Young’s modulus has less impact on the opening of bedding planes at a junction, but it greatly
influences shear displacements.
Applications/Significance/Novelty: This model estimates the fracture width distribution and the shear displacement
along the bedding planes near junctions between the fracture tip and bedding planes and enables assessment of the
pumping rate required to ensure the fracture width at junctions and along intersected bedding planes is sufficient for
proppant transport.
The Impact of Natural Fracture Thickness on Hydraulic Fracture Interaction Mechanics
1 2 2 3
W. Wang; ; J. Olson; ; M. Prodanović; ; R. Schultz; ;
1. Shell International Exploration and Production Company, Houston, TX, United States.
2. Hildebrand Department of Petroleum and Geosystems Engineering, The University of Texas at Austin, Austin, TX,
United States.
3. Orion Geomechanics, Houston, TX, United States.
Objectives/Scope: Field observations indicate natural fractures(NFs) thickness varying from millimeter up to
centimeter scale. Depending on its thickness and mechanical properties, the induced hydraulic fracture(HF) may
across/divert along NFs. Commonly used fracture crossing/diverting criterion treats these NFs as frictional interfaces
without considering the cement fillings. The objective of this paper is to evaluate the NF thickness effect on HF
propagation. In addition, modulus contrast, differential stress, and rock-cement interface properties are analyzed
combined with the effect of NF thickness.
Methods/Procedures/Process: In our finite element method simulation, HF propagation and interaction with NFs are
modeled using the cohesive element approach. The simulation results are validated against the analytical fracture
crossing criterion proposed by Renshaw and Pollard (1995), which treats the NFs as frictional interfaces. Then,
different mechanical properties are assigned to regions that representing NFs for the purpose of investigating the
effect of NF thickness.
Results/Observations/Conclusions: The results show that the HF crossing behavior is strongly influenced by the NF
thickness and the rock-cement interface friction coefficient. If the NF thickness is considered, the commonly used
fracture crossing criterion by Renshaw and Pollard (1995) will overestimate/underestimate the HF crossing scenario
depending on the NF stiffness. For a compliant NF, as the NF thickness increases, the shear stress along the rock-
cement interface is increased comparing with the zero-thickness case. Thus, HF diverting along the frictional interface
is likely to occur for having thick and compliant NF. For a stiff NF, the probability for HF crossing a thick and stiff NF
increases by more than 20% when the differential stress ratio is in the ranges from 1.04 to 4.
Applications/Significance/Novelty: One application of this work is to investigate the HF height growth due to the effect
of bedding-parallel veins, which are abundant in unconventional resource plays such as the Vaca Muerta formation.
Field observations indicate the thickness of these bedding parallel veins changes from 1 mm to 15 cm, with an
average of 5 cm. When the rock−cement interface coefficient of friction is 0.5 and the stress ratio is 1.15, our results
suggest HF will cross over complaint NFs that are less than 5 mm in thickness. Whereas the Renshaw and Pollard
(1995)’s crossing criterion predicts HF will cross all NFs regardless of their thickness, which leads to an
overestimation of the crossing scenario.
Field Scale Proppant Transport Simulation and Its Application to Optimize Stimulation Strategy
1 1 1
R. Kou; ; G. J. Moridis; ; T. Blasingame; ;
Objectives/Scope: The understanding of proppant transport plays a critical role in estimating propped fracture
dimensions and performance. Existing models generally assume a vertical planar geometry, whereas the reality in the
subsurface may be much more complex. We use the discrete element method to simulate field scale proppant
transport in complex fracture networks. Our results show that sharply-angled fractures reduce fluid velocity and
increase the particle-wall interaction. The combined effects can cause early settling of proppants, thus limiting their
efficient placement and the fracture effectiveness.
Methods/Procedures/Process: To calibrate our numerical model, we conducted two validation simulations that
describe particle settling tests and laboratory proppant transport experiments. Through scoping calculations, we
determined the correct drag force model and matched both analytical solutions and experimental data for a wide
range of flow regimes that included three proppant sizes (20-30 mesh, 30-40 mesh and 50-70 mesh) in two types of
fluids (water and oil).
For the main component of our study, we simulated proppant transport in a 3 dimensional, field-scale fracture network
using our benchmarked models. In our search for an optimal stimulation strategy, we also experimented with various
perforation and pumping strategies, and then compared the proppant distribution results.
Results/Observations/Conclusions: By analyzing the velocity and trajectory of proppant particles during transport, we
identified two different stages of the proppant transport process — a "suspension" stage and a "settling" stage. During
the suspension stage, the fluid drag and the gravitational forces dominate, driving proppants further into the fracture.
When the proppants reach a flow stagnation area, or if/when the proppants collide with the fracture boundaries (side
walls or bottom of fracture), the proppant particles lose momentum and then accumulate into “dunes” and become
"settled." Finally, we observed that proppant transport in a sharply-angled fracture network leads to local flow velocity
reduction and stronger particle collision interactions, leading to dramatic reductions in the transport efficiency.
Applications/Significance/Novelty: This work provides a better understanding of proppant transport behavior in sharply
angled fracture networks. A significant difference from prior (laboratory) experiments is the capability of our model to
simulate proppant transport at field-scale flowrates. Such a capability is critical for understanding the proppant
transport behavior by ensuring that the correct Reynolds number and flow regime are used in the calculations. To the
best of our knowledge, this is the first study that emphasizes the importance of particle Reynolds number to proppant
transport process.
Lastly, using the discrete element method and high-performance (parallel) computing, we were able to represent
various perforation and pumping schemes, which is essential for developing an optimized stimulation strategy.
Hydraulic Fracture Test Site – Project Overview and Summary of Results
2 1
J. Ciezobka; ; J. Courtier; ;
1. Laredo Petroleum, Tulsa, OK, United States.
2. Gas Technology Institute, Des Plaines, IL, United States.
Objectives/Scope: The Hydraulic Fracture Test Site (HFTS) is a $25 million research project sponsored by a public-
private partnership with the Department of Energy, Gas Technology Institute (GTI), Laredo Petroleum and multiple
industry sponsors. Laredo Petroleum hosted the test site, which is located in the Midland basin of West Texas. Field
data acquisition to review fracturing and fracture diagnostics, as well as environmental and research tasks y at the
HFTS have been completed. As part of the HFTS experiment approximately 600 feet of core was obtained by drilling a
one of a kind core well through the created hydraulic fractures at the test site. The presentation will provide an
overview of the experiment, discuss field operations and risks involved and discuss research results.
Hydraulic Fractures in Core From Stimulated Reservoirs: Core Fracture Description of the HFTS Slant Core, Reagan
County, Midland Basin, Texas
1 1 1
J. F. Gale; ; S. J. Elliott; ; S. E. Laubach; ;
Objectives/Scope: The geometry and extent of the hydraulic fracture network produced during well stimulation are
usually not well known. Indirect methods such as microseismic monitoring may provide information on the network,
but are not definitive. For this reason the Hydraulic Fracture Test Site project was designed to recover core from a
stimulated volume, where the core provides direct information about the hydraulic fractures. Here we present the
methods and some results from a fracture description of the HFTS slant well (Sugg A 171 6TW), which cored through
a stimulated reservoir.
Methods/Procedures/Process: The examination was done on whole 4-inch diameter core prior to slabbing, with the
core contained within an aluminum sleeve. A CT scan of the core was made available. We developed criteria for
distinguishing between hydraulic, natural, drilling-induced and core-handling fractures by examining the orientation
and surface features of all fractures.
Results/Observations/Conclusions: More than 700 fractures total were found in 600 ft of core, including hydraulic
fractures, two sets of calcite-sealed natural fractures, drilling-induced and core-handling fractures. Hydraulic fractures
were found to be most abundant in the section of the core that was closest to the stimulated wells.
Applications/Significance/Novelty: The benefits from an experiment of this type are: 1) Hydraulic fractures are
observed directly, allowing fracture density and spatial distribution to be quantified. 2) Twist hackles on hydraulic
fracture surfaces may provide information about the direction of fracture propagation. 3) The degree to which natural,
sealed fractures have been reactivated may be assessed. 4) Proppant distribution in the fractures can be examined.
5) Details of the observed hydraulic fracture network can be used to verify, or provide input data for, models of
hydraulic fracture growth and proppant distribution. 6) The findings can provide ground truth for indirect diagnostic
techniques such as microseismic monitoring. 7) Fractures in core and image log can be compared, calibrating other
horizontal image logs.
Assessment of In Situ Proppant Placement in SRV Using Through-Fracture Core Sampling at HFTS
1 1 1
D. Maity; ; J. Ciezobka; ; S. Eisenlord; ;
1. Gas Technology Institute, Des Plaines, IL, United States.
Objectives/Scope: The aim of this study was to develop robust methods for detecting and quantifying the subsurface
proppant distribution using the cored samples obtained as scrapings from the cut core tubes. The broad objectives
were firstly, to determine the spatial distribution of proppant in the created SRV along the cored interval; secondly, to
determine if the zones of interest were sufficiently propped; thirdly, whether the distribution can be tied with other
independent observations such as fractures; and finally, to see if cluster/ well spacing are optimal.
Methods/Procedures/Process: 213 core sludge samples scraped from the interior of core barrel sleeves and exterior
of the core itself from all available core sections were collected. These were washed, wet sieved and dried. After
sufficient agitation, fractions were weighed and the 20-120 fraction was used for analysis. Representative subsamples
were extracted from this fraction and spread over 8.5 × 11” area. Samples were scanned using a high resolution
transparency scanner and the images were processed using an image processing algorithm which uses a Circular
Hough Transform along with other attributes associated with proppant sand particles (blue hue, darkness, size,
roundness, translucence, etc. Multiple sub-samples were analyzed for each cored section to check for data
redundancy.
Results/Observations/Conclusions: Proppant transport in terms of vertical distribution seems to be defined by stress
contrast at bed boundaries such as carbonate layers and other zones with significant brittleness contrast. Propped
height has been found to be around ~50ft. at a lateral offset of ~ 100 ft. from the hydraulically fractured well. Proppant
presence seems to be governed by the nature of fracturing, complexity and whether there are enough arrest features
to allow for proppant accumulation during flow. As we move further away from the treatment well, smaller proppant
size fractions tend to dominate. Broadly, proppant density is found to be related to presence of hydraulic fractures.
Some indication of heel-side proppant distribution bias was also observed.
Applications/Significance/Novelty: To the best of our knowledge, this study is the first of its kind in its scope and
methodology. It will provide a useful template for future analysis involving direct sampling of stimulated rock where
proppant is present. The results are very significant in terms of our understanding of proppant distribution in the
subsurface as they get distributed during hydraulic fracturing. This study also highlights various factors that seem to
impact the proppant distribution and ties in independent observations from petrophysics as well as geomechanical
analysis of identified fractures to help us better understand the proppant distribution behavior as observed. Moreover,
there is scope for expanding this technique to analyze samples from drilling mud returns for improved completion
designs.
Analysis and Distribution of Proppant Recovered From Fracture Faces in the HFTS Slant Core Drilled Through a
Stimulated Reservoir
1 1
S. J. Elliott; ; J. F. Gale; ;
Objectives/Scope: We quantified the amount of proppant in open fractures within a slant core drilled through
hydraulically fractured Wolfcamp Shale. We address basic questions regarding proppant behavior in the subsurface in
order to improve understanding of hydraulic fracture growth and patterns. For example, where is the proppant found;
in hydraulic fractures, in reactivated natural fractures, or is it transported into drilling-induced fractures during
recovery? Is there variation along the length of the core examined? Do the majority of proppant grains remain whole
or tend to get crushed in the subsurface?
Methods/Procedures/Process: We collected all material from every open fracture within the core (~450 samples),
which consisted of drilling mud, aluminum shavings from the core barrel, shale, lithics, calcite, and potentially proppant
pumped during the hydraulic fracturing of adjacent wells. Samples were washed with detergent to remove drilling mud,
and were then sieved into 4 size fractions: >500 µm, 250-500 µm, 120-250 µm, and 63-120 µm, with all material <63
µm removed. When dry, samples were weighed and imaged using a DSLR macrolens. Image analysis was performed
to determine proppant proportions, as well as area, perimeter, and major/minor axes of each particle. SEM imaging
was conducted on a representative suite of samples to resolve proppant surface composition and overall percentages
of crushed grains.
Results/Observations/Conclusions: We saw no evidence of embedment on fracture faces, however, proppant packs
0.5-1 cm thick were found in two locations with complex fracture geometry. A few other fractures had sand patches,
one of which had abundant sand patches on the fracture face. Pumped proppant was 40/70 and 100 mesh (149-420
µm), yet the majority of recovered proppant is in the two smallest size fractions: 120-250 µm and 63-120 µm,
suggesting some grain crushing. Furthermore, some of the pumped 40/70 mesh included grains coated in colored
resins, with different colors pumped in different stages in the adjacent wells. We’ve noted all recovered colored grains
(red, yellow, and blue) in an attempt to diagnose the pumping stage linked to specific fractures.
Applications/Significance/Novelty: Retrieved samples were labeled with respect to the master fracture database where
fractures were identified in the core and on CT scans of each 3-ft interval, allowing us to code them as either hydraulic
fracture, natural reactivated fracture, or drilling/recovery induced fracture. From this we were able to not only quantify
the amount of proppant within each fracture, but also analyze the distribution of grains among different fracture types.
Ultimately, the majority of fractures contain at least some proppant, however the extent to which the proppant was
introduced during the core drilling process is unknown.
Natural and Hydraulic Fracture Density Prediction and Identification of Controllers
1 1 1
W. Campbell; ; J. Wicker; ; J. Courtier; ;
Objectives/Scope: The Hydraulic Fracture Test Site (HFTS) provides a rare insight into the lateral distribution of a
subsurface natural fracture network and the associated controlling factors of natural fracture distribution relating to
changes in both lithology and mechanical rock properties. Analytical modeling demonstrates the occurrence of
hydraulic fractures, their interaction and relationship to the natural fracture system appears directly related to changes
in lithology, fluid saturations, and mechanical rock properties at the distance the core intersected the hydraulic volume.
Non-linear multivariate models accurately honor distribution of hydraulically stimulated reservoir volume and can give
higher control in the distribution of the natural fracture volume. Model results improve confidence in developing a
highly accurate discrete fracture network for hydraulic fracture modeling and reservoir simulation. Moreover, analytical
modeling in this location provides fundamental insights into the long-standing debate on how hydraulic fractures
interact with natural fracture systems.
Inter-well Communication Study of UWC and MWC Wells in the HFTS
1 1 1 1
T. Wood; ; R. Leonard; ; C. Senters; ; C. Squires; ;
1. ProTechnics a Division of Core laboratories, Fort Worth, TX, United States.
Objectives/Scope: The focus of this paper is to examine the water, oil and proppant communication between wells at
the Hydraulic Fracturing Test Site (HFTS) in the Permian Basin. Unique tracers were employed during the stimulation
of the HFTS pad in conjunction with many additional diagnostic technologies to more effectively evaluate well-to-well
communication, fracture behavior, proppant transport and reservoir drainage.
Methods/Procedures/Process: Water and oil tracers were utilized on four UWC and four MWC horizontal wells on the
HFTS pad. Each traced well was divided into three or four uniquely traced segments. The segments consisted of
consecutive stage groupings that were stimulated with the same water and oil tracers. During the production phase of
the project, oil and water samples from a combined total of 29 horizontal and vertical wells were analyzed for the
presence of the tracers from the eight wells. Additionally, proppant tracers were employed in the stimulation of two
UWC and one MWC horizontal wells. After all the wells were completed, a spectral gamma ray log was run on each
well to identify the proppant coverage, cluster efficiency, NWB fracture behavior and/or any proppant communication
that occurred.
Results/Observations/Conclusions: The results of this project provide a better understanding of the fluid and
hydrocarbon communication that is taking place between and within the UWC and MWC wells over an initial short-
term period as well as the long term. Additionally, the target zone, stimulation design, production methodology and
stimulation sequencing is examined and presented for comparison with other diagnostic data from this project.
Applications/Significance/Novelty: The insights from this two year study highlights the continued need to optimize
completion designs, even while targeting different parts of the same formation. Additionally, future fracture modeling
and design criteria can be enhanced with the hydraulic fracture and propped fracture heights and half-lengths
determined from fluid and proppant communication to the offset wells.
Well Interference Diagnosis Through Integrated Analysis of Chemical Tracer and Pressure Interference Tests
1 1 1 1 1
A. Kumar; ; P. Seth; ; K. Shrivastava; ; R. Manchanda; ; M. Sharma; ;
1. Petroleum Engineering, University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: Communication between wells through connected fractures results in tracer and pressure
interference. Understanding the connectivity between fractured horizontal wells in a multi-well pad is important for infill
drilling. Inter-well tracer and pressure interference tests involve two or more fractured horizontal wells and provide
information about hydraulic fracture connectivity between the wells. In this work, we present a workflow based on
combined analysis of tracer and pressure interference data to analyze the degree of interference between fractured
horizontal wells in a multi-well pad.
Methods/Procedures/Process: We analyze well interference using tracer and pressure data in an 11-well pad in the
Permian Basin. For this interference study, a fully coupled 3D geomechanical reservoir model was developed to
simulate production and tracer transport in complex fracture networks. A displacement discontinuity method is used to
generate the complex fracture network representing the interaction between hydraulic and natural fractures. This
model captures the effect of fracture closure due to changes in in-situ effective stress during production. Fracture
closure directly affects the degree of interference between wells. The impact of natural fracture density, well spacing
and induced un-propped (IU) fracture closure on well interference was analyzed through simulation results and field
data.
Results/Observations/Conclusions: Changes in pressure and tracer concentration in the monitor wells were used to
identify and evaluate interference between the source and monitor wells. Extremely low tracer recovery and weak
pressure response signify the absence of connected fractures and suggest that interference through matrix alone is
insignificant. Combined tracer and pressure interference data suggest connected fracture pathways between the
communicating wells. The degree of interference is quantified in terms of 1) tracer breakthrough and pressure
response times, 2) tracer recovery and magnitude of pressure response, and 3) number of tracer peaks and pressure
derivatives. An increase in well spacing, IU fracture closure decreases interference, whereas, an increase in natural
fracture density increases interference.
Applications/Significance/Novelty: We present a new approach to analyze well interference which includes the effect
of geomechanics on flow through complex fracture networks. Tracer and pressure interference analysis provides a
unique tool for understanding the degree of connectivity between communicating wells, which can be useful for
optimizing infill drilling, well spacing and fracture sizing in future treatment designs.
Case Histories of Integrating DAS Fiber-Based Microseismic and Strain Data for Monitoring Horizontal Hydraulic
Stimulations
1 1 1
R. Hull; ; R. Meek; ; H. Bello; ;
Objectives/Scope: Invited paper - no abstract.

Methods/Procedures/Process: Invited paper - no abstract.
Results/Observations/Conclusions: Invited paper - no abstract.
Applications/Significance/Novelty: Invited paper - no abstract.
Integrated Statistical Workflow for Optimum Well Spacing With Data Analytics, Pilots, Geomechanical-Reservoir
Modeling, and Economic Analysis
1 1 1 2 2 1 1 1
R. Cao; ; C. Chen; ; R. Li; ; T. Firincioglu; ; C. Ozgen; ; T. Croft; ; A. Girardi; ; N. Chowdhury; ;
2. NITEC LLC, Denver, CO, United States.
Objectives/Scope: The statistical nature of subsurface parameters results in statistical outcome in well spacing pilots,
which makes it very difficult to draw reliable conclusions with limited number of trials. To handle the challenges and
significant uncertainties in optimum well spacing for development of unconventional plays, a novel integrated
statistical workflow with data analysis, pilots, detailed geomechanical and statistical reservoir modeling are proposed
and being applied to the unconventional Liquid Rich Shale development at Permian basin.
Methods/Procedures/Process: First, the data analytics is performed are multiple pilot results with detailed information
(production, Microseismic, BHP, fluid sampling, etc.) and public data. A well spacing trial with good quality of data is
selected as the modeling target. A full-scale 3D multi-well reservoir model with geomechanical effects was built to
history match the oil, gas, water productions, and pressure. As it is still too computational expensive to forecast with
multiple realizations of full-scale models, statistical studies with hundreds of realizations are performed with a
simplified reservoir model to investigate the production interference for different well spacing and subsurface
parameters. The modeling results are carefully calibrated with the field data at different time and length scales.
Results/Observations/Conclusions: The result shows that the single well spacing pilot has significant uncertainties.
The probability distribution of different well spacing, P10/P50/P90 interference factors are used to capture the
uncertainty. EUR reduction factors obtained from this workflow are integrated with economic criteria to determine
optimum well spacing. The field data provide a basis for the modeling process. The full-pad integrated Geomechanics
and reservoir model is very helpful to understand the hydraulic fracturing and production process. The multi-realization
forecast method is very useful to quantify the statistical distribution of production interference among wells. The results
also show that different Decline Curve Analysis (DCA) parameters should be used when evaluate EUR with different
well spacing.
Applications/Significance/Novelty: Appropriate well spacing decision is critical for unconventional asset development.
Operators generally use three methods to identify optimum well spacing: trial and error, pilots, and modeling. The field
trials and pilots often involve significant capital investment and require significant time before conclusive results can
be observed. On the other hand, modeling is cost effective and time efficient but has large uncertainties in the results,
which must be carefully calibrated with the field measurements to narrow down the uncertainty range. In this paper,
these three methods are systematically integrated to study well interference and identify optimum well spacing for
Wolfcamp development at Permian basin. This is the first integrated statistical workflow proposed in the industry.
Myths and Facts of Forecasting Horizontal Well Production in Unconventional Reservoirs – Are We Complicating a
Simple Analysis?
1, 2
V. Muralidharan; ; K. Joshi; 1, 2;
1. Occidental Petroleum Corporation, Sugar Land, TX, United States.
2. SPE, Houston, TX, United States.
Objectives/Scope: Accurate prediction of long-term performance of multi-frac horizontal wells (MFHW) targeting
unconventional reservoirs is still a challenge due to variability in reservoir properties, well spacing, stimulation design,
and production optimization strategies. The use of type curves that normalizes the aforementioned variables has been
adopted by operating companies mainly for reserves booking and as a benchmark for well performance.This paper
presents a simple yet powerful methodology that accurately creates a generic type curve that can be easily scaled
based on one-year cumulative production.
Methods/Procedures/Process: The proposed method employs an interrelation between one-year cumulative
production and contacted oil-in place (OIP) from flowing material balance to accurately explain the dynamic behavior
of MFHW. The case studies include synthetic simulated cases and actual data from horizontal wells in multiple basins
with at least two years of production history and varying stimulated rock volumes (SRV). The population of wells
include nano- to micro-darcy rocks in Wolfcamp, Bone Springs, Brushy Canyon, Spraberry, Wichita Albany and San
Andres. These formations have wide ranges in initial reservoir pressure, rock and fluid properties, well spacing, SRV
size, flowback and artificial lift strategies. EURs per well were generated using production history match and decline
curve analysis.
Results/Observations/Conclusions: The larger the OIP within the SRV, the higher the cumulative production.
Therefore, well performance is directly proportional to SRV OIP. Well spacing affects the SRV which is translated to
the first year cumulative production. The ratio between the dynamic production from early- to late-time is the same
independent of rock and fluid properties, stimulation design, well spacing, and production optimization strategies. The
results point out that one year cumulative volume is a good proxy for well performance driven by subsurface dynamic
and static properties. The correlations developed should be a valuable supplementary tool when analyzing well
performance by arriving at a reasonable range of EUR for a specific one year cumulative volume.
Applications/Significance/Novelty: The main application of this paper is to simplify the type curve process by creating a
universal type curve that removes any bias when conducting decline curve analysis and narrows uncertainty in the
EUR. While this paper does not eliminate the need to do more thorough analysis in unconventional wells, the results
show that a simplistic parameter like one year cumulative volume can be extremely useful in predicting long term well
performance with an acceptable level of certainty. This paper has numerous applications in evaluating best practices
related to SRV drainage height, well spacing, frac hits and interference, clean-out candidates, landing point
sensitivities, field development plan to name a few. The result of this paper might remove bias in MFHW production
forecasting.
An Integrated Geomechanics-Reservoir Simulation Workflow for Completion Design Optimization
1 1 1 1 1 1 1
T. Yeh; ; J. Jennings; ; D. Cakici; ; J. Chavarria Guerra; ; M. Durand; ; B. L. Williams; ; T. Chen; ; R.
1 1 1 1
Casillas; ; V. Jain; ; R. Li; ; T. Bai; ;
Objectives/Scope: Achieving optimal completions designs has been one of the most challenging topics in the recent
development of unconventional plays. This is largely due to the complexity of the completion design parameters as
well as the uncertainties in the subsurface. To devise a robust field development strategy a better understanding of
the hydraulic fracturing process, geomechanical and fluid flow behavior is crucial. In this study we present a novel
workflow that integrates geomechanics and reservoir simulation to expedite our understanding of production
performance under subsurface uncertainties.
Methods/Procedures/Process: In this workflow, an advanced geomechanics simulator is used to simulate the
hydraulic fracturing process and calibrate the models to the injection data. A collection of models is generated based
on the identified subsurface uncertainties. These fracture geometries scenarios along with the modeled properties
such as fracture permeabilities and proppant mass distributions are incorporated into a reservoir simulator for flow
simulation. A multiple-realization history matching is performed to further calibrate the dynamic parameters to the
production data. Finally, a posterior distribution for the dynamic parameters for each scenario and a set history-
matched models are obtained. These models are then used for production forecast as well as development planning
analysis.
Results/Observations/Conclusions: The workflow was demonstrated using a pilot well in the Lower Vaca Muerta
(LVM) formation. Thirty-three distinct fracture geometry scenarios were generated. The fracture distributions shows
high uncertainty in the fracture height, half length, and possible contributions from Upper Vaca Muerta (UVM). Such
uncertainty was reduced in the history matching exercise where the production data was used to calibrate effectively-
propped fractures along with other dynamic parameters. Analysis shows that despite the differences in the fracture
geometries, the ranges of EUR only varies moderately. However, comparison between the most optimistic and
pessimistic fracture geometry realizations shows that the pressure depletion distributions can vary significantly and
hence influence development decisions.
Applications/Significance/Novelty: The proposed workflow has a wide range of applications. In particular, it has strong
advantages over tradition modeling approaches (e.g. bi-wing models) since 1) it considers multiple fracture geometry
scenarios to reflect the impact of subsurface uncertainties and, 2) it incorporates realistic hydraulic fracture geometries
directly generated from a geomechanics simulator in flow simulations. Thus, the workflow is best suited for estimating
production performance under large subsurface uncertainties, optimizing completion design and studying complex
fluid flow behavior such as well interference.
Nanoparticle-Encapsulated Acids for Stimulation of Calcite-Rich Shales
1 1 1 1
S. Tong; ; R. Singh; ; K. Panthi; ; K. Mohanty; ;
1. Petroleum Engineering, The University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: The goal of this work was to develop novel encapsulated acids (EA) using surface-modified silica
nanoparticles for acid stimulation of natural micro-fractures in shales. Direct acid injection leads to fast surface
dissolution of calcite and unwanted reaction in well bore / hydraulic fractures. The performance of these EA was
compared with conventional acid-in-oil emulsions in terms of thermal stability and interaction with shales. The effect of
the size of nanoparticles, acid concentration, and shale calcite content was also investigated.
Methods/Procedures/Process: First, the surface-modification of bare silica nanoparticles were performed to achieve
the desired surface wettability. The synthesized NP was used to encapsulate acids of varying concentration to obtain
EA. Second, the thermal stability of these EA was studied at 125 C. Third, the reaction of shale chips with varying
calcite-content with EA was investigated. Fourth, the EA was injected in a fractured shale cores, and fracture closure
was simulated by increasing the overburden pressure. This resulted in the mechanical crushing of the EA which
released the acid. The permeability of the cores before and after the EA treatment was measured. Finally, the surface
etching on the fracture face due to acid was quantified by measuring the surface roughness using an optical
profilometer.
Results/Observations/Conclusions: The mixing of surface modified nanoparticles with acids under high shear rates
resulted in the formation of encapsulated acid (EA) by self-assembly of the nanoparticles. These EA particles were
found to be a robust stimuli-responsive acid carrier system where the release of acids could be triggered using
mechanical crushing or addition of surfactant. The thermal stability tests showed the robust stability of EA even at 125
C as compared to acid-in-oil emulsions. The bulk shale reaction tests showed minimal interaction with acids until the
encapsulation was broken via mechanical crushing. The fracture core experiments showed that these EA could
improve the permeability of the cores depending on the calcite content of the shales.
Applications/Significance/Novelty: Acid treatment is a common stimulation technique for calcite-rich shale reservoirs.
Retarded acid systems such as acid-in-oil emulsions are typically used to minimize wellbore damage and long-
distance propagation of acids from the wellbore. However, the stability of these emulsions under harsh reservoir
conditions is still a challenge. This study proposes a novel and robust acid carrier system using surface-modified
nanoparticles for acid treatment of shales and systematically compares its performance with conventional systems.
Post-Frac-Hit Mitigation and Well Remediation of Woodford Horizontal Wells With Solvent/Surfactant Chemistry Blend
1 2 2 2 2 2 2 2
C. Swanson; ; W. A. Hill; ; G. Nilson; ; C. Griman; ; R. Hill; ; P. Sullivan; ; C. Aften; ; J. C. Jimenez; ;
2 2 2
G. Pietrangeli; ; D. C. Shedd; ; J. Pursley; ;
2. Flotek Industries, Houston, TX, United States.
Objectives/Scope: A proposed chemistry treatment based on a patented solvent/surfactant blend is presented that
addresses problems in two sets of wells in the Anadarko Basin’s Woodford productive trend: (1) wells experiencing
frac hits from nearby offset wells and (2) wells with low initial production rates related to sub-optimal stimulations. The
treatment is designed to reverse the negative effects of frac hits and boost production levels in poorly stimulated wells
to closer to predrill estimates.
Methods/Procedures/Process: In Grady County, Oklahoma, four hydraulically stimulated wells were selected for
testing. Two of the wells were impacted by frac hits while the remaining two did not meet pre-drill production forecasts
due to inadequate stimulation jobs. An emerging chemistry formula, a combination of the patented solvent/surfactant
blend, hydrofluoric acid and hydrochloric acid was developed to treat both sets of wells. The chemical blend was
designed to alleviate breakdown pressures, increase relative permeability, reduce scale and sludge formation, and
control fines migration and clay swelling. The blend also included carefully selected solvents to facilitate the removal
of high-viscosity emulsions, which would subsequently help increase flow capacity within the reservoir.
Results/Observations/Conclusions: An evaluation of the four wells uncovered important insights related to their
treatment. The formation rock’s response to the treatment may depend on how long each well has been in production,
pressure changes in the reservoir and each well’s stimulation design. The well’s position in the reservoir relative to
natural fractures, reservoir mineralogy and clay content also likely influence production gains. Careful consideration of
these reservoir attributes, combined with the composition of the remediation chemistry, may yield further production
improvements in future wells.
Based on the increased percent uplift in production observed for the four treated wells, the remediation treatment’s
total project economics were positive.
Applications/Significance/Novelty: Chemical remediation treatments like this have broad application—both along the
Woodford trend and in other unconventional shale plays with diminished production resulting from frac hits or poor
stimulations. Varied and customized chemistry formulations will likely be required in wells with different geological
compositions, such as in stimulated wells with high clay content. Detailed analysis will also be needed as the
mineralogy changes across a given portion of the reservoir’s producing trend. Such analysis may include life of
production, reservoir pressure depletion, structural characterization and the stimulation plan of infill wells in the field.
Ion Tuned Water Can Greatly Enhance Alteration of Carbonate Surface to Water-wet
1 1
H. Ding; ; S. Rahman; ;
1. School of Petroleum Engineering, University of New South Wales, Sydney, NSW, Australia.
Objectives/Scope: In this study, we aim to investigate the change of wettability of carbonate rocks to a favorable state
by ion tuned water. The wettability change of carbonate surface is mainly attributed to surface charge modification.
2+ 2+ 2-
Results of this study showed that this aspect is correlated to potential determining ions (PDI, Ca , Mg and SO
4
2+ 2-
), amongst Mg and/or SO are most effective to alter the oil-wet carbonate rock into water-wet which in turn
4
enhances the oil recovery. Therefore, the optimized composition of ion tuned water has the potential to greatly
enhance water-wetness of carbonate.
Methods/Procedures/Process: Atomic force microscopy with “soft tip” (an oil drop is attached to a tipless cantilever) is
used to measure the forces of oil tip and cleaved calcite in salt solutions (NaCl, MgCl , CaCl , and Na SO at same
2 2 2 4
ionic strength 0.34M). Surface charge change of carbonate rocks in the presence of PDI is estimated by zeta potential
measurements in 8 ion tuned waters. Contact angles on carbonate surfaces with different roughness are measured
with sessile drop method in same series of ion tuned waters. Contact angles are also extracted from micro-CT
imaging and compared with macroscopic contact angles. At last, series core flooding experiments combined with
micro-CT imaging have been carried out both with 6mm core and 1 inch core with two optimized ion tuned waters.
2- 2+
Results/Observations/Conclusions: From AFM force measurements, we determined that SO and Mg generate
4
less adhesion and stronger repulsion. In contrast, huge adhesion and ‘jump in’ are observed in CaCl , indicating oil is
2
2- 2+
more likely to adhere onto calcite surface. We manipulated seawater containing more SO , more Mg , or less Ca
4
2+
(overall 8 kinds of ion tuned waters). Zeta potential results show that four waters have comparable efficiency in
modifying the calcite as well oil into more negative. Further contact angle data show that only SW4SO and SW1/4Ca
can modify the water-wet to a largest magnitude. Core flooding and micro-CT imaging show that SW1/4Ca has the
best performance in enhancing oil recovery than SW4SO.
Applications/Significance/Novelty: In this study, multiple methods have been adopted to study the fundamental of
wettability alteration on carbonate surfaces. The novelties are: (1) novel technique, atomic force microscopy with "soft
tip" technique, (2) newly topic, we focus on carbonate rocks which limited work has been found, especially from
microscopic view (AFM, zeta potential and micro-CT), (3) new finding, we suggest an optimum ion tuned water for
enhancing oil recovery which is comprehensively determined from the aforementioned studies. It is more reliable than
those simply obtained from core flooding and/or spontaneous imbibition and/or macroscopic contact angle
measurements. The results therefore have more significance to guide field application, both from theoretical and
economical aspects.
Synthetic Well Log Generation Using Machine Learning Techniques
1 2 2
O. Akinnikawe; ; S. Lyne; ; J. Roberts; ;
1. E&P Advanced Analytics and Data Science, Devon Energy Corp E&P, Oklahoma City, OK, United States.
2. Reservoir Technology and Optimization , Devon Energy Corp E&P, Oklahoma City, OK, United States.
Objectives/Scope: When using wireline log to characterize formation properties for an area we often run into
incomplete datasets. One way to address this lack of data is to create synthetic curves to use in the analysis. The
objectives of this paper is to show workflows that can be used to generate synthetic photoelectric (PE) and unconfined
compressive strength (UCS) logs using machine learning techniques.
Methods/Procedures/Process: Synthetic PE logs were generated using wireline logs from over a hundred wells that
included gamma ray, density, neutron, resistivity logs and volume of clay. Similarly, UCS data were generated using
wireline logs and core rebound hammer data from fourteen wells including gamma ray, density, porosity, neutron,
shale volume, kerogen volume, compressional slowness, shear slowness, Young’s modulus and Poison’s ratio. The
data was randomly partitioned into a 70:30 split for training and validation data set respectively. Model competition
among a suite of machine learning algorithms such as Regression, Artificial Neural Networks, Decision Trees,
Gradient Boosting and Random Forest was used to select the best algorithm based on the least average squared
error of the validation dataset.
Results/Observations/Conclusions: The use of Variable Clustering to remove collinearity, decrease variable
redundancy, and choose the best variables for analysis and Cluster analysis to identify factors that differentiate data
segments from the population improved prediction of target variables. Results show that neural networks and random
forests generated the best prediction of UCS and synthetic PE logs compared to other machine learning algorithms
used.
Applications/Significance/Novelty: Modern well log data sets usually include PE logs which provides important
information about the lithology of the formations the wellbore intersects. However, many legacy wells do not have PE
logs that we need in order to understand the lithology of these formations. Similarly, it is important to obtain UCS data
from mechanical failure tests done to core samples to determine the strength of the rock. Obtaining and testing core
samples for the entire zones of interest is both expensive and time consuming. Synthetic well logs can be a reliable
and cheaper alternative to predicting PE and UCS values rather than running a new set of logs or coring and testing
the whole zone of interest.
Novel Monitoring Technology Helps to Make Informed Decisions and Maximize the Efficiency of Completion Strategy
1 1 1 1 2
A. Fedorov; ; S. Parkhonyuk; ; R. Korkin; ; A. Kabannik; ; M. Nikolaev; ;
1. Engineering and Manufacturing, Schlumberger, Novosibirsk, Russian Federation.
2. Schlumberger, Houston, TX, United States.
Objectives/Scope: In this paper, we will describe the novel cost-effective, small-footprint method for hydraulic
fracturing monitoring. The focus will be made on a practical case studies, where the method was used to address
most common issues of the well completion process.
Methods/Procedures/Process: There are events that can diminish the effectiveness of multistage fracturing and can
include such as ball landing failure frac hits and failure of bridge plug resulting in the ineffective execution of
completion strategy for operators and non-optimum reserves recovery.Casing leaks may prevent proper well
perforation. Methods such as microseismic monitoring, radioactive tracers did not cover all the described issues.
Sometimes they are not cost effective and can provide results only after the job has been performed.
This means that there is a need for the monitoring method that allows proactive changes on the well site to address
completion issues.
High-frequency pressure monitoring (HFPM) method uses proprietary algorithms to determine the location of the
downhole completion features.
Results/Observations/Conclusions: Effectiveness and applicability of the method were tested on the number of wells
across major US shale plays.
Case studies will cover the topic of:
ball failure during fracturing treatment,
successful isolation of the previous interval while no surface pressure signature was observed
detection and mitigation of casing leak
diverter efficiency confirmation on troubled completion.
Applications/Significance/Novelty: Proposed method demonstrates excellent performance in the above-mentioned
case studies. It can be easily deployed. It does not require any interruptions in the well stimulation process. At the
same time, it makes possible informed decision making on the well site.
High-Quality 3-D MicroCT Imaging of Source Rocks – Novel Methodology to Measure and Correct for X-Ray Scatter
1, 2
A. Katsevich; ; M. Frenkel; 1; Q. Sun; 3; S. Eichmann; 3; V. Prieto; 1;
1. iTomography Corporation, Houston, TX, United States.
2. University of Central Florida, Orlando, FL, United States.
3. Aramco Services Company: Aramco Research Center – Houston, Houston, TX, United States.
Objectives/Scope: Micro Computed Tomography (microCT) of cores yields valuable information about rock and fluid
properties at pore-scale for conventional rock and at rock heterogeneity scale for unconventionals. High level of
uncorrected X-ray scatter in CT data leads to strong image artifacts and erroneous Hounsfield Unit (HU) values
making reconstructed images unsuitable for accurate digital rock (DR) characterization (e.g., segmentation, material
decomposition). MicroCT scanners typically do not include scatter correction. We propose a methodology to correct
for scatter and ultimately improve image quality.
Methods/Procedures/Process: Widely used approaches for scatter estimation, based on Monte-Carlo simulation and
simplified analytical models, are time-consuming and may lose accuracy when imaging complex unconventional shale
cores. We propose a more practical approach to perform scatter correction from direct scatter measurements. We use
a Beam Stop Array method where an X-ray transparent support holds highly-absorbing lead beads between the
source and core during a scan. Data at the projection center of beads gives an estimate of scatter that is then applied
across the detector. Subtracting scatter from projection data yields corrected data used for 3D image reconstruction.
To test the methodology, we developed and implemented a fast scatter compensation algorithm into our 3D microCT
image reconstruction workflow.
Results/Observations/Conclusions: The proposed method was tested using both physical phantoms and real shale
core samples from the US and Middle East. Our experiments confirmed that the amount of X-ray scatter increases
significantly with increased size of scanned cores. For high-precision reconstructions, we need to conduct two core
scans: first – to estimate scatter, and second – to obtain a complete data set. For less-demanding applications, the
microCT data behind the beads is recoverable by interpolation. We demonstrated that the developed methodology
allows for accurate estimation and removal of scatter from the projection CT data, enabling reconstruction of high-
quality images required for performing DR analysis. Ultimately, these results expand operational envelop and value of
microCT imaging in the E&P workflow.
Applications/Significance/Novelty: Accurate correction for X-ray scatter is one of the most challenging problems in
microCT imaging. Amount of scatter depends on scan geometry, X-ray source energy, core size, detector type, and
other factors. MicroCT scanners use flat-panel detectors, which do not allow collimation (unlike curved detectors used
in medical CT) and are not typically equipped with scatter correction, leading to significant amount of scatter in the
acquired raw CT data. The presented scatter correction methodology is general and can be utilized with any microCT
scanner employed by the petroleum industry to improve image quality and derive accurate HU values. This is of
significant importance for quantitative characterization of highly-heterogeneous rock with fine structural changes as is
the case for shale.
Quantitative Analysis and Feature Detection for Scanning Electron Microscopy Images Using Machine Learning and
Image Processing
1 1 2
X. Tian; ; H. Daigle; ; F. Xu; ;
2. RIPED, CNPC, Beijing, China.
Objectives/Scope: Microfractures are important mechanical discontinuities in shales and are important for fluid flow
during production. Understanding their properties is crucial for accurate shale production prediction and implementing
effective stimulation strategies. Scanning electron microscope (SEM) images are useful for characterizing shale
microstructure, but manual image analysis is often challenging and time consuming. We present an alternative
method for quickly characterizing microfractures and obtaining pore structure information from SEM images using
machine learning algorithms.
First, we use an image clustering approach to get a statistical observation of shale micro-structure from SEM images
of Eagle Ford shale and a siliceous oil-bearing shale. Next, we slice around 100 overview SEM images into around
1500 smaller sub-images for better image processing. The sub-images are separated into a training set (70% of sub-
images) and a test set (30% of sub-images) for the machine learning models.
We compare the performance of two classic machine learning algorithms which are most widely used for image
classification. They are support vector machine classifier (SVM) and convolutional neural networks (CNN). We apply a
histogram of oriented gradient (HOG) computation to SEM sub-images to extract better input features for SVM
modeling.
First, we obtain areal pore size distributions and fracture size distributions using image clustering for SEM images.
The average areal porosity is around 2.3% for the overview images. 251 relatively large fractures are detected in
these images. Most of the large fractures develop along organic matter clusters.
The machine learning models can automatically detect whether fractures exist in an SEM image. The result from
machine learning models are compared with manual analysis results. The CNN has the highest accuracy, a
combination of SVM and HOG also produces good prediction accuracy, but the combination of SVM and PCA doesn’t
work well for this problem.
Applications/Significance/Novelty: Machine learning is a promising method for detecting features in SEM images.
Once the convolutional neural network algorithm is properly trained with manually labeled samples, it can be used to
automatically analyze other SEM images. This approach can be used for quick quantitative and qualitative analysis for
vast amounts of images, obtaining areal pore size distribution information, and detecting features including
mineralogy, fracture, organic matter, etc.. With this method, SEM images having objective features in them can be
quickly selected from all the images and used for more detailed analysis.
A Methodology to Characterize Well Performance in the Delaware Basin: A Wolfcamp Case Study
1
K. Mohan; ;
1. Delaware Basin, ConocoPhillips, Houston, TX, United States.
Objectives/Scope: This paper highlights a detailed and systematic approach to characterizing well performance of
hydraulically fractured Wolfcamp horizontal wells in the Delaware Basin.
Methods/Procedures/Process: The workflow outlined in this paper integrates fracture treatment analysis, Rate
Transient Analysis (RTA), Numerical simulation coupled with Assisted History Matching to characterize and
understand well performance. The Wolfcamp formations produce significant amounts of water and flowing bottom-hole
pressures fall rapidly below the reservoir dew point making the application of RTA challenging. One way to deal with
the multiphase flow effects is to convert the oil and water streams into an equivalent gas stream. By doing this, we can
use RTA to get preliminary estimates of the created fracture surface area. These fracture properties are then used to
constrain the inputs of a numerical model that better deals with the above described issues.
Results/Observations/Conclusions: In this paper, we present two field cases where we have successfully applied this
workflow to characterize well performance. The method demonstrates that analytical techniques should be used
before undertaking detailed numerical simulation, in order to constrain the ranges of fracture dimensions and
permeabilities that explain observed performance.
Applications/Significance/Novelty: The applications of this approach are pretty significant. This allows engineers to not
only characterize and quantify well performance but also aid other engineering disciplines like completions on the
optimal cluster spacing between fractures.
Pressure and Rate Transient Behavior of a Horizontal Well Intercepting Multiple Hydraulic Fractures Within a Fractal
Reservoir
1 1 2
A. R. Valdes-Perez; ; T. Blasingame; ; L. Larsen; ;
2. Universitetet i Stavanger, Stavanger, Norway.
Objectives/Scope: The effective reservoir "system" created by a “multi-fractured horizontal well” (or MFHW) is a
challenging system to model for flow diagnosis given three factors: (1) the heterogeneous nature and low/ultralow
permeability nature of the reservoir, (2) the geometry of the well and the fractures, and (3) the properties of the
petroleum fluids. In this paper, we will address the first two of these points in by providing semi-analytical reservoir
models and diagnostic interpretation relations to evaluate the effectiveness of MFHW scenarios.
Methods/Procedures/Process: Our approach is to extend the models for horizontal wells intercepting a single finite
conductivity hydraulic fracture to consider the fractal
reservoir scenario. The well performance of this "fractal" solution is then compared to the so-called "anomalous
diffusion" case, as we believe that the fractal reservoir model is more likely in theory and in practice than the
"anomalous diffusion" case.
Results/Observations/Conclusions: In this paper, we show that the pressure transient response of a “multi-fractured
horizontal well” (or MFHW) within a fractal reservoir can exhibit a maximum of four (4) distinct periods of flow: (1)
fracture (dominated) flow, (2) early fracture-reservoir interaction, (3) late fracture-reservoir interaction, and (4)
reservoir-dominated flow.
We provide various diagnostic plots and illustrative interpretations of field data to demonstrate the potential utility of
models for multi-fractured horizontal well in a fractal reservoir scenario.
Applications/Significance/Novelty: The following contributions are derived from this work:
* We have developed a solution for the pressure and rate transient performance behaviors of a horizontal well
intercepting multiple hydraulic fractures within a fractal reservoir.
* The fracture-reservoir interaction period can be used to evaluate the effectiveness of multi-stage hydraulic fracture
treatments in a horizontal well.
Improved Rate-Transient Analysis Coupled With Pore Mechanics and Multi-Scale Fluid Dynamics in Heterogeneous
Fractured Liquid-Rich Shale
1 1
B. Yuan; ; C. R. Clarkson; ;
1. Geoscience, The University of Calgary, Calgary, AB, Canada.
Objectives/Scope: The industry is increasingly reliant on rate-transient analysis (RTA) to extract valuable information
about the reservoir and fractures. However, currently-available RTA models can lead to incorrect analysis of liquid-rich
shale (LRS) because they largely ignore the complex pore-fracture mechanics and fluid dynamics in multi-scale pores
which are difficult to constrain. The aim of this work is to develop simple-yet-rigorous RTA methods that integrate
macroscopic production with microscopic petrophysics & micro-seismic data
Methods/Procedures/Process: The following mechanisms are first included to improve RTA in LRS: 1) oil transport in
hydrophobic organic pores with near-wall retardation, and in hydrophilic inorganic pores with slip flow; 2) both pore
shrinkage and bond breakage using permeability function of stress-dependent pore sizes and coordination numbers;
3) a fractal random-fracture-network algorithm to interprete micro-seismic data. A novel unified apparent permeability
and total compressibility model are developed for the generalized diffusivity equation, where pseudo-pressure and
pseudo-time are defined to achieve analytic solutions. Dynamic-Drainage-Area (DDA) model is corrected using
concept of maximum pressure response. A correction factor is also applied to improve the analysis of specialty plots
for fracture diagonise.
Results/Observations/Conclusions: Through interpreting published experimental data, in inorganic pores, water film
and slip flow are found to enhance oil transport; for organic pores, oil adsorption cause retardation of oil transport.
Contrary to the negligible pore-connection loss in sandstones, the coordination number in shale can be reduced by
50%, with 98% permeability reduction. The novel apparent permeability and total compressibility model herein are
functions of pore size, in-situ stress, and reservoir heterogeneity. The complexity and connectivity of fracture networks
are quantified using L-system fractal theory. Novel DDA model is verified with a synthetic fine-grid simulation. Field
examples are examined to improve linear-flow analysis over previous RTA models.
Applications/Significance/Novelty: The major contribution of this work lies on the first attempt to integrate macroscopic
production with microscopic petrophysics & micro-seismic data, as an approach to extend lab-scale findings to field-
scale applications. Moreover, the resulting novel models, with consideration of complex pore-fracture mechanics and
multi-scale fluids dynamics, are of interest to those petroleum scientists & engineers eager to achieve more realistic
characterization of their wells and reservoir analysis.
Transient Linear Flow Analysis of Multi-Fractured Horizontal Wells Considering Three-Phase Flow and Pressure-
Dependent Rock Properties
1, 2
H. Behmanesh; ; H. Hamdi; 2; S. A. Ghaderi; 2; C. R. Clarkson; 2;
1. Anderson Thompson Reservoir Strategies, Calgary, AB, Canada.
2. University of Calgary, Calgary, AB, Canada.
Objectives/Scope: Rate-transient analysis (RTA) methods have historically been used to characterize multi-fractured
horizontal wells (MFHWs) during transient linear flow in order to estimate the total effective fracture area. A primary
complication in linear flow analysis is the incorporation of multiphase flow, as well as pressure-dependent rock
properties, into the calculations. A new linear flow analysis technique is presented in the current study which can be
applied to tight/shale systems with multi-phase flow and pressure-dependent rock properties.
Methods/Procedures/Process: In recent work (SPE 172928-MS), the authors developed a semi-analytical model for
analysis of early-time production data of tight oil reservoirs during transient linear flow. The model assumes two-phase
flow of oil and gas in the reservoir. However, there are numerous documented tight oil reservoirs that exhibit high
water-oil ratios (WORs) throughout their production history. This fact necessitates the inclusion of water production
into RTA to ensure accurate hydraulic fracture characterization. In the current work, the model development is
revisited to account for all flowing phases (oil, gas and water). The proposed technique is fundamentally based on the
application of the Boltzmann transformation technique.
Results/Observations/Conclusions: In the current work, it is demonstrated that, through application of the Boltzmann
transformation, the highly nonlinear partial-differential equations (PDEs) governing three-phase flow through porous
media can be converted to three nonlinear ordinary-differential equations (ODEs). The proposed approach has two
important advantages: firstly, the solution of the set of ODEs provides the relationship between saturation and
pressure, which is used to evaluate pseudovariable parameters; secondly, similar to the approach of Perrine (1956),
the solution to the individual ODEs can be combined through defining total mobility and total flow rate to determine the
linear flow parameter, xf*sqrt(k).
Applications/Significance/Novelty: The presented model provides a theoretical framework for analyzing production
data considering a wide variety of reservoir fluid systems. The robustness of this innovative approach is tested
through comparison with more rigorous numerical simulation. Through comparison with numerical simulation, it is
concluded that the estimated pressure-saturation relationship is robust. Average relative errors in xf*sqrt (k) estimation
using the new methodology are calculated to be less than 15% for all the simulated cases presented in this study. The
new technique should serve as a useful tool for petroleum engineers responsible for forecasting tight oil wells
exhibiting the complexities of multi-phase flow and pressure-dependent rock properties.
Understanding Production Drivers in the Vaca Muerta Shale Using an Integrated Reservoir Simulation Approach
1 2 3 1
R. Altman; ; D. Pederiva; ; R. Mehranfar; ; M. Frydman; ;
1. Schlumberger, Buenos Aires, Argentina.
2. Wintershall, Buenos Aires, Argentina.
3. Schlumberger, Mexico City, Mexico.
Objectives/Scope: Investment decisions are often made in newly explored shale acreages where there is no seismic
data and few wells. Production data is used to compare prospective areas, yet it is well known that there is production
heterogeneity even where wells are in close proximity. This is caused by multiple production drivers, some related to
rock quality and others on completion and production techniques. In this work, an integrated approach was used as
part of a reservoir simulation workflow to understand what caused the production differences between wells in
different areas of the Vaca Muerta shale.
Methods/Procedures/Process: The case study includes three vertical wells (A,B,C) located in different areas of the
Vaca Muerta shale within 25km of one another. Initially, the productivity of these wells was compared. The results
implied differences in production performance but gave no indication of why the wells were different. This was
followed by adopting a consistent reservoir simulation history match workflow across all wells. All models were run
with dual porosity and compositional PVT, and all petrophysical and geomechanical input data used were based on
consistent core-calibrated interpretations. Prior hydraulic fracturing modeling helped establish limits to fracture
geometries used in the simulation model. Water was incorporated in all models to account for flowback and moveable
formation water.
Results/Observations/Conclusions: Simulation history matching parameters were compared between the three wells.
Despite well C having a smaller hydrocarbon-filled porosity than wells A and B, the best production performance was
from Well C, implying that there were other production drivers acting. The history matches showed an enhanced
hydraulic fracture surface area in well C compared with wells A and B, even when normalized for the amount of
proppant and fluid pumped. This ease of generating hydraulic fractures was consistent with the lower minimum
horizontal stress observed in well C. However, by decoupling production drivers in the reservoir simulation, the PVT
(lighter oil) was shown to play a bigger part in the enhanced performance of well C than the generation of larger
hydraulic fracture surface area.
Applications/Significance/Novelty: Reservoir simulation applications in shale reservoirs tend to focus on EUR,
quantifying drainage areas for optimal well placement, and completion and stimulation optimization. There are few
examples that demonstrate the simulation application shown in this work which is to understand why shale wells are
producing differently. Reservoir simulation history matching is a powerful integration tool. When used consistently
across shale wells and with sufficient production and well data, can help reveal differences in production drivers
between wells. This work shows that a highly integrated approach is required to decouple production drivers,
understand why wells in different areas are producing differently, and provides a foundation for making better
business decisions.
Reservoir Characterization to Understand Optimal Well Spacing – A Wolfcamp Case Study
1 2 1 1
N. Bansal; ; T. Blasingame; ; J. Han; ; Y. Shin; ;
1. Subsurface and Technology , Anadarko Petroleum Corporation, Midland, TX, United States.
Objectives/Scope: Potential oil resources and unique nature of The Delaware Basin (West Texas (USA)) make
commercial development in the area attractive and challenging at the same time. Establishing the optimal well spacing
has been recognized by operators in the region as one of the key challenges — especially balancing between the
technical aspects of optimal hydrocarbon recovery and the economic aspects of optimal financial metrics. It is
extremely important to first understand the basic nature of this reservoir itself and what makes it so unique, before an
attempt can be made to tackle optimum well spacing.
Methods/Procedures/Process: This study provides an integrated approach to characterize the Wolfcamp Reservoir.
Facies-based modeling is used to represent the geological and petrophysical heterogeneity as well as to understand
the source and dynamic behavior of the reservoir fluids (hydrocarbon and water). Numerical simulation is then used
for history-matching and forecasting. Given the non-uniqueness of reservoir modelling results, these outputs are
cross-validated with complementary data and analysis (e.g. production data analysis, interference tests, microseismic,
and geochemical data analysis). Reservoir modeling outputs are then used to calculate the EUR degradation as a
function of well spacing, and then incorporated into a systematic economic evaluation of various field development
scenarios.
Results/Observations/Conclusions: As part of the work performed in this study, comprehensive analysis has been
performed on both single- and multi-well pads to establish the advantages of each scenario. Single-well models have
been instrumental in providing an understanding of fluid movement in subzone targets, the influence of relative rock
properties of these intervals, and their impact on depletion profiles over time while multiwell models have provided key
insights into understanding well-to-well interactions over time and the impact of these interactions on long-term
forecasting. Validating such modeling results with other methodologies provides robust solutions and improved
predictability both on a local and regional scale.
Applications/Significance/Novelty: In this work we provide an integrated modeling workflow for the Delaware Basin
Wolfcamp shale that explores vertical and lateral well interactions over time. The workflow requires an interdisciplinary
understanding of the reservoir, which is critical to develop well spacing solutions for these types of unconventional
reservoir systems.
Modeling Well Performance in Piceance Basin Niobrara Formation Using Embedded Discrete Fracture Model
1 2 3 4 1
Y. Xu; ; W. Yu; ; N. Li; ; E. Lolon; ; K. Sepehrnoori; ;
1. Petroleum Engineering, University of Texas At Austin, Austin, TX, United States.
3. Black Hills Exploration & Production, Denver, CO, United States.
4. Liberty Oilfield Services, Cibolo, TX, United States.
Objectives/Scope: Production results from six Mancos wells drilled into the Piceance Basin Niobrara Formation show
large performance differences among the wells despite the similarity of completion designs. The differences are
attributed to well interference and shared production through fracture hits. The objective of this study is to develop a
faster numerical model to history match field production data, simulate well interference, and quantify its effect on well
performance.
Methods/Procedures/Process: A numerical gridding technique called Embedded Discrete Fracture Model (EDFM),
where both hydraulic and natural fractures are embedded explicitly in the structured matrix grid, was used in this
study. As the case in previous works, the complexity in fracture geometries can easily be handled without creating grid
refinement or performing unstructured gridding around the fractures. For this study, an EDFM processor was used in
conjunction with a commercial reservoir simulator to perform production history matching on the six wells
simultaneously by varying fracture half-length, fracture conductivity, and fracture height, and honoring other diagnostic
test results (such as PLTs, DFITs, etc.). The model was then used to estimate gas recovery factors from the subject
wells.
Results/Observations/Conclusions: This study shows that the EDFM method reduces the total number of simulation
cells to handle well interference through fracture hits, resulting in a significant improvement in computational efficiency
(about four times faster than traditional local grid refinement techniques). The convenience of modifying fracture
properties in the EDFM method facilitates the history-matching process for complex fractures efficiently. Furthermore,
we investigated the impact of natural fracture properties such as fracture number, length, and permeability on gas
production. The sensitivity results of various parameters (i.e., well spacing, fracture spacing, fracture orientation, and
landing zone) are presented to identify opportunities to increase gas recoveries and reduce development costs.
Applications/Significance/Novelty: Although the model had been improved in recent years, this study was the first
attempt to demonstrate the applications of the EDFM for multi-well pad development in a specific shale play. Due to its
convenience of modifying fracture properties such as fracture length, fracture number, fracture height, and fracture
conductivity, the EDFM method speeds up the history-matching and sensitivity study processes for complex fractures.
Predicting Pore-Pressure From On-shore Seismic Data in the Delaware Basin
1 2 1 3 4
M. Rauch-Davies; ; B. Schmicker; ; S. W. Smith; ; S. Green; ; J. J. Meyer; ;
1. RTO, Devon Energy, Oklahoma City, OK, United States.
2. Devon Energy, Oklahoma City, OK, United States.
3. IKON Science, London, United Kingdom.
4. IKON Geoscience, Adelaide, SA, Australia.
Objectives/Scope: Based on wireline logs, core data and pressure information obtained directly during drilling, the
various shale units within the Wolfcamp Formation in the Delaware Basin are known to be variably pressured with
depth, and the pressure can change laterally within the same rock formation. Zones with anomalous high pressure
can generally be linked to better producing wells. Unknown overpressured areas are also considered a drilling hazard
and being able to predict these cells is of high interest.
Methods/Procedures/Process: Pore-pressure prediction with on-shore seismic data is not trivial as the relationship
between porosity and overpressure is complicated by a relatively complex geological history. In these settings, the
typical variation observed in seismic velocities may not relate directly to changes in pressure; for example, the
presence of gas, and the presence of TOC can both act to slow the velocity which mimics a pressure response that is
actually erroneous. A well-based workflow was developed which was able to predict the pore pressure and construct a
geomechanical model which matched the wellbore measurements. This model was tested on wells with the requisite
log dataset and was able to replicate the observed mechanical wellbore behavior, highlighting the accuracy of the
pore pressure prediction.
Results/Observations/Conclusions: Pore pressure is a critical input to a geomechanical model and can impact the
mechanical behavior of the well. Consequently, comparing the predictions of different geomechanical models can be
used to help calibrate the pore pressure model. A pre-stack facies-based seismic inversion process, capable of
producing physical estimates of impedances and density, was used to invert the seismic dataset for facies and elastic
properties. This technology requires meticulous pre-processing that accurately preserves amplitude, phase, and
bandwidth, and was integrated into DEVON’s pre-drilling workflows in 2015.
Applications/Significance/Novelty: The inclusion of facies in the inversion process ensures that the distribution of
laterally discontinuous units are defined only by the seismic reflectivity and not biased by any interpolation
assumptions. The inversion approach was calibrated using a set of facies-dependent elastic property trends, rather
than a single set of trends for the whole inversion window. The predicted elastic impedance properties could be
expected to honor the rock physics relationships observed in the well log data with greater fidelity. This is important
when considering the subsequent geological characterization and geomechanical analysis using the property
volumes. The well based workflow was applied to inversion results to understand the efficacy of using seismic
inversion for pressure prediction.
Production Metric Analytics in the Wolfcamp Formation
1
C. P. Ross; ;
1. Cross Quantitative Interpretation, LP, SANTA FE, NM, United States.
Objectives/Scope: Successfully discriminating future lateral wells with above type-curve performance from those with
below type curve performance is important for financial planning and field development with unconventional resource
assets. To address this, a predictive analytics approach is utilized that will leverage an asset’s financial position
through infill drilling prioritization.
Methods/Procedures/Process: While most lateral wells drilled into unconventional shale reservoirs or benches in the
Permian Basin will be economical, some wells will have higher yields or better performance than others due to
variations in shale mineralogy. With modern high-fidelity seismic data, we can systematically identify variations in
silica content, effective porosity and total organic carbon content. However, determining the best combination of these
mineralogies in terms of higher yields can be problematic, since the variability of actual shale reservoir can
demonstrate significant variations across the extent of the asset. To that end, a predictive analytic production model is
constructed using multi-attribute seismic data and normalized production metrics from lateral wells.
Results/Observations/Conclusions: For this example, the seismic data covers approximately 55 square miles and
includes 20 lateral wells within the stratigraphic unit used for training and validation. The production metric used is
MBOE cumulative production for a six-month interval normalized to 4800 ft lateral length. Half of the wells were used
as out-of-sample wells; the remaining ten wells were used for training the seismic attributes. Training and out-of-
sample wells are reviewed to assess the validity of the model, which to date agrees with new drilling, indicating a
robust answer that can now be used for future drilling locations, leases, and acquisitions and divestitures can be
evaluated in a similar manner to better leverage the asset and improve business intelligence.
Applications/Significance/Novelty: Predictive analytics is a proactive methodology to better anticipate future outcomes
(i.e. good, better and best lateral locations) using multiple weighted data sets on an asset scale encompassing tens of
square miles. This predictive analytics approach incorporates multi-attribute 3D data with production monitoring of
lateral wells to build a production metric model, which objectively determines which intervals will better perform or
underperform within an unconventional reservoir. A production metric model can be updated over the life time of the
asset to continually improve business intelligence.
Geophysical Data Processing, Rock Property Inversion, and Geomechanical Model Building in a Midland Basin
Prospect, Midland/Ector Counties, Texas
1
S. Singleton; ;
1. Independence Resources Management, Houston, TX, United States.
Objectives/Scope: Current state-of-the-art for development of unconventional prospects is to integrate geology,

geophysics and engineering for a comprehensive reservoir description that increases the probability of drilling
economically successful wells. The objective of the current project is to build a high quality integrated data set to
define the structure, stratigraphy, and rock mechanics of a stacked reservoir sequence in a Midland Basin prospect,
and to use this reservoir description to (1) analyze past completions and production, and (2) high-grade multiple
stacked options for new lateral locations.
Methods/Procedures/Process: The data set consists of portions of 3 seismic data sets that were merged and
processed azimuthally and anisotropically. The data was migrated via PSTM. Well control consisted of 35 DTC, 8
DTS, 3 Sonic Scanner, 2 lateral FMI, and 3 cores. Nine newly-drilled laterals are on production along with a large
number of vertical legacy wells.
Prestack inversion produced AI, SI and density with good accuracy although Vp/Vs was not stable. This volume was
successfully recreated via neural net, which was also used to calculate PHIT and SWT. PR, YM and brittleness were
created using standard equations. Structural attributes consisted of ant track, coherence, Kmin and Kmax curvature. A
geocellular grid was built and populated with facies, VCL, PHIE, PHIT, SWE, SWT and SOE by cokriging with seismic.
Results/Observations/Conclusions: The geocellular model has been shown to have good reliability in predicting rock
properties of laterals that weren’t used in building the model. Principal stresses were added using Stoneley wave and
DFIT calibration. A major concern is the non-virgin stress field due to many legacy vertical producing wells. The
derived regional stress field is updated with each new lateral completed. The objective is to acquire enough data to
perform production history matching.
A primary use of the MEM is completions frac modeling. These are key in determining the most optimal vertical
landing point off of a specific pad. Using these models, we have identified carbonate frac barriers to avoid as well as
fracable limey zones to target, and clayey zones to avoid versus fracable shale zones to target.
Applications/Significance/Novelty: Use of our geomechanical model, fully backed by log, seismic and engineering
data, has led to a sea change in the way we approach potential drilling targets within our lease acreage. We recognize
that error bars at this point are likely large, particularly considering the likelihood of local variations in pore pressures
and Shmin magnitudes and orientations, but with time we will succeed in reducing them to a manageable level. One
key method we hope will increase resolution and reduce error is production history matching but this has its own set of
issues to overcome, namely numerous offset well frac hits, pump shutdowns, equipment maintenance, etc., all of
which lead to shut-ins of varying intensity and duration, and thus creating production curves showing a multitude of
interruptions.
Revealing Geologic Features in the Midland Basin Through Frequency Decomposition
1
R. Moore; ;
1. GeoTeric, Houston, TX, United States.
Objectives/Scope: Using seismic data to determine geologic features is a long draw out process which requires
careful interpretation of several surfaces. In order to help make the process more efficient, attribute volumes and RGB
(Red, Green, Blue) blends of frequency bandpass volumes can reveal the geology before interpretation begins. The
data in this study is multi-client data presented with permission from Geophysical Pursuit, Inc. The area of interest is
Dawson County, Texas, in the Midland Basin. The zones of focus will be the San Andres formation down to the
Ellenburger (Wright, 2008).
Methods/Procedures/Process: Frequency decomposition is a method of filtering the seismic into the frequency
bandpass volumes and recombining 3 select volumes into red, green and blue channels for the low, mid and high
frequencies respectively. This interplay of colors allows for more geologic detail to been seen than amplitude alone.
There are several types of frequency decomposition available, this study will focus on a continuous wavelet transform
(CWT) and a matching pursuit algorithm (Lowell et al, 2014). Frequency data reveals another dimension of information
within the seismic data. Besides geologic features, frequency data can also tell us about the thickness of events
(Paton and McArdle, 2014).
Results/Observations/Conclusions: Through the generation of CWT RBG blends, several key geological features can
be identified. In the San Andres formation a channel system can much more easily be identified on the RGB color
blend than the reflectivity data alone. The RGB blend highlights the direction of flow from the north to the southeast.
The frequency decomposition blends also detail infomration on the thickness of the units, the individual channels are
higher frequency, thinner, whereas the combined channel is lower frequency, thicker.
Southward prograding clinoforms are more apparent on the CWT RGB color blend than on the reflectivity data. The
clinoforms can be seen as almost blue lines, moving through the slice, showing the direction of progradation.
Applications/Significance/Novelty: Understanding the geologic features within the seismic data before beginning the
interpretation will garner more precise interpretations. This paper will describe the different techniques of frequency
decomposition, when to use each type and how the information in the RGB blends can shape the interpretation.
The frequency decomposition RGB blends allow for a quick visual understanding of the geologic data contained within
seismic data, by knowing the geology before interpreting, quicker more efficient horizons and faults can be generated.
Acknowledgements
This project would not be possible without Geophysical Pursuit Inc and their contribution of the multi-client data that
was used in this study.
Strategies for Improving the Performance of Child Wells in the Permian Basin
1 1 1 1 2 2
R. Manchanda; ; P. Bhardwaj; ; J. Hwang; ; M. Sharma; ; M. Maguire; ; J. Greenwald; ;
1. The University of Texas At Austin, Austin, TX, United States.
Objectives/Scope: A Permian Basin case study is presented to provide a better mechanistic understanding of parent-
child well fracture interference. In the example field scenario presented in this paper, the parent well was produced
and refractured before the child well was fractured. The key objective of this study was to explain and remedy the 40%
reduction in expected EUR of the child wells compared to the parent wells. Another objective was to understand why
sand flowback in the child wells was observed to be 4 times that in the parent wells.
Methods/Procedures/Process: A 3-D reservoir-scale poro-elastic geomechanics software (Multi-Frac) was used to
model the depletion of the parent well and calculate the induced stress changes. The impact of induced unpropped
(IU) fractures around the parent well on the induced stress changes was analyzed. The production of the parent well
was history matched to estimate SRV and fracture dimensions and capture the stress reorientation and stress
changes induced by reservoir pressure depletion. Different mechanisms that could cause child well underperformance
were evaluated. Strategies for improving the hydrocarbon production for the parent-child well pairs were suggested.
The simulation results were corroborated with various field diagnostic data (microseismic, tracer, and production data).
Results/Observations/Conclusions: From this case study, we concluded that the key mechanisms that control the
performance of the child well are the extent of the depletion region around the well, poro-elastic stress changes
induced by depletion and the stress changes induced by refracturing. The poro-elastic capabilities of Multi-Frac enable
calculation of the stress changes in the reservoir induced by depletion and re-fracturing. These stress changes explain
the observed fracture hits from child wells. The lower stresses induced by the depletion are also used to explain the
greater proppant flowback in the child well.
Applications/Significance/Novelty: This work combines the learnings from numerical simulations and field data to
mechanistically understand the reasons for underperformance of child wells in a pad. The three main factors that
induce interference between parent wells and child wells are reservoir depletion, reservoir total stress changes and
reservoir effective stress changes. Each of these factors has been quantified in this work. Modifications to parent well
fracturing treatment design, refracturing treatment design and child well fracturing treatment designs are suggested to
improve the productivity of both the parent and child wells.
Estimation of Seismic Attenuation in the Delaware Basin Using Peak Frequencies
1 1 1 1 1
R. Keller; ; M. Drwila; ; L. Eisner; ; Z. Jechumtalova; ; D. Anikiev; ;
1. Seismik s.r.o., Prague, Czechia.
Objectives/Scope: Induced seismicity in Delaware Basin, USA, has recently gained attention of the geoscience
community. Regional knowledge about seismic attenuation in this sedimentary basin is generally poor. The peak
frequency method was applied to the selected seismic events from the Delaware Basin to estimate effective quality
factors of the basin sediments. The observed attenuation can be characterized by Q factors of approximately 90 and
100 (with an uncertainty of 30) for P-wave and S-waves, respectively. The ratio between Q and Q being close to
p s
one is common for saturated sedimentary basins.
Methods/Procedures/Process: Recently Eisner et al. (2013) developed a robust methodology to obtain effective Q
factor using the peak frequency of a direct wave of a microseismic event. The quality factor Q can be calculated using
measured peak frequency and traveltime of the direct wave of shear source seismic events. These relations are
suitable only in case when the peak frequency is below the corner frequency, otherwise the peak frequency value is
not driven by propagation effects and cannot be used in the attenuation analysis properly. Peak frequencies of direct
waves can be reliably obtained by measurement of the first half-period of the signal, as shown by Wcislo and Eisner
(2016). This approach has an important advantage as it minimizes the influence of waveform distortion caused by
later arrivals.
Results/Observations/Conclusions: Most of the observed effective values of Q factor can be explained with a
homogenous isotropic attenuation model. The observed attenuation in the sedimentary basin can be characterized by
Q factors of approximately 90 and 100 for P-wave and S-waves, respectively. The ratio between Q and Q is
p s
common for sedimentary basins as they are less attenuating of S-waves (Pham et al, 2002). The standard deviation
representing the uncertainty of the determined Q factors is approximately 30 for both Q and Q .The ratio between P-
P S
and S-wave attenuation factors is consistent with the previously observed attenuation factors from other sedimentary
basins and may indicate liquid saturation in the sedimentary rocks.
Applications/Significance/Novelty: While velocity structure of the basin is discussed in various earlier studies, little is
known about seismic attenuation in this basin. Proper estimation of the Q factor is important for source mechanism
inversion (e.g., Eisner et al., 2011), computation of moment and magnitude (Lay and Wallace, 1995) and network
design (e.g., Hallo, 2012). Thsi si the first study discussing the attenuation int he basin. Analysis of attenuation in
Delaware basin significantly differs from the attenuation in the basement.
Development of the Stacked Pay in the Delaware Basin
1 1
F. Alimahomed; ; C. Defeu; ;
1. Schlumberger, Midland, TX, United States.
Objectives/Scope: The Delaware basin is one of the most prolific basin in the United States, with operators targeting
multiple reservoirs in one section, using a stacked and staggered pattern. With the expectation that wine-racking is the
future of the industry, we have developed workflows to understand the growth of the hydraulic fractures and their
production interference with the offset wells. The integrated workflows are centered on building a calibrated 3D model,
to perform predictive modeling to determine the optimum well spacing, both vertically and laterally.
Methods/Procedures/Process: Technology integration plays a significant role in identifying the key drivers of
production. Our workflows involve building a geomodel using high-tier logs and utilizing an unconventional fracture
model to simulate hydraulic fractures to understand the stimulated reservoir volume (SRV). The fractures are then fed
to a reservoir simulator to perform production history matching. This calibrated model is then used for predictive
modeling of the various combinations of stacked & staggered laterals. With the use of 3D finite element
geomechanics, inter and intra stage stress shadow is considered. The workflows involve spacing wells at varying
distances and depths, and understanding the fracture interference and pinch points vertically to optimize the stacking
and improve hydrocarbon recovery.
Results/Observations/Conclusions: Pilot well logs show that high stress barriers of varying degrees exist across the
basin. Knowing the stress regime local to a field or section is important in determining the optimum landing locations
and completion designs. Typical pump schedules for the zones of interest were selected based on current industry
practices. Fully 3D planar fracture simulations were performed on the pilot well at various landing locations. Based on
fracture heights and the interference between the zones and, the production forecasts, the best producing targets
were selected for the well spacing modeling. The production history matching was performed using a P50 type curve
for each reservoir in the area. Well spacing sensitivity was performed to determine the optimum number of wells per
zone per section.
Applications/Significance/Novelty: The workflows address many questions that operators in the Delaware basin are
asking today. How many wells do I need to drain the entire section effectively? How far apart should do I space my
laterals? Do pinch points impair long-term production?
Material Balance Approach to Determine Drainage Volume for Multi-Fracture Unconventional Oil Wells
1
A. Gherabati; ;
1. Bureau of Economic Geology at University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: The purpose is to optimize well spacing and completions in shale oil plays by estimating drainage
volumes. Drainage volume is a useful measure to compare the success of completions and spacing. Stimulated rock
volume determined from microseismic and effective propped volume determined by proppant tracer logs yield static
volumes; however, microseismic overestimates and tracer logs underestimate the reservoir volume that contributes to
flow. This work presents a modified material-balance method to provide a dynamic drainage volume for horizontal
wells using production data and fluid properties.
Methods/Procedures/Process: Drainage volume is defined by the propagation depth of pressure disturbance created
by production. The well-reservoir description consists of a hydraulic fracture in the center of an expanding volume
where pressure increases away from the hydraulic fracture. The difference between the initial solution GOR (R ) and
si
cumulative GOR is used to estimate pressure distribution within drainage volumes from which the size of expanding
volumes are calculated. Material-balance equations for liquid and gas are solved simultaneously through an iterative
process to estimate the pressure distribution and drainage volume. Pressure distribution has a Wattenbarger linear
flow solution shape, and an adjustment factor is used to modify the pressure distribution to match observed
production.
Results/Observations/Conclusions: Oil, water, and gas rates are used to calculate the drainage volumes of 200 wells
in the Midland Basin. The model is evaluated by comparing results to numerical models describing wells with high
watercut and wide ranges of GOR. We present an analysis of the effect of spacing and completion variation, such as
stage length, on the relation between drainage volume, slurry volume, and amount of proppant. Analysis shows that
average drainage-volume pressure at abandonment is a function of initial pressure and bottom-hole pressure, and that
higher proppant concentration increases the size of the drainage volume. Wells with smaller completions are
performing independently of spacing, but spacing of less than 500 feet results in overlapping drainage volumes in
larger fracture treatments.
Applications/Significance/Novelty: The effect of solution gas drive is determined by estimating the gas saturation when
reservoir pressure drops below bubble point. Sensitivity analysis on gas saturation shows that ignoring gas saturation
can erroneously double the size of the drainage volumes. The material-balance method yields a drainage volume that
is only a function of completion size, rock mechanics, and time; therefore, it facilitates the application of data analytics
to understand the impact of completion parameters. Analysis suggests that spacing and completion can be optimized
to avoid wasting economic resources and increasing the net present value. Because of the modest data requirement,
efficiency, interpretability, and dynamic nature, material-balance methods can be applied to a large number of wells.
Characterizing Well Spacing, Well Stacking, and Well Completion Optimization in the Permian Basin – An Improved
and Efficient Workflow Using Cloud Based Computing
1
P. Pankaj; ;
Objectives/Scope:
In a multi-well environment, the formula for improving the recovery efficiency per rock volume depends on the well
spacing, stacking and the completion strategy. Operators in the multi-benched Permian basin have been actively
pursuing various trials of different combinations of vertical and horizontal spacing and completions of the wellbores.
The study presented in this paper tries to achieve a prescription for successful exploitation of the cube of the
unconventional reservoir rock through cloud based multi-variate simulation modeling.
Methods/Procedures/Process: A multi-layer Wolfcamp earth model was calibrated. Reservoir characterization for
petrophysical and geomechanical properties and discrete natural fracture network (DFN) were the fundamental steps
to build the calibrated earth model. Over 300 multi-threaded streamlined cloud based complex hydraulic fracture
simulations, use of unstructured gridding, fine resolution numerical simulations and finite-element geomechanical
simulations served as the ingredients to derive the optimal solution space. Optimal well landing was achieved by using
a full-3D hydraulic fracture simulator. Effect of varying proppant/ft design (1000lbs/ft-5000lbs/ft), cluster spacing and
stage spacing and various wells spacing (300ft-1500ft) configurations, vertically stacked and staggered configurations
are studied.
Results/Observations/Conclusions: From the study, it is demonstrated that there are four elements of the recipe to
maximize the recovery factor: optimal well landing, optimal well completion, optimal well spacing and optimal time of
completion. The parent-to-child affects production impairment upto 15% in 1year which was exemplified though the
means of finite-element simulations capturing the stress magnitude and direction re-orientation. Stimulation sequence
such as zippering and non-zippering the wellbores for completion was also found to be critical. Multiple sensitivities
have therefore allowed to draw the envelope for optimal strategy of asset development in the reservoir volume.
Applications/Significance/Novelty: While the cloud computing served as the enabler, the methodology discussed in the
case study provides an integrated workflow to optimize the completion strategy in a multi-layered unconventional
formation such as in Permian. It helps to derive a structured approach to minimize the development cost, increase well
completion effectiveness and minimizes the by-passed left-over hydrocarbon in the reservoir.
Advanced Modeling of Production-Induced Stress Change Impact on Wellbore Stability of Infill Well Drilling in
Unconventional Reservoirs
1 1 1 1 1
W. Zheng; ; L. Xu; ; P. Pankaj; ; F. Ajisafe; ; J. Li; ;
1. SIS, Schlumberger, Houston, TX, United States.
Objectives/Scope: Drilling and completion of long, highly-deviated wells have become normal in most of the mature
petroleum reservoirs, especially for unconventional plays. Field observations indicate that infill wells could pass
through depleted zones where geomechanical property changes are induced by stimulation and production of the
offset wells. This often leads to wellbore instability during infill well drilling. To better understand and tackle this
problem, this paper presents a comprehensive modeling study and sensitivity analysis to provide recommendations
for operators on infill well planning.
Methods/Procedures/Process: In this paper, an integrated workflow coupling reservoir and geo-mechanic model is
introduced, to systematically study the depletion effects and key parameters controlling the process. A base case
study is carried over a well completed in Permian Basin with geo-mechanical model created from well logs. The
integrated workflow with finite element computation was applied to predict the induced stress change after stimulation
and production. Infill well wellbore stability and performance were analyzed for different scenarios: 1. Different well
spacing 2. Different drilling time. Sensitivity studies were made by altering geo-mechanical and reservoir parameters:
1. Stress anisotropy 2. Rock strength 3. Pore pressure. Based on the results, dominant factors for wellbore stability
were estimated.
Results/Observations/Conclusions: In a normal stress environment, due to the change in the stress state from
production and the increase in differential stress around the wellbore, the knowledge of equivalent mud weight
required to reduce the wellbore stability issues on the lateral can be critical in order to drill and complete the well
successfully. Need to change mud weight and possibility of wellbore breakouts and mud losses while drilling the infill
well can be imminent. In addition, the capability of the workflow to model the pressure depletion and associated stress
conditions with respect to time provides us insights on stress magnitude, stress anisotropy, stress reorientation, and
impact of timing and spacing of infill wells on the safe mud weight window to optimize the field-scale development
strategies.
Applications/Significance/Novelty: The study presents a reference to strategic planning for the operators drilling their
infill wells on an existing well pad. Preventative methods such as higher mud weight could be applied in advance to
avoid wellbore failure. The study provides us insights to understand the controlling factors over stress changes and
their effects over depletion magnitude during infill drilling. Therefore, guiding us to understand the key triggers for
wellbore stability in unconventional formations. This, in general, will help improve the overall economical returns from
reservoir development stand point.
Understand the Early Indicators for Long-term Performance of Unconventional Wells
1 1
R. Cao; ; H. Liu; ;
Objectives/Scope: Measuring and comparing unconventional well performance is critical for oil and gas activities and
decisions. It is challenging to forecast the unconventional LRS well performance at Delaware basin due to the
heterogeneous reservoir properties, complex multiphase flow, various well spacing, completion designs, and operating
conditions. The objective of this work is to identify the representative early performance indicators for unconventional
wells at Delaware Basin, and understand the key physical parameters that determine these behaviors.
Methods/Procedures/Process: First, two datasets are selected. The first set of data contains over 300 wells with
detailed reliable daily production data and forecasted EUR. The second set of data have over1500 wells obtained from
public database with monthly production data. After that, the correlation coefficient of early production to EURs or 3-
years production are used as the measure for degree of representative. Finally, the wells are examined in various
ways to understand physics behind the results. Reservoir and fluid parameters such as reservoir pressure, PVT
envelopes are linked to the production. RTA are used to compare relative productivity of different wells. A carefully
calibrated reservoir model is also used to understand the well performance at different flowback and operating
conditions.
Results/Observations/Conclusions: The results indicate that the IP30 almost have no correlation with EURs and
should not be used as the metrics for long-term well performance. The IP90 is not a good metric either. The IP180 or
IP365 are recommended as the performance indicator based on this study. The well production is usually wellbore or
facility constrained at the early stage. The pressure and fluid properties are key differentiators for unconventional well
performance at Delaware basin. The reservoir pressure is usually 2000-5000 psi higher than the bubble point, which
enables very high potential flowback rates at early stage. RTA is a good method to compare well performance but
requires detailed production and pressure data.
Applications/Significance/Novelty: A representative early performance indicator is very important for exploration,
commercial deals, big data analysis, field development planning, etc. This study shows that the widely used IP30 and
IP90 production are not good metrics for unconventional well performance at Delaware basin. It is a novel study to try
to understand this from multiple aspects such as reservoir driving mechanism, fluid properties, operating conditions.
Depth of Investigation for Linear Flow: Theory and Practice
1 1
A. Ravikumar; ; J. Lee; ;
Objectives/Scope: The idea of radius of investigation has been used in conventional pressure transient analysis for
various applications such as calculation of distance to a flow barrier. In multistage fractured horizontal wells reservoirs,
linear flow predominates for most of the production period, prompting the development of an analogous concept. In
this work, we present the derivation of the depth of investigation for linear flow, and analyze the difference between
time of detection and time of arrival for constant rate and constant pressure production modes.
Methods/Procedures/Process: We derive the unit impulse function for linear flow rigorously using the method of
Green’s functions, and then obtain the expression for depth of investigation using the concept of maximum pressure
disturbance. Similar concepts that have historically been used to develop depth of investigation type expressions,
such as the successive steady states method, probe radius method, and methods based on asymptotic solutions
analogous to wave propagation are explored and compared. Using these ideas, we analyze the discrepancy between
time of arrival of the pressure pulse at the boundary and the time of detection of the boundary at the well.
Results/Observations/Conclusions: We show that the unit impulse function method is the most reliable method for
determining depth of investigation (and the distance of investigation, in general). We also show that the depth of
investigation is the same regardless of the producing conditions held at the well (i.e., constant bottomhole pressure or
constant rate). The discrepancy between the time of detection of the boundary and time of arrival of the pressure
pulse at the boundary is purely a function of the resolution of the pressure measurements at the well.
Applications/Significance/Novelty: Distance of investigation can be used to estimate drainage areas and map
depletion levels in the reservoir. It can therefore be used to calculate optimal well spacing, and, in the case of
multistage fractured horizontal wells, optimal fracture spacing. In this work, we demonstrate the determination of
permeability from the end of linear flow once the depth of investigation has been estimated, and how this method can
lead to errors if end of linear flow is not properly identified.
Rapid Field-Scale Well Spacing Optimization in Tight and Shale Oil Reservoirs Using Fast Marching Method
1 1 1 1
A. Iino; ; T. Onishi; ; F. Olalotiti-Lawal; ; A. Datta-Gupta; ;
Objectives/Scope: An immediate and pressing need in the development of unconventional reservoirs is an innovative
workflow that facilitates fast optimization of wells placement. The optimal well placement design must ensure minimal
interference between wells while maximizing the recovery/NPV. However, due to highly uncertain and poorly
understood complex fracture geometries, optimal well spacing designs based on classical analytical techniques can
be unreliable. We propose a rapid workflow to optimize well placement in unconventional reservoirs using the Fast
Marching Method (FMM).
Methods/Procedures/Process: Our approach is to partition the reservoir model into independent sub-domains
associated with individual wells based on the Diffusive Time of Flight (DTOF) computed as a solution to the Eikonal
equation using the FMM. Multiphase production rates are then computed for each well by utilizing the DTOF as spatial
coordinate to reduce the full 3-D numerical model to 1-D equivalent models resulting in orders of magnitude speed up
in computation time. The computational efficiency of our approach enables intensive simulation runs to determine the
point of diminishing value of additional well placement to obtain the optimal well spacing.
Results/Observations/Conclusions: We illustrate the power and utility of our optimization workflow with synthetic and
field examples with multiple multistage hydraulically fractured wells. First, we demonstrate the reliability of our
technique by confirming good rank correlation between our FMM-based simulation workflow and full scale simulation.
Then we apply our optimization strategy to obtain the optimal well spacing while accounting for uncertainties in
complex fracture geometries and conductivities. Finally, we show the robustness of our well spacing optimization
workflow with an application to a shale oil reservoir in the Eagleford.
Applications/Significance/Novelty: A unique and distinctive feature of our workflow is the computational efficiency that
results from the significant reduction in simulation complexity through FMM-based 1D transformation. This enables
rapid evaluation of well placement designs for large field models using an exhaustive search method which can be
computationally prohibitive using classical numerical simulation techniques.
Physics-Driven Optimization of Drained Rock Volume for Multistage Fracturing: Field Example From the Wolfcamp
Formation, Midland Basin
1 1 1 1
S. G. Parsegov; ; K. Nandlal; ; D. S. Schechter; ; R. Weijermars; ;
Objectives/Scope: Geomechanical modeling of Hydraulic Fractures (HF) in 3D is used to develop algorithms for
estimation of rock permeability enhancement due to leak-off into secondary fracture systems. Instead of using slow
numerical simulations for transient fluid flow we propose a meshless model based on Complex Analysis Methods
(CAM) to solve linear differential equations.
Our systematic study highlights trade-offs between fracture design inputs and the total drained rock volume. Field data
from the Wolfcamp Formation, Midland Basin in West Texas, provides a real-world case to demonstrate our workflow.
Methods/Procedures/Process: Our workflow starts with the analysis of publicly available logs, geological settings, and
calculation of geomechanical attributes with QC against treatment pressures. Then in a commercial planar-3D
fracturing simulator, we obtain the 3D geometry and conductivity of each HF.
After pressure history matching for each stage, we identify unstimulated zones due to stress shadowing and inefficient
proppant transport. Fluid pressure evolution and related leak-off are used to predict enhanced permeability in the
proximity of HF planes. Subsequently, reservoir simulation with CAM identifies poorly drained regions between
fracture planes.
Recommendations are made for design optimization, infill drilling and future refracturing based on history matching of
production and undrained zones identification.
Results/Observations/Conclusions: Special care should be taken to extract static Poisson’s ratio and Young’s modulus
from dipole sonic logs (DTC/DTS) and check it against treatment pressures. Near-wellbore tortuosity can be reduced
by early pumping of small mesh proppant to erode perforations and limit pressure dependent leak-off through major
natural fractures. Recent streamline simulations of Weijermars et al (2017a, b) confirmed that the steepest pressure
gradients occur near HFs tips, and in between such fractures fluid flow diminishes rapidly forming “dead zones”.
Recommendations are made for well and stage spacing, number of clusters and pumping schedule. We show why HF
can “hit” offset wells and, at the same time, a part of the reservoir may remain understimulated in the absence of
proppant in HF between wells.
Applications/Significance/Novelty: We provide recommendations for HF redesign using pressure depletion and
drained volume simulations. The novelty of the approach is in the proposed algorithm of permeability improvement
between fracture planes. Fluid flow modeling algorithms based on Complex Analysis combines high-speed calculation
of the physics-driven HF simulator with high resolution. It allows fast modeling of drainage contours and production
allocation between individual fractures. Recommendations are made referring to well spacing, clusters, stages
spacing, and proppant pumping schedule taking into account stress shadowing between prior and later fractures
emanating from perforation clusters and stages.
Our work independently confirms results of Narasimhan et al (2017), Raterman et al (2017), and Grechka et al (2017).
Experimental and Numerical Investigation of Fracture Toughness of Anisotropic Shale
1 1 1
Y. Suo; ; Z. Chen; ; S. Rahman; ;
1. School of Petroleum Engineering, University of New South Wales, Sydney, NSW, Australia.
Objectives/Scope: Hydraulic fracturing is a common technique for shale gas reservoir development. Fracture
toughness is a key parameter, which is often used to predict fracture growth and propagation. Shale has laminated
matrix structure with a natural characteristic of anisotropy and plays an important role in controlling the fracture
toughness. In this paper, we present a series of experiments and numerical simulations studies of fracture toughness
of shale and investigate fracture propagation behavior.
Methods/Procedures/Process: The samples were prepared from 0°, 30°, 45° and 90° angles with respect to the
bedding plane The shale samples were saturated in different concertation KCL solutions which is considered to be a
good source of inhibiting agents for clay expansion. The laboratory results of toughness have shown to be highly
variable with respect to both bedding plane and brine concentration and that the angle of 45° and KCl concentration of
4% exhibit the highest fracture toughness value.
Results/Observations/Conclusions: The CCNBD shale samples were tested at anisotropic inclination angles of 0°, 30°,
45°and 90° in the lab. It shows that the shale sample has a strong anisotropy and 45° anisotropic inclination angle
provides the greatest resistance to fracture.
It was also shown that saturation with water based fracturing fluids also affected fracture toughness and 4% KCL
solution is suitable to saturate.
Using the experimental results and numerical simulations, the T-Z criterion curve is demonstrated to be an effective
tool for determining accurate rock failure using the CCNBD specimens
Applications/Significance/Novelty: Results of this study illustrate that shale has relatively high anisotropy due to the
existence of the laminated structure and micro fractures and these characteristic features are strongly reflected in
fracture tough ness estimation. Based on the results a model for estimating fracture toughness is also presented.
The Elastic Moduli Change After Carbon Dioxide Flooding Into Limestone: An Experimental Study
1 1 1 2 3
Y. Zhang; ; M. Lebedev; ; M. Sarmadivaleh; ; H. Yu; ; S. Iglauer; ;
1. Curtin University, Kensington, WA, Australia.
2. Northwest University, Xi'an, China.
3. Edith Cowan University, Perth, WA, Australia.
Objectives/Scope: CO EOR or geo-storage in deep saline aquifers or oil and gas reservoirs is a key technology to
2
mitigate anthropogenic greenhouse gas emissions. Technically, CO is captured from large emitters, e.g. coal-fired
2
power plants, and injected deep underground into geological formations. One potential host rock for CO is thus
2
porous limestone; however, the geomechanical response of the limestone when it is exposed to the acidic CO -fluids
2
(CO saturated brine or wet supercritical CO ) is only poorly understood.
2 2
Methods/Procedures/Process: We thus measured the geomechanical properties of three Savonniere limestone cores
before and after injection of supercritical CO2 at representative flow conditions: a fully brine saturated storage
reservoir at approximately 1000m depth (323K temperature, 10 MPa pore pressure, 5 MPa effective stress) was
-6
simulated. 10 pore volumes of supercritical CO were injected at a capillary number of 10 , which mimics storage
2
conditions.
Results/Observations/Conclusions: Our measurements showed that the limestone partially dissolved due to acid
injection (as substantiated by porosity, permeability, NMR-T2, ultrasonic and micro x-ray tomography measurements;
e.g. porosity increased from 24% to 30%, permeability increased from 20mD to 40mD), and this dissolution
significantly weakened the geomechanical strength of the rock (the Unconfined Compressive Strength reduced by
40%, and the Young’s modulus by 30%).
Applications/Significance/Novelty: We conclude that CO injection into limestone reservoirs can have a dramatic
2
impact on the geo-mechanical strength of the rock; and this drastic reduction may have serious consequences for
reservoir integrity and may pose a geohazard.
Pulsed Power Plasma Stimulation Technique – Experimental Study on Single Pulse Test for Fractures Initiation
1 1 1 1
Y. Xiao; ; W. House; ; E. Unal; ; M. Soliman; ;
1. Petroleum Engineering, University of Houston, Houston, TX, United States.
Objectives/Scope: The primary aim of the study is to investigate the potentials of PPPS (Pulsed Power Plasma
Stimulation) as a cost-effective and water saving near wellbore stimulation technique that does not require the use of
significant amount of water or and no chemical additives. The technique allows the stimulation of the wellbore by
creating self-propped short fractures and/or changing the in-situ permeability by several orders of magnitude.
Methods/Procedures/Process: During these experiments, a pulsed power plasma was created by electrical discharge
with 2 kJ stored electrical energy in water-filled wellbore, which also generates pressure shock waves.
Using the plasma to generate an exothermic chemical reaction can multiply the produced energy by up to two orders
of magnitude. The output mechanical energy was augmented by a factor of 16 without considering heat loss by using
an aluminum foil-water system. When aluminum and water are mixed in stoichiometric ratio, the augmentation factor
could be even higher.
Two groups of test samples with uniform dimension were built under different conditions: water-to-cement ratio,
wellbore size, and completion method. This experimental work focused on the performance of fracture initiation during
single-pulse test.
Results/Observations/Conclusions: The experiments show that multiple fractures emanating from the wellbore can be
created. The extent, number and orientation of these effects depends strongly on the electrical energy transfer
efficiency and the properties of target samples. Experiments using 2 kJ of stored electrical energy achieved peak
pressures of almost 200 MPa. These shockwave pressures were generated by typical plasma durations of 6 micro-
seconds. No evidence of a separate thermite component was obtained.
Fracture propagation depended on pre-existing micro-fractures, bore size and bore tube perforations. Small-bore
vertical models with perforations that do not penetrate the rock produced the most significant effects. In all cases,
significant multiple microfractures were created and propagated.
Applications/Significance/Novelty: Hydraulic fracturing involves a low rate of loading and the maximum treatment
pressures will only slightly exceed the minimum in-situ stresses. Unlike quasi-static hydraulic fracturing, the process of
PPPS involves a faster dynamic loading of energy. It is characterized by reaching a peak pressure that exceeds the
maximum in-situ stress within a few microseconds but not beyond the level of rock yield strength.
The PPPS technology could be advanced for use in horizontal wells or re-energization of existing fractures system. As
low oil prices have pushed the fracking industry into a corner, the PPPS technique has great potentials to work as a
cost-effective complementary stimulation method for hydraulic fracturing or even an alternative for small treatments,
especially in water-scare regions.
An Integrated Geological Modeling Approach to Assess Potential of Field Wells for Application of a Surfactant
Imbibition Process in an Ultra-Tight Rock Formation
1 1 1
S. L. Detwiler; ; A. Roth; ; D. Wang; ;
1. Geology, University of North Dakota, Grand Forks, ND, United States.
Objectives/Scope: As part of our investigations involving surfactant formulations to enhanced oil recovery from Bakken
shale, our laboratory experimental studies (at reservoir conditions) noted incremental oil recovery values up to 20%
OOIP. In order to scale the laboratory results to a real-world field application in the Bakken Formation, actual
geological structures must be incorporated into the modeling.
Methods/Procedures/Process: Our target tight Bakken Formation is complex and consists of various tight rocks
(limestone, siltstone, dolomite, shale) and clays (kaolinite, chlorite, illite, smectite). In this study, a new methodology
for geological evaluation and modelling was employed that combined laboratory surfactant imbibition data, a well log
interpretation program, an online geological mapping tool, and functional reservoir numerical simulation approach.
Results/Observations/Conclusions: Our multi-disciplinary approach incorporated reservoir characterization of 30 target
well locations, well trajectory, formation porosity distribution, permeability distribution, and water and oil saturation
distributions, formation pressure distributions from lateral and vertical directions, and contour mapping of the top and
bottom of the Bakken Formation. Our simulation model predicts that a field application of surfactant imbibition could
potentially enhance oil recovery by 8-10% OOIP over imbibition processes that use brine alone.
Applications/Significance/Novelty: In this paper, our integrated approach details conditions needed for successful
application of the surfactant imbibition process and identifies several promising well locations. Our approach can
readily be adapted for application in other tight formations.
Characterization of Thermal Evolution of Pores and Fluids in Shales Using NMR 2-D Measurement
1 1 1 1
H. Jiang; ; H. Daigle; ; B. Zhang; ; X. Tian; ;
1. Hildebrand Department of Petroleum & Geosystems Engineering, UT Austin, AUSTIN, TX, United States.
Objectives/Scope: The evolution of pores and fluids due to thermal effects is a key factor for predicting shale gas
production. However, different fluid types and a wide range of pore sizes pose difficulties for characterization. We
experimentally changed the maturity of shales using pyrolysis. Pyrolysis and nuclear magnetic resonance (NMR)
techniques were applied for better understanding the thermal evolution of pore/fluid system. The main objective is to
quantify thermal effects on the pore and fluid distribution within the same formation, and to compare results among
different formations.
Methods/Procedures/Process: Shale samples used in the experiment were from the Eagle Ford (EF) and a siliceous,
oil-bearing shale from the northern Rocky Mountains (NRM). For better heat transportation and fluid evaporation,
samples were crushed and sieved to between 20-30 mesh. Initial fluid conditions at room temperature (20°C) were
determined by NMR T / T measurement. The samples were heated in an argon atmosphere for 24 hours at four
1 2
temperature stages (110°C, 250°C, 450°C, and 650°C). For each stage, NMR T and T characterization was
1 2
performed on the samples. Further, the T andT distributions were mapped to T /T ratio (R) and T secular (Ts) for
1 2 1 2 2
better quantification of different fluid distributions. We divided the R / Ts distributions into 7 regions according to fluid
properties and pore size.
Results/Observations/Conclusions: As temperature increased, fluid volume decreased for both lithologies. However,
the change of fluid volume between each thermal stage almost monotonically decreased. Over 68% of fluid
distribution was contributed from low- and mid-viscosity fluid in small pores. A difference metric was used to quantify
difference between two R / Ts distributions. The largest difference occurred between 20°C and 110°C, while the
smallest value occurred between 250°C and 450°C. The difference metric between two samples decreased as
temperature increased, indicating a convergent evolution due to thermal effects. This suggests that thermal effects on
different shale types differ only at low temperatures, and the pore structure and fluids react in similar ways at high
temperatures regardless of rock type.
Applications/Significance/Novelty: Pyrolysis of shale samples allows us to control the maturity of the source rock.
NMR T ratio and T secular mapping provides a quantitative way to differentiate fluids associated with different pore
2 2
sizes. The work characterized different fluid evolution under thermal effects. The evolution depends on the initial fluid
distribution, mineralogy, and organic matter content at low temperature, but is relatively independent of these
properties at high temperature. This study provides important information for understanding how thermal maturation
affects pore and fluids distributions in shales.
Determination of Confined Fluid Phase Behavior Using Modified Peng-Robinson Equation of State
1 1 1
G. Yang; ; Z. Fan; ; X. Li; ;
1. The University of Kansas, Lawrence, KS, United States.
Objectives/Scope: Confinement effect of nano-pores in shale gas/oil reservoirs causes the fluids behaving distinctively
from that of bulk fluids. The unique structural and thermodynamic properties of the confined fluids, such as oscillatory
density profile and critical property shift, disable the commonly used equations of state (EOS) to be applied in such
reservoirs. Molecular simulation can be an alternative, though it is impractical for field-scale application due to the
expensive computation and time. It is important to correct the existing EOS to be viable for unconventional reservoir
fluids.
Methods/Procedures/Process: Techniques have been developed to investigate the phase behavior and properties of
the confined fluids by coupling critical property shift and capillary pressure in the Peng-Robinson EOS (PR EOS). The
PR EOS is modified by introducing a new term to consider the molecule-wall interaction resulted from the pore
proximity. Based on 66 data points collected from both experiments and molecular simulations, a correlation which is
a function of dimensionless pore size and wall chemistry has been developed and incorporated in the modified PR
EOS to correct the shifted critical properties. The capillary pressure effect is also considered in phase equilibrium
computation to account for the strong capillary force in nano-pores, where the pressure in vapor phase is larger than
that in liquid phase.
Results/Observations/Conclusions: The newly modified PR EOS has been validated by accurately reproducing the
measured data and it has a good agreement with the result from molecular simulation, yielding an overall error of
5.6%. It is found that the confinement effect caused by molecule-wall interaction is almost negligible when the pore
size is over 50 nm, where the modified EOS can be degraded to the original form. In addition, the liquid-vapor two-
phase region on a P-T diagram tends to shrink with a decrease in pore size. The bubble point temperature of octane
decreases by 4.5% when the pore size is reduced to 4 nm. Such finding is consistent with the previously measured
data. Critical property shift and capillary pressure are both important factors that control the phase behavior deviation
of confined fluids.
Applications/Significance/Novelty: The proposed techniques can be applied to guide the development of shale gas/oil
reservoirs. In practice, the widely used PR EOS for the conventional reservoirs is well modified to be viable for shale
oil/gas reservoirs by coupling both critical temperature/pressure shift and capillary pressure. This work lays a solid
foundation for the investigation of the thermodynamic properties of confined fluids.
Simulation of Gas Adsorption and Capillary Condensation in Shale Nanopores Using Lattice Boltzmann Equation with
Non-ideal Equation of State
1 1 1
R. Xu; ; M. Prodanović; ; C. J. Landry; ;
1. Hildebrand Department of Petroleum and Geosystems Engineering, The University of Texas at Austin, Austin, TX,
United States.
Objectives/Scope: Study of gas adsorption in shale organic nanopores has recently gained significant attention
because it is directly related to reserve estimation and production forecast of shale gas reservoirs. Existing theoretical
applications such as density functional theory result from simplified thermodynamic principles, which do not capture
the kinetics that is essential to reproduce the real-life gas adsorption behavior in nanoscale confinements.
Methods/Procedures/Process: In this study, we present a novel 3D lattice Boltzmann model that properly accounts for
long-range interaction between fluid particles as well as fluid-solid interaction by using a modified pseudo-potential
model first proposed by Shan & Chen. Non-ideal equation of state (EOS) such as Peng-Robinson EOS is realized with
proper choice of the potential function. When EOS is chosen appropriately, gas adsorption and phase separation
occur spontaneously as the bulk pressure is increased. We apply the above model to study and compare nitrogen
adsorption isotherms, which are frequently used for pore size distribution and surface area characterization of
nanoporous media, and methane adsorption isotherms that commonly occur in organic shale.
Results/Observations/Conclusions: For the first time, we show the detailed adsorption and capillary condensation
behavior in a single pore with different geometries and in 3D porous medium extracted from unconventional reservoir
core images. The density profile and the phase interface evolvement is visualized during the adsorption process. The
simulated isotherm is compared to density functional theory and the improvements of our model are demonstrated. By
varying the pore size, shape, and the strength of surface forces, we are able to obtain various types of adsorption
isotherms as defined by IUPAC.
Applications/Significance/Novelty: The model developed in this study not only provides a tool to study gas adsorption
and phase separation behavior in complex geometries under nanoscale confinement, but also serves as a
fundamental model based on which other complex physics can be incorporated to study multiphase flow behavior in
shale nanopores.
Fracture Closure (FC) Determination From Two Successive DFITs (Diagnostic Fracture Injection Test) in One
Formation
2 1
G. Liu; ; Y. Rizwan; ;
1. Engineering, TU Delft, Delft, Zuid Holland, Netherlands.
2. University of Houston, Houston, TX, United States.
Objectives/Scope: DFIT is a reliable technique to evaluate formation & get essential parameters for the hydraulic
fracture designs.Several methods to pick Fracture Closure Stress(FCS) are published in last few years,such as the
holistic method and the variable fracture compliance method, but they are inconsistent in identification of final FCS
because of distinct assumptions of closure mechanism.In this paper,a field case,designed with 2 successive injections
in one single formation through same horizontal well,is used to see if the different methods can pick consistent closure
stress from multiple injections.
Methods/Procedures/Process: Fracture Closure Stress might be the most important parameter, obtained from a DFIT
test, followed by reservoir permeability&pressure.
Two successive injections are recorded in the same formation and same well. The first injection was a small volume
DFIT, pumped at a low rate. After shut-in, pressure falloff was recorded for a relatively long time. Then, a much larger
volume was injected in the second injection (breakdown/ step-rate injection) at a higher rate, which is to create a
larger hydraulic fracture. The second falloff was then recorded after the shut-in, and compared with the first one to
pick a consistent fracture closure.
Results/Observations/Conclusions: Due to small volume&low rate of the first injection,the 1st falloff presents fairly
normal leakoff with little abnormal leakoff behavior.Besides, small fracture size&long monitoring time enable fast
fracture closure,after-closure pseudo-linear flow&even pseudo-radial flow,which was rarely observed in the tight
formation.While, because of much larger injection volume&higher rate in the 2nd injection,the 2nd falloff shows much
more complicated behaviors potentially including tip extension,pressure dependent leakoff (PDL),& transverse storage
or height recession.The final FCS picked with holistic method is proved to be identical to that picked from the first
falloff,while the closure picked with variable fracture compliance method is not consistent with & larger than that from
first falloff.
Applications/Significance/Novelty: The case with two successive injections is very informative and could be used as
benchmark for comparing different methods of pressure decline analysis. The paper provides a field case in a tight
gas sand to support that the final fracture closure can be consistently picked with the holistic method instead of
variable fracture compliance method in conventional tight gas reservoirs.
A New Model for Pressure Transient Analysis of Fractured Horizontal Wells in Shale Gas Reservoirs With Continuum
and Discrete Fracture Networks
4 1 1 2 3 1 1 1 1
Q. Liu; ; Z. Chen; ; X. W. Liao; ; H. Zhang; ; W. Yu; ; X. D. Shen; ; X. Shang; ; J. L. Zhang; ; H. Chu;
;
1. State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum, Beijing, China.
2. Research Institute, Shengli Oilfiled Co. Ltd., Dongying, China.
Objectives/Scope: Recently, model developments for pressure transient analysis (PTA) of fractured horizontal wells in
shale gas reservoirs have gained tremendous attention. Microseismic data shows that some shale gas reservoirs
comprise complex hydraulic/natural fracture networks. It is neither practical nor advantageous to simulate a huge
number of these fractures with numerical models. In this paper, we develop a new semi-analytical model for PTA in
fractured horizontal wells with considering complex natural and hydraulic fractures, which provides guidelines for
fracturing design and performance forecast.
Methods/Procedures/Process: Our semi-analytical model has the capability to analyze the pressure behavior by
considering complex fracture networks, non-Darcy flow, and stress dependent permeability of natural fracture
networks. The proposed model includes three domains: (1) matrix, (2) natural fracture networks, and (3) discrete
hydraulic fracture networks. By applying the nodal analysis technique, we eliminate the flow interplay at fracture
intersections and establish the diffusivity equations for gas flow in hydraulic fracture networks. The pressure transient
solution of these diffusivity equation is obtained by using Laplace transforms and Stehfest numerical inversion. We
verify the proposed model by performing a case study with a numerical simulator for complex fracture networks.
Results/Observations/Conclusions: We find that there are some interesting flow behaviors in shale gas reservoirs like
bilinear flow, formation linear flow, pseudo boundary-dominated flow, etc. The pseudo boundary-dominated flow
provides us the information about how large the area covered by hydraulic fracture networks. When compared with
the impacts of geometries of hydraulic fracture networks on pressure depletion, the influences of fracture properties
are stronger. Furthermore, results indicate that fracture property has the influence on the starting time of some flow
regimes. These phenomena offer a possibility to evaluate the properties of hydraulic fracture networks. Based on that,
we develop a diagnostic tool to distinguish the complex fracture geometries in Marcellus shale reservoir.
Applications/Significance/Novelty: Some shale gas reservoirs comprise a huge number of natural and hydraulic
fractures. It is neither practical nor advantageous to use numerical models to analyze the pressure transient
behaviors. This paper innovatively develops a semi-analytical model for shale-gas horizontal wells with continuum and
discrete fracture networks, which provides a throughout understanding of transient pressure behaviors in shale gas
reservoirs and guidelines for the producer optimize field development and well economics.
A Chemical Blend for Stimulating Production in Oil-Shale Formations
1 1 1
K. Mohanty; ; C. Miller; ; S. Tong; ;
Objectives/Scope: Shale reservoirs are fine grained, multi-mineral, sedimentary rocks with ultra-low porosity and
permeability. Hydraulic fracturing increases the stimulated reservoir volume (SRV) and helps drain out the oil from the
SRV. Oil production from such matrix declines rapidly during the primary production. The goal of this work is to
rejuvenate old wells by increasing the SRV and the effective oil permeability further by chemical stimulation while
preserving fracture surface integrity and hardness. This work is aimed at a West Texas oil shale.
Methods/Procedures/Process: Chemical blend components were selected to increase effective permeability and SRV
in a West Texas shale. Three types of experiments were conducted: shale plate, flow-cell, and propped fractured-core
experiments. The surface hardness and roughness were measured before and after the chemical treatment in all 3
experiments. Plate experiments provided kinetics of reactions and wettability. Flow-cell experiments provided matrix
permeability before and after. Micro-CT was used to detect microchannels. Fractured-core experiments provided
conductivity of the fracture before and after.
Results/Observations/Conclusions: The chemical blend resulted in increased surface hardness, surface roughness,
water-wetness and permeability. The surface hardness increased from 210 to 280 MPa. Interaction with the shale
increased the pH from 2.4 to 4.5. The presence of chemicals delayed acid reactivity, dissolved and precipitated
minerals selectively, which increased the roughness of the surface. Contact angle changed from 150 degrees to about
60 degrees, which may improve oil relative permeability. The flow cell experiments showed an increase in the
permeability of the shale plate by a factor of two or more. Micro-CT imaging showed formation of new flow channels.
In the fractured core experiments, the fracture conductivity was maintained during chemical blend treatment, possibly
due to lack of mud generation.
Applications/Significance/Novelty: A chemical blend is developed that can be used to rejuvenate old oil shale wells
periodically. It can also be applied as a pad during hydraulic fracturing. The blend combines weak acids, surfactants
and solvents to improve effective oil permeability and SRV. This mixture can create microfractures and in situ fracture
surface roughness while preserving the fracture integrity and surface hardness. It is environmentally friendly and cost
effective.
Comprehensive Workflow for Lab to Field-Scale Numerical Simulation to Improve Oil Recovery in the Eagle Ford
Shale by Selective Testing and Modeling of Surfactants for Wettability Alteration
1 1
I. Saputra; ; D. S. Schechter; ;
Objectives/Scope: Surfactant has been widely used as a hydraulic fracturing fluid additive. Lately it is believed to
improve shale oil well recovery by enhanced spontaneous imbibition, driven by surfactant-induced wettability
alteration. Most work done on this method is limited on lab experiments, requiring a method to upscale lab results to
field dimension to complete the assessment. This work provides the missing step, utilizing numerical simulation to
perform the upscaling. The improved workflow is provided to demonstrate logical reasoning for surfactant application
as additive for completion fluid.
Methods/Procedures/Process: This work consists of three parts, lab experiments, lab-scale modelling, and field-scale
modelling. Measurements of various surfactant-related properties; i.e., interfacial tension (IFT), contact angle (CA),
zeta potential, adsorption isotherm, combined with spontaneous imbibition experiments are presented on the first part.
These properties are combined with numerical model constructed from CT-Scan images and are used for the lab-
scale modelling to obtain a surfactant-specific set of Pc(Sw) and kr (Sw) curves through history matching of oil
recovery on imbibition results. For the final step regarding field-scale modelling, a synthetic, mechanistic model of a
single stage hydraulic fracture is used to compare oil recovery on the different fluid systems used.
Results/Observations/Conclusions: Four fluid systems are tested in this work, the base case is tested with no
surfactant and the three other cases tested with three different surfactants. Wettability alteration and IFT reduction are
observed on all surfactant cases. Surf1 alters wettability the most followed by Surf2 then Surf3. Surf2 has the lowest
IFT followed by Surf3 then Surf1. For imbibition experiments, addition of surfactant improves the recovery from base
case recovery of 4% to up to 35%. History-matching provides the Pc(Sw) and kr(Sw) of each fluid system and a
correlation of Sw(Pc=0) and CA is observed. Upscaling results are in agreement with lab experiments where Surf1
shows highest recovery followed by Surf3 then Surf2. For the field model, Surf1 improves initial oil rate by 38% and 3-
yr cum. oil by 16%.
Applications/Significance/Novelty: This manuscript provides a comprehensive workflow utilizing basic fluid-fluid and
rock-fluid interactions observed in the lab and up scaled to reservoir dimensions utilizing fine grid simulation with
precise property distribution as determined by CT Scanning. Evaluating surfactant properties for field applications as
completion additives is often difficult when performed independently in light of the heterogeneous nature of the shale
reservoir where each well accesses different region with different petrophysical properties. The significance of this
work is the ability to isolate the difference in surfactant-related properties which allows us to provide an independent
comparison of well performance between different surfactant used and compared with slick water fracs.
Comprehensive Study of Gas Cycling in the Bakken Shale
1 1 1 1
A. Sanaei; ; A. Abouie; ; M. Tagavifar; ; K. Sepehrnoori; ;
1. Petroleum and Geosystems Engineering, University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: Economic production from unconventional reservoirs requires long horizontal wells with an
extensive hydraulic fracture network. However, the ultimate recovery factor under primary depletion still remains less
than 10% of original oil in place. Recently, CO2 huff-and-puff technique has shown promising results in improving the
recovery from liquid-rich shale reservoirs by reducing oil viscosity, swelling oil volume, and vaporizing and extracting
hydrocarbon from the matrix. This paper discusses the technical and economic potential of CO2 huff-n-puff in
Methods/Procedures/Process: In this work, CO huff-n-puff process in several Eagle Ford reservoirs is studied by
2
using a commercial reservoir simulator. Fluid flow in hydraulically fractured wells coupled with natural fractures in the
reservoir is modeled using Embedded Discrete Fracture Model (EDFM). In addition, to understand the main
mechanisms of CO2 injection, the results of a CO core flood is compared with the simulation model. Moreover, we
2
performed history matching on reservoir performance using the available production data. Next, we used the history
matched model and estimated incremental oil recovery due to CO2 huff-n-puff considering operational parameters
such as well spacing, and fracture overlap, injection and production rates, depletion state of the reservoir, number of
cycles, and cycles duration.
Results/Observations/Conclusions: Several mechanisms are contributing to incremental oil recovery in CO2 gas
cycling. repressurization and oil swelling are the main mechanisms in this process. However, diffusion and relative
permeability hysteresis do not have significant effect on this process. The sensitivity studies indicated that
heterogeneity, fracture connections, and fluid phase behavior are the critical factors for successful field
implementation. The study of operational parameters indicated that there is an optimum time for the start of gas
cycling and duration of the cycles. The first cycle should be the longest to reach the saturation pressure and ensures
oil swelling. Moreover, injecting the same volume of gas with lower rate results in lower rate of return compared to gas
cycling with higher injection rates.
Applications/Significance/Novelty: The presented simulation workflow provides fundamental understanding of the key
parameters affecting the CO huff-n-puff performance such as reservoir properties, starting time of CO injection, CO
2 2
injection rate/pressure, soaking time, optimum cycles duration, and number of cycles. In addition, the interaction of
2
injected CO with the rock formation was included for the first time, which might results in higher fracture conductivity
2
based on formation rock mineralogy. The outcome of this study results in fundamental understanding of huff-n-puff
feasibility and design, well selection, and optimized operating conditions.
Simulation of Wettability Alteration Effect on Well Performance in Highly Fractured Reservoirs
1 1
J. Li; ; K. Wu; ;
1. Petroleum Engineering , Texas A&M University, College Station, TX, United States.
Objectives/Scope: Surfactant added in completion fluids has been applied in the unconventional reservoirs for
improving oil recovery through wettability alteration and enhancement of spontaneous imbibition. However, few
simulation runs have considered the impact of complex fracture networks on the wettability alteration due to the
difficulty in handling the connection between fractures and matrix. In this paper, we present an efficient method named
Embedded Discrete Fracture Model (EDFM) to model wettability alteration in reservoirs with complex fracture
networks.
Methods/Procedures/Process: In EDFM, the reservoir is discretized with structured grids and additional grids are
introduced for both hydraulic and natural fractures. Each fracture plane is physically embedded inside matrix grid and
discretized by cell boundaries, but actually represented by those additional grids. Those additional fracture grids are
connected with corresponding matrix grids through non-neighboring connection (NNC) transmissibility. Based on the
concepts of EDFM, we modify the wettability condition of the activated fracture grids and matrix grids around the
activated fractures to be water wet while the inactivated fractures and nearby matrix grids remain to be oil wet.
Results/Observations/Conclusions: We verified our model against Local Grid Refinement (LGR) in both oil wet and
water wet conditions. Both cases give a perfect match in well productivity and pressure response. Based on our
model, we present several numerical studies to investigate the effect of wettability alteration using four surfactants in
different concentrations in highly fractured reservoirs. Our results show that complex fracture networks will amplify the
impact of wettability alteration on enhanced oil recovery and vice versa. The surfactant types and surfactant
concentration also influence the effect of wettability alteration. Different surfactants change the initial wettability into
different states. Higher concentration not only gives more favorable wettability condition but also enlarge the water wet
area.
Applications/Significance/Novelty: The successful development of EDFM preprocessor through coupling a reservoir
simulation model makes the simulation of surfactant injection in complex fractured reservoirs possible. The
modification of EDFM to handle wettability alteration is creative. The results provide critical understandings about
wettability alteration in highly fractured reservoirs and helpful insights about surfactant added into fracture fluid for
enhancing oil recovery in unconventional reservoirs.
An Efficient Method to Determine Wormhole Propagation During Matrix Acidizing
1 1 1 2
Z. Fan; ; X. Li; ; R. D. Ostermann; ; J. Jiang; ;
1. University of Kansas, Lawrence, KS, United States.
2. Tarim Oilfield Company, Korla, China.
Objectives/Scope: The wormhole geometry has profound influences on the efficiency of the matrix acidizing in
carbonate formations. Although numerous attempts have been made to experimentally optimize the wormhole
efficiency, the difficulties of quantitatively describing the wormhole geometry still impede the optimization of the
acidizing treatments. The present work is aiming to efficiently and accurately depict the dynamic wormhole geometry
under various conditions.
Methods/Procedures/Process: In this paper, an analytical method has been proposed to describe the dynamic growth
of wormholes during the matrix acidizing process. More specifically, the growth of wormholes within a given time is
represented by a continuity equation by considering the transient interstitial velocity of the injected acid and the
location of the wormhole tip defined in a Cartesian coordinate system. The measured relationship between the
consumed acid and increased pore volume is used to determine the radius of the grown wormhole within the given
time. As such, the parameters describing wormhole geometry can be analytically calculated by integrating the
continuity equation, the relationship, and proper initial conditions.
Results/Observations/Conclusions: Subsequently, the analytical method is validated through two laboratory acidizing
experiments with various operational conditions. It has been found that the wormhole geometry calculated from the
analytical method has a good agreement with that observed by the CT scan. The transition of the wormhole radius
during the acidizing process has been well captured while theoretical wormhole length has an absolute average
relative deviation of 8.6% and 4.5% with the observed values in two experiments, respectively. The coefficients of
derived formulas reveal the influences of operational conditions during acidizing on the wormhole growth. Since the
branchings of the observed wormholes are missed in the calculations, the new method is more appropriate to the
dominant wormholes with few branches.
Applications/Significance/Novelty: The newly proposed method provides an efficient tool to characterize the dynamic
wormhole geometry during the matrix acidizing in carbonates. The effect of the acidizing conditions on the wormhole
growth have been integrated and well reflected by the coefficients in the proposed method. With a well-defined
geological model, it is possible to determine the field-scale wormhole geometry, which would generate significant
insights to optimizing the matrix acidizing strategies and predicting the production performance.
The Impact of the Fracturing Additives on the Near Fracture Face Matrix Permeability for Shale and Low Permeability
Sand Formations
1 1 1 1 1
A. A. Al-Ameri; ; T. D. Gamadi; ; L. Heinze; ; I. Ispas; ; S. Gorell; ;
1. Texas Tech University, Lubbock, TX, United States.
Objectives/Scope: Hydraulic fracturing is a well-known and commonly used method of EOR. The polyacrylamide
polymer based fracturing fluid known as Slickwater is widely used as a pad fluid. The polymer adsorption near the
fracture face is one of the factors that affects fracturing fluid flowback. This study investigates experimentally the
reduction in the permeability caused by the polymer adsorption in shale and low permeability sand formations.
Moreover, the effect of adding surfactants to enhance the flowback is also investigated.
Methods/Procedures/Process: Constant liquid rate flooding apparatus was used for the purpose of measuring the
effective permeability. Firstly, a brine of 3 wt. % KCl is injected into intact samples of shale outcrops (Eagle Ford and
Marcellus) and sand outcrops (Scioto and Kentucky) to measure the base effective permeability. Then the samples
o
were injected with slickwater fluid (0.1wt. % friction reducer) at 200 F. The samples are kept in the apparatus under
high pressure and temperature for several hours to simulate the shut-in time stage. Lastly, the brine permeability is
measured after reversing the flow direction. The permeability reduction from the base value is attributed to the
polymer adsorption. The upper procedures are repeated for injecting slickwater with 2 gptg nonionic surfactant.
Results/Observations/Conclusions: The base permeability is reduced by 15% and 10% because of the polymer
adsorption for Eagle Ford and Marcellus samples respectively. Injecting slickwater with surfactant resulted in a
reduction of 6% from the base permeability for Eagle Ford and an increase of 11% from the base permeability for
Marcellus.
For Scioto and Kentucky samples, the polymer adsorption led to a decrease of about 80% from the base permeability
and adding surfactant led to about 20% reduction from the base permeability.
The Residual Resistance Factors, RRF for Eagle Ford and Marcellus are (1.2 & 1.1) respectively when injecting
slickwater. Adding surfactant to slickwater led to decrease RRF to (1.06 & 0.9) for Eagle Ford and Marcellus
respectively.
Applications/Significance/Novelty: This study adds to the literature new data about the permeability reduction and RRF
(mobility of water before polymer adsorption to water mobility after polymer adsorption) due to polymer adsorption in
shale and sand reservoirs. Moreover, the data include the enhancement in permeability by adding surfactant. The
results are important for any flowback simulation.
Surfactant Based EOR for Tight Oil Unconventional Reservoirs Through Wettability Alteration: Novel Surfactant
Formulations and Their Efficacy to Induce Spontaneous Imbibition
1 1 1 2 1 1 1 1
A. Katiyar; ; P. D. Patil; ; N. Rohilla; ; W. Yu; ; S. Falcone; ; B. Brown; ; S. Duckworth; ; P. Rozowski; ;
1. Enhanced Oil recovery, The Dow Chemical Company, Sugarland, TX, United States.
2. Polyglycol surfactant and fluids R&D, The Dow Chemical Company, Freeport, TX, United States.
Objectives/Scope: For fractured oil wet shale reservoirs, chemical flooding with surfactants has been shown to
enhance oil production from water flooding (huff and puff method) due to spontaneous imbibition of water into oil-rich
matrix, leading to oil displacement from tight pores and increased recovery. The objective of this work is to identify the
key mechanisms responsible for surfactant formulations and to characterize the interactions that can induce imbibition
into an originally mixed-wet, tight, hydraulically fractured shale reservoir via laboratory evaluations.
Methods/Procedures/Process: Wettability alteration formulations are first developed and screened for aqueous
stability and phase behavior at reservoir conditions. Tailored hydrophobic surfaces are used to evaluate selected
formulations for their potential to shift rock wettability through contact angle measurements. Water-oil interfacial
tension behavior is investigated to elucidate the mechanism of oil recovery. Adsorption behavior of the chemicals is
investigated on various rock minerals to identify formulations with minimal retention and component separation under
static conditions. Finally, the candidates are comparatively tested for oil recovery performance in spontaneous
imbibition experiments at simulated reservoir conditions.
Results/Observations/Conclusions: The key factors considered in developing an optimal surfactant formulation are
favorable phase behavior in reservoir fluids at reservoir conditions, the ability to shift the rock surface wettability to a
more water-wet state, and minimal chemical losses on reservoir minerals. Also, the formulation should not form a
stable emulsion which may pose an oil-water separation difficulty in oil production. The designed formulation was
tested in an Amott cell apparatus to demonstrate the spontaneous imbibition in tight shale cores, and show a
significant incremental oil recovery. Such surfactant formulations can be used in changing wettability of shale rock as
well to improve relative mobility of oil in tight pore to improve oil recovery using huff and puff technique.
Applications/Significance/Novelty: As part of our rigorous approach to design optimal surfactant formulations for
improving oil recovery, we perform surfactant phase stability, surfactant retention and rock-surfactant interaction
related to wettability alteration. Surfactant formulations designed in this work are unique and provide high temperature
o o
stability (above 70 C and up to 120 C) and high salinity tolerance (> 12 % TDS). This study also highlights the
underlying mechanism of surfactant induced wettability alteration of shale rock due to novel surfactant formulations.
Geomechanical Simulation of Different Conceptual Models for Microseismic Interpretation
1 1
S. Maxwell; ; A. Pirayehgar; ;
1. IMaGE, Calgary, AB, Canada.
Objectives/Scope: Despite the wide-spread use of microseismic imaging of hydraulic fractures, there is no consensus
on how the microseismicity specifically links to the hydraulic fracture network. Microseismic can occur from localized
pressure increases (wet) or more remotely associated with stress changes (dry). With the recent increase in use of
moment tensor mechanisms, two conflicting end-member conceptual models have emerged: a view that the
microseismically active fractures represents the entire flow system versus aseismic tensile hydraulic fractures
activating microseismic on pre-existing fractures.
Methods/Procedures/Process: A coupled hydraulic-geomechanical-microseismic, discrete-element model is used to
investigate characteristics of the end-member conceptual models along with intermediate cases. Using a physics
based simulation, the appropriateness and characteristics of various models are compared. Input characteristics such
as variations in stress, pre-existing fracture density, interconnectedness, initial hydraulic conditions of sealed and
open are simulated. The resulting hydraulic fracture system is compared with typical published field conditions,
including the simulated intensity of microseismicity and magnitudes, net pressure, ability to accept proppant and other
diagnostics such as pressure hits and tiltmeter responses.
Results/Observations/Conclusions: Simulations demonstrate that when flow is constrained to interconnected, pre-
existing, microseismically-active fractures, the resulting number and magnitude of associated microseismicity is
significantly higher than observed in most reservoirs (except with occurrence of induced seismicity). By definition, the
simulation also requires interconnected pre-existing fractures. Atypically high net injection pressures were also found.
Fracture apertures were small, resulting in increased probability of a screen-out during proppant stages. Alternatively,
including an aseismic tensile fracture resulted in reasonable microseismic magnitudes and net pressures, plus can
accept proppant. Slowly opening, aseismic tensile fractures are consistent with reported tiltmeter results.
Applications/Significance/Novelty: Helping identify a conceptual interpretation model, will enable microseismic to come
closer to fulfilling its potential, including a quantitative interpretation utilizing moment tensor data. A geomechanically
validated conceptual model will also help promote the technology and application thereby increasing its value. As far
as we are aware, this is a one-of-a-kind, unique study that is a critical step in microseismic technology development,
that would be of interest to anyone involved in interpreting microseismic data.
Modeling Distributed Fiber Optic Sensor Signals Using Computational Rock Mechanics
1 1 1 1
C. S. Sherman; ; R. J. Mellors; ; J. P. Morris; ; F. J. Ryerson; ;
1. Lawrence Livermore National Laboratory, Livermore, CA, United States.
Objectives/Scope: Distributed fiber optic sensors provide a new perspective on the subsurface environment and allow
observations of previously unrecorded signals, such as the strain field associated with fracture opening during
stimulation and high-resolution images of near-field micro-seismicity. Utilizing GEOS, a flexible multi-scale, multi-
physics simulation environment, we are developing the capability to extract maximum information from the signals to
constrain key subsurface parameters such as fracture length, width, and aperture. This will allow optimization of the
stimulation procedure.
Methods/Procedures/Process: The distributed fiber optic subsurface signals are simulated in two ways: 1) a massively
parallel multi-physics 3D THMC code (GEOS) that is capable of modeling hydraulic stimulation of heterogeneous
reservoir with a pre-existing discrete fracture network, and 2) a parallelized 3D finite difference code (SW4) for high-
frequency seismic signals. Models, geometry and parameters for the simulations are derived from fiber deployments.
The combination mimics both the low-frequency strain signals generated during the fracture process and high-
frequency signals from microseismic and perforation shots. The same GEOS framework can be used to simulate the
spatial and temporal distribution of microseismic moment and borehole pressure allowing integration of key
stimulation-induced observations.
Results/Observations/Conclusions: Results (e.g. strain response as measured by distributed fiber sensors) are
compared with available data from distributed fiber optic deployments and demonstrate that quantitative interpretation
of the fiber data provides valuable constraints on the fracture geometry and microseismic activity. Features observed
in fiber data can be replicated in the synthetics. Parameters such as fracture length, width, height and aperture can be
estimated, as well as the effect of pre-existing natural fracture networks on observed strain signals. The effect of
operational changes, including injected fluid viscosity, can be estimated.
Applications/Significance/Novelty: This work demonstrates that quantitative constraints on fracture location and
geometry can be estimated from fiber optic data using state-of-the-art computational codes combined with realistic
models of the subsurface. Specific features in the strain response can be isolated and used in an inversion or
interpretation of field data to improve understanding of the subsurface environment.
Estimation of Fracture Geometries From a Network of Poroelastic Pressure Responses in Offset Wells
1 1
S. Spicer; ; E. Coenen; ;
1. Reveal Energy Services, Houston, TX, United States.
Objectives/Scope: It is now common practice for operators to collect pressure data from isolated “monitor” stages on
offset wells during treatment of adjacent wells. This enables engineers to measure poroelastic responses that
correspond to changes in effective stress caused by fractures being created in each treatment stage. This paper
describes a method to solve for both monitor and treatment fracture geometries given a network of observed pressure
responses using a fully-coupled, three-dimensional, finite element effective stress model and will address
reproducibility and uniqueness of the solution.
Methods/Procedures/Process: Estimation of fracture geometries from a network of poroelastic pressure responses is
performed using a multi-objective, constrained, non-linear optimization approach common in the domain of operations
research. The objective functions that are to be minimized are constructed using an observation graph that drives the
structure of the solution vector. Iterative steps are performed within bounded convex manifolds defined by the
structure of the objective functions, respectively, where limitations on the physical nature of fractures (e.g., that they
must originate along a wellbore) can be expressed in both linear constraint matrices and nonlinear constraint
functions. Advantages and disadvantages of sparse vs dense matrix constructions (large v. small scale) are
discussed.
Results/Observations/Conclusions: A series of increasingly complex “model problems” in both 2D and 3D are
considered. For each problem, a set of known observed poroelastic pressure responses are calculated using a finite
element model. The modeled observations are then used as input to the fracture geometry estimation method.
Determined geometries are compared to the known model geometries. We demonstrate that the method can recover
geometries across a range of model scenarios, including stress barriers due to pad operation order, observation from
toe-preparation stages, and observations in different zones. Solution convergence and uniqueness are shown to be
driven by the manifold space of the effective stress environment induced by the treatment fracture, subject to
constraints on the feasibility of the solution.
Applications/Significance/Novelty: This method for estimating fracture geometries can be applied to offset pressure
observations taken from isolated monitor stages during pad development in a wide variety of settings. These
diagnostic measurements can be used to inform the completions design process — augmenting and enhancing
methods currently in practice. In addition, the consistency of data acquisition and fracture geometry estimation
enables the method to be used as part of full-field “fracture assurance” programs. Continuous statistical surveillence of
completions design, well construction, well spacing, and downspacing decisions can be realized - yielding useful
insight in a volatile comodity pricing environment. We present recent examples from pad development in the Eagle
Ford, Permian, and Stack plays.
Paradigm Shift in Wolfcamp Shale Improves Well Performance by 70%
2 2 2 2 2 2 3
R. Sethi; ; E. Ejofodomi; ; E. Aktas; ; J. Padgett; ; B. Mackay; ; A. Mirakyan; ; B. McCrackin; ; C.
3
Douglas; ;
1. IPS, Schlumberger, Houston, TX, United States.
3. Manti Tarka Permian, Houston, TX, United States.
Objectives/Scope: Determine reason for excessive flowing well pressure decline and low production. Develop an
integrated completion strategy to improve production of newly drilled Wolfcamp Shale wells.
Methods/Procedures/Process: The integrated, workflow is based on the seismic-to-simulation workflow that models
complex hydraulic fractures and their interaction with pre-existing natural fracture networks and the resulting
production impact. This process was applied on an horizontal well in the field and comprised three main steps:
modeling the created complex hydraulic fracture systems, matching the observed production response, and
developing an optimized completion strategy to be applied on future wells, including a new intrinsic rock-fluid
interaction process to identify the optimum flowback aid additive. The integrated workflow was applied on a producing
Wolfcamp horizontal well in Ward County TX with the results compared to the existing completed well in the area.
Results/Observations/Conclusions: The results revealed significant degradation of the hydraulic fracture systems
within the first 2 months of production. Three possible causes were identified: improper flowback procedure,
inadequate completion strategy, and resistance to flow due to rock-fluid interactions. Advanced flowback analysis
indicated no proppant mobilization. The calibrated fracture system indicated a highly conductive system but with
limited surface area. Production sensitivity analyses demonstrated that propped surface area was a significant
production driver hence an optimized completion strategy was developed. A new fluid compatibility test revealed the
flowback additive negatively affected the fracture flow back. An optimum flowback aid along with a clay stabilizer were
determined.
Applications/Significance/Novelty: The new design was executed on a newly drilled well. The first year of production
showed a 70% increase in cumulative oil with 50% less pressure degradation compared to the offset. Additionally,
tracer data indicated that all of the treated stages were contributing to flow. The innovative integration of
unconventional fracture modeling with rock-fluid compatibility testing is a step change in completion optimization and
provides the ability to properly understand and predict the well performance. The positive impact of these results
provided an excellent platform for efficiently determining the optimum completion strategy including fluid additives to
maximize production in the Wolfcamp shale and serves as a model that is readily applicable to other unconventional
basins.
Production Optimization Using Machine Learning in Bakken Shale
1 1 1 1
G. Luo; ; Y. Tian; ; M. Bychina; ; C. Ehlig-Economides; ;
1. Department of Petroleum Engineering, University of Houston, Houston, TX, United States.
Objectives/Scope: Researchers from both industry and academia have studied the tight oil resources intensively in the
past decade since the successfully development of Bakken Shale and Eagle Ford Shale and made tremendous
progress. It has been recognized that locating the sweet spots in the regionally pervasive plays is of utter significance.
However, we are still struggling to determine whether the dominant control on shale well productivity is geologic or
technical. Given certain geological properties, what is the best completion strategy?
Methods/Procedures/Process: Most of the previous studies either analyze the completion data alone or divide the
entire play into different data clusters by map coordinates and depth, which may neglect the heterogeneity in
thickness and reservoir quality parameters. In our study, we first conducted stratigraphic and petrophysical analyses,
using the regional variation in depth, thickness, porosity, and water saturation to capture the regional heterogeneity in
the Bakken Shale Petroleum System. We selected approximately 3,000 horizontal wells targeting Mid Bakken
Formation and 200 horizontal wells for Three Forks with detailed completion records and initial production dates
during 2013 and 2014. Completion data inputs include lateral length, number of stages, fluid volume, and proppant
loading.
Results/Observations/Conclusions: Then we built the neural network model to identify the relationship between the
First year oil production with the geological and completion parameters. We separated the data into three sets for
training, calibration, and validation. After we trained the model using the training set, we cross validated the model to
test its robustness. The model identified the relative importance that each parameter exerts on the first-year
production. Then, with given geological input, we provided the completion strategy that optimizes the first-year
production.
Applications/Significance/Novelty: The developed technique provides a method to identify the best well location,
understand the effectiveness of the completion strategy, and predict the well production. While the data used came
from wells in the Bakken shale, the methodology applies in a similar way to other tight oil plays.
An Industrialized Approach to Stage-by-Stage Completion Designs
1 1 2 2 2
W. D. Logan; ; P. Dalamarinis; ; B. Williams; ; K. Urbis; ; S. Lipari; ;
1. C&J Energy Services, HOUSTON, TX, United States.
2. Silverback Exploration, San Antonio, TX, United States.
Objectives/Scope: To date engineered completion designs have focused on the placement of clusters and stages.The
work presented in this paper addresses the importance of understanding lateral rock properties as it relates to fracture
propagation. A novel approach for rock characterization is demonstrated that predicts the type of hydraulic
fractures(planar vs complex fracture systems) that will be created during pumping execution. These predictions are
then used to tailor completion designs for each stage, altering pump rates, cluster spacing, proppant volume/type,
based on the anticipated fracture propagation.
Methods/Procedures/Process: MSE calculations has been shown to be quite effective for engineered completions
over short intervals of the well. However, inconsistencies in the field data prevents the utilization of MSE as a
consistent approximation of rock properties along the well. A new modeling approach addresses this issue by applying
a dynamic analysis on the drilling data and well bore system. This method delivers a consistent Synthetic Unconfined
Compressive Strength (SUCS) profile of the rock, demonstrating high correlation with Sonic Data and other evaluation
methods. This work correlates SUCS with fracture treatment pressures and micro-seismic measurements from an
offset well. The results provide a qualitative understanding of rock deformation that are applied on 3 Wolfcamp wells.
Results/Observations/Conclusions: The strong correlation between Synthetic UCS and Sonic UCS allowed the
operator to discontinue sonic logs.The results were then used to guide a stage-by-stage engineered hydraulic fracture
design. Parameters such as cluster spacing, pump rate, proppant type/concentration were varied by stage to help
mitigate issues such as increased treatment pressures and screen-outs, resulting in cost savings of close to a quarter
of a million dollars per well for these 3 wells. This lead to increased production and lower depletion rates, compared
with offset wells that have been stimulated with both engineered and geometric completion designs, in the same field.
The best producer of the 3 wells delivered a 49% uplift in overall production.
Applications/Significance/Novelty: The majority of the engineered completions have focused on perforation placement
in similar/minimum stress rock for improved efficiencies during fracture initiation. In a few cases operators have
engineered other aspects of the completion (proppant, fluids, cluster spacing and pump rates) to take full advantage of
rock properties during the fracture propagation phase. This work done in this paper industrializes this process, using
geomechanical rock profiles derived from drilling data. The predominant type of rock in a stage guides the completion
designers’ understanding of the fracture systems that will be created. The use of the SUCS profile of the rock and its
anticipated performance during fracture has been applied to more than 40 wells in several basins with comparable
results .
A Simple and Cost-Effective Workflow for Engineered Perforations
1 2 3 4 4
B. Cottingham; ; J. Truax; ; M. Padgham; ; X. An; ; S. Denney; ;
1. Completions, Linn Energy, Edmond, OK, United States.
2. Petrophysics, Linn Energy, Oklahoma City, OK, United States.
3. Reservoir, Linn Energy, Oklahoma City, OK, United States.
4. Baker Hughes, Oklahoma City, OK, United States.
Objectives/Scope: In the current oil and gas industry, unconventional-reservoir development is largely driven by
horizontal wells and hydraulic fracturing. More specifically, managing risk is an important aspect for cost-effective
economic workflows. We address large uncertainties around cluster efficiency and how we manage it. In order to
increase our cluster effeciency, we have developed a workflow to take currently-acquired drilling and geological data
and apply them to a process for selectively engineering our perforations. In addition, we enhance our workflow with
the application of LWD technologies.
Methods/Procedures/Process: In order to face the challenge of cluster efficiency and combat heterogeneity within our
completion stages, we have developed a cost effective workflow to selectively engineer our perforations to optimize
our completions. The workflow that we have developed incorporates data that we acquire through the drilling process
for little expense and incorporates it into our analysis. This analysis uses gamma ray, drill cuttings, rate of penetration
(ROP), mud-log gas shows, and gas chromatography to correlate each foot of the lateral to our understanding of the
reservoir and geo-mechanics. In our case study we applied LWD technologies to further enhance this workflow. The
goal of the workflow we employ, is to place perforations in each stage in as similar of reservoir rock as feasible.
Results/Observations/Conclusions: Given the completion challenges and the complexity of the reservoir and geologic
system, we determined to apply a high-definition LWD electrical imaging tool to enhance our engineered completion
workflow for our case study well. The application of this technology allowed us to engineer a completion to target
natural fracture networks, select similar rock quality for perforation clusters, and avoid geo-hazards in our design.
Through this application we were able to optimize our completion, decrease overall treating pressure, and increase
our capital efficiency through geo-hazard mitigation.
Applications/Significance/Novelty: When completing unconventional horizontal wells that commonly porpoise through
differing lithology and rock mechanics, taking an engineered approach will assist in mitigating risks and enhance the
value of the well. In this paper and case study our workflow addresses large uncertainties around perforation cluster
efficiency, geo-hazard mitigation, and productivity enhancement.
Investigation of the Shale Electrical Resistivity Reversal Commonly Observed at the Wet- to Dry-Gas Transition
2 1, 2
A. P. Byrnes; ; S. P. Cumella; ; M. D. Sonnenfeld; 2;
1. Consultant, Evergreen, CO, United States.
2. Whiting Petroleum Corp, Denver, CO, United States.
Objectives/Scope: Shale reservoirs exhibit evolution of resistivity with a noted reversal typically occurring at the wet- to
dry-gas transition. This study documents this phenomenon and investigates mechanisms that cause and influence the
reversal. In many shales, resistivity evolves from low resistivity at low thermal maturity, increases with increasing
maturity, commonly peaking in the wet-gas window, and then decreases at the wet- to dry-gas transition. Resistivity
changes can be related to change in water saturation and distribution, or resistivity of the water, mineral matrix, or
pore network.
Methods/Procedures/Process: These, in turn, are related to evolution of mineral and fluid properties, fluid-rock
properties, porosity, and pore-size distribution. Two key variables are wettability and capillary pressure (Pc) which
influence water saturation (Sw) and distribution. Thermal maturation of kerogen from oil generation onset through wet
gas involves both expulsion of polar hydrocarbons from the kerogen that can partially wet adjacent mineral surfaces
and the development of partial to fully oil-wet porosity within the organic matter (OM). Increase in oil wetness
decreases Pc and the continuity of the water-wet surface which increases resistivity. Further thermal evolution causes
cracking of polar surface-wetting, heavy hydrocarbons to lighter, non-polar, non-surface wetting hydrocarbons.
Results/Observations/Conclusions: This decreases oil wetness and associated electrical resistivity. Concurrently, free
hydrocarbons undergo cracking from low API/low GOR to high API/high GOR. The associated increase in interfacial
tension and decrease in contact angle act to increase Pc. At constant hydrocarbon pressure, an increase in Pc
increases Sw and BVW which decreases resistivity. Observed decreasing resistivity with increasing hydrocarbon
saturation can be related to the greater increase in intra-OM porosity development, which is hydrocarbon-filled. Other
mechanisms that can act to decrease resistivity to varying degrees include: increasing pyrite content, increasing
aromaticity and graphene within kerogen, and increasing brine salinity associated with water vaporization in
overpressured expelled gas.
Applications/Significance/Novelty: Conversely, mechanisms that can act to increase resistivity include: porosity
occlusion, pore-throat size decrease, fresh-water expulsion during clay transformation from smectite to illite, decrease
in clay cation exchange capacity, and increasing hydrocarbon pressure. Resistivity reversal is a significant
phenomenon and is exhibited on well-log cross sections for the Niobrara Formation in the Piceance, Sand Wash, and
Denver basins (Al Duhailan & Cumella, 2014) and is observed in other shale formations (e.g., Woodford, Eagle Ford,
Marcellus, Utica, Vaca Muerta). Understanding and quantifying this phenomenon is important for accurate log
saturation determination and may be used to map the wet-to-dry-gas transition for shale reservoirs.
Observations of Induction Dielectric Measurements and Their Role in Determining Thermal Maturity of Organic
Mudrocks
1 1 1 1
J. C. Rasmus; ; D. Homan; ; G. Wang; ; N. Uschner; ;
1. Schlumberger, Richmond, TX, United States.
Objectives/Scope: Observations of wireline Induction measurements in gas bearing organic mudrocks have shown
that the quadrature or out-of-phase conductivity behaves strangely.
Induction measurements in over-mature Utica gas bearing organic mudrocks have shown an additional strange
behavior where the in-phase conductivity is abnormally high. No amount of alterations in normal saturation
relationships can be used to compute a reasonable saturation in these cases.
The objective of the study was to conclusively determine the cause of the abnormal in-phase and out-of-phase
conductivities.
As kerogen matures it preferentially loses hydrogen and oxygen as compared to carbon atoms. The remaining carbon
atoms therefore become increasingly aromatic. These aromatic carbon rings coalesce into a graphitic type structure.
Laboratory tri-axial measurements of artificial sand-packs laced with commercial graphite show that both the in-phase
and out-of-phase conductivities are altered in the same manner as the log measured values due to the presence of
just a few weight percent of graphite flakes. The presence of pyrite particles produces an opposite effect on the in-
phase conductivities.
Lab and log measurement comparisons show that the abnormal response of the log measurements is due to the
presence of graphite in the over-mature organic matter.
The petrophysical properties of kerogen change with maturity, especially the density and hydrogen index. The
accepted practice of computing the kerogen volumes is to have the petrophysicist assign a maturity to the kerogen
and then assign the kerogen petrophysical properties based on this maturity. To avoid this circular logic, we devised a
method to compute a volume of immature kerogen and a volume of fully mature kerogen or graphite. This avoids
having to arbitrarily choose the kerogen maturity first.
We also discovered a way to map the induction out-of-phase conductivities and graphite volumes to the lab based
kerogen maturity (Ro) values. This mapping is used to compute the kerogen maturity from the logs on a depth by
depth basis.
Applications/Significance/Novelty: The log based kerogen maturity index is used to identify the best reservoir quality
rock in the reservoir. The kerogen maturity can be seen to vary within but also between stratigraphic units. This is due
to mineralogical and diagenesis factors that overprint the effects of pressure and temperature.
The out-of-phase conductivity is primarily a function of the formation dielectric permittivity. We propose, but have not
fully developed, a petrophysical model to explain the formation conductivity and dielectric permittivity in organic
mudrocks. The result is a conductivity and dielectric corrected “host” conductivity to use in conventional saturation
relationships allowing an accurate gas saturation to be computed using all of the log measurements including
resistivity.
Quantifying the Influence of Rock Fabric, Composition, and Electrical Double Layer on the Broadband Dielectric
Dispersion of Organic-rich Mudrocks
1 1
A. Posenato Garcia; ; Z. Heidari; ;
States.
Objectives/Scope: Dielectric measurements are significantly affected by rock fabric, composition, and electrical double
layer. Interpretation of these measurements poses challenges, especially in reservoirs with complex structure and
composition such as organic-rich mudrocks. We develop a simulation framework to calculate multi-frequency
permittivity of rocks to (a) quantify the sensitivity of permittivity to spatial distribution of clay/pyrite and to the double
layer properties in mudrocks and (b) improve rock physics models for interpretation of dielectric measurements for
assessment of hydrocarbon content.
Methods/Procedures/Process: We consider the quasistatic approximation to Maxwell’s equation including the
dispersive contribution of electrical double layers. We mathematically demonstrate that this approximation is valid for
up to 1 GHz for the length-scale of investigation. We apply a finite volume method to solve the Kirchhoff’s current law
in the frequency domain. We assign a permittivity to each voxel of the three-dimensional pore-scale image (obtained
from a Scanning Electron Microscope), depending on the voxel composition. The electrical double layer is modeled as
complex impedance at the interface between solid rock components and electrolyte. Then, we calculate the
permittivity of the rock samples from the relationship between the volume integral of the complex electric current flux
and the electric field.
Results/Observations/Conclusions: We verified that the dielectric dispersion estimated by our numerical simulations
agreed with core measurements. Then, we performed a sensitivity analysis of permittivity to rock fabric and
composition on several synthetic samples. We demonstrated that directional connectivity of pyrite/clay network can
alter the effective permittivity by more than 100% in the frequency range of kHz to MHz. The polarization of the
electrical double layer can modify the effective permittivity by orders of magnitude and cause a detectable low-
frequency conductivity dispersion. Finally, we confirmed that conventional models for interpretation of permittivity
measurements (e.g. Hanai-Bruggeman, Complex Refractive Index) fail to accurately represent the dielectric
dispersion in the entire frequency spectrum.
Applications/Significance/Novelty: Recent publications introduced numerical models to calculate effective permittivity
at pore-scale domain. However, these models were limited to two-dimensional images or a single frequency. In this
paper, we propose a new three-dimensional numerical simulator to calculate dielectric dispersion in organic-rich
mudrocks in the frequency range of mHz to GHz. Our model can describe the wide-band effect of rock
texture/composition and the low-frequency enhancement of permittivity due to the double layer. Hence, the new
simulator can be used (a) as part of an inversion or upscaling algorithm and (b) to develop new dielectric models to
improve interpretation of multi-frequency dielectric measurements for characterization of fluid saturations as well as
rock-fluid interfacial interactions.
Multiscale Pore Systems in Shales and Their Effect on Laboratory Measurements and Transport Properties
1 1 1 1
H. Daigle; ; C. Jiang; ; X. Tian; ; H. Jiang; ;
Objectives/Scope: Characterizing the pore structure of shale is essential to understanding fluid transport through the
matrix and optimizing any stimulation plan. Organic shales are heterogeneous at multiple scales, and the
characteristic length scales or correlation lengths are often longer than the scale of samples used for laboratory
analysis. Using laboratory data to make predictions at the wellbore scale therefore requires careful upscaling. We
show how percolation theory may be used to this end, and demonstrate how the upscaled pore structure may be used
to predict cm- to m-scale transport properties.
Methods/Procedures/Process: We used samples of Barnett and Eagle Ford shales, and a siliceous, oil-bearing shale
from the northern Rocky Mountains. We performed high-pressure mercury intrusion (HPMI), low-pressure nitrogen
sorption, and scanning electron microscopy (SEM) and x-ray microcomputed tomography (micro-CT) analyses on
samples. We determined the properties of the pore network (size distribution, connectivity, spatial correlation) by
constructing representative pore networks that allowed reconstruction of the HPMI and nitrogen sorption data. We
then upscaled the results determining the correlation length with a percolation-based scaling function. The results
were compared with length scales and representative elementary volume (REV) size from images and core-scale
permeability measurements.
Results/Observations/Conclusions: Based on the HPMI and nitrogen sorption measurements, pore networks tend to
be very poorly connected at the micron scale, with average coordination numbers between 2 and 3. Clusters of
connected pores are typically a few hundred microns in size, which is consistent with REV estimates from SEM
images. However, percolation scaling indicates that most samples correlation lengths that are closer to millimeter
scale, which is consistent with micro-CT images. The pore network at both scales must be considered when predicting
transport properties. The scale mismatch is less pronounced in samples with less organic matter, highlighting the role
of organic-hosted porosity in controlling the pore structure in organic shales.
Applications/Significance/Novelty: Our work has significant implications for using laboratory measurements to predict
reservoir properties. While samples are relatively homogeneous at the scale of SEM images or HPMI/nitrogen
sorption measurements, organic-rich samples in particular have longer-range correlations that are not captured at this
scale and yet exert significant control on transport properties. This will affect production from a fracture-stimulated well
since induced microcracks and their interactions with the in situ pore structure are extremely important for moving
hydrocarbons toward the main induced fracture system, as demonstrated by previous researchers. Multi-scale
characterization is therefore necessary to gain a full understanding of the shale matrix.
NMR Considerations in Shales at Elevated Temperature
1 1 1
A. Chakravarty; ; C. Rai; ; C. Sondergeld; ;
Objectives/Scope: - Downhole logging tools capture NMR data at reservoir temperature whereas benchtop NMR core
analysis is routinely carried out at ambient laboratory temperature.
-This difference in measurement conditions makes the comparison of log data and laboratory core data risky.
-The purpose of present study is to analyse low field (2 MHz) benchtop NMR response of Hexadecane and paraffin
wax saturated shale plugs at elevated temperature.
Methods/Procedures/Process: - Oil rich shales - Vaca Muerta and Utica shale plugs used in the study.
- Plugs were cleaned and dried prior to saturation.
○
-T spectra at dry and saturated states were collected at room temperature and at elevated temperature (110 C).
2
Results/Observations/Conclusions: - Heavy hydrocarbon i.e. paraffin wax in present study has a viscosity comparable
○
to water at 110 C.
- Wax is almost completely undetected by low field benchtop NMR at room tempature.
○
- At elevated temperature (110 C), the hexadecane saturated response is similar to paraffin wax saturated response.
- The NMR porosity decreases with increasing echo spacing.
Applications/Significance/Novelty: -The present study shows that heavy hydrocarbons (C24+ in the present study) -
which are mobile at reservoir conditions - will go almost completely undetected by laboratory NMR unless the
measurements are made at elevated temperature.
-The observations show the importance of carrying out elevated temperature NMR measurements in the lab for
improved correlation with log data, especially in liquid rich shales.
Measurement of Natural Gas Isotherms and Imaging Gas in Shale Using NMR
1 1 1
D. Veselinovic; ; M. Dick; ; D. Green; ;
1. Green Imaging Technology, Fredericton, NB, Canada.
Objectives/Scope: Despite the downturn in oil and gas prices, shale reservoir production continues and is expected to
grow as prices normalize. The total gas in place is a measure of the total natural gas content in a shale which consists
of both free gas in the porous spaces of the shales and adsorbed gas on the surface of the shale matrix. The total gas
in place is dependent on the pore pressure and temperature and is vital to the profitable development of a shale
reservoir. The goal of this work has been to employ NMR T distributions to derive gas isotherms.
2
Methods/Procedures/Process: Gas isotherms are derived from NMR data as follows. First, the total gas content is
derived from the distributions at a series of pore pressures. Next, the volume of absorbed gas is estimated using the
pore volume of the rock and the Van der Waals gas equation. The pore volume of the rock was measured in a
separate experiment using the fully water saturated core. The adsorbed gas content is then calculated by subtracting
absorbed gas content from the total gas content.
Results/Observations/Conclusions: The NMR method has been successfully employed to derive gas isotherms. With
the success of the isotherm measurement via NMR, experiments have expanded to include imaging natural gas in
rock cores as a function of time and position. These experiments are difficult as the density of natural gas is
approximately 1400 times lower than water leading to a significant decrease in the signal to noise ratio when imaging
gas as compared to water. Nonetheless, these experiments are valuable for determining where in the core gas is
located and how the gas fills the core over time (i.e. into cracks first followed by occupation of pores).
Applications/Significance/Novelty: Traditionally, gas isotherms are measured by exposing the core to helium and
methane at ever increasing pressure while tracking the volume of gas absorbed and adsorbed. These experiments
involve destruction of the core and provide no information on the pore size distribution. The NMR method proves
advantageous as the NMR analysis can be completed without destruction of the shale core. In addition, NMR pore
distributions are obtained at each pore pressure as part of the experiment. These distributions can be further mined
for information on adsorption versus absorption.
High-Field (400 MHz) T2 Measurements Using a Custom-Built NMR Probe, Eagle Ford Shale, Gonzales and La Salle
Counties, Texas
1 1 2
B. McDowell; ; A. N. Tutuncu; ; Y. Yang; ;
1. Geology and Geological Engineering, Colorado School of Mines, Golden, CO, United States.
2. Chemistry, Colorado School of Mines, Golden, CO, United States.
Objectives/Scope: The objectives of this study were: (1) designing and constructing a high-field (400 MHz) NMR probe
capable of analyzing whole core plugs; (2) measure T2 relaxation distributions using the newly-built probe to calculate
pore size distributions and total porosity in the Eagle Ford Formation; and (3) determine the effect of fluid composition
and salinity on diffusion relaxation in the Eagle Ford Formation.
Methods/Procedures/Process: A custom-designed NMR probe was built to study the advantages of 400 MHz NMR in
porosity characterization of the Eagle Ford Shale. Design and construction was composed of three primary phases:
(1) construction of the structural framework; (2) construction of the radiofrequency circuit; and (3) testing/redesign until
the proper frequency and a stable spectrum were achieved. A series of preserved core plugs were measured for T2
distributions at a variety of saturation states (as-is, dry, deionized water, 8 wt.% KCl, 17.9 wt.% KCl). High-field
measurements were then compared to 2 MHz benchtop NMR spectrometer tests conducted at a commercial
laboratory. An additional study was undertaken to observe the effects of salinity and brine type on the self-diffusion
coefficient in crushed samples.
Results/Observations/Conclusions: An NMR probe was successfully constructed to measure high-field T2 distributions
on 1-inch diameter core plugs. Although the results that have provided better understanding, only a slight difference in
resolution was observed between the T2 distributions in the low-field (2 MHz) and high-field (40 MHz) experiments.
The magnetic field inhomogeneities caused by the inductor along with the effects of paramagnetic mineral (i.e., pyrite)
content of the core plugs have been ascribed for the lower resolution. Furthermore, these effects may be attributed to
the salinity and brine composition dependence. Measurements of the self-diffusion coefficient show NaCl and KCl
brines have an increasing self-diffusion coefficient with increasing salinity; the opposite is true for LiCl brines.
Applications/Significance/Novelty: This research study provides a proof-of-concept that uncrushed core plugs can be
measured using high-field NMR and offers a step-change in the volume of rock measured by high-field NMR
spectrometers. It also confirms the difficulty in measuring higher-resolution NMR data using high-frequency NMR. This
study provides an early step in the usage of high-field NMR acquisition for increasingly smaller pore sizes and
quantifying the effects of fluid and rock properties on T2 distributions. It also highlights that restrictive diffusion due to
salinity and brine composition may be a source of error when calculating T2 distribution in high-frequency
environments.
Integrating Microseismic, Geomechanics, Hydraulic Fracture Modeling, and Reservoir Simulation to Characterize
Parent Well Depletion and Infill Well Performance in the Bakken
1 1 1
C. L. Cipolla; ; M. Motiee; ; A. Kechemir; ;
1. Hess Corporation, Houston, TX, United States.
Objectives/Scope: North Dakota Bakken-Three Forks development has proceeded in two phases; the first phase
consisted of drilling one well per drilling spacing unit (DSU) to hold the lease. The initial well (H1) depletion can
significantly affect infill well hydraulic fracture geometry and impact well spacing decisions. The objectives of this work
were to characterize the effect of H1 well depletion on infill well fracture geometry and well performance. This paper
documents the multi-disciplinary data integration that was required to calibrate the hydraulic fracture and reservoir
simulation models.
Methods/Procedures/Process: This project included an existing Middle Bakken (MB) H1 well, two infill Three Forks
(TF) wells and one infill MB well. Microseismic depletion delineation (MDD) was performed on the existing H1 well to
estimate the drainage pattern prior to the completion of the three offset wells. The MDD consisted of low-rate injection
of 5000 bbls of water and was monitored using microseismic. H1 bottom-hole pressures were monitored during the
MDD and offset well completions. Fracture geometries were determined using microseismic for the MB infill well and
one of the TF infill wells; with the microseismic array in the other TF lateral. The MDD results were used to calibrate
the H1 reservoir simulation model and geomechanical modeling was used to predict the stress changes due to the H1
depletion.
Results/Observations/Conclusions: The microseismic data showed fracture height covering both MB and TF
formations and very asymmetric fracture growth toward the H1 depletion. Asymmetric growth was expected for the
offset TF well that was 750 ft from the H1 well, but the severe asymmetry observed during the completion of the offset
MB infill well that was 1500 ft from the H1 well was very surprising. Guided by the microseismic data and H1 “frac
hits”, the results from the geomechanical modeling were combined with advanced hydraulic fracturing modeling to
evaluate the effect of H1 depletion on infill well fracture geometry, accurately predicting the asymmetry. The fracture
geometries for the offset wells were discretely gridded in the reservoir simulation model and history matching was
used to characterize well performance.
Applications/Significance/Novelty: The reservoir simulation model accurately predicted the production behavior of the
infill wells, while also matching the 3-fold increase in H1 production due to the numerous “frac hits” during the infill well
completions. The integration of MDD, microseismic fracture geometry measurements, geomechanical modeling,
advanced hydraulic fracture modeling, and detailed reservoir simulation history matching provided valuable insights
into the depletion patterns of H1 wells, MB-TF connectivity, the impact of H1 depletion on infill well fracture geometry,
the effect of “frac hits” on H1 well performance, and infill well performance. The calibrated hydraulic fracture and
reservoir simulation models were essential components of subsequent studies to optimize Bakken development.
Interdisciplinarity (Operators’ Forum only): The paper and presentation will highlight the integration of geophysics
(microseismic), completion engineering (hydraulic fracture modeling), geomechanics, and reservoir engineering
(reservoir simulation). Although not the focus of this paper, the starting point for the work was a detailed geo-model
that included rock mechanical and stress data required for hydraulic fracture and geomechanical modeling, in addition
to the standard reservoir properties required for production modeling. There will be two or three speakers for this
presentation.
An Integrated Approach to Optimizing Completions and Protecting Parent Wells in the Montney Formation
1 1 1 1
J. Nieto; ; G. Janega; ; B. Batlai; ; H. Martinez; ;
1. Subsurface, Canbriam Energy Inc, Calgary, AB, Canada.
Objectives/Scope: The key to optimizing completions is a thorough understanding of the subsurface. This paper
describes the multidisciplinary integration at Canbriam which has lead to the creation of a traffic light protocol to select
which intervals in a horizontal well to complete or leave out. The work integrates the 'static' earth model, including the
discrete fracture network (DFN) with 'dynamic' hydraulic frack modeling to assist in prediction of interwell
communication, which in turn helps in the selection of Parent wells for pressure pump-in protection. These pressure
pump-in results are shown.
Methods/Procedures/Process: The geological layering, structure and petrophysics are all integrated with the 3D
seismic data. This enables the extraction of inverted rock properties together with fault and fracture information along
each wellbore. Cemented sliding sleeve completions are then displayed along the wellbore with all geological,
petrophysical and seismic attributes overlaid.
The traffic light protocol uses a combination of parameters to select which sleeves to open (green) and which to leave
closed (red)
The DFN is determined from coherency, ant tracking and AvAz work and overlaid on each wellbore
The young's modulus and porosity from logs and 3D inversion are plotted on the wellbore
The microseismic data are overlaid
Frack simulation is then performed using DFIT derived permeability in the Static earth model
Results/Observations/Conclusions: Results have shown that using one indicator is inconclusive, so multiple attributes
are used in the completions optimization traffic light protocol. For example, where 2 orthogonal faults intersect the
borehole, coupled with large contrasts in Young's modulus in that interval in the well, casing crimping could occur.
There could also be increased likelihood of Induced Seismicity as monitored by our permanent array. The results of
this IS are shown in the paper. Further, we have determined that interwell communication is predictable and can be
mitigated by frack simulation, understanding of the subsurface static model, pressure interference monitoring and
trends in array microseismic. Parent well pump-in protection has been shown to be very effective when subsurface
knowledge is applied.
Applications/Significance/Novelty: The applications are twofold; Firstly, to improve frack efficiency by not wasting sand
and water in a fractured or contrasting mechancial properties zone as little new rock will be broken. Fracking in such
zones could also cause casing crimping and induced seismicity.
Secondly, to use all information to predict inter-well communication and give the best possible chance of Parent well
protection.
This level of integration is uncommon especially as it relates to parent/child protection. Full use is made of the 3D
seismic data to characterize our Montney reservoir, notably to produce a DFN which is used to control permeability in
Frack simulation. Additionally, the gas and water pump-in protection techniques which are described are new, only a
handful of Operators have tried them in Canada.
Interdisciplinarity (Operators’ Forum only): Please consider this abstract for the Operator's forum. Alternatively
sessions 5 or 6
Tank Development in the Midland Basin, Texas: A Case Study of Super-charging a Reservoir to Optimize Production
and Increase Horizontal Well Densities
1 1 1 1 2
J. Thompson; ; N. Franciose; ; M. Schutt; ; K. Hartig; ; J. McKenna; ;
1. QEP Resources, Denver, CO, United States.
2. MicroSeismic, Inc., Houston, TX, United States.
Objectives/Scope: Simultaneous operations including drilling, completion, and production on tightly spaced horizontal
wells are the unavoidable future in unconventional development. This next wave of development has just begun in the
Permian Basin, where operators are shifting from parent well development to increased well densities. However,
increasing well densities over multiple stratigraphic horizons commonly results in production downtime, less effective
stimulated rock volumes of child wells, and drilling hazards. The key to maximizing corporate value will be managing
densities and production optimization
Methods/Procedures/Process: Here we present a Midland Basin case study of a novel multi-disciplinary approach
aimed at the optimization of both surface and subsurface operations for tightly-spaced and stacked intervals. Our
methodology, termed ‘Tank Development’, aims at the exploitation of a block of rock at one time to maximize reservoir
potential. The fundamental principle of Tank Development is the maximization of well productivity by ‘super charging’
the reservoir. Super charging of reservoir energy is accomplished by sequencing hydraulic fracture operations and
simultaneously bringing all wells online to create a more complex fracture network that increases well recovery
efficiency. Our pressure wall concept minimizes well interference while optimizing stimulation, production, and
operational efficiency.
Results/Observations/Conclusions: Surface gauge data, production, and pump intake pressures indicate that an
effective reservoir pressure wall was formed during the development sequence which minimized production losses
from frac hits and maximized effective stimulations. We also used a surface microseismic array to monitor four
horizontal wells in the Spraberry Shale formation. Our results indicate that as Tank Development advances there is
evidence of breaking more rock than when the technique was not employed. The microseismic data also suggests an
increase in near-wellbore fracture complexity for wells that were stimulated later in the Tank Development sequence.
Production of two 16 wells/mile density tests with Tank Development show similar well performance to an 8 wells/mile
non-Tank Development density test.
Applications/Significance/Novelty: Our Tank Development methodology has proved essential for maximizing asset
value during simultaneous development of multiple horizons with horizontal wells in tight proximity to one another. Our
multi-disciplinary development approach is key to optimizing near wellbore fracture complexity and the conservation of
completion energy required for increased well densities. In addition, Tank Development prevents waste by minimizing
frac hits and pressure depletion during parent-child development. The surface and subsurface efficiencies gained
through Tank Development maximize recovery efficiencies and corporate returns.
“Fit-for-Purpose” Treatment of Produced Water for Hydraulic Fracturing
1
R. R. Sharma; ;
1. Global Pdroduction, ConocoPhillips, Houston, TX, United States.
Objectives/Scope: ConocoPhillips is committed to creating a responsible water management plan in developing

assets in the Permian Basin of West Texas. Sourcing water for the large multi-frac stimulations in West Texas is a
well-known constraint on oil and gas development activities in the area. Additionally, large volumes of produced water
(PW) have traditionally required clean up and disposal in injection wells. Rather than viewing these constraints only as
challenges, ConocoPhillips realizes the tremendous opportunity to address both issues by treating and reusing PW for
hydraulic fracturing operations.
Methods/Procedures/Process: ConocoPhillips executed a multi-well reuse pilot program in the Permian Basin. A “fit-
for-purpose” treatment scheme was deployed to selectively remove free oil, suspended solids hydrogen sulfide, and
iron as well as to inactivate microorganisms. While gravity separation was adequate for de-oiling and filtration for
suspended solids, oxidation followed by coagulation and clarification was necessary for dissolved iron removal. The
treated water iron concentration was below 5 mg/L whereas the turbidity was consistently below 10 NTU, making it
suitable for well completions. The treated PW was compatible with the new salt-tolerant friction reducer, achieving
similar friction reduction compared to a standard friction reducer and fresh water.
Results/Observations/Conclusions: The six month pilot operation, demonstrated that PW reuse is technically feasible
and can be a cost-effective solution compared to other available water sourcing alternatives. The technical approach
developed allowed use of salty produced water (>220,000 mg/L TDS) for completion activities utilizing a new type of
friction reducer. Based on the success of the pilot program, a full-field produced water reuse program has been
initiated early in the appraisal and development phase of the project to reduce implementation risks and improve
economics in a safe and environmentally responsible manner.
Applications/Significance/Novelty: Water management strategy development has become a truly collaborative, cost-
effective, and integrated process that addresses the full life-cycle of water.
Strong collaboration led to success in these projects, and they were designed and executed through a significant team
effort made by completion engineers, water treatment engineers, the asset development team, and the operations
team. This paper provides an overview of the plan implemented for safe pilot operation, procedures for treatment of
produced water and water quality results.
Produced Water Treatment R&D: Developing Advanced, Cost-Effective Treatment Technologies
1 1 1 1
E. Frye; ; E. Folio; ; O. Ogunsola; ; E. Melchert; ;
1. U.S. Department of Energy, Washington, DC, United States.
Objectives/Scope: Unconventional oil and gas (UOG) resources are vital to national economic and energy security,
and support energy independence. The Department of Energy (DOE) conducts early-stage research on UOG
resources with the goal, in part, to develop and advance technology that can improve production, while protecting
water resources reducing water use, and addressing induced seismicity. This paper presents a summary of results,
insights, and findings of the produced water treatment subset of the DOE R&D portfolio from work in 2008-2016.
Methods/Procedures/Process: The major research thrusts of the DOE portfolio in produced water management
include collecting data and information on key water quality parameters; developing more cost effective and efficient
technologies to treat water, and developing tools to help identify and implement appropriate technologies for specific
water streams. This review paper will focus on the treatment technologies studied in this portfolio. DOE has
investigated the use of several advanced treatment methods to clarify (remove oil/grease, solids, microorganisms,
etc.), soften (remove divalent cations), and desalinate produced waters. Technologies such as forward osmosis,
humidification-dehumidification, plasma, magnetic ballast clarification, vortex-generation, and ultra- and nano-filtration
membranes were investigated.
Results/Observations/Conclusions: The insights and technology advancements from the 20 DOE-funded produced
water treatment projects reviewed can help overcome common challenges with treating produced water, like cost,
mobility, flexibility (for varying quantities and contaminants), and durability (e.g. prevent fouling). Example results
include:
antifouling membranes that removed organic material and bacteria with less sensitivity to water variability;
comparison of two technologies for water with high barium and radium concentrations; and,
a prototype forward osmosis desalination system with 90% salt rejection.
Collectively, this R&D advanced state-of-knowledge to help reduce cost, increase efficiency, and increase field-
readiness of promising treatment processes. Additional results will be covered in the paper.
Applications/Significance/Novelty: Treated produced water can also be a valuable resource, for future oilfield and
other beneficial use. The major barrier to use (vs. disposal) of produced water is the high treatment cost. Produced
waters vary greatly, so an array of technologies will be needed. The technological advances can reduce reliance on
freshwater and reduce cost in UOG development, which can help enable the development of U.S. resources.
Collective insights from DOE’s R&D portfolio can inform current and future R&D in this area, and help operators
identify options to responsibly manage their produced water.
Tailoring Treated Brines for Reuse Scenarios
1 1 1
A. Hakala; ; M. Wenzlick; ; N. Siefert; ;
1. Research and Innovation Center, National Energy Technology Laboratory, Pittsburgh, PA, United States.
Objectives/Scope: Treatment coupled with injection for additional fracturing jobs is a viable alternative option for
managing produced brines from hydraulic fracturing operations. This study focuses on calculating the amount of
minimum work required to treat brine compositions typical for the Marcellus Shale and the Barnett Shale, along with
estimates of the actual electricity usage and the total costs to tailor-make brines using commercially-available
equipment. The study also addresses the potential to form mineral scale in wells based on the composition of the
tailor-made brine.
Methods/Procedures/Process: Two types of calculations are applied in this study: geochemical modeling for
calculating mineral saturation indices and the activity of water, and theoretical minimum work of separation
calculations. The theoretical minimum work of separation is the least amount of energy per volume of product water
required to chemically separate the brine into a product water stream and a concentrate stream, according to the
Second Law of Thermodynamics. Using reported costs from literature and estimates from private industry, correlations
3 3 3
between the theoretical minimum work of separation (kWh/m ) and the energy use (kWh/m ) and cost ($/m ) to
dewater a brine using mechanical vapor recompression (MVR) has been determined.
Results/Observations/Conclusions: Based on acquired data from case studies, the cost to dewater produced brines
using MVR ranges from $2.00 to $6.00/kWh of theoretical minimum work, with an average of $2.70/kWh. A typical
Marcellus Shale well produces between 500,000 and 1,500,000 gallons of water over ten years of operation with a
weighted average TDS of 125 g/L. To concentrate produced water from 125 g/L to 250 g/L while generating fresh
3 3
water, costs range from $9 to $28/m of generated fresh water with an average cost of $13/m . With a 10% electricity
3
efficiency of MVR, the expected energy requirement to dewater the brine is 49 kWh per m of concentrate (24 kWh
3
per m of inlet produced brine). Wellbore mineral scale can form depending on the sulfate and organic acid
concentrations in the injected treated brine.
Applications/Significance/Novelty: Challenges with dewatering produced brines include: removal of suspended solids;
removal of organic material; removal of scale-forming ions, and; the high energy cost of overcoming the osmotic
pressure to dewater the brine. Knowledge of the energy requirements and costs for treating produced waters to TDS
levels suitable for injection as fracturing fluid base water, without the need to add fresh water for dilution or additional
chemicals for mineral scale prevention, provides a base understanding for future produced water management
strategies. For an average produced water volume of 1,000,000 gallons/well over ten years, costs average $2,500 per
well per year to generate tailor-made brines and fresh water from produced waters in the Marcellus Shale using MVR.
The Water Challenge Program – Permian Basin Pilot Results
1
A. Wilcox; ;
1. Energy Production, HARC-EFD, The Woodlands, TX, United States.
Objectives/Scope: We need water to provide energy, and we need energy to provide water. Drilling, completions and
production requires and yields water. Sometimes, that water can be recycled/reused. Sometimes it can’t. Industry has
been doing this a long time and knows the value of water. The Water Challenge (WC) provides a mechanism through
which O&G companies make and track commitments to reduce potable water usage, increase the use of produced
water in operations and research beneficial use of produced water, providing a transparent manner to document how
industry is performing.
Methods/Procedures/Process: The Environmentally Friendly Drilling Systems (EFD) Program works with industry,
academia, regulators, environmental organizations and other stakeholders to provide unbiased science addressing
environmental/societal aspects associated with O&G operations. The WC Program provides a mechanism for
companies to undertake and document voluntary water conservation efforts through the development and
implementation of cost-effective most applicable practices that address environmental issues. The Water Challenge
Pilot focused on the Permian Basin to identify program improvements, provide industry and state regulator feedback
on the WC structure and begin recruiting participants for the full Water Challenge program.
Results/Observations/Conclusions: The WC Pilot project team conducted comprehensive literature/case study
reviews, met with key stakeholders and toured multiple operator sites in the Permian Basin in order to document
industry’s water management commitments, most applicable practices, successes, challenges and opportunities.
There is no ‘one size fits all’ for water management.
Key results as of November, 2017 are:

1. A major challenge identified centers on logistics.
2. Opportunities:
2.a. Common infrastructure
2.b. Regional collaboration
2.c. Identification of key performance indicators
3. Beneficial reuse of water is very important to industry and warrants further research.
4. Value refers to monetary aspects as well as operational reputation.
Applications/Significance/Novelty: By recognizing and documenting most applicable practices (MAPs), the WC
Program promotes adoption of expanded, voluntary commitments to address water management while offering
flexibility so companies can utilize paths that best fit with their capabilities and sustainability/operating priorities.
Deliverables from this 1 year Pilot include a comprehensive list of MAPs detailing methods being successfully utilized
in the Permian Basin and a website to share this information with all stakeholders.
Optimal Planning for Wastewater Disposal Facilities: Application of Geographic Information System and Data
Analytics
1
A. Jamali; ;
1. Petroleum Engineerg, Texas Tech University, Lubbock, TX, United States.
Objectives/Scope: Oil and gas wells produce substantial amounts of wastewater in forms of drilling fluids, hydraulic
fracturing flowback water, and formation water. Operators usually truck the produced wastewater to injection wells or
disposal facilities. Therefore, wastewater transportation and disposal generate lifetime operating expenses and
improving the efficiency of this process can yield significant capital cost savings. This paper introduces a novel data-
driven procedure to optimize the location and capacity of the disposal facilities and minimize wastewater
transportation costs.
Methods/Procedures/Process: We propose a three-step data analysis procedure: data integration, predictive
modeling, and facility planning. In the first step, we collected, processed, and cleaned data from regulatory agencies
and public domains. This step delivered an integrated database of oil and gas production, also wastewater volume,
origin, and transportation data. In the second step, we used predictive modeling tools, such as decline curve,
correlation, regression, and spatial analyses, to forecast wastewater production on a well-by-well basis. In the third
step, we used the predicted wastewater volumes to evaluate future regional disposal requirements and applied road
network analysis to compute trucking distances and estimate optimal disposal facility size and locations.
Results/Observations/Conclusions: We applied our three-step procedure to the Pennsylvania section of the Marcellus
shale. We observed four classes of producers and created a customized solution method for each to alleviate the
drawbacks of missing data. This process generated state-wide wastewater production maps for Pennsylvania over the
next ten years. Next, origin-destination cost network analysis revealed an excessive 30-50 miles of wastewater
trucking over the past three years. Our optimal solution combined the predicted wastewater production maps and the
location and capacity of current disposal facilities to propose five new state-wide disposal hubs. Our results show that
development of injection and disposal facilities in the proposed locations can reduce trucking distances by 50%.
Applications/Significance/Novelty: Wastewater transportation and disposal are among the most crucial facets of oil
and gas operations. When planning for wastewater management, it is essential to understand the current disposal
practices and future disposal requirements. For the first time, this paper brings the problem of excessive transportation
to the forefront and offers the necessary tools to alleviate it. The analysis finds the optimal facility locations to minimize
wastewater transportation time and cost. Also, reducing the road traffic generated from wastewater disposal
contributes significantly to developing a better rapport with environmental agencies. Finally, the proposed methods
and procedures are developed into an integrated data analysis framework which is applicable to oil and gas reservoirs
across the world.
Replacing Freshwater With Seawater: Problems, Solutions, and Applications
1 2 2
T. Almubarak; ; J. Ng; ; H. Nasr-El-Din; ;
1. Saudi Aramco EXPEC ARC , Bryan, TX, United States.
Objectives/Scope: In this review, a complete picture of seawater as an alternative to fresh water will be presented.
This includes examining the reason for using seawater, the challenges faced, the technologies developed, and many
applications of seawater based treatment fluids. Through this, readers should be able to gain a complete picture of the
problem at hand and the solutions available to tackle it.
Methods/Procedures/Process: In the oil and gas industry, significant volumes of water are used to carry out various
treatments such as water injection, matrix acidizing, and multistage hydraulic fracturing. Fresh water is clean and
contains low salt content, making it the ideal water source to mix these treatments with ease. However, as
unconventional and tighter formations are developed, the use of water in such treatments exponentially increases.
Trucking, piping and shipping freshwater for oil and gas purposes becomes both uneconomical and unethical,
especially for offshore operations.
Results/Observations/Conclusions: Seawater has been gaining attention as a viable alternative to freshwater in the oil
and gas industry. Ideally, the most cost effective way is to use raw seawater. However, raw seawater contains ions
and microorganisms that can introduce or exacerbate scaling, corrosion, bacterial problems, and most importantly
hinder desired fluid performance. Therefore, seawater is typically treated to remove adequate quantities of these
components. Since treatment of seawater is increasing in popularity, it is important to realize which components are
necessary to remove, to achieve a balance between treatment cost and treatment benefit.
Applications/Significance/Novelty: As the world’s population grows, its thirst for water continues to increase. Yet, the
total amount of freshwater that is available naturally does not replenish quickly enough to match this growth.
Consequently, increase in water consumption and continuous drainage has resulted in a growing shortage for
industrial use.
In Situ Fractionation in Liquids-Rich Shales and Its Implications for EOR: Experimental Verification and Modeling
Study
1 1 1 1 1 1 1
S. T. Dang; ; A. Tinni; ; F. Perez; ; D. Devegowda; ; T. Trong; ; C. Sondergeld; ; C. Rai; ;
Objectives/Scope: Hydrocarbon production from unconventional shale reservoirs is often characterized by anomalous
flow behavior that are typically not addressed with commercially available modeling and forecasting tools.
Compositional changes with periodic sampling of the produced fluid is a commonly reported observation. This occurs
even when the well flowing pressures are higher than the fluid saturation pressures. While there are several plausible
explanations for this behavior, this study is the first to combine a pore-scale modeling study with experimental
verification of in-situ fractionation.
Methods/Procedures/Process: The experiment involves periodic measurement of the concentration of the various
hydrocarbon species in the produced fluid for two different shale core samples saturated with a gas condensate. The
fluid is produced by decreasing the pressure for a fixed period of time. The two samples are characterized by different
TOC values of 1.3 wt% and 7.2 wt% respectively. The molecular simulation (MD) study mimics kerogen and models
multicomponent adsorption and non-equilibrium fluid movement.
Results/Observations/Conclusions: The low TOC sample did not show any anomalous behavior in terms of fluid
compositions. The high TOC sample however produced a significantly leaner fluid than the original reservoir fluid
although the samples were produced above the condensate dew-point pressure. The in-situ fractionation was
observed at pressures over a 1000 psi above the dew-point pressure. This result mirrors field observations from wells
in liquids-rich shale resource plays. The non-equilibrium molecular dynamics study echoes the experimental
observations and shows that heavier components are transported relatively slower in comparison to lighter species
due to molecular sieving, varying diffusion coefficients and multicomponent adsorption.
Applications/Significance/Novelty: This study is the first study that combines experimental verification of field
observations and modeling studies that indicate in-situ fractionation and progressive changes in produced fluid
compositions. The in-situ fractionation results in lower recoveries of the intermediate and heavier hydrocarbon
species. Therefore, EOR candidate selection is simplified by the identification of wells producing at a lower CGR in the
vicinity of high CGR wells. Additionally, the remaining reservoir fluid has a composition different than the original
reservoir fluid and merits careful selection of a suitable solvent based on the residual fluid composition and rock-fluid
interactions. This study provides fundamental insights and verifies commonly observed aspects of gas condensate
wells in shales.
Minimum Miscibility Pressure Calculation for Oil Shale and Tight Reservoirs With Large Gas-Oil Capillary Pressure
1 1 2 3
K. Zhang; ; B. Nojabaei; ; K. Ahmadi; ; R. Johns; ;
1. Virginia Polytechnic Institute and State University, Blacksburg, VA, United States.
2. Pometis Technology, Houston, TX, United States.
3. Pennsylvania State University, University Park, PA, United States.
Objectives/Scope:
Shale and tight rocks are associated with tiny pore throats, on the order of nanometers, and subsequently large
capillary pressure. The calculation of the minimum miscibility pressure (MMP) in nanopore space is complex because
the phase compositions from flash calculations are affected by capillary pressure. This paper examines the effect of
capillary pressure on the calculation of MMP using cubic equation-of-state (EOS) and three techniques: the method of
characteristics (MOC), multiple mixing cell method, and slimtube simulation.
We used three techniques of method of characteristics, multiple mixing cells and slimtube simulation with capillary
pressure, to estimate MMP in nanosized pores of shale and tight rocks. Ternary and quaternary mixtures of
hydrocarbons and real reservoir fluid from the Bakken and Eagle Ford formations are considered. Using MOC,
capillary pressure changes both liquid and vapor compositions and alters the tie lines. The reason for the change in
the MMP is illustrated graphically with ternary and quaternary diagrams. The modified mixing-cell as well as a slimtube
simulation tool are used to estimate MMPs of CO2 with Bakken and Eagle Ford oil. We use an upgraded flash
calculation in the mixing-cell procedure to estimate MMPs with large capillary pressures for real reservoir fluids.
The results show that high capillary pressure changes liquid and vapor phase compositions significantly. This change
tends to decrease the CO MMP. For displacement of an oil consisting of C , C and C by CO , the MMP
2 1 4 10 2
decreases by about 10% for capillary pressure corresponding to a pore size of 10 nm. The MMP decreases more if
the fluid is initially close to the bubble point pressure and the displacement is a combined condensing/vaporizing (CV)
drive. The results from multiple mixing cell show that for Eagle Ford oil, the CO MMP is reduced when the effect of
2
capillary pressure is accounted in phase behavior. The results from our slimtube simulations also agree well with the
MMPs calculated from the method of characteristics and mixing cell.
Owing to poor primary recovery, gas injection through multistage hydraulically fractured horizontal wells is being used
by the operators to increase recovery. MMP is a key parameter governing gas floods. Understanding the impact of tiny
pore throats on PVT properties and the resulting MMP calculations is important in the design of gas floods in very tight
rocks. To the best of our knowledge, this is the first time that multi-contact MMP is estimated using the methods that
account for capillary pressure effect on phase behavior. We demonstrate the reason for the reduction in MMP with
capillary pressure conclusively for condensing/vaporizing displacements, and that MMP does not change for
vaporizing drives.
Solution Gas Drive in Tight Oil Reservoirs: New Insights From Capillary Evaporation Experiments
1 1 2 2
E. Barsotti; ; M. Piri; ; J. Chen; ; S. Althaus; ;
1. Petroleum Engineering, University of Wyoming, Laramie, WY, United States.
2. Aramco Services: Aramco Research Center-Houston, Houston, TX, United States.
Objectives/Scope:
Current PVT analysis used in developing tight oil reservoirs largely omits the impact of nanopores which may
drastically alter the phase behavior of confined fluids. Nanopores constitute a majority of the porosity in tight
formations. Thus, phase diagrams that overlook their effects may introduce significant uncertainties into reservoir
development strategies and production forecasts. This study investigates the effect of nanoconfinement in predicting
the onset of gas exsolution in the design of solution gas drive processes for further mobilization of oil from previously
fractured reservoirs.
Methods/Procedures/Process: To probe the mechanisms of gas exsolution in nanopores, we performed capillary
evaporation experiments utilizing a novel gravimetric apparatus. First, three holders, packed with silica nanoporous
powders, were fully saturated with a liquid hydrocarbon mixture containing n-decane and propane. The mediums were
then subjected to a pressure reduction procedure, using varying rates, until the fluid in the nanopores exsoluted gas.
To determine the phase transition of the fluid, we used an ultra-high-resolution balance while monitoring the
composition of the fluid with a gas chromatograph. The nanoporous powders were custom-synthesized to establish
narrow pore size distributions such that the three mediums were characterized by pore diameters of 2.78 nm, 3.70
nm, and 6.32 nm.
Results/Observations/Conclusions: The results provide insight into the fundamental mechanisms of solution gas drive
in ultra-tight oil reservoirs. Of particular interest is comparison of the measured phase transitions and fluid
compositions from the experiments involving different rates of pressure depletion as they relate to different production
rates from tight reservoirs. These changes have direct implications for not only the amount of oil produced but also the
composition of the produced fractions owing to the fact that nanopore confinement can alter the phase behavior of the
fluids as well as their compositions as they are produced.
Applications/Significance/Novelty: Our results highlight the contribution of our methodology to advance more realistic
studies for the investigation into the combined effects of confinement, wettability, and fluid chemistry on gas exsolution
from oil in ultra-tight reservoir samples under application relevant conditions. This methodology progresses the
accuracy of PVT analysis for use in unconventional reservoirs.
Flow Behavior From Organic- and Mineral-Hosted Porosity Systems—From Pores to Production
1 1 2
R. M. MacDonald; ; S. Geetan; ; D. Klemin; ;
1. EP Energy Corp, Houston, TX, United States.
Objectives/Scope: The Eagle Ford Formation source rocks consist of multiple porosity types: organic, mineral, and a
combination of the two. Flow observed at the well level is a summation of influxes from billions of these diverse pores.
Using density functional hydrodynamics implemented in the Direct HydroDynamic (DHD) simulator, we previously
modeled fluid flow only from organic pore spaces within the Eagle Ford. We continue our study to expand our
understanding beyond just organic pores and have modeled flow through organic, mineral, and fracture porosity, as
well as a combination of all three pore types.
Methods/Procedures/Process: The DHD simulator performs direct numerical simulations of multiphase transport with
allowances for phase transitions and nanoscale phenomena. The simulator was used to model gas expulsion drive
occurring in the organic pores and to calculate theoretical recovery. The DHD simulator was also used to gain insight
into enhanced oil recovery (EOR) scenarios by investigating the behavior of organic pore systems during gas
injection. Simulations showed good uplift in recovery due to gas injection. The effectiveness of the EOR injectant, is a
function of the in-situ oil. In cyclic gas injection (huff-n-puff), oil release increases as the number of injection-
production cycles increases. The theoretical maximum EOR recovery approaches 30% for well-constructed injection
scenarios
Results/Observations/Conclusions: Expanding the organic pore studies to include the interaction with mineral and
combined pores provides insight into the dynamic behavior of Eagle Ford Formation source rocks. We now are able to
demonstrate the connection between various porosity types and the dynamic fluid flow behavior as a function of pore
pressure. The outcome of the simulation represents dynamical hydrocarbon release data (including release rate of
chemical components of studied fluid) and gas/oil ratio (GOR) data as a function of pore pressure. It shows dynamical
flow behavior within the source rock during initial and EOR production, and provides insight on the pore-scale
mechanisms that are affecting release.
Applications/Significance/Novelty: Predicting unconventional well performance is problematic. In decline curve
analysis (DCA), a changing “b” factor with production is a commonplace. In rate transient analysis (RTA), multiple
models are often needed to match early and late time performance: the tale of two wells. The GOR behavior is
anything but conventional. DHD simulation data provides insight into understanding these production mechanisms and
can serve as the basis for laboratory experimentation.
Experimental and Theoretical Investigation of the Confinement Effect on Gas Properties in Nano-scale Porous Media
1 1
S. Salahshoor; ; M. Fahs; ;
1. Petroleum & Geological Engineering, University of Oklahoma, Norman, OK, United States.
Objectives/Scope: The shift in critical properties of the confined gas has been explored in many studies through
several mathematical methods, from the modification of the existing models to developing new ones. All these studies
confirm that the properties of the confined gas inside nanopores are altered from the properties of the bulk gas.
However, limited experimental data are available in the literature to evaluate and validate these mathematical models
and predictions. The present study provides a deep experimental insight into this effect for gases entrapped inside
minute pores of nanoscale porous media.
Methods/Procedures/Process: A semi-automated apparatus to monitor and compare the behavior of gases confined in
nanopores and the bulk gases is used in this study. A well-packed bed of barium-titanate nanoparticles with pore sizes
of fifty to a few nanometers is synthesized in the laboratory to represent a tight formation. While heating up through an
isochoric process, pressure responses of the gas inside this nanoporous bed are obtained simultaneously with the
pressure responses of the bulk sample that is undergoing the exact same process. Pressure-temperature
measurements are then used to demonstrate the deviation of the gas behavior inside the nanopores. These
experimental data are utilized to propose an accurate correlation for gas compressibility factor as a predictor of
confinement effect on gas properties.
Results/Observations/Conclusions: Various modifications on equations of state from the literature are investigated
based on the experimental results to evaluate their accuracy in estimating the pore size effect on gas behavior and
illustrate their deficiency in capturing the extent of this effect. A new correlation for gas compressibility factor is
proposed as an efficient way to incorporate the confinement effect on gas properties into real gas equations. Both
hydrocarbon and non-hydrocarbon gases, including methane, ethane, nitrogen, and helium are used in exploring this
effect over a wide range of pore sizes to provide a systematical analysis of the effect of pore size on gas behavior for
various gases. Reproducibility of the data is examined thoroughly and uncertainty analysis for the entire experiment is
included.
Applications/Significance/Novelty: This study shows that the gas compressibility factor is the most promising way to
capture the effect of pore size and elevated molecule-wall interactions on the gas behavior and is a practical
parameter for experimental quantifications. The modified correlation for the compressibility factor encompasses all
other consequences of the pore size reduction, including changes in critical properties and density variations, over a
wide range of pressures and temperatures. The data and analysis presented in this study provide an alternative
approach to capture the pore size effect on confined gases that is anchored in experimental findings, also help with
understanding existing models and evaluating the accuracy of mathematical predictions in determining the confined
fluid behavior
The Effect of Mineral Composition on Shale Oil Recovery
1 1
T. Hoffman; ; A. Fakhry; ;
1. Petroleum Eng., Montana Tech, Butte, MT, United States.
Objectives/Scope: Identifying contact angles in porous media is essential for characterizing multiphase flow of fluids in
reservoirs. Traditional methods to measure contact angles assume a homogenous structure of reservoir rock;
however, microscale pictures by Scanning Electron Microscopes (SEM) show that rock composition varies even inside
a single pore. In addition, the preferentiality of oil layer formation is different according to the minerals constituting the
reservoir rock. As a result, contact angles have heterogeneous behavior at the pore-scale.
Methods/Procedures/Process: For the purpose of this research, contact angles are measured on the pure minerals
that make up the main components of a Bakken reservoir rock. Investigations of different minerals show that each
mineral has a different contact angle from the other minerals at the same medium properties. Altering medium
properties, such as salinity, also shows different contact angle behavior according to the mineral tested. However,
these separate contact angles alone do not explain how this heterogeneous mineral composition would affect the
reservoir. Therefore, a pore scale network model is utilized to study the effect of heterogeneity of mineralogical
content on oil recovery.
Results/Observations/Conclusions: The model is built to simulate the flow of fluids in a mineralogically heterogeneous
pore network system. It incorporates the contact angles measured on the pure components; thus, fractional wettability
caused by the changes in mineral composition is integrated into the model. The most practical benefit of the model is
that it was used to calculate relative permeability and capillary pressure curves, which are difficult to obtain for
Applications/Significance/Novelty: In general, this research gives more attention to mineral properties of reservoir rock
which leads to a better understanding and characterization of the reservoir. It became clear that the mineralogical
content plays a significant role in low salinity flooding where contact angles generally decreases with water compared
to contact angles measured in brine. These results will lead to better understanding of heterogeneous reservoir
behavior, as well as the effect of Improved Oil Recovery (IOR) projects, such as waterflooding, on oil production in
shale formations.
Determining the Impact of Mineralogy Composition for Multiphase Flow Through Hydraulically Induced Fractures
1 1, 2
J. E. Santos; ; C. J. Landry; ; M. Prodanović; 1, 2;
1. Hildebrand Department of Petroleum and Geosystems Engineering, University of Texas at Austin, Austin, TX,
United States.
2. Center for Petroleum and Geosystems Engineering, University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: The objective of this work is to correlate the reduction of oil relative permeability in multiphase flow
through hydraulically induced fractures with different mineral arrangements and surface roughness. High permeability
channels are created by hydraulic fracturing to increase production; however, the reduction in oil mobility in these
fractures as a result of wettability alteration has received limited attention. Since analytical solutions to quantify this
problem work only for simple geometries, we propose direct simulations based on realistic fractures to quantify their
flow properties.
Methods/Procedures/Process: Utilizing the Shan-Chen multiphase model of the lattice-Boltzmann method (LBM) we
can simulate a pair of fluids that honor the behavior of oil and water in the subsurface. A surface interaction parameter
is adjusted to mimic the time-dependent microscale wettability of the different minerals present in fractures. The
domains were extracted from 3D micro-CT scans and digitally synthesized fractures. We then map the mineralogy of
surfaces obtained by SEM imaging and integrate them with contact angle measurements on individual minerals to
provide input for the simulations. We also account for the possible change in contact angle over time due to wettability
alteration. The simulations are processed to obtain the absolute and relative permeability. These were carried out at
the TACC cluster.
Results/Observations/Conclusions: Published analytical solutions for multiphase flow properties of fractures fail to
account the complexity of the surface wetting heterogeneity and its effect on production from fractured reservoirs.
Based on our results, it’s been shown that the wettability and surface roughness could significantly decrease the oil
flow through fractures. The simulations quantify the effect of different mineralogy arrangements and wetting state on
the macroscopic flow. From these measurements, we derive relative permeability correlations based on fracture
aperture, surface roughness, and the spatial distribution of minerals.
Applications/Significance/Novelty: This paper provides the-first-of-a-kind workflow to integrate information from SEM
images, micro-CT scans and laboratory measurements with a direct simulation procedure to quantify multiphase flow
in rough fractures. Correlations were proposed based on different fracture apertures, surface roughness and mineral
composition to quickly assess the impact of these parameters on macroscopic flow.
Manager of Geophysics, Cimarex
1
D. Klepacki; ;
1. Cimarex, Houston, TX, United States.
Texas State Seismologist
1
A. Velasco; ;
1. State of Texas, Austin, TX, United States.
Science Writer and Reporter, Dallas Morning News
1
A. Kuchment; ;
1. Dallas Morning News, Dallas, TX, United States.
Integrating Poroelastic Effects of Wastewater Injection and Rupture Dynamics to Understand Induced Seismicity
1 1
D. Szafranski; ; B. Duan; ;
1. Department of Geology and Geophysics, Texas A&M University, College Station, TX, United States.
Objectives/Scope: The main objective of this study was to develop a tool that might be used in injection operation
planning and development of risk mitigation schemes. Moreover, we performed numerical simulations for different
scenarios to find the main physical factors leading to triggering the slip on preexisting faults that might produce
significant induced seismic events. Finally, we also studied rupture propagation patterns and their effect on the ground
motion recorded on the Earth surface.
Methods/Procedures/Process: In order to simulate the effect of water injection into the subsurface on the stress state
and natural fault activation, we use linear theory of poroelasticity. To get a basic insight into the sensitivity of the
solution to the variability of different properties of the rocks and injected fluid, the geometry and the parameters of the
injection operation, we started investigations with a simplified homogeneous, isotropic, 3D model and analytic
solutions of the poroelasticity problem for a point source and a constant injection rate. We combined this with
numerical simulations of dynamic rupture propagation via an explicit finite element method, treating the faulting
boundary with the traction-at-split-node scheme to study rupturing process of the triggered earthquakes.
Results/Observations/Conclusions: Results of our simulations show that dependent on relative position between
injection well and fault plane as well as rocks and fluid properties, we observe first seismic event after different time
period. Continuing the injection, the first event is then followed by increasingly more energetic earthquakes. The
highest impact on the timing of first slip and the released seismic moment by consecutive events is imposed by shear
modulus and permeability of the formation. We also observed that for stronger events, the ruptured fault area is limited
to a region of elevated ratio of shear and effective normal stresses, and rupture propagation is halted outside of this
zone. This study lays a solid base for the investigation of faults behavior in more complicated cases.
Applications/Significance/Novelty: Understanding of stress and pore pressure field development with time and its
effect on the fault activation helps in development of schemes for seismic risk mitigation, which is especially important
in the areas where significantly increased seismic activity was observed (e.g. Oklahoma). Combining poroelastic
effects due to water injection with dynamic rupture simulation allows us to study in details induced earthquake source
physics and predict potential ground motions on the surface of the Earth. It also provides a tool to estimate possibility
of fault activation given a certain set of medium and injected fluid properties as well as injection scheme. Finally, the
method helps to assess how uncertainty of a given medium parameter affects the possibility of strong earthquake
triggering.
Source Mechanisms of Hydraulic-Fracturing Induced Event Sequences in the Fox Creek Area
1 1
H. Zhang; ; D. W. Eaton; ;
1. Department of Geoscience, University of Calgary, Calgary, AB, Canada.
Objectives/Scope: The objective of this study is to undertake a detailed investigation of 17 earthquakes with
magnitudes M > 1.5 that occurred during a multiwell hydraulic-fracturing completion program west of Fox Creek,
Alberta in 2016. By using microseismic data recorded by a dense shallow well array, we apply P-wave polarity-based
focal-mechanism inversion and 3C P-wave amplitude-based moment-tensor inversion to the 17 induced events. To
provide insights into the geomechanical significance of the events, comparisons between the two sets of inversion
results are made.
Methods/Procedures/Process: The focal-mechanism inversion is based on the polarities of observed P-wave first
motions on the vertical component. In the inversion, the focal mechanism is estimated using a grid search method that
minimizes the misfit between predicted and observed polarities based on the theoretical radiation pattern for a double-
couple source.
We performed the moment-tensor inversion based on the 3C amplitudes of the direct P-wave in a 1-D layered
isotropic medium. The relationship between observed data and moment-tensor elements can be represented in matrix
form (Aki and Richards, 2002). For this overdetermined problem, 6 independent components of moment tensor are
retrieved by a generalized least-squares inversion that minimizes the misfit between observed and predicted 3C
amplitudes.
Results/Observations/Conclusions: Two groups of source mechanisms were identified: group I, predominantly double-
couple (DC) N-S or E-W trending strike-slip events; and group II, a set of events exhibiting significant non-DC
components (up to >60%) with NE-SW trending strike-slip (with low dip angle) or NW-SE trending dip-slip (thrust-
faulting with high dip angle). For group I, nodal planes of the two sets of inversion results are similar. For group II, the
nodal plane solutions exhibit slightly larger difference between the two sets of results.
In addition, the relative locations of the events within each group can provide aid to choose the real rupture plane from
the two inverted nodal planes.
Applications/Significance/Novelty: Understanding the source mechanisms of induced events can provide aid in
estimating time-dependent seismic hazard in areas where large-scale fluid injection is ongoing.
The source mechanisms can provide information about the fault planes, and the non-DC components contain
information about complex source characteristics suchs as multiple intersecting fractures, dilatant jogs create at the
overlapping areas of multiple fractures, or non-planar fault geometry.
Case Study: Fault Slip and Casing Deformation Induced by Hydraulic Fracturing in Sichuan Basin
4 1 2 3
M. Zoback; ; Z. Chen; ; L. Zhou; ; R. Walsh; ;
1. CNPC Drilling Research Institute, Beijing, China.
2. Petrochina Southwest Oil&Gas Field Company, Chengdu, China.
3. Now at Decision Geomechanics LLC, San Fransico, CA, United States.
4. Stanford Univeisity, Palo Alto, CA, United States.
Objectives/Scope: As of March 2016, casing deformation occurred in 32 of 101 wells during fracturing in Changning
and Weiyuan shale gas blocks in Southwest China. The analysis of data shows faults at locations of sheared casing,
suggesting that hydraulic fracturing induced fault shear is likely the cause of casing deformation. Here we build a
probabilistic geomechanical model in which we observe that the sheared fault had a 75% probability of doing so under
the bottom hole pressure it was exposed to during stimulation. We then interpret other faults and assess their Fault
Slip Potential within the same model.
Methods/Procedures/Process: The deformed casing location is almost in the fault zone, which indicates that casing
deformation was caused by hydraulic fracturing induced fault slip. From the microseismic monitoring data the dip
direction and dip of the fault was fit as N57°E and 70°respectively. Image and wire line log data, and minifrac test data
from this and nearby wells are analyzed to set up geomechanial model parameter distributions. Using these
parameters, Mohr-Coulomb fault activation was explored. A Quantitative Risk Analysis was undertaken to analyze the
probability of induced fault slip and the sensitivity of the slip to model parameters. Lastly this analysis was repeated for
each fault interpreted from the ant tracked image, colored by its potential for induced slip under the same pressure.
Results/Observations/Conclusions: The results show that in the current in-situ stress conditions, faults in the block are
at critical stress states and easily activated. By propagating uncertainties in the geomechanical model as uniform
distributions through Monte Carlo modeling, we observe with hindsight that the fault that sheared had a 75%
probability of doing so with 2,500 PSI added to it. Applying this analysis to mapped faults shows which segments
would ideally not be intersected by future production wells (red) and which pose less risk (green). The uncertainty
analysis can also be used to prioritize gathering further information, by focusing on the parameters that the answer is
most sensitive to, in this case Shmin and natural pore pressure.
Applications/Significance/Novelty: The procedure will provide method to understand the cause the casing deformation
and some measures of mitigation of fault activation.
Statistical Controls on Induced Seismicity: A Physics-based Data Mining Approach Integrating Seismic and Well Data
1 1 1 1
S. Sinha; ; Y. Wen; ; R. A. Pires de Lima; ; K. Marfurt; ;
1. University of Oklahoma, Norman , OK, United States.
Objectives/Scope: In this paper we have quantified the effect of different variables such as injection pressure,
geomechanics, local geology etc. on induced seismicity in Oklahoma. For the study, we have used seismic attributes
such as curvature and coherence to identify the geological controls. For geomechanics, we have carried out a
geostatistical seismic inversion based model. We have used the public information database to extract the properties
of injection wells such as pressure, rates etc.
Methods/Procedures/Process: We use geomechanical parameters from a geostatistical seismic inversion and extract
average properties of Poisson’s ratio and Young’s modulus using an in-house developed trace averaging algorithm
around the injection wells.
We use the pressure transient analysis on injection wells to extract permeability. We obtain porosity from well logs.
Other inputs include seismic attributes and well injection pressure, rates and Euclidean distance from the mapped
fault planes.
We use a machine learning workflow including principal component analysis, factor analysis, feature construction and
prediction algorithms such as support vector machines, neural networks etc. to predict the induced seismicity. Our
typical outputs consist of earthquake magnitude and frequency.
Results/Observations/Conclusions: We observed including geomechanical parameters in the study improved the
prediction of earthquakes much better than without geomechanics as documented by previous studies. We obtained a
prediction rate of approximately ~ 85% in our study on the earthquake events.
Feature construction play an important role as normalizing the rates by some of the pressure transient variables
improved our results. In our study support vector machines constantly performed better than its counterparts such as
neural networks. Then, we perform sensitivities based on our models to predict thresholds on different variables to
avoid induced earthquakes.
Applications/Significance/Novelty: In our study we have established a physics based statistical relationship between
human induced seismicity and the key controls such as injection rates, water salinity etc. Our work can be utilize to
develop robust water disposal well planning to avoid induced seismicity.
Modeling of Azimuthal Gamma Ray Tools for Use in Geosteering Unconventional Reservoirs
1 1 1 1
H. Wang; ; E. Stockhausen; ; D. Wyatt; ; D. Gulick; ;
1. Chevron U.S.A., Inc., Houston, TX, United States.
Objectives/Scope: More and more these days, horizontal well targets in unconventional reservoirs are being designed
to stay within a narrow stratigraphic zone. Some target windows are on the order of 10-foot thick. These require
geosteering. It can be quite challenging to stay inside the target with a standard omni-directional gamma ray tool. The
lack of azimuthal sensitivity in such tools make it difficult to determine whether the wellbore is approaching the top or
bottom of the target window during and even after drilling the lateral. Thus, geosteerers often fail to keep the well path
within the target zone.
Methods/Procedures/Process: To help solve this problem, the industry is just beginning to turn to azimuthal GR tools
to assist in determining the relative stratigraphic position of the drilling assembly as it cuts up-section or down-section
when approaching or crossing bed boundaries. In real-time, only up and down looking azimuthal GR curves are
available from some service providers, while others have real-time multi-sector data and some can even provide real-
time GR images.
Not all azimuthal GR tools are created equal. There are a large variety of commercial tools available to choose from.
Results/Observations/Conclusions: One can find azimuthal GR (AGR) tools in almost any section of the bottom-hole
assembly (BHA), such as contained within the drilling rotary steerable system (at- or near-bit), measurement-while-
drilling (MWD), or logging-while-drilling (LWD) system. At- or near-bit AGR sensors are only a few feet from the drill bit
for early decision making. MWD AGR sensors are mounted close to the cylindrical axis (center) of drilling collar for
more robustness and lower cost. LWD AGR sensors are mounted further off-axis into the body of drilling collar to be
close to the formation and gain higher sensitivity. LWD AGR tools can have one to four sensors. When there is more
than one, sensors are mounted evenly around the circular body of drilling collar.
Applications/Significance/Novelty: Detailed tool design, such as the amount of shielding materials and the placement
of GR sensor within the tool cross section, can lead to different tool performance and azimuthal sensitivity. Here we
compare the expected performance of different AGR tool concepts based on Monte Carlo nuclear modeling. In
addition, high quality gamma ray images can be used for dip picking if provided with an accurate depth of image. We
will discuss a method for picking apparent dips by converting the gamma ray images into true-vertical-thickness
reference space, with case studies from several unconventional reservoirs.
Monitoring Wellbore Quality in Real-Time Using a Geometrically Derived Tortuosity Metric
1 1 1 2 2 2 2
J. D'Angelo; ; P. Ashok; ; E. van Oort; ; M. Shahri; ; T. Thetford; ; B. Nelson; ; M. Behounek; ;
1. University of Texas, Austin, TX, United States.
2. Apache Corp, Houston, TX, United States.
Objectives/Scope: Poor wellbore quality can negatively affect operations from drilling to production in unconventional
wells. Tortuosity is a geometric concept that provides an indicator of wellbore quality. Several indices exist for
quantifying tortuosity, yet none of them capture the overall tortuosity of the well. Building on prior work, a previously
published original tortuosity index (OTI) was modified to better characterize the overall tortuosity of a well in real-time.
This paper documents the modifications and field verification relating to this new index, which has now been applied
on 20 drilling rigs.
Methods/Procedures/Process: The OTI characterized tortuosity only for independent well sections and could not
distinguish between planned and unplanned tortuosity. Dividing the well into sections makes real-time decision-
making for directional drilling difficult because overall tortuosity effect on the well was unclear. Handling planned
trajectory features such as build sections in the same way as unplanned features such as doglegs gave misleading
information on well risks. Therefore, the OTI was modified (yielding an MTI) to analyze the entire well and to
distinguish planned from unplanned tortuosity. Objective checks were made to test the validity and gauge the
performance of the new index, including comparing MTI values to torque and drag (TD) values and trends for both
artificial and actual wellbore trajectories.
Results/Observations/Conclusions: Both the OTI and MTI were compared on two sets of unconventional wells. One
set was generated artificially to make tortuosity analysis obvious. The planned well had no tortuosity and the “as-
drilled” well was literally the planned well with tortuosity superimposed onto it. Both the OTI and MTI performed in a
similar manner for these tests and agreed with intuitive and TD ranking of the wells. The second set consisted of real
well trajectories made available by unconventional operators. For wells where most of the tortuosity was planned, the
OTI failed to properly indicate TD trends. The MTI, on the other hand, was able to distinguish between planned and
unplanned tortuosities, correlated strongly with TD trends, and performed superior at ranking wellbore quality.
Applications/Significance/Novelty: The significance of the MTI is that it provides an overall representation of the
wellbore quality without the costs of complex modeling or additional sensors. The MTI can be used for proactive
decision making while drilling to increase wellbore quality and decrease the likelihood of problems such as high torque
and drag, poor casing runs, poor zonal isolation, and hydraulic fracture reorientation. To make this representation
useable, the MTI has been mapped to ranges of acceptable and unacceptable TI values. This was done by comparing
the TI’s of several wells with corresponding torque and drag analyses and historical records. The MTI algorithm has
already been converted to a real-time program and implemented on 20 U.S. drilling rigs.
Cement Placement in Severe Doglegs and Its Impact on Well Integrity: A Numerical Assessment
1 1
A. Dahi Taleghani; ; H. Yu; ;
1. Energy and Mineral Engineering, Pennsylvania State University, State College, PA, United States.
Objectives/Scope: Horizontal drilling with hydraulic fracturing technology has played a critical role in shale revolution,
while the integrity of the drilled wells have been a concern in some parts of the world. In the dogleg section of these
directional wells, stiff straight casing strings should deform to fit the borehole trajectory, which lead to its complicated
contact geometry in the hole. The severity of dogleg and casing properties may make the formed annulus space so
complex that prevent full placement of the cement around the casing, which can endanger well integrity in some
occasions.
Methods/Procedures/Process: We developed a 3D large deformation finite element model (FEM) to simulate the
running of a casing string into the dogleg. The deformation of casing in presence and absence of friction reducer
additives have been calculated by this model to determine the annulus geometry supposed to be cemented. Then, we
utilized a computational fluid dynamics (CFD) model to simulate cement displacement. Different cement and casing
specifications are considered to investigate the impact of incomplete and asymmetric cement placements on fluid
migration and casing failure during hydraulic fracturing and production. Cohesive zone methods (CZM) is used to
simulate the bonding between the casing and the cement sheath in different well integrity scenarios after cement
placement.
Results/Observations/Conclusions: The FEM results show the complicated contact geometry of casing and borehole
at dogleg, which may worsen in casing strings with lower thickness or large curvature doglegs and show the
magnitude of stress in casing can dramatically change by friction. Complicated geometry of the annulus made the
displacement of drilling mud by cement in the dogleg very difficult. For slugging cement flow, decreasing the pumping
rate or increasing the effective viscosity of the displacing fluid helps development of a steady displacement interface.
An upward tending spike of cement and an azimuthal counter-current flow above or below the advancing interface
was observed along the wide side when displacing mud through dogleg. The CZM results show the impact of
incomplete cement on casing debonding initiation.
Applications/Significance/Novelty: This paper provides a comprehensive workflow to anticipate challenges for cement
placement before running the casing based on geosteering data provided by the driller. Hence, operators may predict
potential integrity issues or casing failures due to the poor cement placement and address these issues by modifying
the cement composition, rimming the dogleg or adding mud friction reducers before running the casing. The outcome
will help the oil and gas industry to prevent any potential damage that might happen due to fracturing treatment
through casing by minimum costs and avoid these integrity problems by using packer and tubing setups or more
sophisticated cement recipes.
Impact of Cyclic Pressure Loading on Well Integrity in Multi-Stage Hydraulic Fracturing
1 1 1 1
D. Barreda; ; M. Shahri; ; R. Wagner; ; G. King; ;
1. Apache Corp., Houston, TX, United States.
Objectives/Scope: During well completions in unconventional formations, hydraulic fracture stimulations commonly
exceed fifty frac stages in a single horizontal lateral. Investigation has found that repetitive pressure cycles during the
completion creates loads that may exceed the failure envelope for the casing grade, connection and cement support.
Problems encountered during this type of event have ranged from casing shape distortion to rare cases of casing
failure within the pay-zone lateral.
Methods/Procedures/Process: Although high pressure pumping, large sand and water volumes and numerous
fracturing stages are an accepted stimulation design in successful unconventional resource development, some well
designs based on a single fracture treatment may not be adequate for repetitive cycles of pressure and erosion
exposure that is now common within the industry.
Results/Observations/Conclusions: In this study, casing failures were identified both during and immediately after
fracturing as a small subset within the group of successful stimulations without failures. Failure events were studied to
determine causes and differences in design or application conditions as compared with the same factors in successful
wells. We have investigated the loads applied on the wellbore; specifically in the curve section above the heel of the
lateral, during multi-stage fracturing operation and analyze them using a fatigue model. The effect of cyclic loading on
casing body, couplings and cement is investigated. A parametric study was performed to evaluate and determine the
important factors controlling the overall well integrity and potential actions to reduce impact of cyclic load application.
Applications/Significance/Novelty: Using a more comprehensive design methodology, changes to the well design,
completion application process and fracture pumping schedule are proposed to minimize the risk of casing failure
while reaching stimulation goals.
Diagnostic Applications of Borehole Hydraulic Signal Processing
1 1 2
C. J. Clark; ; J. L. Miskimins; ; D. L. Gallegos; ;
1. Petroleum Engineering, Colorado School of Mines, Golden, CO, United States.
2. Geology & Geophysics, Texas A&M University, College Station, TX, United States.
Objectives/Scope: To achieve high stimulation efficiency in horizontal multi-stage fracturing treatments, it is important
to ensure that stage-to-stage isolation is maintained. However, this is often not confirmed until after the job is
completed and the isolating bridge plugs are tagged during drillout operations, preventing the operator from identifying
failures in real-time. By applying signal processing techniques to post-fracturing “water hammer” events, the location
of bridge plugs in a wellbore is accurately and simply measured, and inefficiencies due to tool failures can be
effectively minimized.
Methods/Procedures/Process: A “water hammer” is a hydraulic impulse created due to the shutdown of pumps at any
point during fracturing operations. These hydraulic impulses oscillate between the wellhead and the isolating tools at
the resonant frequencies of the wellbore, which are controlled primarily by hydraulic length. In this work, a signal
processing technique called the Fourier transform is applied to pressure data collected at the wellhead, to identify the
contributions to pressure oscillations at the wellhead on the frequency spectrum. The resonant frequencies of the
wellbore are then identified as strong peaks in the frequency spectrum. By understanding the boundary conditions that
exist downhole, the resonant frequencies can be normalized to accurately determine depth of isolation tools.
Results/Observations/Conclusions: The methods were applied to 27 stages in five separate horizontal wells in the
Denver-Julesburg basin. The predicted depth of the isolating plugs was identified and indicated an average error of
26’ off of the depth the plugs were set at according to the wireline counter. The maximum deviation between the
hydraulic impulse measurement and the setting depth of a bridge plug was 137’, and was verified as an observed tool
failure during drillout operations when no tag was recorded at the setting depth. Studies are continuing to constrain the
limitations of hydraulic impulse measurements, but on the sample set it proved to be an accurate and reliable method
of confirming isolation at the end of a fracturing stage.
Applications/Significance/Novelty: With increasing interest in “engineered completions” and the related treatment
designs, any failure in bridge plug isolation can be extremely detrimental. There are currently no methods on the
market to verify isolation tool depth within the wellbore after a stage has been pumped that do not require additional
time, equipment, or well entry. Using hydraulic signal diagnostics, tool failure conditions can be verified in the field and
inefficiencies mitigated. Further applications include inexpensive field testing of dissolvable bridge plugs, which are
becoming increasingly important as the industry makes extended laterals the new norm. With additional study, these
methods might also be used for reliable detection of sealing ball action, or live depth verification during wireline runs.
Enhancing Placement of Microproppant in Microfractures for Increasing Stimulated Reservoir Volume in Shale
Reservoirs
1 1 1 1
L. Xu; ; K. He; ; P. Lord; ; P. Nguyen; ;
1. Halliburton, Houston, TX, United States.
Objectives/Scope: Based on microseismic data, only a small fraction of fractures in a complex fracture network will be
propped, despite injecting large quantities of small-diameter fracturing sand per fracturing stage. Because most
natural and induced fractures have narrow gaps much smaller than 50 microns, current small-sized sand cannot
access these small microfractures. This paper presents a method for enhancing placement of microsized proppant
(MP) into microfractures to help ensure they remain open for enhancing production of hydrocarbons.
Methods/Procedures/Process: Microfluidic-based rock-on-a-chip cells containing various configurations of a single
main channel (~50 microns) simulating a dominant fracture connecting with microchannels that have channel sizes in
the range of 7 to 8 microns were used to evaluate fluid injection through microfractures. Flow testing was performed
by injecting fluid compositions containing a low concentration of MP with a D50 of ~5 microns, with and without a
weakly emulsifying surfactant, to determine the impact of surfactants for enhancing placement of MP and its transport
behaviors into the microchannels. The main channel comes with various configurations such as S, H or L shapes. And
the secondary channels are typically placed perpendicular to the main channel.
Results/Observations/Conclusions: Inclusion of surfactant showed that in addition to lowering surface tension it also
improved suspension of MP in fracturing fluids, thereby allowing more successful migration of MP into these
microstructures. However, at a similar injection rate but without containing surfactant, MP tended to form aggregates
near or at the entrances of microchannels, hence limiting placement of microparticles deep inside the narrow
channels. Field treatments employing MP introduce it in the pad fluid stages at a low particle concentration to treat
natural and induced secondary fractures in the far-field regions, followed by injection of larger-sized fracturing sand
(100-mesh and larger) to prop the primary fractures as part of multistage hydraulic fracturing treatments in the Barnett
and Woodford shales.
Applications/Significance/Novelty: By adjusting the interfacial tension (IFT) and optimizing surfactant and MP particle
interaction, application of surfactant can help to enhance imbibition of fracturing fluids into the shale matrix, altering
the stress load of shale fabric structures to cause stress and tensile failures, thereby inducing generation of
microfractures. Combining a weakly emulsifying surfactant and MP in a single treatment could enhance generation of
microfractures while helping ensure these microstructures remain open to maintain well production.
Is Conductivity Still Important in Unconventional Reservoirs? A Field Data Review
1 1 1 2
R. Shelley; ; B. Davidson; ; K. Shah; ; T. Palish; ;
1. StrataGen, Houston, TX, United States.
2. CARBO Ceramics, Houston, TX, United States.
Objectives/Scope: The science of hydraulic fracturing (HF) has long understood the impact of fracture conductivity on
well success. As HF was applied to unconventional reservoir development, increasing contact area was deemed a
first-order priority. While conductivity still remained important, the downturn in the industry led many operators to
replace premium (high conductivity) designs with large volume designs employing low quality (conductivity) proppant.
These high contact area designs have now seen 2-3 years of production results that can now be analyzed and
compared to higher conductivity designs.
Methods/Procedures/Process: This paper will review the science of finding the optimal balance between reservoir
contact and fracture conductivity in the context of unconventional reservoir development. The authors will use actual
completion/fracture designs to pressure-match fracture models which are then history-matched to actual production
through the use of numeric reservoir simulators. In addition, actual field production will be compared in several
unconventional basins to validate the results found in the history-matched models. These basins include the Bakken,
Eagle Ford, Haynesville, Permian and Utica. In addition to production comparisons, economics will be determined of
each design to identify the completion designs which provide the highest return on investment.
Results/Observations/Conclusions: The results of the modeling and history match will confirm the importance of
increasing contact area, which are also confirmed in the actual field results. However, the modeling and field results
will also show that while increasing the intensity of the completions (more stages, larger proppant volumes, etc)
increases well productivity, deploying higher conductivity fracture designs for a portion of the treatment can further
increase productivity. Field data will prove that these design modifications can provide an optimal balance between
short term productivity (initial production) and long term value of the asset. In some cases, less proppant and water
can be pumped while still maintain or even increasing productivity. These combined factors will show a higher return
on investment.
Applications/Significance/Novelty: Operators continue to look for ways to maximize the value of their assets through
optimal completion design. Current challenges include the EHS impacts of larger proppant volumes and increased
water usage and handling, as well as stemming the significant declines seen in today’s unconventional developments.
In most cases more efficient fracture designs can lead to increased drainage areas and the potential to decrease the
number of wells required to drain a section. This paper will be useful to completions and reservoir engineers as they
work to optimize their development projects.
Novel 3-D Field-scale Characterization of Reservoir Fractures Using Surface Seismic Data by the Double-Beam
Method and Field Applications
1 1
Y. Zheng; ; H. Hu; ;
1. Department of Earth and Atmospheric Sciences, University of Houston, Houston, TX, United States.
Objectives/Scope: Field-scale characterization of natural/induced fractures is critical in the reservoir development and
sweet spot identification. Commonly used fracture characterization methods are based on equivalent seismic
anisotropy, which is valid when fractures are uniformly distributed in space. However, this condition is not always
satisfied as fractures show different scales. We propose to use our new double-beam (db) method to invert for
spatially dependent fracture network information, in terms of fracture orientation, spacing and compliance.
Methods/Procedures/Process: The input for the db method is the surface seismic data (shot gathers). It creates a
focusing source beam and a focusing receiver beam from the surface to the fractured reservoir location. The
interference pattern of these two focusing beams produces the db output which includes fracture orientation, density,
and compliance as a function of space in the reservoir. The db can also distinguish multiple crosscutting fracture sets.
These information can be used to build reservoir flow model for best selection of drilling site.
Results/Observations/Conclusions: The db method has been rigorously tested blindly using 3D modeled datasets. The
db method can provide field-scale fracture distribution information with spatially dependent fracture orientation,
density, and compliance. The compliance is directly related to the flow property of the reservoir. The db method
serves a link between seismic data and reservoir modeling.
Applications/Significance/Novelty: We also apply the db method on two field datasets acquired in the Marcellus shale
and an unconventional play in Texas. The db obtained fracture distribution maps at the field scale, consistent with
wellbore image logs where such information is available.
Improving Marcellus Imaging Through the Use of FWI and Joint Tomographic Inversion for Velocity and Epsilon
1
D. McCann; ;
1. Geokinetics, Houston, TX, United States.
Objectives/Scope: Since the onset of unconventional opportunities geophysicists have been working to improve the
return on investment of those assets. Numerous authors have demonstrated the value of prestack depth migration
(PSDM) for resource plays. There are two areas that PSDM data has proven to be useful. These are in reducing
structural risk and producing superior gathers for elastic inversion. The challenge has been in the definition of the
subsurface anisotropic model. This paper addresses two novel approaches that significantly improve the quality of the
anisotropic model.
Methods/Procedures/Process: We will demonstrate the use of Full Waveform Inversion (FWI) to build the near surface
model. This tool is critical in areas of velocity reversals where other tools fail. These include areas of shallow coals or
Karsting.
To complete anisotropic model building we will demonstrate a new joint tomographic inversion tool for velocity and
epsilon. This tool performs the inversion in an uncoupled manner which provides an accurate model with fewer
iterations. The benefit is that it uses all the offsets which yield a superior velocity model while deriving an epsilon
model.
Results/Observations/Conclusions: For every asset that we work there is a time value of money. As such any
reduction in the cycle time has a direct, positive, impact on the return on investment. When we talk about cycle time
reduction there are several considerations. By estimating a better near surface velocity field with FWI we reduce
errors that would impact tomography. With a joint tomographic inversion, we can reduce the number of iterations
thereby saving up to 30% in the cycle time of a project. Using improved velocity determination tools, we can reduce
the structural and rock property risk of an unconventional asset. The key tools that yield this result are land FWI and
joint tomographic inversion.
Applications/Significance/Novelty: It is not often that we can improve our technology and reduce cycle time. The two
technologies highlighted in this paper provide for an accurate anisotropic model while reducing cycle time. The unique
workflow proves an effective means to circumvent the low frequency requirement of FWI. With an improved shallow
velocity field, we prevent shallow errors from imprinting on the target.
For many years the industry has tried to develop an effect joint tomographic inversion. By parsing the tomographic
solution, we can solve for both velocity and epsilon in a single update. This provides an improved velocity and epsilon
field while reducing cycle time.
Quantitative Interpretation Efforts in Seismic Reservoir Characterization of Utica-Point Pleasant Shale – A Case Study
1 1 2 3
S. Chopra; ; R. K. Sharma; ; H. Nemati; ; J. Keay; ;
1. Reservoir Services, Arcis Seismic Solutions/TGS, Calgary, AB, Canada.
2. Arcis Seismic Solutions, TGS, Calgary, AB, Canada.
3. TGS, Houston, Houston, TX, United States.
Objectives/Scope: In the present study, the objective was to identify the sweet-spots within the Point-Pleasant interval
in eastern Ohio using 3D seismic and other relevant data in terms of organic richness and brittleness. It was intended
to determine brittleness using the available XRD data, as well as the rock-physics parameters from the seismic
prestack impedance inversion. The organic richness was sought in terms of a TOC volume, which could correlate with
the available core-log petrophysical modeling results. Finally, the results were to be correlated with the available
production data.
Methods/Procedures/Process: We tried to bring in data from core, geochemical and geomechanical analysis, and
integrate that with surface seismic data. The density and TOC measurements made on the core samples in the Point
Pleasant interval were crossplotted to determine a relationship between them. Density was determined using
probabilistic neural network analysis (as no appropriate angle information available), and used the determined
relationship to transform it into TOC volume. The brittleness index (BI) was derived, first, using the rock-physics
parameters such as Young’s modulus and Poisson’s ratio from simultaneous inversion followed by a procedure
proposed by Grieser and Bray (2007). Second, it was derived using the mineralogical data available from XRD data
and regional petrophysical modelling.
Results/Observations/Conclusions: We have characterized the Point Pleasant formation using 3D surface and its
integration with core, geochemical and geomechanical data. This has been done by deriving rock-physics parameters
through simultaneous inversion and neural network analysis. We find that the Point Pleasant (PP) formation does not
seem to follow the commonly followed variation in terms of low Poisson’s ratio and high Young’s modulus for brittle
pockets. Instead, by restricting the values of Poisson’s ratio and examining the variation of Young’s modulus, we
could determine the brittleness behavior within the PP interval. Combining the brittleness behavior with the organic
richness determined through the TOC content, we could pick sweet spots in the Point Pleasant interval which match
the production data.
Applications/Significance/Novelty: With mineralogical BI computed, the available regional petrophysical modelling
allowed us to determine the volumes of individual minerals present in the Utica play. A mismatch between higher
values of BI and the peak initial production rate lead (PIPR) led us to conclude that a formation with higher brittleness
is not always a good candidate for fracking. Frackability index (FI) was then used to get a somewhat better idea about
favorable fracturing sweet spots. A reasonable match of higher FI values was noticed with higher PIPR which lent
confidence in our workflow. Further, TOC volume was brought into the analysis and it was found that the whole Point-
Pleasant interval could be treated as organic-rich, and thus zones with higher FI and TOC should be considered for
further development.
Applications of a Novel Hybrid Model for Unconventional Reservoirs
1 2 2 2
J. A. Ayoub; ; B. Blakey; ; S. Krishnamurthy; ; M. Thambynayagam; ;
1. J. Ayoub Consulting LLC, Houston, TX, United States.
2. Emerson Automation Solutions, Houston, TX, United States.
Objectives/Scope: The effect of the dimensions, positions and orientations of fractures on the performance of
horizontal wells is a subject of considerable interest to shale oil and gas producers. The performance of the majority of
unconventional wells is evaluated using empirical models (DCA) and trial-and-error experience rather than by
reservoir physics-based approaches. Such approaches require numerical simulation that can take considerable time.
In this paper, we illustrate use of a new generation, analytical-numerical hybrid model respects the physics but that is
much faster to setup and run.
Methods/Procedures/Process: The model uses the principles of source functions to solve analytically for complex
fracture systems and for fractures with varying conductivities. In addition, it uses an efficient numerical scheme to
handle multiphase flow, hence the use of the term hybrid. It eliminates the complexity of classic numerical reservoir
simulation and the sensitivity of results to the gridding which can be problematic when modeling features like
fractures. The model results were validated using published solutions, comparisons with numerical simulation results,
and by comparison with field data.
Results/Observations/Conclusions: Faster set-up and computing time are observed compared with numerical models.
The effect of angle between fracture planes and well axis on the productivity is presented and compared to the
orthogonal fractures case. In particular, it is shown that as neighboring wells are more closely spaced, there is a
threshold after which colinear fractures outperform orthogonal fractures. Examples with more complex fractures are
then analyzed and their impact on production quantified. Finally, field data are history matched illustrating the ease of
model calibration and use for production forecasting with a physics based model.
Applications/Significance/Novelty: A novel approach to model productivity and EUR of unconventional wells including
complex fracture configurations is illustrated with numerous examples. The advantages of this hybrid model in
comparision with common numical simulation include ease of setup, faster run times and correct treatment of the
underlying physics without gridding. The approach is in many ways superior to the common alternative: those without
experience in numerical modeling can set up and compute scenarios quickly and with greatly reduced risk of
numerical error. Applications of this new model include selection of best well placement, completion strategy and lift
strategy, and real-time comparision of measured and forecasted performance for well diagnosis.
Multiscale Modeling to Evaluate the Mechanisms Controlling CO -Based Enhanced Oil Recovery and CO Storage in
2 2
the Bakken Formation
1 1 1 1 1 1 1
J. Torres Rivero; ; L. Jin; ; N. Bosshart; ; L. J. Pekot; ; J. A. Sorensen; ; K. Peterson; ; P. Anderson; ; S.
1
B. Hawthorne; ;
1. Energy & Environmental Research Center, Grand Forks, ND, United States.
Objectives/Scope: A research project has been conducted to integrate well logs and advanced laboratory-based data
from Bakken core samples into geocellular and simulation models at multiple scales. The objective of the efforts was
to improve the accuracy of modeling approaches for predicting the potential incremental oil production through CO -
2
based enhanced oil recovery (EOR) in an unconventional tight oil reservoir and to evaluate the ability of organic-rich
shales to store and/or contain CO and possibly produce oil.
2
Methods/Procedures/Process: Unique data sets applied to the modeling included results from advanced scanning
electron microscopy techniques, computerized tomography (CT) scanning, and CO permeation and oil mobility
2
experiments performed on both reservoir rocks and source shales. Plug- and core-scale models were used to
simulate and history-match the CO permeation and oil mobilization experiments. Larger-scale models, such as near-
2
wellbore- and reservoir-scale models, were used to simulate and predict CO behavior under conditions that are more
2
representative of what might be expected in the field. Reservoir pressure, temperature, and injectivity data from a CO
2
injection field test in a Bakken well were also applied to the reservoir-scale modeling.
Results/Observations/Conclusions: The plug- and core-scale modeling efforts were able to reasonably reproduce the
oil recovery results observed in the lab. The simulation results from models with and without CH adsorption settings
4
suggest that integration of CH adsorption and core CT data yielded better reproduction of the experimental results.
4
However, perfect matches were never achieved. Lack of data on capillary pressure effects and relative permeability
may be a possible reason for the imperfect matches. The reservoir-scale efforts included modeling of huff ‘n’ puff
scenarios. The highest incremental improvement in recovery factor observed was approximately 5%. This number
could be increased by conducting more huff ‘n’ puff cycles over the lifespan of an operating well and/or by optimizing
the operational parameters.
Applications/Significance/Novelty: These results will ultimately be used to better estimate the EOR potential and CO
2
storage resource of unconventional tight oil formations. While a 5% improvement in ultimate recovery may be
perceived as a relatively small increase over primary production, the implications for an incremental recovery of 5% for
production throughout the Bakken play is enormous. With estimated Bakken original oils in place (OOIPs) ranging
from 300 billion to 900 billion barrels, realization of even this small improvement would yield billions of incremental
barrels of oil and extend the lifetime of the play by decades.
Impact of Cluster Spacing on Infill Completions in the Eagle Ford
1 2 2
S. Evans; ; J. Magness; ; S. Siddiqui; ;
1. formerly Halliburton, Houston, TX, United States.
Objectives/Scope: The purpose of this study was to use a uniquely robust dataset to better understand one of the
biggest challenges in unconventional reservoirs: optimal cluster spacing. The dataset consisted of microseismic
monitoring (MSM), permanently installed fiber optics (FOs), and a cased-hole pulsed neutron log for reservoir
characterization. The data was then incorporated into fracture and reservoir simulators to obtain a calibrated reservoir
model. This model was then used to run multiple sensitives to determine which cluster spacing maximized production.
Methods/Procedures/Process: The workflow begins by building a calibrated fracture model for all stimulation stages. A
single stage was pressure matched to the treatment data, and the geometry was matched to the MSM and FO results.
The remaining treatment stages used a mass balance approach to calibrate the fracture geometry to the MSM and FO
data, which created significant time savings compared to pressure matching each individual stage. Using the MSM
and FO data allowed for a unique scenario wherein fracture geometry was able to vary on a cluster level as well as
have asymmetry induced by offset depletion. The fracture geometry was then incorporated into a reservoir simulator
and production history matched. The history matched reservoir model was then used to run sensitivities to determine
optimal cluster spacing.
Results/Observations/Conclusions: Optimal cluster spacing depends on many factors, including reservoir type,
reservoir quality, stimulation volumes, and distance between wellbores. This study investigated a single well in the
condensate window of the Eagle Ford. Findings revealed that approximately 40-ft cluster spacing is optimal when
completing a single wellbore with 1,600 lb/ft of proppant. This cluster spacing distance decreases as wells are drilled
closer together. With many wells in the Eagle Ford being drilled 330 ft apart, reduced cluster spacing can be employed
to keep fractures more contained and to shorten fracture half-lengths to help mitigate interference to offset wellbores.
Applications/Significance/Novelty: The data used and workflow applied to this study were a new approach to analyze
cluster spacing. Traditionally, cluster spacing is determined through an iterative process of varying cluster spacing on
a batch of wellbores, followed by measuring the production impact. This traditional method of field experimentation
requires significant time and capital investment. By acquiring adequate data on a single wellbore, and applying the
correct tools and analysis of that data, conclusions on cluster spacing, as well as many other completions variables,
can be made much more rapidly.
Numerical Modeling and Optimization of Condensate Banking Treatment in the Hydraulic-fractured Shale Gas
Condensate Reservoir
1 1 1 2 1
W. Liu; ; R. Ganjdanesh; ; A. Varavei; ; W. Yu; ; K. Sepehrnoori; ;
1. The University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: As the bottomhole pressure drops below dew point during gas production, condensate banking
forms around fractures and compromises the gas productivity. Dimethyl Ether (DME) has been mentioned to efficiently
mitigate the condensate banking. But few work has been done to testify its efficiency in unconventional reservoirs.
Besides, the cost of injecting DME in field scale is one of the challenges hindering its further application. The study
testifies the efficiency of DME and proposes an optimum injection strategy evaluated the operation cost based on the
actual data in Duvernay Shale Formation.
Methods/Procedures/Process: A robust algorithm capable of three phase flash calculation is essential during phase
behavior modeling due to the solubility of solvent in different phases, which most of commercial simulators cannot
handle. Also, most methods using density for phase identification usually do not consider composition change and
cause the discontinuity between timesteps. It would further lead to a physically incorrect result and slow down the
numerical simulator. In this study, an in-house compositional simulator is developed to simulate the process of DME
partitioning into condensate/gas/aqueous phases. The relative permeability option is implemented with an innovative
method that the relative permeability is interpolated based on the Gibbs free energy of each phase at reference
composition.
Results/Observations/Conclusions: The study shows the process of banking formation and mitigation in hydraulic
fractures are accurately simulated with the proposed model. The production rates with and without solvent treatment
are compared to determine the benefit of DME injection. Then the field data cases are tested. The result indicates the
eligibility of DME erasing the blockage in field scale in the first hundreds of days after injection. Since the banking
resumes increasing hundreds of days after injection, the cycle period of treatment is investigated. To further evaluate
the cost and profit during operation, slug size and solvent mixture are also considered. Design of experiment and
response surface methodology are used to obtain an optimal injection strategy combining the factors mentioned
above.
Applications/Significance/Novelty: The study proposes an innovative solution (DME) to mitigate the condensate
blockage in hydraulic fractured shale with a robust numerical algorithm capturing the whole process accurately. It
helps enhance the understanding of the complex phase behavior in fractured shale gas condensate reservoir. DME
has a higher vapor pressure and larger density so it improves the flowback and does not have to be pressurized
before injection, which saves the operational cost compared with CO2 injection. The practical optimum strategy in this
study makes it possible to inject DME for the future shale gas production economically.
RTA-Assisted Production Forecasting in Shale Reservoir Development
1
C. Aniemena; ;
1. BP Plc, Houston, TX, United States.
Objectives/Scope: Analog decline curve analysis is the foremost method for production forecasting. The typical basis
for analog well selection are proximity, thickness, porosity and saturation. In this work RTA is used to show that in
shales, the rock mambers may undergo varying degrees of permeability enhancement despite having similar classic
petrophysical metrics thus resulting in signifiantly different production performance. By incorporating effective
permeability variation into the analog well selection process, the accuracy of production forecasts of planned wells is
improved.
Methods/Procedures/Process: In development shale plays, there is often an abundance of parent wells previously
drilled for the purpose of acreage capture . These wells often have sufficient production data with established linear
flow signature and evidence of boundary influenced flow. The presence of boundary influenced flow allows the
application of linear flow analysis towards the determination of system permeability with reduced uncertainity. By
conducting linear flow analysis on several parent wells across the plays, the effective rock permeability is deduced
and mapped across the play. This effective permeability is significantly larger than matrix permeability from core data
and is held to be infuelced by preexisting natural fractures.
Results/Observations/Conclusions: 1. RTA results in wells with boundary influenenced flow in the woodford shale
show that hydraulic fracturing enhances the permeability of rock bodies adjacent to the main natural fractures by more
than 10 times in majority cases.
2. Effective permeability can vary up to 100% across areas of a shale play with similar classic petrophysics (porosity,
saturation, thickness). This phenomenon is not experience in RTA results for tight sands and suggests natural
fractures may be contributing to permeability enhancements in shale reservoirs.
Applications/Significance/Novelty: 1.Effective oermeability maps from RTA provide a valuable data set for high grading
shale assets and selecting appropriate analogs for new well performance prediction and reserves estimation.
2. Improved decline type curve analysis improves the efficiency of the well planning process and capital allocation
decisions
3. The use of RTA for field level effective permeability mapping is a novelty in shale development according to author's
literature review. The method avoids the need for sohphisticated well logs with may become time and cost prohibitive
to execute at field scale. RTA relies of production and pressure data that are typically available in the natural course of
asset development.
Integrating Model Uncertainties in Probabilistic Decline Curve Analysis for Unconventional Oil Production Forecasting
1, 2
A. Hong; ; R. B. Bratvold; 1, 2; L. W. Lake; 3;
1. Department of Petroleum Engineering, University of Stavanger, Stavanger, Rogaland, Norway.
2. The National IOR Centre of Norway, Stavanger, Rogaland, Norway.
States.
Objectives/Scope: Decline curve analysis (DCA) for unconventional plays requires a decline curve model that can
capture the characteristics of different flow regimes. Thus, various models have been proposed. Traditional
probabilistic DCA uses a single model that best fits the data. However, several models might fit the data almost
equally well, and the one best fitting the data might not best capture the flow characteristics. Therefore, the uncertainty
in the “best” model remains. This work aims to integrate the model uncertainty in probabilistic DCA for unconventional
plays.
Methods/Procedures/Process: Instead of identifying the “best” model, we propose to regard any model as potentially
good, where goodness is characterized by a probability. The probability of a model is interpreted as a measure of the
relative truthfulness of this model to the other models. This probability is further used to weight the model forecast. A
Bayesian framework is used to assess the model probabilities for given data. Multiple samples of the model parameter
values are obtained using maximum likelihood estimation with Monte Carlo simulation. Thus, the unique probabilistic
forecasts of each individual model are aggregated into a single probabilistic forecast, which incorporates the model
uncertainty.
Results/Observations/Conclusions: We demonstrate and conclude that using the proposed approach can reduce the
risk of using a single decline curve model that best fits the data but might not best represent the uncertainty in
production forecasting given the data. The proposed approach performs well in propagating the model uncertainty to
the uncertainty in production forecasting; that is, we end up with a forecast that represents the uncertainty given
multiple possible models conditioned to the data. The field data show that an analytically derived decline curve model
is more likely to be a good model for unconventional plays than the other three empirically derived models.
Applications/Significance/Novelty: The novelties of this work are: (1) probability is used to describe the goodness of a
model, (2) a Bayesian approach is proposed to integrate the model uncertainty in probabilistic DCA, (3) an illustrative
example demonstrates the value of using this approach to consider multiple models in probabilistic DCA for
unconventional plays, and (4) the approach is applied to actual field data to identify the most probable model given the
data.
SCOOP/STACK Periphery Plays: A Multidisciplinary Approach
1 1 1 1 1
D. Yee; ; G. Johnston; ; S. Ahmed; ; J. Wakter; ; D. Howard; ;
Objectives/Scope: Drilling targets in the SCOOP/STACK play in the Anadarko Basin have changed over the past
decade. Back in 2010, operators initially targeted the deep Woodford Shale. The Meramec was first tested in 2013
and is the primary target today. With land in the STACK tied up by many long-time operators, activity is shifting
towards the periphery areas and pushing out the perimeter of the play. Understanding the lithology, structural
implications, production and completion trends are key to drawing the boundaries between the STACK and these
bordering plays.
Methods/Procedures/Process: This study takes a statistical approach to delineate the periphery areas versus the
STACK, correlating the geological model to the drilling, completions and production data. Well log data is used to
create regional stratigraphic correlations for the Late Devonian Woodford Shale and the conformably overlain
Mississippian Meramec and Osage formations. Statistical analytics were completed on smoothed Oklahoma
production, drilling, completions and authorization for expenditure data. Correlations were derived using up-to-date
EURs, peak rates and breakevens.
Results/Observations/Conclusions: RSEG defines the periphery plays as the Shelf, Northwest Extension and the
Merge. The Meramec calcareous shale in the STACK transitions into a blockier carbonate towards the northwest and
northeast shelf, while localized faulting and structural activity are present in the Merge. The lithological trends and
associated petrophysical characteristics have been mapped and reveal rock type is a key driver of success. An
understanding of the hydrocarbon windows is important for operators to determine best drilling practices over their
acreage to maximize well recovery.
Applications/Significance/Novelty: The SCOOP/STACK play accounts for approximately 90% of rigs running in the
Anadarko Basin and has the second-highest number of rigs running in the Lower 48, trailing only the Permian. The
lowest breakevens in the SCOOP/STACK are found in the oiler parts of the play, a reflection of commodity prices.
Looking at zonal development and land valuations (based on developed dollar per acre), the core STACK becomes
most favorable. With this data set and comprehensive analytics, we explore new methodologies to understand
variations in drilling and completions practices and how they affect production results. A multidisciplinary approach is
the best way in such a rapidly evolving play to understand the intricacies of targeting peripheral areas and maximizing
success.
Using Data Analytics to Maximize Value Within the Denver-Julesburg Basin
1 1 1
K. Repchuk; ; A. P. Reimchen; ; D. Gregoris; ;
Objectives/Scope: Unconventional resources require an integrated and innovative approach incorporating multiple
data sets to characterize and derive insights into mechanisms that make these plays successful. This study evaluates
geological and completion drivers impacting well productivity within the Denver Julesburg (DJ) Basin of eastern
Colorado and southeastern Wyoming, focusing on the Wattenberg Field and peripheral regions. Understanding these
complex relationships is required to maximize acreage value in already-delineated areas and to explain well
performance variability across the play.
Methods/Procedures/Process: Approximately 4,000 horizontal wells were analyzed across the study area. Completion
data including lateral length, fluid (gal/ft) and proppant intensities (lbs/ft) were put through a comprehensive scrubbing
process to identify and fix incorrect data. Type curves were built off grouped wells by geographical area, operator,
vintage and geological zone. Bottom-hole locations for the horizontal wells were used to derive geological parameters
from maps generated with over 2,000 digital wells logs, including isopach, dominate mineralogy volumes, effective
porosity and geothermal gradient. Multivariable regressions were subsequently run for the Niobrara and Codell
formations within four regions in the Wattenberg plus the remaining wells outside the main field.
Results/Observations/Conclusions: Overall, fluid intensity, effective porosity and geothermal gradient had the most
impact on well productivity for the Niobrara Formation. In the Codell Formation, the top three variables were isopach,
fluid intensity and effective porosity. The modeling suggests doubling fluid intensity from 1,000 to 2,000 gal/ft results in
productivity increases of 5% to 30% on a total EUR basis, and 10% to 50% on an oil EUR basis. Conversely, the
results are inconclusive on the impacts of increasing proppant intensities, with the majority of regressions exhibiting
non-statistically significant values. Altogether, the modeling suggests the majority of operators in the basin could
benefit from adopting a more fluid-intensive completion design.
Applications/Significance/Novelty: Results of this study will help elucidate key geological and completion design
factors influencing production, as well as illustrating how acreage valuations change with differing completion designs.
The outlined workflow can be applied to other basins to derive similar conclusions on well productivity drivers.
Are Redox-Sensitive Geochemical Proxies Valid in Mature Shales?
1 1 1
D. R. Lindsey; ; S. M. Rimmer; ; K. B. Anderson; ;
1. Geology, Southern Illinois University Carbondale, Carbondale, IL, United States.
Objectives/Scope: The objective of this study is to establish the validity of redox-sensitive geochemical proxies in
paleodepositional environmental analyses of mature shales. A new method for artificial maturation, open-system
hydrous pyrolysis (OSHP), attempts to mimic natural source-rock maturation by maintaining a small continuous flow of
only liquid phase water in the shale reactor. This method allows mobilized materials to migrate out of the reactor,
enabling a net change in elemental mass balance during pyrolysis. OSHP will allow the use of redox proxies along a
maturation gradient to be assessed.
Methods/Procedures/Process: Samples of the Pennsylvanian-age Hushpuckney Shale were collected from the IRC
core of southwestern Iowa. Samples were crushed to < 2 mm and split into aliquots for geochemical, petrographic,
and OSHP analyses. Trace and major elements are determined on pelletized powders using a Bruker handheld XRF
instrument. Geochemical analyses include Rock-Eval pyrolysis and carbon and sulfur analyses; petrographic analyses
include maceral characterization and mean vitrinite reflectance (carried out in the Organic Petrology Lab at SIUC).
Run conditions for OSHP include pyrolysis for 72 hr at 285-345°C at a constant 1 ml/min flow rate, without allowing for
the presence of a gas headspace. All analyses are performed before and after OSHP pyrolysis to determine maturity
and geochemical changes.
Results/Observations/Conclusions: Previous studies have shown that the organic matter in the Huspuckney Shale in
this core is predominantly terrestrial, thus facilitating assessment of maturation using vitrinite reflectance. In addition to
the anticipated increase in maturation (as indicated by T and vitrinite reflectance), the decrease in total organic
max
carbon content, and decrease in HI with increased temperature, the stability or loss of trace elements can be
assessed.
Applications/Significance/Novelty: Results from this study will provide some insight into the utility of geochemical
proxies used to interpret depositional condition of mature black shales. If no geochemical alteration is observed
following OSHP, then the utilization of C-S-Fe and trace metals as proxies for EOD assessments is only strengthened.
If OSHP yields heterogeneity in the geochemistry of shale samples, then the utilization of certain redox proxies in
source-rocks that have undergone maturation is open to question. Furthermore, the new concept of OSHP more
closely simulates natural maturation than conventional hydrous pyrolysis and importantly allows mobilized materials to
migrate out of the reactor, enabling a net change in elemental mass balance, and a proper characterization of certain
proxies in mature shale.
Interpretation of High Resolution XRF Data From the Bone Spring and Upper Wolfcamp, Delaware Basin, US
1 1 2
B. Driskill; ; J. Pickering; ; H. Rowe; ;
1. Shell Exploration and Production Company, Houston, TX, United States.
2. Premier Oilfield Laboratories, Houston, TX, United States.
Objectives/Scope: The Wolfcamp and Bone Spring comprise a heterolithic mix of turbidites, mass flow deposits,
hemipelagics, and ashes. These rocks have different origins, multiple sources and in some cases unique elemental
signatures. While XRF proxies used for interpreting aspects of a mudrock such as sediment sources and depositional
environment are fairly well understood, the same interpretive frame does not work for non-mudrocks. This study
compares high resolution XRF profiles across heterolithic intervals, integrate with lab and core observations, and
determine guidelines for interpreting non-mudrocks.
Methods/Procedures/Process: High resolution XRF at 2 inch scale was acquired over several cores. The XRF data
were compared to XRD, organic geochem, thin section, whole-rock and plug CTs, and visual core descriptions to build
an integrated interpretation for depositional environment, depositional process, lithology, mineralogy, diagenesis, and
texture. These interpretations were compared to the XRF signatures of identified facies and facies associations.
Various typical XRF elemental proxies found in literature were calculated as were enrichment factors. Other
unpublished proxies based on the integrated view from the core data were calculated. The dataset was statistically
analyzed to provide a validated interpretation.
Results/Observations/Conclusions: Accepted interpretations of mudrock XRF based on raw elements, elemental
ratios, and enrichment factors were challenged in an interpretation of the heterolithic upper Wolfcamp and Bone
Spring. Many of the more common interpretation guides did not work well for the mudrocks, probably because the
studied section was deposited in a very different environment and with different processes than more typical
unconventionals such as Eagle Ford. The non-mudrock portions of this section did not follow any of the established
mudrock rules. The XRF data did, though, contain details about source (e.g. terrestrial or non-terrestrial) and other
factors. However, the proxy thresholds and enrichment factors are very different than those for mudrock in general
and the interbedded mudrocks specifically.
Applications/Significance/Novelty: While many papers have been published giving guidelines for interpreting mudrock
XRF data, none have specifically addressed interpretation of interbedded non-mudrock units. While XRF is used in
conventionals it is primarily just for correlation. This is the first published study that uses very high resolution sampling
(2-inch) to fully characterized all layers in an extremely heterolithic and thinly bedded interval and give interpretive
guidelines for the non-mudrocks.
Field Test of CO Injection in a Vertical Middle Bakken Well to Evaluate the Potential for Enhanced Oil Recovery and
2
CO Storage
2 1 1 1 1 1 1 1
J. A. Sorensen; ; L. J. Pekot; ; J. Torres Rivero; ; L. Jin; ; S. B. Hawthorne; ; L. Jacobson; ; T. Doll; ; S.
1 2
Smith; ; M. Flynn; ;
2. XTO Energy Inc., Houston, TX, United States.
Objectives/Scope: Research was conducted to examine the ability of the Bakken Formation to store CO and improve
2
oil production through CO injection. Previous lab- and modeling-based work suggests CO can permeate the Bakken
2 2
matrix, largely through diffusion, to mobilize oil. Those studies showed that CO will preferentially mobilize lower
2
molecular weight hydrocarbons. In 2017, an injection test was conducted in a vertical well completed in the Bakken.
The objectives of the test were to determine the injectivity of an unstimulated Bakken reservoir and the ability of
injected CO to mobilize oil.
2
Methods/Procedures/Process: The test was conducted in a virgin reservoir. The well completion program did not
include a stimulation operation. Upon perforation, the well did not flow to surface, but oil samples were collected
before injection. Approximately 99 tons of CO was injected over 4 days. The CO was allowed to soak for 15 days.
2 2
Reservoir pressure and temperature were monitored during all stages of the test using downhole gauges. During the
flowback period, gas composition was monitored, and fluid samples were collected. Preinjection and postinjection oil
samples were analyzed for oil composition to determine the molecular weight distribution of the hydrocarbons. Pulsed-
neutron logs were also run before and after injection to evaluate the vertical distribution of the CO in the near-
2
wellbore environment.
Results/Observations/Conclusions: Injectivity of the unstimulated Middle Bakken matrix was found to be low, with
stable CO injection rates between 6 and 12 gallons per minute and bottomhole pressure during continuous injection
2
ranged from 9400 to 9470 psi. During flowback, a mix of CO and hydrocarbon gas was produced for 10.5 hours. The
2
well then flowed oil to surface briefly, during which time fluid and gas samples were collected. Analyses of the
preinjection and postinjection oil samples indicate that the composition of the postinjection oil samples had greater
amounts of lower molecular weight hydrocarbons than the pretest oils. Interpretation of the results from the field test
suggest that although matrix injectivity is low, injected CO can penetrate the Middle Bakken to mobilize oil from the
2
matrix.
Applications/Significance/Novelty: Past CO injection tests in the Bakken were all conducted in horizontal wells.
2
Those tests showed injectivity into a stimulated Bakken well is not a problem, but incremental oil was not produced,
and there is no evidence in publically available data to suggest that a scientific approach was taken to evaluate the
mechanisms controlling the ability of CO to mobilize stranded oil. The use of a vertical well in a virgin reservoir
2
reduced uncertainties associated with horizontal wells, such as the unknown distribution of rock properties and
hydraulically induced fractures along the wellbore. The data generated by this test support the findings of previous lab
studies which indicate that diffusion of CO into the rock matrix may play a significant role in the use of CO for EOR
2 2
in the Bakken.
Miscible EOR Process Assessment for Unconventional Reservoirs: Understanding Key Mechanisms for Optimal Field
Test Design
1 1
V. Sahni; ; S. Liu; ;
1. Occidental Petroleum, Houston, TX, United States.
Objectives/Scope: The miscible Enhanced Oil Recovery (EOR) mechanisms in unconventional reservoir(s) may be
substantially different than those associated with miscible gas EOR in conventional reservoirs. Due to potentially low
throughput through low permeability rocks, Huff-n-Puff process is being considered in the industry for application in
unconventional reservoirs. The objective of this study was to use learnings from literature and Occidental’s long
experience with successful applications of miscible gas EOR processes in conventional reservoirs to design an EOR
scheme to test in unconventional reservoirs.
Methods/Procedures/Process: The methodology followed consisted of literature review and a process that involved
Design of Experiments (DOE) using a mechanistic compositional simulation model. This approach, often used to
scope “green field” oil and gas developments, was adopted to better understand miscible CO Huff-and-Puff process
2
physics and design an injection-soak-production scheme for a field implementation in an unconventional reservoir.
Results/Observations/Conclusions: An integrated workflow applied for process assessment indicates that CO2 Huff-
and-Puff may be a technically feasible EOR method for unconventional reservoirs.
This study indicates that vaporization of lighter oil components and IFT reduction at pressures above minimum
miscibility pressure (MMP) are the main recovery mechanisms. Moreover, the presence of hydraulic and natural
fractures may provide a large contact area for injected gas to penetrate into the ultra-low permeability matrix. The
effect of Molecular Diffusion was also investigated but did not have a major impact on recovery for the modeled
conditions at the field scale.
The optimization scheme indicates that oil recovery is proportional to mass of CO injected and soak times do not
2
greatly affect the EOR oil in this process.
Applications/Significance/Novelty: The workflow utilized suggests Huff-and-Puff to be a technically feasible EOR
method for unconventional reservoirs. This approach also helps explain key EOR mechanisms and demonstrates how
to design an optimal scheme for field trials before the tests are initiated. Huff-and-Puff is a single well processes
where well-to-well connectivity is not required. If successful, the process has the potential to increase the recovery
factors currently being achieved in the development of unconventional resources.
Huff-n-Puff Gas Injection Performance in Shale Reservoirs: A Case Study From Duvernay Shale in Alberta, Canada
1 1 1 1 1 1 2
C. R. Clarkson; ; H. Hamdi; ; A. Ghanizadeh; ; S. A. Ghaderi; ; A. Vahedian; ; N. Riazi; ; A. N. Esmail; ;
2. Encana, Calgary, AB, Canada.
Objectives/Scope: Despite the use of multi-fractured horizontal wells for primary recovery operations in tight and shale
reservoirs, oil recovery factors are still very low, barely exceeding 5-10% of OOIP. Recently, laboratory results,
augmented with pilot data, have pointed to the technical viability of Enhanced Oil Recovery (EOR) techniques to
improve oil recovery from such formations. In this paper, the feasibility of cyclic gas injection (i.e. huff-n-puff) in the
Duvernay Formation in Alberta, Canada is investigated using compositional numerical simulation calibrated to primary
production.
Methods/Procedures/Process: In this study, compositional numerical models are used to accurately simulate the
miscible huff-n-puff (HnP) process by lean and rich gas injection. A robust fluid model is constructed using collected
fluid samples, and tuned to several lab experiments. Stress-dependent permeability (SDP) data are collected using a
series of steady-state gas (N ) permeability tests performed at varying effective stresses for quantifying the impact of
2
the loading/unloading stress cycles on the system permeability. The fluid model, log data, and lab measurements are
used as input to the model which, in turn, is used to history-match the primary production data. The calibrated model
is subsequently used to optimize the HnP operational conditions to maximize the oil recovery.
Results/Observations/Conclusions: The history-matching resulted in a calibrated Limited Enhanced Fracture Region
(LEFR) model along the propped hydraulic fracture. To assess the value of lab-measured SDP data, another history-
matching trial with unknown SDP curves was performed. Interestingly, the history-match-derived SDP curves closely
matched the lab-measured depletion SDP curves. The calibrated numerical model was then used for the HnP
optimization – lab-measured SDP curves were also used for this purpose. The optimized results indicated that, in a 20
year time span, a 1.5 to 2 times increase in oil recovery can be obtained. The combination of higher injection rates,
shorter injection, soaking, and production times, and higher producing pressures leads to optimum recoveries from
this volatile oil reservoir.
Applications/Significance/Novelty: This study presents a comprehensive workflow with practical guidelines for
integrating lab measurements and field observations for the successful calibration of an LEFR reservoir model in the
Duvernay Formation. The feasibility of using cyclic gas injection in ultra-low permeability formations is investigated
with various injection fluids considered. The history-matching and optimization steps are conducted using novel
algorithms to minimize the number of simulation runs while achieving a balance between computation times and the
quality of the history-match.
Laboratory Investigation of EOR Techniques for Organic Rich Shales in the Permian Basin
1 1 1 2 2
S. Liu; ; V. Sahni; ; J. Tan; ; D. Beckett; ; T. Vo; ;
1. Occidental Petroleum, Houston, TX, United States.
2. Core Laboratories , Houston, TX, United States.
Objectives/Scope: The commercial production from light oil, organic-rich shales in the Permian Basin to date has
largely come from a solution gas drive recovery mechanism effected by horizontal drilling and multi-stage hydraulic
fracturing. These onshore, capital-intensive developments feature steep production declines and low expected
ultimate recoveries. This study involved laboratory experiments introducing miscible gases into core samples to
investigate Enhanced Oil Recovery (EOR) mechanisms for Permian Basin shales to provide information to design field
tests for a Huff-and-Puff recovery process.
Methods/Procedures/Process: Fluid and core samples were collected from a new well completed only in the nano-
Darcy organic rich shale zone. The fluid characterization and testing program consisted of Pressure Volume
Temperature (PVT) tests, swelling tests and minimum miscibility pressure (MMP) for three different gases. With the
data collected from the fluid characterization program, the laboratory experimental set-up and procedures were
tailored to investigate the efficacy of a novel EOR process using shale core plugs.
Results/Observations/Conclusions: The fluid characterization program indicated miscible gas injection to be a
promising EOR technique for the oil sample collected from the shale. The three gasses that were used to estimate
MMP at reservoir temperature were carbon dioxide (CO2), methane (C1) and field produced gas. All three gasses
show miscibility at initial reservoir pressure conditions. CO2 was found to be first contact miscible at the lowest tested
pressure condition. PVT data was tuned to an equation-of-state (EOS) model that was used in modeling studies to
better understand EOR mechanisms and design Huff-and-Puff field trials. The core experiments, conducted with CO2
at reservoir conditions, showed favorable results, including good oil recovery and CO2 utilization in up to seven
consecutive Huff-and-Puff cycles.
Applications/Significance/Novelty: The study describes a comprehensive laboratory testing program to investigate key
recovery mechanisms and help define a potentially attractive EOR process for Permian Basin organic rich shales.
The study should help accelerate learnings about EOR for organic rich shales and assist in the design of field trials.
Recovery Mechanisms During Gas Injection for EOR in Organic Rich Shale Reservoirs
1 1 1
F. Tovar; ; M. Barrufet; ; D. S. Schechter; ;
1. Texas A&M University, Houston, TX, United States.
Objectives/Scope: Laboratory experiments of gas injection in organic rich shales have resulted in significant oil
recovery. In a previous publication, we presented an operational philosophy to maximize recovery factor; consisting of
the use of a huff and puff injection scheme at the highest possible pressure, regardless of the MMP value. This paper
focuses in the impact that poor fluid transport in organic rich shale have on the recovery mechanisms under gas
injection, and lays out the rationale behind the proposed operational philosophy.
Methods/Procedures/Process: We used CT-scanning data from nine core-flooding experiments conducted injecting
CO in organic rich shale sidewall cores, two more injecting N , and 3 test of CO injection in Berea sandstone. The
2 2 2
core plugs were re-saturated with crude oil in the laboratory, and the experiments were performed at reservoir
pressure and temperature using a novel design that mirrors gas injection through a hydraulic fracture. CT-scanning
was used to visualize the compositional changes with time and space during gas injection. The overall difference in
composition between the oil injected and the oil recovered was obtained using gas chromatography.
Results/Observations/Conclusions: A peripheral slow-kinetics vaporizing gas drive is the main production mechanism.
Gas injection is done through a fracture due to the poor fluid transport of shales. Gas flows preferentially through the
fracture due to its high permeability, avoiding the formation and displacement of a miscible front along the matrix to
mobilize the oil. Instead, the gas in the periphery of the matrix vaporizes the light and intermediate hydrocarbons
making recovery a function of the fraction of oil that can be vaporized into the volume of gas in the fracture at the
prevailing thermodynamic conditions. The gas-oil mass transfer is sufficiently fast to yield significant recovery within
24 hours, but slow enough to create a compositional gradient inside the matrix that exists until the end of the
experiment.
Applications/Significance/Novelty: The peripheral and the slow kinetics aspects of the recovery mechanism, are a
consequence of the poor fluid transport through the rock. The combination of different injection gases and reservoir
rocks in our experiments enabled us to effectively isolate the effects of transport properties and phase behavior. In this
work, the operational guidelines regarding injection pressure and soaking time derived directly from experimental
observations are now explained in the light of the new production mechanisms using ternary diagrams, thus providing
the necessary understanding to successfully conduct gas injection processes in organic rich shale reservoirs. Given
the vast volume of crude oil trapped shale reservoirs, this is a major step towards securing energy sources for the
future.
The Influence of Organic Matter on Supercritical CO Migration in Organic-Rich Shales
1 1 1 2 1 2 3 4
S. Smith; ; B. A. Kurz; ; J. A. Sorensen; ; S. B. Hawthorne; ; H. Sanei; ; O. Ardakani; ; J. D. Walls; ; L. Jin;
1
; S. Butler; 1; C. Beddoe; 1; B. Mibeck; 1;
2. Department of Geoscience, Aarhus University, Aarhus C, Denmark.
3. Geological Survey of Canada, Calgary, AB, Canada.
4. Ingrain-Halliburton, Houston, TX, United States.
Objectives/Scope: The Bakken petroleum system is a world-class oil play with oil-in-place estimates in the hundreds
of billions of barrels. Despite the resource potential, recovery factors are typically low, ranging from 4% to 10%. Efforts
to evaluate mechanisms to increase oil recovery have focused on the use of supercritical CO for enhanced oil
2
recovery (EOR) with the added benefit of geologic carbon sequestration. This work describes a series of laboratory-
based tests to evaluate the efficacy of CO for EOR and to better understand the controlling mechanisms of CO
2 2
permeation in Bakken rock samples.
Methods/Procedures/Process: To evaluate the efficacy of CO to improve oil recovery, static extraction tests were
2
performed on rock samples representative of the Bakken reservoir (siltstone) and source rocks (organic-rich shales).
A dynamic CO injectivity test was also performed on an unfractured Bakken shale sample that was confined at
2
reservoir pressure. In conjunction with the extraction tests, a suite of standard and advanced characterization
techniques were used to better understand the rock fabric, pore and fracture networks, and the organic content of the
samples, including field emission scanning electron microscopy (FESEM) imagery, focused ion beam SEM (FIBSEM),
organic petrography and reflectance measurement using advanced microscopy, and extended slow heating (ESH)
Rock-Eval analysis.
Results/Observations/Conclusions: The extraction test results demonstrated recovery of residual oil in the tight
siltstone samples comprising the reservoir (recovery values approaching 100%) and also in the shale source rocks, in
which oil recovery from 11-mm-diameter rods ranged from 12% to 65% after
24 hours. Analysis of samples pre- and post-CO extraction using the Rock-Eval ESH method confirmed the results of
2
the CO extraction tests. FESEM imagery of the Bakken shale samples revealed the presence of organic-associated
2
fractures that could provide a mechanism for CO transport into the OM (organic matter) and subsequent extraction of
2
crude oil. Image-based analysis of the FESEM and FIBSEM imagery suggested that the majority of the porosity in the
shales was OM-hosted, and that much of the pore networks were connected.
Applications/Significance/Novelty: The results of this work revealed that CO is able to permeate samples of
2
unconventional reservoir and shale source rocks much more readily than expected, resulting in significant recovery of
residual oil. The mechanism of permeation within the organic-rich shales appears to be within OM-hosted fracture
networks and possibly through nano-scale porosity associated within the kerogen and bitumen of the samples. The
implications of this work suggest that organic-rich source rocks may be a viable target for CO -based EOR, with the
2
added benefit of long-term CO storage via adsorption and absorption.
2
Ethane Flooding as an Alternative to CO Injection in Tight Formation: A Bakken Case Study
1 1 12 1
B. N. Yolo; ; H. Jabbari; ; W. Yue; ; Y. Agbor; ;
1. PETROLEUM ENGINEERING, University of North Dakota, GRAND FORKS, ND, United States.
Objectives/Scope: Over the last few years, many studies have shown promising results on CO injection in tight
2
formation which may lead to a significant increase of the recovery factor. However, disadvantages of CO injection
2
include corrosion on equipment, reaction with formation minerals, limited gas source, and high price. Ethane
production in Bakken has drastically increased together with the massive oil production. It has become a viable EOR
solvent due to its advantage in miscibility, solubility, operation, and pricing with supplies from development of Bakken
Shale and other shale gas formations.
Methods/Procedures/Process: In this work, we first use Molecular Dynamics Simulation to study the interaction
between ethane and Bakken oil. Next, we conducted core flooding experiments on middle Bakken core samples, each
flooded separately with CO and ethane. Injectivity of both injectants are recorded and compared. Oil recovery factors
2
are recorded for each flooding experiment. We then used CT scan to determine the vertical and areal sweep efficiency
of the samples. A core scale simulation model is built up to mimic the injection process in the experiment. After which,
we use the experimental results to calibrate our simulation models to further evaluate the efficiency and behavior of
ethane as an EOR agent in the Bakken. Finally, we provide an economic evaluation of using both gases as injectants
in Bakken.
Results/Observations/Conclusions: We have studied the feasibility of a more profitable alternative of CO as an EOR
2
agent in the Bakken formation and the results from our study show that ethane is more favorable than CO as an
2
EOR agent. Ethane injection achieved higher recovery factor than carbon dioxide injection. The diffusivity coefficient
provided by our MDS simulation were higher for ethane than CO , with a much lower minimum miscible pressure with
2
the Bakken in-situ oil. Ethane is more available in the Bakken than CO and at a lower cost. Additionally, ethane is
2
safer than CO in terms of corrosion, environment, and management.
2
Applications/Significance/Novelty: This combined work of both experimentation and simulation provides a pioneer
study in evaluating the performance of Ethane injection in Bakken formation. The experiments give clear insight into
the differences of recovery factor and recovery efficiency we can expect from injecting CO and Ethane into Bakken
2
formation. The calibrated simulation provides details of oil composition change when injecting both gases. By laying
the theoretical ground for Ethane injection, this study provides data and proof for engineers and researchers to design
more economically profitable EOR strategies.
Full-Field Strain Measurement on Naturally-Fractured Rocks
1 1 1
F. Nath; ; M. Mokhtari; ; S. Jiang; ;
1. University of Louisiana at Lafayette, Lafayette, LA, United States.
Objectives/Scope: Strain is a key parameter in the evaluation of rock mechanical properties and it is usually measured
using strain gauges. This laboratory method, however, has inherent uncertainty in the attachment of sensitive gauges
to the rock surface and it is limited to one-point strain measurement. To overcome these challenges, the application of
optical technique with image processing is proposed. Optical technique has the advantage of being a non-contact
method. It also provides full-field strain measurement which is essential for heterogeneous samples having complex
strain map.
Methods/Procedures/Process: Buda limestone samples with extensive calcite-filled fracture system is investigated in
this paper. Random speckle pattern was generated on the rock samples. The samples were diametrically compressed
while the camera recorded the images of the speckled surface. The images were then processed to calculate the
strain changes all over the sample from beginning until the rock failure.
Results/Observations/Conclusions: The results of full-field strain measurement over time illustrates the points of
maximum strain accumulation in the samples. Therefore, the fracture initiation point and fracture propagation can be
predicted using this optical technique. This is especially important on the tested samples with random natural-fracture
pattern since the complexity of induced fracture pattern can be illustrated. We discuss the role of natural fracture
bonding, location and orientation on the interaction of induced fracture with natural fracture.
Applications/Significance/Novelty: One of the main questions in hydraulic fracturing community is still over a better
understanding on the geometry of generated fractures. Answering this question has economic benefits to optimize the
number of fracturing stages and to avoid over-fracturing or under-fracturing. The conventional rock mechanics
laboratory technique using strain gauge provides only a single strain measurement and it goes through tedious strain
gauge attachment process. Using the proposed optical technique, new insights on the fracture initiation point and
fracture propagation over time is provided. Moreover, the results on Buda limestone samples with extensive natural
fractures help a better understanding on the interaction between induced fractures and natural fractures.
Rock Dilation and Its Effect on Fracture Transmissivity
1 1 1
H. Zhou; ; Q. Zhao; ; G. Grasselli; ;
1. Civil Engineering, University of Toronto, Toronto, ON, Canada.
Objectives/Scope: Dilation of rock describes the increase in joint aperture during shear, which can greatly increase the
transmissivity. Source mechanisms computed from recorded microseismic events during hydraulic fracturing show a
significant amount of shear failure during treatment. Therefore, the ability to properly account for shear-induced
dilation is of considerable importance in assessing the transmissivity of fracture networks and hydrocarbon
deliverability of stimulated reservoirs.
Methods/Procedures/Process: Conventionally, researchers have described dilation using the dilation angle. The
physical behaviour behind such an angle is that the asperities on the surfaces, which can be quantitatively described
by roughness, prop open the joint walls. This paper details the development of an empirical relationship, based on
laboratory experiments, that evaluates joint dilation as a function of surface roughness, emphasizing the effects of
normal stress and material strength. Additionally, numerical modelling of a shear experiment is conducted to
understand the effect of dilation, asperity damage, and gouge material formation on joint transmissivity.
Results/Observations/Conclusions: Dilation responses showed that the rate of dilation decreased with shear slippage
and the dilation was asymptotically bounded by a threshold defined by asperity height, material strength, and normal
stress. The effect of roughness was most prominent in the convergence rate of the dilation curve, e.g. rougher
surfaces with steeper asperities result in higher dilation rates. Using the hybrid finite-discrete element method (FDEM),
aperture changes and asperity damage were observed in the model. Results showed that permeability increased at
the onset of dilation. As shear progressed, increasing amount of gouge material filled the dilated region and restricted
flow. This indicates that joint transmissivity calculated solely from dilated joint opening would be overestimated.
Applications/Significance/Novelty: By establishing a dilation relationship describing changes in joint opening during
shear, one can better understand the additional influence of shear on the rock mass. The numerical model illustrates
that the dilation of the joint, damage of asperities, and gouge material filling have distinct implications on
transmissivity. With further insight on joint dilation under shear and its effects on permeability and transmissivity,
assessment of hydrocarbon deliverability of stimulated reservoirs can be more accurate.
Effects of Porous Properties of Rock on Near-Wellbore Hydraulic Fracture Complexity
1 1 1
Y. Feng; ; E. E. Podnos; ; K. Gray; ;
States.
Objectives/Scope: Using a fully coupled fluid flow and geomechanics model, this paper demonstrates that a near-
wellbore fracture trajectory is a strong function of pore pressure and porous properties of the rock, beyond rock
mechanical properties and in-situ stresses.
Methods/Procedures/Process: A 2D, fully coupled hydro-mechanical model for growth of a hydraulic fracture in the
near-wellbore region is presented. The model is developed with XFEM-based cohesive zone methodology. Rather
than focusing on factors usually considered (e.g. injected fluid viscosity, rock elastic properties, and injection rate), this
paper takes into account the effects of pore pressure and rock porous properties (e.g. permeability and leak-off
coefficient) on near-wellbore fracture trajectory, factors which are rarely considered in previous near-wellbore fracture
models. The capabilities of the model are illustrated using numerical examples and validated against published
experimental results.
Results/Observations/Conclusions: This study demonstrates that the proposed near-wellbore fracture model retains
the advantage of XFEM in capturing arbitrary fracture paths while considering fluid flow in the fracture and poroelastic
behavior of the rock. The validation studies show good agreement between fracture trajectories obtained from
simulations and experiments. The simulation results indicate that porous properties of rock can strongly influence
near-wellbore fracturing. In general, an increase in permeability or leak-off coefficient results in relatively sharper
fracture curving when a fracture initiates from a perforation misaligned with the maximum principal stress direction. In
addition, a larger initial pore pressure tends to cause sharper fracture curving as well.
Applications/Significance/Novelty: Existing studies based on linear elastic fracture mechanics have shown that near-
wellbore hydraulic fracture complexity is only dependent on far-field stresses, rock mechanical properties, and
injection parameters. A more accurate and realistic hydro-mechanical model proposed in this paper further illustrates
that near-wellbore fracture complexity is also highly related to porous nature of rocks.
A Novel Method for Experimental Characterization of the Poroelastic Constants in Unconventional Formations
1 1 1 1 1 1
D. Gokaraju; ; M. Aldin; ; S. Govindarajan; ; A. Thombare; ; O. Abdulbaki; ; R. Patterson; ;
1. MetaRock Laboratories, Houston, TX, United States.
Objectives/Scope: The characterization of the poroelastic response of a rock is key to understanding the impact of
pore pressure change during depletion/injection on insitu stress. Accurate characterization of these constants plays a
crucial role in many applications such as refracturing, identification of stress barriers and subsidence prediction. This
paper evaluates the existing methods of Biot estimation and proposes an alternative way to directly measure the
poroelastic constant. The poroelastic constants are then compared against mineralogical data and other rock
properties to evaluate their sensitivity.
Methods/Procedures/Process: The proposed new method comprises of directly measuring Biot constant using real-
time continuous monitoring of P-wave arrival time. The method involves introducing a small pore pressure increase
(~300-500 psi) and monitoring the resultant change in P-wave velocity. The confining pressure is then increased until
the original P-wave velocity is achieved as shown in Fig 1. The proposed method mitigates uncertainty related to
sensitivity due to pressure change inherent in the other methods.
Results/Observations/Conclusions: The efforts to determine Biot in the past have focused on estimating bulk and grain
modulus using volumetric strain measurements or using permeability. The stress variation is relatively large in these
experiments to measure a response in the strain gauges. When tight mudrocks are subjected to such large stress
changes, hysteresis plays a significant role. If a highly sensitive gauge is used instead of strain gauges or
permeability, these issues can be addressed. Using a high precision system with a resolution of 1 nanosecond to
measure P-wave velocity enables the measurement of Biot’s coefficient with much higher accuracy and over a small
pressure perturbation without exceeding the insitu stress conditions.
Applications/Significance/Novelty: The method employs supercritical gas as pore fluid to access the pore space which
shortens the duration of the experiment and avoids slippage and frictional effects. Using a high precision system with
a resolution of 1 nanosecond to measure P-wave velocity enables the measurement of Biot’s coefficient with much
higher accuracy and over a small pressure perturbation without exceeding the insitu stress conditions. Finally, the
poroelastic response characterized above is tied back to the rock fabric and the rock’s mineral constituents. The
measured parameter reduces uncertainty in establishing hydraulic fracture geometry, depletion strategy and
refracturing program design.
Layered Modulus Effect on Fracture Modeling and Height Containment
1 1 2
K. Yue; ; J. Olson; ; R. Schultz; ;
1. Petroleum Engineering and Geosystems, University of Texas at Austin, Houston, TX, United States.
2. Orion Geomechanics , Houston, TX, United States.
Objectives/Scope: Hydraulic fracture height growth is one of the critical factors in the success of hydraulic fracturing
treatments. It has been well documented that modulus contrast between adjacent layers is generally considered of
secondary importance in terms of direct control of fracture height containment. However, arrested fluid-driven
fractures at soft layers are often observed in outcrops and hydraulic fracture diagnostics field tests. The objective of
this study is to investigate hydraulic fracture height containment due to modulus contrast in layered reservoirs.
Methods/Procedures/Process: In order to illustrate the effect of modulus contrast on fracture height containment, this
study proposed a new approach, which is based on the effective modulus of a layered reservoir. In this work, two-
dimensional finite element models were utilized to evaluate the effective modulus of a layered reservoir, considering
the effect of modulus values, fracture tip location, height percentage of each rock layer, layer location, the number of
layers, and the mechanical anisotropy. Then, the effect of modulus contrast on fracture height growth was investigated
with an analysis of the stress intensity factor, considering the change of effective modulus as the fracture tip
propagates from the stiff layer to the soft layer.
Results/Observations/Conclusions: This study showed that the details of layering did not impact the value of effective
modulus and the only important parameters were fracture tip locations, modulus values, and the height percentage of
each rock layer. In addition, this study empirically derived two approximations of effective modulus depending on
fracture tip location, namely the modified height-weighted mean and the modified height-weighted harmonic average.
By combining linear elastic fracture mechanics with the appropriate effective modulus approximations, the results
indicate that hydraulic fracture propagation will be inhibited by the soft layer due to a reduced stress intensity factor.
Applications/Significance/Novelty: The averaging methods developed in this work can significantly improve material
balance in hydraulic fracturing simulation. This study also suggested that soft layers inhibit hydraulic fracture
propagation in layered reservoirs. As a result, hydraulic fracture height containment within a stratified rock stack can
be better evaluated by comparing the modulus contrast between adjacent layers.
Stimulation Mechanisms in Unconventional Reservoirs
1 1 2
Z. Ye; ; A. Ghassemi; ; S. Riley; ;
1. Petroleum and Geological Engineering, University of Oklahoma, Norman, OK, United States.
2. Devon Energy, Oklahoma City, OK, United States.
Objectives/Scope: Shear stimulation has been recognized as an important factor in enhancing unconventional
reservoir permeability during hydraulic fracturing or re-fracturing jobs. However, the fundamental mechanisms and
their contributions to surface area generation and permeability increase are not well understood. Thus, the objective of
this paper is to shed light on this issue using laboratory shear stimulation tests on shales. In particular, we address the
possibility of flow rate increase through fracture propagation and coalescence due to slick-water treatment under
representative stress conditions.
Methods/Procedures/Process: Cylindrical 2-inch diameter by 4-inch length Eagle Ford shales, with a single and
multiple pre-cut fractures (1.6 mm width), are used to conduct the injection tests under triaxial condition. For a given
confining pressure and overburden stress, water is injected into the fractures to induce fracture propagation and
coalescence. The acoustic emission (AE) sensors attached on sample were used to monitor the fracturing process,
and the flow rate, stress and displacement are also recorded.
Results/Observations/Conclusions: The laboratory scale shear stimulation tests show that pre-existing fractures can
propagate and coalesce to generate a fracture network at treatment pressures below the minimum principal stress. In
addition, significant flow rate increase can be observed after fracture propagation and coalescence. The fracturing
process during water injection can be correlated with the acoustic emission events. Furthermore, both wing cracks
and shear zones can be noticed on the tested samples, which indicate that fractures propagate in both tensile and
shear modes during shear stimulation.
Applications/Significance/Novelty: This is the first time such an experiment has been conducted on shale fractures.
Most experimental studies that have considered the behavior of pre-existing fracture propagation have used artificial
material or granite or sandstone rather than real shale rocks and have been mainly limited to uniaxial or biaxial
compression. The effects of water injection pressure applied inside the fractures and confining pressure on fractures’
cracking behavior are not well understood. The results in this paper explain the process of shear stimulation and
permeability enhancement in shale reservoirs, and clearly show that mixed-mode stimulation caused by shear slip can
be used to achieve economic flow rates.
Hydraulic Fracture Propagation in a Vertically and Laterally Heterogeneous Stress Media in the Permian Basin
1 1 1 1 2
G. Izadi; ; L. Cruz; ; C. Barton; ; T. Hoeink; ; B. Elliott; ;
1. Baker Hughes, a GE company, Palo Alto, CA, United States.
2. Devon Energy Corporation, Oklahoma City, OK, United States.
Objectives/Scope: The objective of this study is to explore the role of vertically and laterally heterogeneous in situ
stress for hydraulic fracture propagation, and to quantify its effect in fracture stimulation efficiency (e.g., proppant
placement, fracture containment, fracture dimension, etc.) from a pad in the Permian Basin, using a 3D fully-coupled
finite-element simulator. These stress conditions are very common in unconventional plays where lithological
variations and variations in natural or induced reservoir fluid pressure are present
Methods/Procedures/Process: In this study, after creating a 1D geomechanical model based on field data, and
populating a reference 3D volume with initial rock and mechanical properties, a number of 3D volumes were
populated using vertically and laterally varying stresses. Multiple scenarios were analyzed for hydraulic fracture
propagation around reduced horizontal stresses that are commonly caused by depletion and natural stress variations.
Then, the fracture containment, lateral and vertical growth, width, and proppant efficiency, were evaluated and
compared to the reference case. The fully-coupled finite-element simulator allows arbitrary assignment of each
individual 3D element with mechanical properties based on field measurements, which provides a clear advantage
over tools limited to pseudo-3D or 2D geometries
Results/Observations/Conclusions: Lateral and vertical in situ stress variations control fracture propagation and
geometry. When multiple wells are completed and depletion drives horizontal stress reductions, hydraulic fractures
grow preferentially toward the depleted zones or regions of low closing stress. Some regions that experienced fracture
growth under initial conditions, become fracture barriers. Likewise, some regions that initially served as fracture
barriers, turn into fracture pathways. Under homogeneous stress, the fractures grows radially until it hits a horizontal
stress barrier. Subsequently, PKN-type propagation creates a symmetric bi-wing fracture. For variable stress cases,
the fracture immediately grows asymmetrically toward the “low” stress zones, creating a vertically confined and
asymmetric fracture
Applications/Significance/Novelty: This study illustrates, using advanced 3D modeling, how lateral and vertical
variations in the 3D stress field have profound implications for hydraulic fracture design (e.g., well, stage, and cluster
spacing, injection fluid and proppant properties). Therefore, the results of this modeling approach can be directly
applied to field development and benefit completion and fracture engineers
A Novel Methodology for Mercury Intrusion Porosimetry Analysis, Data Reduction, Blank Correction, and
Interpretation for Shales
1 1 1 1 1
K. E. Gorynski; ; T. Dewane; ; T. Smagala; ; M. H. Tobey; ; J. Mansoori; ;
1. Encana Services Company Ltd, Denver, CO, United States.
Objectives/Scope: Current mercury intrusion porosimetry (MIP) laboratory analyses and data reduction techniques are
not tailored for rocks with nanometer-scale pore geometries. The volume of injected mercury (Hg) measured during an
MIP experiment is related to 1.) conformance of Hg around sample material, 2.) compression of the penetrometer
glass, sample volume and grain volume, 3.) thermal expansion of Hg, and 4.) pore intrusion. Accurately allocating Hg
volumes to each of these processes is essential for correct sample bulk density, grain density, porosity, and pore-
throat-size measurements.
Methods/Procedures/Process: For matrix porosity rocks with moderate- to high-porosity and pore throats greater than
~100nm, this process can be relatively trivial. Solvent-cleaned whole plugs or chips can be placed directly in a
penetrometer before running MIP. Conformance for these samples occurs at low pressures (10-30 psi) and is easy to
identify in the data as Hg fills the penetrometer and eventually encapsulates the rugose surface of the rock. At the low
to moderate pressures required to overcome capillary forces and intrude Hg in to these rocks, sample compression,
penetrometer compression, and Hg thermal expansion are low and can be accounted for with minor blank corrections.
Results/Observations/Conclusions: Shales, which have tight, organic-hosted porosity with small pore throats (<100
nm), require crushing prior to MIP to increase sample surface area allowing for effective cleaning of hydrocarbon and
water from the sample, and to help access small disconnected pores. However, using crushed material for MIP adds
significant complexity to the interpretation. Conformance for crushed samples occurs at much higher pressures
(~1000-10,000 psi), and appears to be related to crush size and sample surface roughness. For most shales, high
pressures (5,000 - 60,000+ psi) are required to overcome capillary forces on their nanometer-size pore throats. At
these pressures, much of the observed Hg intrusion is related to sample compression and Hg thermal expansion.
Applications/Significance/Novelty: Here we present a novel methodology for MIP analysis, data reduction, blank
correction and interpretation for shales. Hundreds of porous and non-porous, crushed and un-crushed samples are
used to quantify the extent and magnitude of conformance, compression, thermal expansion, and intrusion. Crushed-
rock MIP-derived porosities, bulk densities and grain densities determined from the presented methodology were
compared to those derived from helium porosimetry measurements. These comparisons validate the presented data
reduction and show that MIP porosities are consistently lower than helium-derived total porosities. At pressures less
than 60,000 psi, MIP is unable to investigate the clay-bound portion of these rocks and its porosities more closely
match other effective porosity measurements.
Application of Integrated Core and Multiscale 3-D Image Rock Physics to Characterize Porosity, Permeability,
Capillary Pressure, and Two- and Three-Phase Relative Permeability in the Codell Sandstone, Denver Basin,
Colorado
1 1, 2
A. P. Byrnes; ; S. Zhang; ; L. Canter; 1; M. D. Sonnenfeld; 1;
1. Exploration & Production, Whiting Petroleum Corp., Denver, CO, United States.
2. DigiM Solution LLC, Burlington, MA, United States.
Objectives/Scope: The Codell Sandstone is an important resource. Core analysis (CA) on these rocks accurately
measures porosity (f), permeability (K), saturation (Sw) and capillary pressure (Pc), but is challenged to measure two-
and three-phase relative permeability (2-P and 3-P Kr). Image-based rock physics (IBRP) provides these data. This
study developed an integrated CA-IBRP workflow, with a robust method for integrating porosity phases at two different
resolutions (e.g., FIB-SEM & µCT), to obtain representative elementary volume pseudo-properties for use in reservoir
modeling.
Methods/Procedures/Process: Using a workflow integrating CA and IBRP (focused-ion beam scanning electron
microscopy FIB-SEM & x-ray micro-computed tomography µCT), we characterized samples representative of the
Codell. The workflow starts with identification of representative lithofacies from core/logs. K, f, Sw, Pc, and effective
gas 2-P Kr (Keg) are measured using CA. µCT is then used to characterize the distribution of the five principal µCT-
image phases; grains(G), intergranular f (IGV), large clay (kaolinite) + intercrystalline f (IX), small mixed-layer clay +
microcrystalline f (MIX), and altered lithics + intragranular f (ING). IGV, IX, MIX, and ING are then imaged with FIB-
SEM at respectively appropriate resolutions and quantified using artificial intelligence-based image analytics.
Results/Observations/Conclusions: Using the images, IGV+IX and MIX+ING permeability were computed using
Navier-Stokes-based Computational Fluid Dynamics (CFD). IGV+IX and MIX+ING capillary pressures were computed
using digital porosimetry. Stationary 2-P & 3-P Kr were computed using image-based CFD at multiple saturations
captured by porosimetry simulation. These FIB-SEM scale phase properties (f, K, Pc, 2-P Kr, 3-P Kr) were assigned
into µCT phases to calculate the corresponding pseudo-properties based on an upscaling model. After cross-
calibrations between IBRP-estimated µCT pseudo-properties and CA properties, and confirmation that µCT volumes
are representative elementary volumes, complete 2-P Kr and 3-P Kr curves across the full spectrum of K-f were
constructed. This workflow provided the suite of properties.
Applications/Significance/Novelty: Multiscale IBRP reproduced CA f, K, K-f, Pc, and effective gas 2-P Kr relationships.
Calculation of complete 2-P and 3-P Kr utilized the same methods as 2-P Kr. Cross-validated IBRP and CA is an
extremely powerful tool for microdarcy rocks with heterogeneity. IBRP provides unique data particularly in assessing
Kr relationships and evaluating sensitivity to controlling variables. The workflow utilized in this study can be applied to
many multiphase/multifacies models. Properties computed herein have been used in reservoir- and well-scale
numerical flow modeling to support exploration, completion and production management decisions.
A New Method for Quantifying Cation Exchange Capacity: Application to Organic-Rich Mudrock Formations
1 2
K. Cheng; ; Z. Heidari; ;
States.
Objectives/Scope: Quantification of Cation Exchange Capacity (CEC) is crucial for reliable formation evaluation of
mudrocks, since it impacts borehole geophysical measurements such as electromagnetic measurements as well as
fluid flow in porous media. In this paper, we (a) introduce a new workflow for CEC evaluation in mudrocks with
complex mineralogy, in the presence of multiple types of clay minerals and (b) cross-validate the effective CEC values
from the new method against those directly-measured from the multiple salinity and the wet chemistry methods in
powdered and intact organic-rich mudrock samples.
Methods/Procedures/Process: The introduced method for CEC estimation in mudrocks incorporates quantitative
characterization of mineral composition. We then perform nuclear magnetic resonance (NMR), N2 adsorption-
desorption measurements, and X-ray diffraction (XRD) measurements to obtain volume of hydration water, surface
area, and interlayer spacing variations, respectively. To assess surface area of each mineral, we perform successive
controlled N2 adsorption-desorption measurements and propose an inversion algorithm to estimate surface area of
each mineral. The results are used as inputs to the introduced model for CEC estimation. We first tested the workflow
on synthetic powdered mudrock samples, consisting of a mixture of minerals with known concentrations and then to
actual organic-rich mudrock samples.
Results/Observations/Conclusions: We cross-validated CEC estimates from the new method with those from the wet
chemistry and multiple salinity methods. In the case of synthetic samples, we estimated CEC of samples with 10% of
clay minerals and 90% of non-clay minerals (including 8% of kerogen) and 40% of clay minerals and 60% of non-clay
minerals (including 12% of kerogen) to be 15 and 45 meq/100g, respectively. The actual mudrock samples (3-8%
TOC) contained 20-40% clay with CEC ranging from 30-50 meq/100g. The CEC estimates from the new method in
both cases of synthetic and actual samples were in agreement with those from direct measurements, with errors of
less than 7 meq/100g. The results confirmed that the introduced method can be used for assessment of CEC in
unconventional formations with complex lithology.
Applications/Significance/Novelty: Assessment of CEC is challenging in formations with complex lithology such as
organic-rich mudrocks. Conventional methods such as wet chemistry, multiple salinity, and membrane potential have
limitations associated with their applications in in-situ condition, low-permeability formations, and empirical
correlations. The new method can overcome the aforementioned challenges and provide reliable estimates of CEC in
organic-rich mudrocks. The experimental results demonstrate the reliability of the introduced method for CEC
assessment, which is promising for in-situ and real-time CEC quantification. The outcomes of this paper can be
applied to variety of formations with complex lithology and can be an onsite supplement application for mud logging in
organic-rich mudrocks.
Impacts of Thermal Maturity and Geochemical Properties on Wettability of Kerogen and Organic-rich Mudrocks
1 1
A. Jagadisan; ; Z. Heidari; ;
States.
Objectives/Scope: Wettability of organic-rich mudrocks has significant impacts on hydrocarbon recovery. Kerogen can
constitute a significant fraction of mudrocks and considerably affects their wettability. Although kerogen is often
assumed hydrocarbon-wet in reservoir characterization, our recent rock physics experiments suggest uncertainty in
this assumption. The objectives of this paper are to experimentally quantify wettability of kerogen and organic-rich
mudrocks as a function of thermal maturity and to quantify the influence of kerogen and mudrock geochemical
properties on their wettability.
Methods/Procedures/Process: We first isolated kerogen samples from different organic-rich mudrock formations. We
synthetically matured the mudrock and kerogen samples and determined their wettability at different thermal maturity
levels using sessile drop method. These direct wettability measurements were cross-validated using indirect NMR-
based wettability estimates. In the sessile drop method, a liquid drop was placed on the surface of kerogen and
mudrock samples (formed as discs) and the contact angle made by the liquid-kerogen/mudrock interface was
measured using a goniometer. We finally determined the concentration and thermal maturity of kerogen using
pyrolysis measurements and quantified the impact of kerogen geochemistry as well as mudrock composition on the
wettability of organic-rich mudrock samples.
Results/Observations/Conclusions: The results indicate that kerogen remains oil-wet in high thermal maturities and
becomes water-wet at low thermal maturities. The X-ray Photoelectron Spectroscopy (XPS) measurements showed
an increase in O/C ratio with decrease in thermal maturity of kerogen, which explains the increase in water-wettability
with decrease in thermal maturity. The wettability measurements in organic-rich mudrock samples suggest that they
become mixed-wet at high thermal maturities and water-wet at low thermal maturities. The results confirmed that
concentration of kerogen and clay minerals as well as geochemical properties control wettability of mudrocks. Results
were in agreement with our observations on experimental electromagnetic, NMR, and elastic measurements on
mudrock and kerogen samples.
Applications/Significance/Novelty: We experimentally quantified and documented the correlation between thermal
maturity/geochemical properties of pure kerogen and its wettability (for the first time, to the best of authors’
knowledge) as well as its impact on wettability of mudrocks. The documented results shed light on the reliability of
typical assumptions about kerogen wettability. The outcomes of this paper also enable quantifying mudrock wettability
when the kerogen concentration and thermal maturity is known. The outcomes have also significant contributions in
understanding flow mechanisms in organic-rich mudrocks as well as in developing reliable rock physics models for
interpretation of borehole geophysical measurements (e.g., electromagnetic, acoustic, and NMR measurements) in
organic-rich mudrocks.
Characterization of Transport Properties of Shale Using Novel Theoretical Pore-scale Dendroidal Model
1 1 1
D. Zheng; ; Y. Zapata; ; Z. A. Reza; ;
1. Petroleum, University of Oklahoma, Norman, OK, United States.
Objectives/Scope: Measurement of transport properties of shale in laboratories can be expensive and inaccurate.
Rigorous pore-scale simulation is prohibitively compute-intensive and is limited to sub-micron scale lacking volume-
representativeness. Thus, theoretical pore-network modeling is a routine method to characterize the transport
properties of shales. Most models for unconventional reservoirs are too simplified to capture realistic pore distribution.
This paper presents an improved dendroidal pore-network model based on the analysis of pore-networks extracted
from scanning electron tomographic images.
Methods/Procedures/Process: We adopted non-plateau characteristic of mercury-injection capillary-pressure (MICP)
vs. saturation curve in our acyclic dendroidal model. We infer pore-body distribution from isothermal
adsorption/desorption measurement and characterize pore-throat distribution from mercury drainage experiment. A
pre-determined critical pore-throat radius determines the number and tortuosity of the main flow paths. Each main flow
path and branched path is composed of a large number of stochastically distributed but constrained pore throats and
pore bodies. The resulting dendroidal model directly provides means to determine intrinsic permeability of shales.
Different second-order correlation equations based on Knudsen diffusion effect under different pore size and pressure
determine apparent permeability.
Results/Observations/Conclusions: We tested the predictive ability of the novel dendroidal model using core samples
from Marcellus and Wolfcamp formations. We employed scanning electron tomographic images, MICP and
adsorption/desorption measurements from these samples. The stochastic distribution of pore-throats and -bodies
provides means for uncertainty characterization of the intrinsic permeability. The error of intrinsic permeability
evaluated is in the order of 0.1 to 1 nanoDarcy and agrees well with published literature. Klinkenberg’s gas slippage
correlation could not be used to predict apparent permeability under these flow regimes in shales. However, second
order correlations considering Knudsen diffusion under different flow regimes provide reasonable prediction of
apparent permeability.
Applications/Significance/Novelty: To the best knowledge of the authors, this novel dendroidal pore-network model
utilizes experimental data from isothermal adsorption/desorption and mercury drainage in constructing realistic pore-
network models. Ability to characterize uncertainty in intrinsic permeability at the pore-scale is a novel feature that was
not exploited much in the pore-network models available in the literature. This provides realistic flavor in the pore-
network distribution within theoretical acyclic model framework. Characterization of various transport properties,
including intrinsic, apparent and relative permeability, and other physical processes like dispersion, capillary pressure
hysteresis, and wettability alteration can be performed with ease using our model.
Probing the Wettability of Mudrocks at the Pore-scale Using Nanoparticle Tracers
1 2 2
C. J. Landry; ; M. Prodanović; ; K. Mohanty; ;
1. Center for Petroleum and Geosystems Engineering, University of Texas at Austin, Austin, TX, United States.
2. Hildebrand Department of Petroleum and Geosystems Engineering, University of Texas at Austin, Austin, TX,
United States.
Objectives/Scope: Our objective is to probe the wettability of mudrocks at the pore-scale. The wettability of mudrock
pore spaces are subject to alteration by hydrocarbon adsorption and deposition. This alteration can lead to clay and
mineral grain surfaces being altered from water-wet to oil-wet, which can significantly alter pore-scale multiphase fluid
distributions which ultimately determine the flow properties of the rock. Specifically, these measurements can be
combined with pore-scale flow simulations to estimate the multiphase flow properties of mudrocks.
Methods/Procedures/Process: The pores of mudrocks are typically smaller than 100 nm, and observing fluid behavior
at this scale can be exceeding difficult. Although it is routine to analyze these pore spaces using scanning electron
microscopy imaging, it is typically not feasible to perform high-resolution imaging with fluids present. Therefore we use
water and oil suspensions of nanoparticles to track the migration of fluid in these nanometer-scale pores. After
spontaneous imbibition the fluids are removed from the sample by evaporation leaving behind nanoparticle deposits
that can be readily identified using cross-sectional broad-ion beam polishing and SEM imaging. The presence of these
nanoparticles indicates previous water or oil presence, and subsequently the wettability of the pore space.
Results/Observations/Conclusions: In this first experiments we use hydrophilic nanoporous ceramics, and show that
the nanoparticle deposits are readily identifiable using SEM. Furthermore, we altered the wettability of the nanoporous
ceramics to hydrophobic through aging and demonstrated that water-suspended nanoparticles did not deposit under
hydrophobic conditions (due to the lack of spontaneous imbibition). We then demonstrate the methodology in mudrock
samples that were determined to be strongly oil-wet by bulk analysis. In these samples the water-suspended
nanoparticles were found deposited only in artifact microfractures and large intergranular pores, but are absent from
the clay and organic matter pore spaces. This suggests the clay pore spaces have been altered to oil-wet in this
mudrock sample.
Applications/Significance/Novelty: Although bulk analysis of wettability is routine, it is questionable that measurements
of fluid behavior at the mm-scale on freshly prepared surfaces reflects the wettability of nanometer-scale pores. Any
mm-scale observation of droplet behavior on a mudrock surface is arguably driven by the interaction between the fluid
and freshly exposed mineral surfaces rather than the pores. Our novel methodology seeks to directly measure fluid
behavior in the pores. This information can then be used for empirical correlations of fluid flow properties, pore
structure, and saturation/charge history. As well as be used directly in numerical pore-scale fluid flow simulations to
estimate macroscopic fluid flow properties of mudrocks.
Measurement and Analysis of Wellbore Micro-Losses and Rock Properties While Drilling: A Novel Approach to
Identification of Fractures in the Osage and Meramec Formations of Anadarko Basin
2 2 1 3
B. Chiniwala; ; A. Palakurthy; ; I. Easow; ; E. Russo; ;
1. Formation Evaluation, Geolog Americas Inc, Katy, TX, United States.
2. Drilling & Formation Evaluation, Geolog Americas Inc., Houston, TX, United States.
3. R&D, Fracture Detection and Formation Evaluation, Geolog S.r.l., Milan, Italy.
Objectives/Scope: The Meramec and Osage formations in the Anadarko basin are known to be tight rocks with low-
matrix porosity and potentially minor or major fracture permeability. The Meramec is predominantly composed of
clastics and carbonates, while the Osage has a shallow limestone facies and a deeper mudstone with chert facies. It
is important to identify the location, frequency, size and type of fractures especially for identifying targets for landing
points and for optimizing completions. These fractures could enhance or impede the production performance of the
well.
Methods/Procedures/Process: While operators are used to deploying traditional techniques for identifying bore-hole
fractures via post-drilling wireline logging runs, alternative techniques have been tested in recent years in various
geological settings, especially in the unconventional wells. One of the methods recently tested and validated in
conventional rocks is fracture detection while drilling, by means of surface measurements. This is made possible with
the help of an advanced flow monitoring sensor, installed strategically on the drilling rig mud returns flow-line. The flow
sensor, either a Coriolis or an Electromagnetic (EM) type is capable of detecting micro-losses in the drilling fluid, by
operating with high sensitivity and accuracy.
Results/Observations/Conclusions: Applying an interpretational model to the derived micro-losses, enables the
identification of open fractures, micro-fractures, and induced fractures, occurring with different apertures and densities.
Real-time quantitative mud-gas provides valuable insight on whether the fractures are hydrocarbon bearing or not,
while running X-ray fluorescence (XRF) on drilled cuttings can be used to complement and confirm the changes in
lithology and corelate fractures with specific trace elements, which in-turn aids in anticipating a tight zone or a fluid-
loss zone. The integration of fracture detection data, XRF, and formation gas is combined with wireline log data to
provide additional insights to differentiate conductive fractures from resistive fractures.
Applications/Significance/Novelty: This paper outlines the details of this technique and its workflow which was
implemented to characterize the reservoir sections in both the Meramec and the Osage formations in vertical and
lateral wells drilled within the Anadarko basin and how this interpretation could subsequently aid in optimizing
completions and production in future wells.
Life Isn’t Perfect: The Petrophysical Difficulties of Modeling the Permian
1 1 1 1 1 1
T. Croft; ; A. Blount; ; M. Durand; ; S. Warneke; ; A. McMullen; ; B. Driskill; ;
1. Permian, Shell, Houston, TX, United States.
Objectives/Scope: This paper will cover major lessons learned while developing the petrophysical models used since
Shell’s entry into the Permian basin in 2012. Shell acquired a stake in a competitive basin with multiple zones with
different maturities from exploration to full development. The drive to ramp up fast on various learning curves added
pressure to evaluate quickly and sometimes inefficiently. Early on, challenging dogma and fully integrating across all
disciplines were areas of needed growth. Learnings from past shortcomings in the Permian are now applied to other
assets as they enter development.
Methods/Procedures/Process: The petrophysics team undertook a multi-disciplinary approach to narrow uncertainty
around key parameters centered on extensive data gathering and QC. For instance, porosity is core calibrated utilizing
facies specific regressions. Saturation is based on the Archie equation with a key input being variable RW mapping
calibrated with flowback sampling. This is tied to a successful downhole pressurized wireline fluid sample taken in the
Wolfcamp silts. Pore pressure is a shear sonic vertical effective stress model calibrated to vertical pilot wells with
cemented downhole gauges. Fracture height is determined from geomechanical models that are tied to production
geochemistry, water tracers, and pressure interference tests. Logs are extensively QC’d and key logs are normalized.
Results/Observations/Conclusions: The ability to determine porosity accurately necessitates having consistent
physical rock measurements and consistent log measurements. Normalizing to a consistent dataset is an ongoing
challenge in the Permian basin due to the numerous generations of logging tools and different labs that have been
used in the past. The other primary input to Archie is the formation water resistivity which varies significantly
horizontally and likely vertically around the field based on a produced water flowback. We have proven it is possible to
acquire downhole fluid samples from some of these low perm silt zones in the Permian which provides end points for
the geochemistry.
Applications/Significance/Novelty: This is an integrated story of the ongoing petrophysical evaluation of the Permian
basin. It will cover the evolution of the petrophysical model through time showing what has worked and what has not
and the key data that was used to make those decisions. Of note, is the downhole fluid sampling tool which Shell has
run in the Permian. This fluid sample was important in characterizing end point geochemistry for production allocation.
The basic log QC, maps ,and plots are significant in that they highlight the need to hire more petrophysicist and also
that data quality frequently limits the complexity and accuracy of models.
What Can Good Data Do For You? Machine Learning Applied to Completion Design Optimization
1 1 1
A. Sommer; ; R. Flumerfelt; ; J. Parkhurst; ;
Objectives/Scope: Invited talk - no abstract.

Methods/Procedures/Process: Invited talk - no abstract.
Results/Observations/Conclusions: Invited talk - no abstract.
Applications/Significance/Novelty: Invited talk - no abstract.
Completion Design and Optimization Program Overview in the Permian Basin
1 1 1
O. Jaripatke; ; G. Barzola; ; R. Flumerfelt; ;
Objectives/Scope: Invited talk - no abstract.

Methods/Procedures/Process: Invited talk - no abstract.
Results/Observations/Conclusions: Invited talk - no abstract.
Applications/Significance/Novelty: Invited talk - no abstract.
Rate of Penetration (ROP) Modeling Using Hybrid Models: Deterministic and Machine Learning
1 1 1
C. Hegde; ; C. Soares; ; K. Gray; ;
Objectives/Scope: This study explores three distinct approaches to ROP modeling: deterministic, data-driven, and
hybrid models. Deterministic or physics-based models rely on a fixed equation derived from drilling physical principles
have been the traditional workhorse of the industry. Newer, and more powerful data-driven models utilize machine
learning and analytics to enhance ROP prediction and optimization. However, the improved predictive accuracy
achieved with statistical techniques comes at the expense of model interpretability. To overcome this disadvantage, a
novel modeling technique is introduced.
Methods/Procedures/Process: Hybrid models combine multiple physics-based models using machine learning
algorithms, resulting in improved model accuracy while maintaining inference capabilities between drilling parameters
and ROP. They provide a means to encode the physics of drilling in machine learning algorithms based on physics-
based models.Hybrid models are approached using the principles of stacking or ensembles. Two strategies are
introduced for creating hybrid models: stacking and ensembles of models. Ensembles of models uses different
machine learning algorithms to combine deterministic models. Stacking utilizes a deterministic model as a weak
learning algorithm within an over-branching machine learning ensemble algorithm.
Results/Observations/Conclusions: While this hybrid model predicts ROP with a greater accuracy than physics-based
models alone, purely data-driven models perform marginally better in most cases. On the other hand, the hybrid
models offer higher interpretability, as they are built from deterministic models.
Applications/Significance/Novelty: Hence, the hybrid models are employed for ROP prediction and optimization by
computing ideal drilling operating parameters – weight-on-bit, RPM and flowrate – for each rock formation in the
vertical section of a Bakken shale horizontal well. The study demonstrates that a lower ROP prediction rates can be
achieved using hybrid modeling. Additionally, a total of 20% of active drilling time can be saved by employing these
models for ROP optimization.
Identification and Evaluation of Viscoelastic Surfactants Including Smart Viscoelastic Systems for Generation and
Stabilization of Ultra-Dry N and CO Foam for Fracturing Fluids and Proppant Transport
1 2 1 2 1 1 1 1 1
S. Alzobaidi; ; M. Lotfollahi; ; C. Lu; ; M. Bloom; ; X. Zhang; ; M. Prodanovic; ; K. Johnston; ; D.
1
DiCarlo; ;
1. The University of Texas at Austin, Austin, TX, United States.
Objectives/Scope: Ultralow water content carbon dioxide-in-water (C/W) foams with gas phase volume fractions above
95% (less than 5% water content) are created and stabilized using a combination of surfactant and nanoparticles or
o
just viscoelastic surfactants at temperatures up to 120 C. These ultra-dry foams are promising candidates to replace
high water-content foam and energized fluids in hydraulic fracturing operations to carry and transport proppant with
minimal water leak-off into the matrix.
Methods/Procedures/Process: To test the texture stability of foam, the ultra-dry foam is created in glass beadpacks at
room and high temperatures, and is captured in high-pressure microscopy sapphire cells. The bulk foam stability is
determined through adding proppant to sapphire cells filled with ultradry foam to determine the effect of proppant on
foam coalescence. The accumulated dropped out particles and the foam height is monitored over the time. The foam
is characterized on the ability of carrying proppant and transport through a fracture. At the outlet and inlet of the tubes,
the injected and produced foam/proppant pass through a sapphire viewing cell to determine the amount of proppant
that exits attached to the foam, and the amount of proppant that moves as a separate phase at the bottom of the view
cell.
Results/Observations/Conclusions: The highly viscous and stable foam (apparent viscosity >100 cP) was created over
o
a wide range of temperatures up to 120 C and gas volume fractions up to 99%. The results show mixing foam with
proppant does not alter foam stability significantly due to much larger proppant size compared to surfactant and
nanoparticle to stabilize foam. The flow experiments indicate ultra-dry foam can carry proppant successfully through
fracture conduit with low settlement.
Applications/Significance/Novelty: Using the ultra-dry foams as a fracturing fluid to carry proppants significantly
reduces the need for large quantities of water and the cost of the water disposal compared to other fracturing fluids
used in conventional hydraulic fracturing practice. Moreover, ultra-dry foams reduce the leak of rate to the matrix
considerably and improve the productivity of the fractured walls.
Degradation Study on Materials for Dissolvable Frac Plugs
1 1 2
S. Takahashi; ; A. Shitsukawa; ; M. Okura; ;
1. Kureha Corporation, Tokyo, Japan.
2. Kureha Energy Solutions, Houston, TX, United States.
Objectives/Scope: Dissolvable frac plugs (DFPs) can eliminate milling out, enable longer lateral and reduce
operational risks. Although DFPs have been successfully used in many cases, uncertainty of their degradation has
also been reported. For example, some DFPs did not degrade under certain conditions even though many conditions
were similar to ones with successful degradation. We report the degradation behaviors of dissolvable materials used
in DFPs to understand the reasons for the uncertainty and to share our knowledge about how to deploy DFPs in the
various conditions to maximize the benefits of DFPs.
Methods/Procedures/Process: The two major dissolvable materials for DFPs; polyglycolic acid (PGA) and dissolvable
Magnesium alloy (Mg) were chosen for degradation study. The influence of various factors such as temperature,
salinity, pH and chemicals added to the fracturing and pump down fluid were investigated to understand the
degradation behaviors comprehensively. The temperature range used in the study was 75-300 degF which covers the
bottom hole temperature range in most of the unconventional areas. The salinity dependence study considers salts
such as sulfate and bicarbonate in addition to chloride since these are also commonly found in actual water sources.
Other chemicals studied are biocides, friction reducers and scale inhibitors which are commonly added to frac and
pump down fluids.
Results/Observations/Conclusions: The study results were as follows. (1) PGA has a simpler temperature
dependence than Mg. Although the degradation rate of PGA is proportional to temperature in most cases, Mg has the
degradation rate peak and shows a sharp rate decrease at high temperature in fresh water. (2) The degradation rate
of PGA is quite stable. It is not affected by salinity level, pH and chemicals. (3) Mg degradation has high salinity
dependence as commonly known. The tests also clarified that each salt (chloride, sulfate, bicarbonate) has a different
contribution to the degradation and a deep analysis of water is needed to accurately predict Mg degradation. (4) Mg
degradation rate is affected by chemicals. Certain chemicals such as biocides and friction reducers may decrease the
rate in fresh water.
Applications/Significance/Novelty: The difference in degradation behaviors for major dissolvable materials was
discussed. The knowledge presented in this paper will assist completion engineers about how to use DFPs in their
well completions by sharing with them the following conclusions: (1) Degradation of Mg-based frac plugs is complex
when they are used in fresh water with chemicals, and in fresh water in high temperature wells. To avoid the
uncertainty, the addition of salt or acid is recommended. (2) Degradation of PGA-based frac plugs is relatively simpler
since it is not affected by salinity, pH or chemicals. The PGA frac plugs have easier usability and can be used in both
fresh water and high salinity water fracturing cases.
Accelerating Well Construction Using a Digital Twin Demonstrated on Unconventional Well Data in North America
1 1 1 2
G. S. Saini; ; P. Ashok; ; E. van Oort; ; M. Isbell; ;
2. Hess Corp, Houston, TX, United States.
Objectives/Scope: For many decades now, navigational aids have been able to not only estimate the time to
destination, but also update it in real-time, accounting for ever changing outside circumstances. The digital drilling twin
aims to achieve the same for the well construction process. This is a complex undertaking wherein real-time data and
physics based models are integrated with a historical database to provide real-time recommendations about future
course of action. This paper details the creation, implementation and testing of a drilling digital twin.
Methods/Procedures/Process: A digital twin combines physics based models, real-time data and data from previously
drilled wells. We demonstrate this concept by creating a twin that makes predictions on time to section depth. Time to
section depth depends primarily on two factors: Rate of Penetration (ROP) when drilling and crew performance when
not drilling. Crew performance can be estimated based on historical data taking into account connection and tripping
times for the various crews at different times in their shift. ROP is dependent on downhole formation, hole and bit
condition and drilling parameters, and can be predicted with a physics based model in combination with data from
offset wells. Once the twin was developed, we tested it on historical datasets.
Results/Observations/Conclusions: Making predictions at every instance in time requires concurrently running multiple
models. Therefore the models and parameter estimations were tailored for concurrent computing and multi-threading.
Before drilling starts, an initial estimate of time to section depth is made and then as drilling gets underway, the
prediction gets updated every second. Data from a 5 well pad was used to validate the twin. Four of the wells were
th
used as a historical data reference, and the twin was re-run on the 5 well. The predictions of the twin were then
th
compared to the actual 5 well. The results clearly demonstrate how a twin could have accelerated the well
construction process, by providing better insight.
Applications/Significance/Novelty: Digital twins lend us the ability to look into the future by combining historical data,
real-time data, and a physics based model. They can help predict and thereby prevent unwanted failures, reducing the
likelihood of NPT events. These time based predictions can also help plan the logistics on a rig, and bring about
significant benefits in terms of reducing well construction time. Additional alternate approaches are suggested to
improve efficiency such as when to trip out and replace a bit.
Oilfield Data Analytics: Linking Fracturing Chemistry and Well Productivity
1 1
D. Khvostichenko; ; S. Makarychev-Mikhailov; ;
Objectives/Scope: Hydraulic fracturing is the method of choice for well stimulation in North America. Large-scale
analysis of the effects of various well stimulation parameters on production is of great interest for elucidating
production trends and treatment efficacy. The objective of this work is to create a suite of tools for fracturing fluid
chemistry and production data analytics based on mining and statistical analysis of data for large datasets of wells in
the US.
Methods/Procedures/Process: Chemistry data in FracFocus Chemical Disclosure Registry coupled with monthly
production data from a commercial production database vendor were used as inputs for data analytics tools. Data
parsing methods and algorithms were developed to extract information on fracturing job types, volumes of used
stimulation materials and nature of chemical additives from noisy FracFocus data. Statistical methods were employed
to infer the individual and combined effects of several chemical technologies on production. The unique analysis
covered the majority of wells available in FracFocus (over 100,000 US wells fractured since 2011) and was not limited
to one operator or service company.
Results/Observations/Conclusions: Results are presented in the form of user-friendly dashboards that allow to
elucidate main factors that affect production by basin, play and on a sub-play level in the US. Detailed geographical
granularity was pursued to focus on the effects of completion parameters on production and to minimize the effects of
reservoir quality in the analysis. A number of production data metrics, such as IP, maximum monthly production, and
aggregated cumulative and best production over specified number of months were considered. Various types of
production normalization (perforated interval, proppant mass, fluid volume) and data manipulation for statistical
analysis (e.g. log and Box-Cox transformations), with proper statistical analysis tools (parametric and non-parametric
ANOVA) were applied to data.
Applications/Significance/Novelty: The uniqueness of this work is the combination of fluid chemistry data, production,
and statistical analysis techniques and application to a large dataset of wells not limited to one operator or service
company. On the fracturing technology side, examples include the analysis of the effects of job type (crosslinked,
hybrid, slickwater), channel fracturing techniques, diverting technologies, and flowback additives on production in
different plays. The results allow to evaluate the most important fluid variables for effective production to enable
informed decision making in asset development.
A Fiber Optic-Assisted Multilayer Perceptron Reservoir Modeling: A Machine Learning Approach in Prediction of Gas
Production From the Unconventional Reservoirs, a Case Study From the Marcellus Shale
1 2 1 3 1
P. Kavousi Ghahfarokhi; ; S. Bhattacharya; ; T. Carr; ; A. Shahkarami; ; J. Elliott; ;
1. West Virginia University, Morgantown, WV, United States.
2. University of Alaska Anchorage, Anchorage, AK, United States.
3. Saint Francis University, Loretto, PA, United States.
Objectives/Scope: This study aims to utilize the high frequency data from a fiber-optic monitoring system during multi-
stage stimulation of a horizontal well drilled in the Marcellus Shale in the Appalachian basin of North America. A
predictive data-driven model is developed to understand the well performance and forecast the gas production using a
multi-scale and multi-sensor dataset comprised of fiber optic, petrophysical, geomechanical, drilling, completion, and
surface data. A spatio-temporal database is constructed and used to perform sensitivity analyses to identify the key
drivers for gas production.
Methods/Procedures/Process: The 28-stage horizontal MIP-3H well, drilled in the Marcellus Shale as a part of
Marcellus Shale Energy & Environment Laboratory (MSEEL) project, contains a plethora of multi-scale and multi-
sensor-based spatio-temporal data, such as Distributed Acoustic Sensing, production logs, geomechanical logs,
surface pressure, and surface temperature. A multi-layer perception neural network was trained using 18 input
parameters to predict the daily gas production from each completed stage. The network was trained on 414 days of
gas production and then tested for subsequent 102 days for validation purpose. After fine-tuning network parameters
the model was iterated 500 times. Radial Basis Function-based Neural Network and Support Vector Regression were
used to identify key performance indicators.
Results/Observations/Conclusions: The data-driven MLPNN model shows accuracy of ~95% in predicting daily gas
production. 102 days of the test data shows that the neural network has a mean absolute error of 9.83 MCF/day for
the gas production. The model can precisely predict the daily gas production using various input parameters, such as
fiber-optic measurements, tubing head pressure, casing pressure, stage length, and well logs. Multiple pattern-
recognition technique-based sensitivity analysis gives a robust understanding of the key drivers for hydrocarbon
production. The results show that temperature, pressure, stage length, gamma, Poisson’s ratio, brittleness, and
minimum horizontal stress are the most important parameters influencing gas production.
Applications/Significance/Novelty: This study has two main novelties. The first deals with efficient use of emerging
downhole fiber-optic technology, and the second novelty concerns applying data analytics tools to predict daily gas
production using multi-stage, multi-scale, and multi-sensor-based large database with high accuracy. Our work will
facilitate further application of continuous stream of high-frequency fiber optic data to better understand reservoir
characteristics and well performance. Incorporating fiber optic data in hydrocarbon production prediction has not been
undertaken previously. Our MLPNN model integrating fiber optic, engineering, geoscientific, and surface data, can
predict daily gas production with high accuracy with less computational time and cost, which can be applied to other
projects worldwide.
Extracted Pore-Network Model for Shales Characterizing Geometry of Void Space
1 1
D. Zheng; ; Z. A. Reza; ;
Objectives/Scope: Extremely low permeability of shale has become a challenge in traditional laboratory core analyses,
therefore, the efforts to better understand shale has led to development of advanced digital rock physics. The details
of information in digital rock physics is overwhelming as we are able to access fine-grain details of core data. It poses
new challenge of how to make use of such abundant information. In this work, we developed an integrated data
analytics toolbox that uses many artificial intelligence algorithms to help us get better understanding of shale’s internal
pore-network structure.
Methods/Procedures/Process: The toolbox incorporates novel rock image-recognition algorithms to detect pore and
pore throats, organic content and key mineralogy. It is also capable of extracting important statistics from 3-D nano-
scale digital images and constructing new 3-D models preserving the underlying pore-network structure. For pore-
body and pore-throat detection, we combined rule-based system with deep convolutional networks. It is able to detect
the surface area and ascertain the corresponding aspect ratio facilitating identification of pore/pore throats based on
different volume threshold values. The toolbox can also perform a series of Bayesian operations on multiple 3-D
models to statistically reproduce the observed characteristics in connected pore volume, surface area, organic content
and mineralogy.
Results/Observations/Conclusions: We have utilized the newly developed artificial intelligence toolbox on multiple
samples of shales from Eagle Ford, Marcellus and Wolfcamp formations. We are able to construct high-resolution 3-D
models of pore-body and -throat network structures captured from the scanning electron tomographic (FIB-SEM)
images. The resulting 3-D pore-network models contain detailed specifications of pore-body and pore-throat locations,
coordination numbers depicting connectivity, lithologic association, organic and inorganic association for each pore-
body and pore-throat. This specific format enables us to use the resulting 3-D models in advanced pore-scale multi-
physics simulator capable of ascertaining the effect of various parameters on multiple physical processes at multiple
scales.
Applications/Significance/Novelty: Data analytics of digital rock images can be a very time-consuming process. The
toolbox we developed enables us to automate the analytical processes that greatly increases the efficiency. To the
best knowledge of the authors, algorithms exploiting deep convolutional networks have not been previously used in
the pore-scale modeling of shale formations. The efficient algorithmic framework further provides means to investigate
different physical phenomena acting at different scales in tight to ultra-tight unconventional formations. Using this new
framework, the scaling behavior of key transport properties of shale formations can also be investigated at multiple-
scales from nano- (pore) to meso- (core or log) to macro-scale (reservoir).
Geological Controls on Liquid Hydrocarbon Permeability of Tight Oil and Liquid-Rich Gas Reservoirs
1 1 1 1
C. Song; ; A. Ghanizadeh; ; A. Vahedian; ; C. R. Clarkson; ;
1. Geoscience, University of Calgary, Calgary, AB, Canada.
Objectives/Scope: Quantification of absolute permeability to liquid hydrocarbons is a key parameter for the evaluation
of production potential in tight oil reservoirs. Focusing on multiple prolific tight oil and liquid-rich gas reservoirs in
Western Canada, the primary objectives of this work are to 1) compare absolute (formation oil) and slip-corrected gas
(N ) permeability values for target intervals at identical stress conditions and 2) examine geological controls
2
(organic/inorganic contents, thermal maturity, porosity, pore size distribution, surface area) on liquid hydrocarbon
permeability of tight rocks.
Methods/Procedures/Process: Using a customized pulse-decay liquid permeameter, permeability tests are conducted
with formation oil on selected intact core plugs from the Bakken, Montney and Duvernay formations at varying “in-situ”
stress conditions. The tests are performed on the same core plugs used previously for pulse-decay gas (N )
2
permeability test at identical experimental conditions – therefore, the impact of heterogeneity on liquid/gas
permeability comparisons is mitigated. The impact of various geological controls on (formation) oil permeability is
examined by integration of the latter datasets with a comprehensive suite of petrophysical (porosity, pore size
distribution, surface area) and geochemical (organic/inorganic contents, thermal maturity) datasets previously
collected on the same samples.
Results/Observations/Conclusions: Experimental observations indicate that 1) absolute (formation oil) permeability
values are consistently, up to 50%, lower than the slip-corrected gas (N ) permeability values and 2) the differences
2
between absolute and slip-corrected gas permeability values are larger for samples with lower permeability values.
Compared to gas (N ) permeability values, the absolute permeability values appear to be more sensitive to variation
2
in effective stress – the latter conclusion is in line with some of the previous observations in literature. The absolute
(formation oil) permeability values collected in this study are compared to those previously obtained for
Canadian/American Bakken Formation and the limited datasets that are available for the Canadian Montney
Formation.
Applications/Significance/Novelty: Absolute permeability, while an important control on oil/condensate flow in tight oil
reservoirs, is notoriously difficult and time-consuming to measure for ultra-low permeability rocks in the laboratory. In
particular, the availability of liquid hydrocarbon permeability data is sparse in literature – these data have been
primarily measured for comparatively high-permeability rocks within the millidarcy permeability range (e.g. Bakken
Formation). Performing absolute (formation oil) permeability tests, the insights from the systematic measurements
reported here could be of significant importance for enhanced oil recovery applications in lithotypes with permeabilities
down to the nanodarcy range.
Impacts of Kerogen and Clay on Stress-Dependent Permeability Measurements of Shale Reservoirs
1 1 1
X. Guo; ; C. An; ; J. Killough; ;
Objectives/Scope: Correct measurement of stress-dependent permeability for shale reservoirs is critical for production
prediction and economic evaluation. However, several phenomena such as stress creep and path-dependent stress
still present difficulties in correctly measuring permeability under varying compaction for core samples. In this study,
an improved permeability model is proposed to investigate the impact of kerogen and clay on stress-dependent
permeability measurements by incorporating the stress creep mechanism and net effective stress coefficient into the
constitutive stress-strain equations.
Methods/Procedures/Process: Firstly, numerous experimental data about stress-dependent permeability
measurement and rock composition from various shale samples were extracted from literature and analyzed.
Secondly, a theoretical model was proposed to explain the stress creep mechanism and quantitatively predict its
impacts on permeability measurement. The available experimental data were then used to validate the model. Thirdly,
a mathematical simulator coupling flow and geomechanics was developed and applied to compare the results of
permeability change and production performance during reservoir depletion for two cases with and without considering
the impact of creep. Based on the percentage of kerogen and clay, the different values of net efficient stress
coefficient (not equal to one) were taken into account.
Results/Observations/Conclusions: The micro-fractures were primarily responsible for permeability reduction at
increasing net effective stress, while the existing micro-fractures embedded in kerogen and clay were observed as the
main reason leading to the creep phenomenon. A mathematical equation was derived to calculate the creep strain and
its impacts on permeability. The predicted creep responses at various stress levels match well with recorded
permeability data. The results showed the permeability and hydrocarbon production could be overestimated by 50% if
the impact of creep is not considered. Compared to different stress levels, the creep impact on permeability was more
severe at low net effective stress. The small effective stress coefficient weakened the influence of pore pressure
change on permeability variation.
Applications/Significance/Novelty: This study shows the significant impact of the stress creep on stress-dependent
permeability measurement and production estimation. Core samples need to be carefully treated before experiments;
otherwise a much longer measurement process is required. Based on the rock composition and distribution, an
improved permeability model was put forward to explain the creep mechanism and to better predict stress-dependent
permeability for shale reservoirs containing rich organic and clay components. The net effective stress coefficient
cannot simply be assumed to be equal to one for rich organic and clay shale reservoirs, especially when pore
pressure is changed to alter the net effective stress in permeability measurement experiments.
A Finite-difference Based, Multi-scale Approach for Electromagnetic Digital Rock Modeling
1 1 2 1
M. Yu; ; Y. Wang; ; J. Chen; ; J. Chen; ;
1. Electrical and computer engineering, University of Houston, Houston, TX, United States.
2. Aramco, Houston, TX, United States.
Objectives/Scope: Processing 3D digital images is the first cornerstone in digital rock physics. A particular difficulty is
that the 3D image of rock sample is too large to conduct numerical simulations directly on the original 3D model due to
the limitation of computer resources. Many common methods are utilized to solve this problem, such as sampling and
sub image extraction. However, using these methods may lose some information of the original image. The method
proposed in this paper effectively overcome this drawback since information in every pixel of the original image is
used to achieve high accuracy.
Methods/Procedures/Process: The idea of multi-scale method is proposed in this paper, which is an efficient approach
that seeks to solve several smaller problems by partitioning the original large model into smaller sub models. The
quasi-static simplification is applied on every single sub model independently, which is governed by the Poisson’s
equation. Finite difference method as well as the periodic boundary conditions are applied to get the electrical
properties in each sub model. Then each sub model will be treated as a uniform element to form the final matrix and
the properties obtained in every sub model will be assigned to the corresponding element in the final matrix. After
applying the same numerical modeling on the final matrix, the electrical properties of the whole strucrure can be
obtained.
Results/Observations/Conclusions: The sparse direct solver is utilized to provide standardized results in smaller
models while the iterative solver is used to provide standardized results in larger models. The performance of the
proposed multi-scale method is pretty stable compared to the iterative solver and the direct solver, which has a
relatively weak relationship with frequency but strongly depends on the number of sub models as well as the size of
each sub model. In the numerical simulations for large-scale 3D complicated models, excellent agreements between
the results of multi-scale method and the iterative solver applied directly on the whole structure have been observed.
More significantly, the CPU time and the memory cost of our multi-scale method are remarkably reduced compared to
the iterative solver.
Applications/Significance/Novelty: When dealing with large-scale models, the average error of our multi-scale method
is less than 5% compared to the results of iterative solver. Additionally, the multi-scale method can achieve 150 times
faster than the iterative solver. Moreover, with the limitation of the computer resources, the multi-scale method is still
able to provide accurate results with high efficiency while the iterative solver is no longer applicable when the model is
very large. With powerful capabilities to make the simulation more efficient and accurate, it can be concluded that our
proposed multi-scale method is very suitable for large-scale 3D models, with a lower memory cost and faster
computational time.
Determination of Shale Matrix Permeability Through Methane Dynamic Production Experiments Using Variable
Pressure Gradients
2, 1
K. Fan; ; Y. Li; 1; D. Elsworth; 2; M. Dong; 3, 1; H. Yu; 2; C. Yin; 4; Y. Li; 4;
1. School of Petroleum Engineering, China University of Petroleum, Qingdao, Shandong, China.
2. Departments of Energy and Mineral Engineering, The Pennsylvania State University, State College, PA, United
States.
3. Department of Chemical and Petroleum Engineering, University of Calgary, Calgary, AB, Canada.
4. Chuanqing Drilling Company, CNPC, Chengdu, China.
Objectives/Scope: Shale matrix permeability is a crucial factor for simulating shale gas flow from matrix to fractures,
evaluating reservoir potential and forecasting gas production. Current methods for measuring permeability with helium
using small pressure gradients (SPG) may lead to erroneous results when applied to actual field production with
variable pressure gradients (VPG). This paper aims to establish a method that measures shale matrix permeability
using real gas as close to the actual gas production situation as possible.
Methods/Procedures/Process: An experimental method for shale matrix permeability measurement is presented
known as a gas dynamic desorption-diffusion production test. The gas production pressure is maintained constant for
each production stage with a designed pressure gradient, and the gas production with time is measured. A
mathematical model accommodating desorption uses pseudo-pressure to accommodate the effect of variations of
pressure-dependent viscosity and compressibility. Approximate analytical solutions allow the matrix permeability to be
estimated by nonlinear fitting to match the approximate solution with the recorded gas production data.
Results/Observations/Conclusions: Analysis and comparison of the SPG and VPG methods reveal that the VPG
method is more applicable for shale matrix permeability measurements. Permeability obtained from the VPG method
decreases when particles are crushed into smaller sizes and is affected both by gas desorption and PVT properties.
The production process is divided into three stages according to desorption rate and its rate of change. Permeability
errors obtained from the SPD method without considering the change of gas PVT is several times smaller or greater
than that from VPG method when it is used for gas production with large pressure gradients. Effects of organic matter
and clay content on shale matrix permeability can be further examined after the conduct of a series of experiments for
different type of shale samples.
Applications/Significance/Novelty: Advantages of this method are twofold: one is that the pressure gradients in the
VPG method can be adjusted according to the actual production drop. The effects of gas PVT properties on gas flow
within a certain pressure drop can be considered as a pseudo-pressure, instead of merely using the average values.
The other is that the reservoir gas can be used in the VPG method instead of helium. Although some approximate
treatments are used for establishing the VPG method, these two advantages make the results obtained from the VPG
method closer to reality and may be directly used for gas production project design and production capacity prediction.
Use of Rate-Transient Analysis Techniques for Evaluating Experimental Core Permeability Tests for Unconventional
Reservoirs
1 1 1 1 1 1
C. R. Clarkson; ; A. Vahedian; ; A. Ghanizadeh; ; B. Zanganeh; ; C. Song; ; H. Hamdi; ;
Integrated Effects of Pore Volume Compaction and Connectivity Loss on Intrinsic Permeability of Shale Samples
1 1
D. Davudov; ; R. G. Moghanloo; ;
1. The University of Oklahoma, Norman, OK, United States.
Objectives/Scope: Shale resources have distinct characteristics than conventional reservoir rocks, such as micro-
scale pores, ultra-low permeability, and complex pore network system. Moreover, experimental results indicate that
permeability reduction owing to increased effective pressure might be as much as two orders of magnitude. Usually
this severe reduction is explained by micro-fracture closure, where impact of pore compressibility and connectivity loss
has often been neglected.
Methods/Procedures/Process: In this study we present novel approach, which permeability reduction is described with
combination of three main phenomena: (1) micro crack closure (2) pore shrinkage and (3) connectivity loss due to
bond breakage between interconnected pores. Based on fractal and percolation theories several permeability models,
which is function of pore radius, porosity, and average coordination number are studied here. Effects of pore
shrinkage and connectivity loss are evaluated through the predicted permeability values as a function of effective
stress. After crack dominated permeability region is identified using Walsh model, experimental data for sandstone
and shale samples are utilized to compare connectivity reduction with effective stress.
Results/Observations/Conclusions: Our results indicate that when effects of both pore shrinkage and connectivity loss
are simultaneously analyzed, connectivity loss (as expected) is insignificant in sandstone samples and that
permeability reduction can be explained only by pore compressibility. However, in shale formations, impact of bond
breakage and connectivity loss on permeability reduction is dominant. The results suggest that average coordination
number can decrease as lower as 50 % of the original value when effective stress exceeds 14,000 psi.
Applications/Significance/Novelty: The result of this study suggests that in shale formations permeability reduction
should be corrected to account for micro-crack closure at early stage and for both pore compressibility and
connectivity loss at later stage. This may well change the industry’s predictions of the reservoir performance in
unconventional shale plays. Improved understanding of connectivity loss leads to better production prediction and
much accurate economic predictions
A Validated Digital Rock Workflow to Accurately Predict Apparent Permeability in Tight Rocks
1 1 1 1 1 1 1
J. F. Bautista; ; D. Freed; ; B. Crouse; ; G. Balasubramanian; ; H. Cheng; ; R. Zhang; ; C. Ghodke; ;
1. Oil and Gas, Exa Corporation, Brisbane, CA, United States.
Objectives/Scope: Core analysis of tight rocks is a challenging task. In terms of fluid flow analysis, conventional
technics to determine absolute permeability do not match field observations and provide a wide range of results. The
low permeability of tight rocks makes the time to perform flow-through experiments impractical. Additionally, fluid flow
in tight pores occurs at different flow regimes as determined by the Knudsen number. To calculate the apparent rock
permeability, we use a lattice Boltzmann solver that accounts for high Knudsen number effects and that's been
validated against experiments.
Methods/Procedures/Process: Digital rock physics studies begin by imaging a rock sample, in this case FIB-SEM
imaging is required to properly resolve the rock’s relevant pore space.The imaging process yields a stack of images
used to obtain a 3D volume of the rock. The pore space of the sample is analyzed to compute the porosity, establish
whether there is a connected porous structure, and determine the percolation pore throat radius. With thees
parameters and information of the reservoir conditions and fluid properties, single-phase fluid flow simulations are
performed to obtain the apparent permeability. Simulations for Knudsen numbers up to 10 are performed on 2D and
3D straight channels for comparison to experiments. Also simulations on a shaly sand and a shale sample are
performed.
Results/Observations/Conclusions: Simulations were performed on a 2D straight channel for Knudsen numbers
-3
ranging from 1x10 to 10. The apparent permeability at Kn=10 is two order magnitudes higher than the Darcy
permeability. The simulations at two different resolutions overlap with experiments up until a Knudsen number of 1. At
this point, the results deviate from experiments possibly due to 3D effects in the experiment. For a 3D tube, the
simulation results match the experiments by Perrier et al. very closely, at Kn=10 the apparent permeability is 60 times
larger than the Darcy value.Simulations on a shale sample with a calcite mineral matrix, Kerogen, pores within the
Kerogen, and intragranular pores, were also performed. An order of magnitude variation was seen for the simulations
at relevant subsurface conditions.
Applications/Significance/Novelty: Accurate assessment of permeability in unconventionals, needed for both static and
dynamic reservoir simulation and modeling, can be critically dependent on properly accounting for the effect of high
Knudsen number, especially for gas flow. By accurately modelling fluid flow in unconventional reservoirs, decisions
like well spacing, lateral length, and well completions can be optimized through the proper prediction of drainage
patterns. The digital rock approach provides a significant amount of data in a very short time, but must include high
Knudsen effects to be reliable for low permeability unconventionals. Accurate flow results for unconventional rocks, as
demonstrated here, can lead to incremental hydrocarbon volumes along with improved returns on hydraulically
stimulated wells.
Considerations in Azimuthal Processing and Velocity Inversion for Unconventional Plays
1 2 2
M. J. Perz; ; W. Keller; ; V. Kriechbaum; ;
1. TGS, Houston, TX, United States.
2. Enervest, Ltd., Houston, TX, United States.
Objectives/Scope: A 3D land seismic survey was acquired over a portion of the Austin Chalk in 2015. An azimuthal
velocity inversion (VVAZ) was performed and the resulting interval attributes showed a strong correlation to historical
gas production, especially when they were combined with additional attributes derived from pre-stack inversion. This
strong correlation serves as implicit validation of the azimuth-preserving, AVO-compliant processing flow which was
used to pre-condition the seismic data and to generate the VVAZ attributes. The primary objective here is to showcase
the key elements of that flow.
Methods/Procedures/Process: Key processing flow elements and methodologies include: (i) multi-domain noise
suppression before migration; (ii) 5D interpolation onto a regularly-sampled set of output azimuth “spokes”; (iii) post-
migration noise attenuation operating piecewise on these spokes; (iii) due care in estimating both the azimuth-variant
time-shifts which form the input to the VVAZ inversion, as well as the maximum incident angle restricting those time-
shifts contributing to the inversion; (iv) comparison of two independent VVAZ inversion methodologies, namely the
Zheng approach and the Generalized Dix Inversion; and (v) testing the impact of the number of azimuths output from
5D interpolation on the quality of the final azimuthal interval velocity attributes.
Results/Observations/Conclusions: The above processing and inversion methodolgy yielded azimuthal velocity
attributes which showed significant correlation to production. The azimuth-spoke configuration proved to be an
effective domain for post-migration noise suppression because it forced the organization of azimuth-dependent
coherent noise trains which were then readily removed using existing processing techniques like Radon demultiple.
An offset-adaptive event flattening algorithm was shown to give superior results in time-shift estimation compared to
existing techniques. Testing on real and synthetic data showed that (i) the Generalized Dix Inversion gave a superior
VVAZ result to the Zheng approach and (ii) creating 6 output azimuths after 5D interpolation gave a high quality VVAZ
inversion.
Applications/Significance/Novelty: The above processing flow can be appplied to any wide-azimuth land 3D survey
acquired for unconventional development. We feel the work is especially significant because the flow has been
validated by correlation of resulting VVAZ attributes with production (such correlation being rarely reported in the
literature). There are several elements of novelty. For example, there has never been a public-domain comparison of
the Zheng and Generalized Dix approaches; also we are presenting a new approach to time-shift estimation. Two
companion papers have been submitted to URTeC 2018: Keller et al. discuss result interpretation and Hilburn et al.
discuss uplift obtained by deriving azimuthal attributes via orthorhombic depth migration instead of via the time-
domain approach discussed here.
Unconventional Play Fracture Characterization Through Orthorhombic Depth Model Building
1 1 2 2 2 2 2
G. Hilburn; ; A. Pendharkar; ; W. Keller; ; R. Mott; ; J. Peinado; ; A. Jumper; ; V. Kriechbaum; ;
1. TGS, Houston, TX, United States.
2. EnerVest, Houston, TX, United States.
Objectives/Scope: Seismic time processing results are often used to characterize fractures in unconventional reservoir
areas. A properly conducted depth processing project, using orthorhombic model building, can be leveraged in the
same way. Orthorhombic tomography derives fast velocity directionality, as well as measures of the degree of
azimuthal anisotropy, which relate directly to fracture direction and density. Interpretable depth models, combined with
related imaging improvements, yield better correlation to production data than time results and can help guide drilling
decisions.
Methods/Procedures/Process: Differential moveout in azimuthally sectored common image gathers can be indicative
of azimuthal anisotropy. This anisotropy may arise due to a variety of situations which lead to preferential stress
directions in the subsurface, and is often theoretically related to aligned fracture series, where propagation velocity
across fractures is slower than along fractures. The tilted orthorhombic model definition considers azimuthal velocity
variation, which is disregarded by simpler transversely isotropic models. Therefore, characterizing orthorhombic
anisotropy in fractured reservoir rock can allow for inferences to be made regarding the fracture directionality and
density. This information can be used to help inform decisions for optimal production.
Results/Observations/Conclusions: A case study in Giddings Field through the central Texas Austin Chalk play has
been examined using the described method. Fracture directionality and density derived through orthorhombic model
building matches well with expectations based on previous fracture studies in the area. In some sections of the
survey, particularly in regions of strong azimuthal anisotropy, large scale fault features in coherence volumes may also
be tied to local variations in the velocity field. Where azimuthal anisotropy is weaker, fracture alignment is known to
be more chaotic, and directionality is less constrained. Analysis of depth model parameters and actual production
data yield statistically significant correlations, better than that obtained using time processing results.
Applications/Significance/Novelty: For onshore processing projects, orthorhombic model building provides an uplift in
imaging which may or may not justify its additional cost. However, due to the sensitivity of unconventional production
to physical reservoir properties, additional understanding of the geology of these situations allows for better guidance
of production decisions. Orthorhombic depth processing models are no longer simply processing parameters, but are
interpretable products which, when accurately determined, yield extra insight into the nature of fractures in reservoir
rock. When combined with the imaging improvement from orthorhombic processing, this suite of results helps build
more complete subsurface understanding and demonstrates the usefulness of seismic data in unconventional plays.
Steps for Improving the Utility of Land Seismic Data for Unconventional Reservoirs
1
C. Stork; ;
1. Land Seismic Noise Specialists, Inc, Denver, CO, United States.
Objectives/Scope: This is a broad talk that aims to give the big picture of why seismic often falls short in helping
unconventional reservoirs and provide the key steps needed to improve the utility of land seismic data. One key
limiting factor with seismic data is that it is often too noisy and cannot produce enough good attributes for proper rock
properties inversion. Processing is not as good as people think it is. Non-experts need to understand noise is a key
problem not to be taken for granted. Things are changing fast in this field, which influences how we use seismic data.
Methods/Procedures/Process: Many rock property aspects go into economically producing an unconventional
reservoir. Seismic data is challenged to help unconventional reservoirs because we generally invert for only 1-2 rock
properties and seismic rock properties do not correlate directly with important reservoir rock properties. But, by
addressing the noise, we can produce 4-8 good seismic rock properties and search for empirical correlations with
reservoir rock properties. Empirical correlations are possible because rock property variations often happen in
combination with other rock property variations for complex reasons specific to a field. Empirical correlation with log or
production data has the advantage of reducing interpretation complexity and being faster than a complete analysis of
rock properties.
Results/Observations/Conclusions: Recent tests with novel seismic acquisition and processing of data from the
Permian Basin show it is possible to significantly improve the noise reduction of seismic data and extract 6-8 good
pre-stack seismic attributes. These additional attributes vastly improve the ability to correlate patterns between
seismic data and log or production data. One attribute can classify 5 patterns, but 4 attributes can classify 5^4=625
patterns. The improved correlations change how well seismic can find sweet spots. However, there is still work to be
done to make this a routine & reliable process.
Applications/Significance/Novelty: We are still searching for how seismic data can reliably predict sweet spots ahead
of the drill bit in a timely fashion. These results show there is promise and point the way.
Coherence Attribute Applications on Seismic Data in Various Guises
1 2
S. Chopra; ; K. Marfurt; ;
1. Reservoir Services, Arcis Seismic Solutions/TGS, Calgary, AB, Canada.
2. The University of Oklahoma, Norman, OK, United States.
Objectives/Scope: The interpretation of discrete stratigraphic features on coherence displays is limited by the
bandwidth of the input seismic data. There are different ways to modify the spectrum of the input data in addition to
simple spectral balancing, using the amplitude-volume technique (AVT), taking the derivative of the input amplitude,
and thin-bed spectral inversion. Spectral voice components, also shows features with significantly more clarity at
certain frequencies. The objective is to compare coherence displays using these different techniques and show their
added value in seismic interpretation.
Methods/Procedures/Process: If the underlying reflectivity can be considered to be random, after spectral balancing,
the spectral contributions of the seismic wavelet are largely removed, allowing the analysis of tuned reflections that
occur at layers exhibiting quarter wavelength thickness. Quantitative measurement of such tuning is achieved through
different spectral decomposition methods where one computes a suite of spectral magnitude and phase components
obtained from the original broadband seismic data. Less commonly used by interpreters, the spectral voice
components often provide additional insight into the subsurface features. Going one step further, coherence computed
from such spectral voice components can highlight discontinuities that are preferentially imaged by a given spectral
component.
Results/Observations/Conclusions: Our comparison of the equivalent time slice displays from the energy ratio
coherence volumes generated on the different versions of the input seismic data allows us to infer, (a) coherence on
spectrally balanced input seismic data yields better lineament detail, (b) coherence on voice components highlights
the discontinuities at different frequencies that show better definition, which can be helpful for their interpretation, (c)
multispectral coherence displays show crisper definition of lineaments and so are useful, (d) coherence run on the
versions of the data discussed above after AVT shows superior definition of lineaments and hence we recommend
should be used in their interpretation.
Applications/Significance/Novelty: The application of amplitude-volume-technique (AVT) workflow entails the
calculation of the root-mean-square of the seismic amplitudes in a definite analysis window and then rotate the phase
o
of the data by – 90 , by using the mathematical operation of Hilbert transform. Such a calculation yields somewhat
higher amplitudes of the frequencies in the bandwidth of the data. We demonstrate the application of the coherence
attribute on input seismic data, on the same data after spectral balancing, spectral magnitude and voice components
obtained after carrying out spectral decomposition, on seismic data passed through the AVT workflow, as well as
multispectral coherence, and compare the results. The latter process as well as the comparison of the other
mentioned above is a new application.
Fracture Productivity Prediction Considering Natural Fracture Formation Proximal to Fault Damage Zone
1 1 1 1
L. Hernandez; ; T. Ramsay; ; J. Li; ; M. Erdogan; ;
Objectives/Scope: Increases in production from unconventional reservoirs can be attributed in some instances to the
impact of well spacing and induced fracture placement. A firm understanding of the abundance and orientation of
natural fractures is often limited yet is important to characterizing unexpected levels of recovery. The uncertainty in the
characterization of natural fracture density and orientation is typically high; thus, value exists in understanding natural
fracture formation given plausible subsurface structural control mechanisms.
Methods/Procedures/Process: Elastic-plastic fracture mechanics are reviewed in the formation of natural fractures
coincident to the fault damage zone. It is evident that disparate fracture orientations occur in the fault process zone
exclusively as a function of energy release rate. A fault likelihood attribute is used to spatially characterize natural
fractures considering the existence of a correlation between faulting and natural fracture formation. Subsurface
consistency between the flow simulation model and the seismic attributes is achieved by way of a petro-elastic model
embedded in the reservoir simulator. After computing time-dependent elastic properties, their dimensionless
representation is used to characterize matrix fracture interaction.
Results/Observations/Conclusions: Primary fracture plane parallelism with respect to the fault damage zone is
anticipated when the existing stress field simultaneously dictates the formation of faults and fractures. Fracture plane
orthogonality relative to the fault zone represents fracture propagation away from the strongly non-proportional plastic
loading zone post-stress field re-orientation. This fracture plane re-orientation is considered as a function of the stress
field orientation after the fracture energy was reached in the formation of natural fractures. An observed correlation
between the pressure profile and the computed dimensionless elastic property suggests the potential for pre- and
post-fracture seismic surveys to better facilitate overall asset management.
Applications/Significance/Novelty: Production from an Eagle Ford reservoir is analyzed under disparate production
scenarios considering the impact of natural fracture density and orientation in the presence of proximal faults. Fault
likelihood and petro-elastic modeling are combined in an integrated multidisciplinary analysis to preserve subsurface
continuity between seismic attributes and flow simulation. Results demonstrate the value of a consistent subsurface
description that promotes increased efficiency in well and hydraulic fracture placement, leading to optimal drilling and
completion costs.
Time-Lapse Petro-Elastic and Seismic Modeling to Evaluate Fracturing Efficiency in Low-Permeability Reservoirs
1 1 2 3 1 1
M. Alfi; ; Z. Chai; ; A. Pradhan; ; T. Ramsay; ; M. Barrufet; ; J. Killough; ;
1. Petroleum Engineering Department, Texas A&M University, College Station, TX, United States.
2. Energy Resources Engineering Department, Stanford University, Stanford, CA, United States.
Objectives/Scope: Hydraulic fracturing in shale reservoirs represents fracture network propagation, which differs from
traditional planar bi-wing fractures that are considered to be developed in the target formation. This complexity is a
result of heterogeneous formations as well as the interaction of induced and natural fractures. Also, the well design
and execution can be less than ideal, resulting in emerging problems of bypassed and stranded pay in fractured wells
with “sweet spots.” This paper discusses simulation-based petro-elastic modeling of the reservoir to identify less-
effective fractured zones.
Methods/Procedures/Process: An essential component of time-lapse seismic reservoir monitoring should entail
validating simulated elastic reservoir properties at two or more points in time. Reservoir elasticity is affected by
lithology, fluid content, and variations in pore pressure. Comparing attributes, such as acoustic impedance, enables
the discrimination of rock and fluid properties and their changes over time. In this study, the fracture network is
accounted for using the corner-point embedded discrete fracture modeling (cEDFM) approach with some designated
bypassed (unfractured) zones. The acoustic impedance of the media is calculated before and after production.
Alteration of such property over time signals fluid movement, phase alteration in that region of the reservoir, and an
effective fracture treatment.
Results/Observations/Conclusions: Results conclude that in places where fracturing is effective, acoustic impedance
decreases with production caused by a decrease in density and P-wave velocity. Evolution of gas phase in liquid-rich
shale further decreases the acoustic impedance. Changes in acoustic impedance are visible if seismic amplitudes are
large enough to overcome the masking caused by noise levels or if modeling is conducted in the dimensionless
impedance space. A sensitivity study on effects of different factors (fracture density, elastic properties of minerals, and
rock petrophysical properties) was performed to find environments where four-dimensional (4D) seismic is effective
and to find an optimum acquisition frequency. The corresponding effects on 4D seismic data were validated using
seismic-forward modeling.
Applications/Significance/Novelty: Time-lapse seismic reservoir monitoring has great potential to image fluid
movements in reservoirs. Based on the numerical simulation, the applicability and influence factors of time-lapse
seismic in identifying bypassed zones are studied, including unfractured and poorly fractured regions or low-
conductivity fractures resulting from proppant or fluid problems. While finding the untouched zones is significant, this
study provides recommendations on favorable settings for applicability of time-lapse seismic surveys and the optimum
acquisition frequency. Although previously used for fluid, rock, and fracture characterization purposes, this study
presents the innovative idea of analyzing fracturing efficiency by monitoring elastic properties of the reservoir after
production starts.
First Unconventional Play From Peruvian Northeast: Muerto Formation
1 1 1 1 2
W. Morales Paetan; ; A. Arguedas; ; J. Rodríguez; ; H. Taipe; ; J. Porlles; ;
1. Petroleum and Natural Gas, Universidad Nacional de Ingeniería, Lima, Lima, Peru.
2. University of Utah, Salt Lake City, UT, United States.
Objectives/Scope: This paper presents the success criteria worked out and the integration of data for a high grading
the local sweetspots of abandoned areas, all for the first set of horizontal wells that allow us appraising this
unconventional place in the northwest from Peru. This will allow Peru to be included in the map of Energy Information
Administration of the United States (EIA), thus Peru can be considered by investors for future projects.
Methods/Procedures/Process: As part of appraising the potential of this formation, as a resource play, a
comprehensive success criterion has been worked out for location selection. An Integration of all available geo-
scientific data, geochemical, a new seismic interpretation, petrophysical analysis as well as electrical logs
interpretations has been carried out. The above data integration and analysis of were combined with the criteria,
leading to the selection of sweetspot for planing the first dedicated horizontal wells targeted on this formation.
Results/Observations/Conclusions: The early Cretaceous Muerto Formation deposited throughout the Lancones Basin
is a TOC content ranging from 0.8 to 4.07% by weight and has an average TOC of about 2.33% by weight. This zone
is favorably located to contain a potential in resources from shale gas, condensate and black oil in the basin. This
formation is located at a depth from 6132 to 22667 ft with type of siliciclastic rocks and carbonates, and the thickness
varies from 380 to 3850 ft. Lancones basin has one of the main source rocks for most of the oil and gas fields in
northeastern Peru. These data allow us to compare with respect to the Vaca Muerta formation with success in terms
of TOC, thickness, sedimentary environment and age, which offers an excellent opportunity to evaluate this formation.
Applications/Significance/Novelty: The use of old data with modern methodologies in discarded areas offers a
successful localization of a first unconventional play and the opening of a new research window.
Currently, Peru is an oil importer. In the past fifteen years, the oil production has been declining and few prospects
were founded, in part because social and environmental problems have caused many projects to be canceled or
postponed. From an international perspective, the Energy Information Administration of the United States (EIA) has a
map with basins of shale oil and gas, but information concerning Peru does not exist yet, which affects investor
choices. However, neighboring countries such as Colombia, Brazil, and Argentina have enough information available
for investors to look at and evaluate.
A Physical Decline Curve for Fractured Horizontal Wells
1 1
V. Artus; ; O. Houzé; ;
1. KAPPA Engineering, HOUSTON, TX, United States.
Objectives/Scope: A new decline curve compatible with the successive emergence of the transient linear, transition
and pseudo-steady state flow regimes is proposed, for fractured horizontal wells in unconventional reservoirs. By
construction, this curve is fully consistent with the predictions of the physical SRV-bounded model.
Methods/Procedures/Process: In order to ensure the physical consistency, the decline curve is obtained by
numerically integrating a base function which reproduces the continuous succession of the expected flow regimes, in
the material balance time domain. 3 parameters are sufficient to fully describe the curve.
In order to generate the curve, internal derivation steps involve an integration in the material balance time domain.
However, from a user’s perspective, the matching process (of the instantaneous rate or of the cumulative) and the
production forecast are performed in the real time domain directly. The actual pressure history, if available, can also
be accounted for during the derivation of the curve for a better accuracy.
Once the decline curve has been matched, the 3 parameters of the curve can be used to infer combinations of the
physical parameters (number of fractures, half length, completion length and permeability).
Results/Observations/Conclusions: In this paper, the main characteristics of the SRV-bounded model are reviewed,
since physical considerations from this model are key to the construction of the decline curve. Then, the calculation
steps leading to the decline curve are detailed. Comparison of the results from the decline curve and the results
obtained with the SRV-bounded model show excellent agreement. Several application examples are shown.
For cases where the production time is short and flow is still in linear mode, we show that the parameters of the
decline curve can be fitted so that a conservative estimate is ensured. Finally, we show how to estimate physical
parameters from the parameters of the decline curve.
Applications/Significance/Novelty: The proposed decline curve is compatible with the successive emergence of
various flow regimes and is fully consistent with the predictions of the physical SRV-bounded model. Its parameters
allow to estimate combinations of the parameters of a physical model. As a consequence, this proposed approach
bridges the gap between empirical decline curve techniques and model-based approaches used for rate transient
analysis.
Variation of Hyperbolic-b-parameter for Unconventional Reservoirs, and 3-Segment Hyperbolic Decline Curve Model
1 1 1 1
J. R. Ewert; ; S. Varma; ; H. Tabatabaie; ; L. Mattar; ;
1. IHS Markit, Calgary, AB, Canada.
Objectives/Scope: The variation of-b-parameter with time for producing wells was studied and resulted in the following
conclusions:
It is extremely difficult to determine-b-parameter uniquely because of the associated noise in the data.
Within the scatter of production data, many values of-b-parameter can give an acceptable production history match.
This non-uniqueness of b, coupled with the data scatter, presents an opportunity for developing a decline
methodology that honors engineering judgement, yet remains both familiar and simple to use.
Methods/Procedures/Process: Because the focus of this study was on unconventional wells, reservoir models of
multiple-fractured horizontal wells in shales were studied in detail. Production forecasts were generated, and the-b-
parameter value derived at every point in time, b(t). Unexpected patterns of b(t) were observed. However, once the
prevailing flow regimes were clearly understood, these patterns could be explained.
Results/Observations/Conclusions: This study showed that:
Clearly, b(t) was not constant; sometimes it was sigmoidal, and often it was multisigmoidal
When noise was added to the synthetic production forecasts to represent real production data, a much simpler
representation of-b-parameter was possible, and resulted in acceptable matches of the noisy production history.
The noise in the data combined with the non-uniqueness of b, allowed us to apply a non-complex solution to a
complex problem. In other words, in this situation, data scatter along with the forgiving nature of b, was “our friend”.
Applications/Significance/Novelty: The commonly used hyperbolic equation assumes that-b-parameter is constant. We
have maintained the simplicity of this equation, and have developed a 3-Segment Hyperbolic Decline procedure. This
proposed method:
has the advantage of familiarity – it uses the well-known traditional hyperbolic equation
allows the-b-parameter to vary in a simple manner for modeling a complex flow situation.
is able to represent production history, within the tolerance of the noisy data.
The time of the transition segment is a function of the reservoir description, and requires further study
Criteria for Proper Production Decline Models and Algorithm for Decline Curve Parameter Inference
1 1 2 1
P. Zhou; ; Y. Pan; ; H. Sang; ; J. Lee; ;
2. Department of Statistics, Texas A&M University, College Station, TX, United States.
Objectives/Scope: The major objective of this paper is to present important criteria that are necessary for a robust and
accurate production decline model. A secondary objective is to present a general method, based rigorously on
fundamentals, that yields a new model which satisfies these criteria. We will show that the resulting model is useful for
forecasting in ultra-low permeability reservoirs, such as shales.
Methods/Procedures/Process: We conducted a detailed study of many production decline models that have been
proposed for ultra-low permeability reservoir production data analysis. We found a set of principles that are physically
important for the accuracy of any production decline model. On the basis of these important principles, we then
developed a general method which can be applied to construct a nonparametric decline model that honors the well
production histories from ultra-low permeability reservoir. Finally, we make forecasting on the future hydrocarbon
production using the optimal solution of initial production rate, initial decline rate and Arps b factor (a function of time)
obtained from the resulting nonparametric model.
Results/Observations/Conclusions: We determined that necessary criteria for a production decline model are (1)
estimated ultimate recovery (EUR) must be bounded at long times; (2) the loss-ratio parameter (inverse of production
decline rate) must be always finite and nonnegative; and (3) consistent with field experience, the derivative of loss
ratio (Arps b factor) must be nonnegative. In addition, we tested the model from our method with synthetic simulation
study and the results illustrate more accuracy compared to modified hyperbolic model with reference to the true
simulation result. We also validated our approach with field data, and the accuracy of the production forecasts indicate
the advantages of this new modelling approach.
Applications/Significance/Novelty: Many empirical production decline models have been proposed in the literature, but
most fail to meet basic necessary criteria for confident forecasting in ultra-low permeability reservoirs. Importantly,
even fewer models have identified and met the criteria necessary for confident application. Our model-building
procedure is a unique and novel approach which fill this gap in knowledge. Our method can also be applied to
conventional reservoirs.
Variable Exponential Decline – Modified Arps to Characterize Unconventional Shale Production Performance
1 1 1 1
D. Devegowda; ; I. Gupta; ; C. Rai; ; C. Sondergeld; ;
1. MPGE, University of Oklahoma, Norman, OK, United States.
Objectives/Scope: Decline curve analyses (DCA) and decline curve models should be capable of responding to the
expected flow regimes associated with multi-stage hydraulically fractured horizontal wells while also reflecting
changes in reservoir and fracture properties such as fracture conductivity and matrix permeability in response to
changes in the subsurface stress changes. This study critically analyzes classical and some of the more recently
developed DCA techniques and proposes a modified, but easy to implement, decline curve model that accounts for
the requirements specified above.
Methods/Procedures/Process: In a shale well, the coupled effects of change in reservoir and completion properties
with time and flow regimes govern the production decline rate. For instance, the decline rate may be a result of
declining fracture permeability (power-law function). This study develops a new simplified decline curve equation by
modifying the existing Arps exponential decline equation, where the constant decline rate is replaced by a power-law
function decline rate derived from analyses of production data from several shale wells. In theory, the basic premise is
similar to the Power-Law model and Stretched Exponential model; in practice, the proposed model is seen to be more
reliable in shales where stress/geomechanical effects dominate, such as in the Haynesville.
Results/Observations/Conclusions: The new decline curve method obtained by modifying existing Arps’ exponential
equation is referred to as Variable Decline Modified Arps (VDMA) model. Twenty wells from the Haynesville shale
serve as test dataset for analyses of the performance of the VDMA model. The results indicate lesser than 3% error in
predicting long-term well production. The fit of the VDMA model was also verified against the radial flow, linear flow,
and pseudo-steady state flow profiles generated using analytical solutions. The initial decline rate and decline
exponent are easily modified to best match the analytical production profiles. The results showed that VDMA model
gives an excellent fit to linear and radial flow regimes, and gives a reasonable fit to pseudo-steady state flow regimes.
Applications/Significance/Novelty: This study demonstrates that the combined effect of flow regimes and changes in
fracture conductivity can be reliably represented with a power-law function. The new simplified decline curve equation
is very robust and easy to apply. While other decline curve models such as the Power-law and Stretched Exponential
decline follow a similar rationale, the VDMA is better able to match well performance in resource plays such as the
Haynesville shale.
A Model-Based Diagnostic Workflow for Time-Rate Performance of Unconventional Wells
1, 2
D. S. Fulford; ;
Objectives/Scope: Analysis of production data is a critical component for performance estimation in ultra-low
permeability unconventional reservoirs. Diagnosis of early-time infinite-acting transient flow is important for accurate
estimation of future unobserved rate-time data. However, well clean-up effects and lack of pressure data may mask
the early-time flow regimes, prohibiting accurate diagnosis of the end of the transient regime by graphical techniques.
A model-based approach can improve forecast accuracy by bounding the range of possible interpretations of
production performance.
Methods/Procedures/Process: The familiar Arps hyperbolic model is re-cast in terms of the colloquially named “linear
flow parameter”, relating the initial rate and initial decline parameters with a functional relationship. Computation of
initial rate can be performed from a diagnostic fit of a half-slope trend line which corresponds to transient linear flow.
Additionally, the early-time masking of linear flow due to well clean-up may be quantified by computing a diagnostic
half-slope trend from the hyperbolic model parameters. Functional relationships are provided to compute additional
diagnostics on the basis of material balance time and a novel transient superposition rate function. Parameterization
of linear flow duration allows creation of a coupled GOR or CGR forecast for secondary phase yields.
Results/Observations/Conclusions: Long well clean-up periods and short duration transient regimes present a
challenge for making consistent production performance interpretations, because graphical techniques assume ideal
conditions that may not exist in field data. Moreover, diagnosis of flow regime change must take place rather late after-
the-fact, as confidence of visual deviation from a straight-line trend is required. An empirical model-based workflow
that maps the fluid flow solution for linear reservoirs to a simple empirical model assists with the creation of
performance estimates that achieve greater consistency with those from more rigorous rate-transient analysis
workflows. The additional diagnostics, which are computed from the model parameters, alleviate some of the
difficulties caused by noisy data sets.
Applications/Significance/Novelty: The workflow is simple and robust, while respecting the physics of fluid flow theory.
Various successful rate-transient analysis concepts and current best-practices for production performance estimation
are encoded into the model parameters, and therefore may be easily applied. The diagnostics that result are not
simply additional plots or manipulation of scales, but functional computations mapped to common Arps hyperbolic
parameters. Given the presentation of numerous diagnostics – for a given model input, an exact set of interrelated
diagnostics is provided – forecast accuracy should improve. The workflow is presented in terms that the typical
reservoir engineer is already familiar with. This enables direct application in day-to-day work for estimation of future
well performance.
Straightforward Representative Fluid Flow Models for Complex Fracture Networks in Unconventional Reservoirs
1
J. A. Acuna; ;
1. ETC, Chevron, Houston, TX, United States.
Objectives/Scope: This paper presents different methodologies to create simple reservoir models that reproduce the
hydraulic behavior of complex networks of fractures. These simple models are constructed quickly using results from
the RTA method presented by Acuna (2017). The simplified models can be run in standard reservoir simulation
software and produce history matches of excellent quality.
Methods/Procedures/Process: The method is based on the RTA methodology recently developed by Acuña (2017).
This approach shows that the Characteristic Flow Volume (CFV) is the fundamental property that defines fluid flow
behavior for complex networks of fractures and that different reservoirs with the same CFV give the same hydraulic
behavior. We use this premise to construct models that share the same CFV of complex fracture systems but look
completely different and are simpler to construct and simulate numerically.
Results/Observations/Conclusions: We show how the simplified models give similar fluid flow performance to complex
fracture networks. The simplified models have the same reservoir volume distribution with distance from the fractures
as compared to the complex networks of fractures. Since the CFV of a given system can be calculated from
production data, it is possible to quickly construct and run models that match well performance with good precision.
Applications/Significance/Novelty: This new approach has been validated numerically with discrete fracture networks
models and it can match not only the transient flow behavior of complex fracture networks but also has the capability
to forecast the long-term performance as accurately as numerical discrete fracture models. We apply the approach to
field cases.
A Novel Production Forecast Model for Hydraulic Fractured Wells Based on Anomalous Transport Phenomenon due
to the Fractal Geometry of the Fracture Networks
1 1 1
S. Liu; ; H. Li; ; P. P. Valko; ;
Objectives/Scope: A novel model for production forecast of the hydraulically fractured horizontal well completed in the
unconventional reservoir is presented, which is based on the anomalous transport description of the flow in the
formation due to the fractal nature of the fracture networks. Comparison with some conventional decline curve models
is made to highlight the advantages and disadvantages of the proposed method.
Methods/Procedures/Process: The new model is formulated with the assumption of anomalous transport, which
results in a flux equation with fractional ordered time derivative and the corresponding “fractional diffusivity equation”.
The constant wellbore pressure solution of this model with several infinite conductivity propped hydraulic fractures
along a horizontal well is expressed explicitly as a function of three undetermined parameters: the anomalous diffusion
exponent (α), characteristic time (τ), and estimated ultimate recovery (EUR). A procedure is developed to fit the model
to production data and obtain α and τ, with which the EUR of the well and the properties of the formation and fracture
network can be determined.
Results/Observations/Conclusions: In this work, we firstly formulate and solve the anomalous transport problem in the
drainage area of one of infinite-conductive hydraulic fractures along a horizontal well. The solution contains three
undetermined parameters: α, τ, and EUR. The relevant features of the proposed model are investigated. Then a
production data-fitting process is recommended based on the feature of the model. Next, the production data for
several wells in a specific unconventional play is fitted to attain the corresponding parameters. Comparison with
conventional rate transient analysis models and other decline models is provided. The advantages and disadvantages
of the suggested approach are summarized.
Applications/Significance/Novelty: The model in this paper has two significant features: the naturally bounded EUR,
and the physically meaningful parameters that distinguish the treatment factors from the formation factors. For
instance, available information on well and fracturing treatments is directly incorporated into the fitting function. The
latter feature puts the production forecast on a more solid base than using just an empirical mathematical formula for
the description of the production history. Also, the newly introduced parameter α explicitly characterizes the fracture
networks, consisting of hydraulic, induced, and natural fractures, which dominate the flow through the complex
fractured formation.
Advances in Borehole Imaging in Unconventional Reservoirs
1 1 2 2
M. Morys; ; S. Knizhnik; ; A. R. Duncan; ; B. E. Tingey; ;
1. PetroMar Technologies, Inc., Exton, PA, United States.
2. Task Fronterra Geoscience, Aberdeen, United Kingdom.
Objectives/Scope: High resolution borehole images have been highly prized by petrohysicists, geologists and
reservoir engineers. They bring significant value to the formation evaluation workflow especially in unconventional
applications. Despite such high value the use of image data has been limited in unconventional reservoirs due to
technical challenges brought by widespread use of Oil Based Mud. Existing LWD imaging technology became
unusable due to lack of electrical conductivity in OBM. Wireline images gained limited traction due to very high cost
and excessive risk of deployment in long horizontal wells.
Methods/Procedures/Process: We present a new approach to high resolution ultrasonic borehole imaging. High image
fidelity and very good signal-to-noise ratio are possible thanks to the latest advances in dynamically focused sensor
design and signal processing taking advantage of immense computing power available in downhole tools.
Implementation in an LWD tool has been used for evaluation of the new technology.
The tools’ design has been optimized for operation while drilling or wiping in heavy oil based mud which until now
posed an insurmountable challenge to borehole imaging while drilling. It is equipped with two depth matched back to
back dynamically focused transducers acquiring image and caliper data with 360 degree borehole coverage.
Results/Observations/Conclusions: This paper discusses tool fundamentals, performance characteristics based on
laboratory and field test data. The tool delivers multiple simultaneously acquired images comprising high resolution
caliper and acoustic impedance images, both from two transducers operating in tandem. With acquisition time under
1ms per pixel the tool can handle high RPM and stick-slip conditions with minimal degradation in image quality.
Operation in high weight mud up to 18 lbs/gal is demonstrated. The sensor resolution allows identification of individual
features separated by approximately 0.25in and detection of fractures and vugs as small as 0.05 in. Field test data
demonstrates the ability to detect natural and drilling induced fractures, faults, breakouts and bed boundaries.
Applications/Significance/Novelty: Field trials and subsequent commercial applications demonstrated consistently
good imaging performance while drilling in horizontal wells using heavy OBM. This is the first demonstrated case of
successful LWD imaging under such conditions. LWD conveyance brings down the logging cost and deployment risks
to the point where the value of image data outweighs the cost for many wells. Applications enabled by the new
technology are also discussed including: geosteering, borehole stability, reservoir quality evaluation and modeling,
completion planning and production enhancement. In particular, the ability to identify fractures and fracture density is
used for identification of like reservoirs for hydraulic fracturing design.
Laboratory Evaluation of Apparent Conductivity of UFP
1 1 1 1 2
J. C. Montalvo; ; D. Hu; ; U. Inyang; ; R. Dusterhoft; ; M. Apostolopoulou; ;
2. University College London, London, United Kingdom.
Objectives/Scope: Presently, hydraulic fracturing operations in North America are used in unconventional reservoirs
such as tight gas and oil reservoirs and organic-rich shale plays that have micro- to sub-microdarcy permeability, and
require significant stimulation to be commercially viable. Hydraulic fracturing helps to unlock these reservoirs by
creating a fracture network to access flow paths to the resources.
Methods/Procedures/Process: This paper evaluates the mechanism by which a non-uniform partial monolayer of small
sized proppants with particle sizes much smaller than 100-mesh can significantly enhance the conductivity of the
complex microfracture network.
To evaluate much smaller proppant particles, a new test design was used to measure fracture conductivity.
Experiments were conducted to measure the effective conductivity of unpropped microfractures and microfractures
propped with very small quantities of ultra-fine particulates (UFP) on outcrop samples from Marcellus, Mancos, Eagle
Ford, and Barnett formations, including an actual Wolfcamp core. Geomechanical testing was performed on formation
samples to determine their mechanical properties.
Results/Observations/Conclusions: Using computed tomography (CT) scans the created fracture faces were studied
pre- and post- UFP treatment to examine UFP distribution. X-ray diffraction (XRD) was used to quantitatively assess
reservoir rock minerology to identify potential rock-fluid interactions that could affect long-term conductivity.
This study evaluates the impact of effective stress on the created microfracture permeability, helping to understand
the influence of UFPs on subsequent microfracture conductivity. The effective conductivity is also influenced by the
UFP concentration, core minerology, and sample anisotropy of the formation samples used for this study.
Applications/Significance/Novelty: This paper summarizes test results complimented by computer simulations to

establish well productivity effects. The experimental results are then compared to actual well production as a means of
validating the study. Field applications of UFP materials have proven to be successful in increasing well productivity
and sustaining higher productivity over a significantly longer time period. This experimental study demonstrates the
impact of small quantities of very small proppant materials on the effective conductivity of microfractures created
during hydraulic fracturing operations. The improved conductivity observed in these tests clearly demonstrates that
these materials can effectively increase the effective permeability of the formation surrounding primary hydraulic
fractures.
Causes of Resistivity Reversal in the Vaca Muerta Formation, Argentina
1 1 1 2
A. C. Ortiz; ; C. Bernhardt; ; F. Tomassini; ; S. P. Cumella; ;
1. YPF, CABA, CABA, Argentina.
2. SPE, Houston, TX, United States.
Objectives/Scope: Recently, the resistivity reversal phenomena described in US shale reservoirs have been identified
in the Vaca Muerta Formation (VM). Archie defined the main factors that can impact on rock resistivity. Understanding
the role of these factors in shale reservoirs will help improve hydrocarbon reserves estimation. This study is focused
on the first T-R cycle. It corresponds to the maximum marine transgression above continental deposits of the Tordillo
Fm and is mainly composed of siliciclastic organic-rich mudstones of homogeneous thickness throughout the basin.
Methods/Procedures/Process: State-of-the-art data from wireline logs and rock samples were analyzed to understand
the resistivity behavior. Data from the dielectric and NMR show relatively constant amounts of total water at different
maturity windows. Formation sigma decrease towards the dry gas. Dielectric and sigma data shows the decrease in
resistivity towards the high maturity zones does not correspond to an increase of water saturation.
Their variability along maturity in pore tortuosity and wetability is observed in Dielectric data & SEM images and may
explain the resistivity trend.
The percentage of pyrite shows an increase towards high maturity.The combination of pyrite, tortuosity, and
wettability pushing the rock resistivity in the same direction may be the main factors to explain the resistivity reversal.
Results/Observations/Conclusions: A combination of different factors could control the rock resistivity. The water
saturation plays a minor role on resistivity. The similarities observed between VM and other US plays indicate possible
common phenomena in all shale plays probably linked to deep transformations that occurred in the rock during the
thermal maturation process.
Applications/Significance/Novelty: Resistivity variation is analyzed at different maturity windows and then compared
with other log information based on different measurement principles taken from latest wireline technology and rock
samples. This gives robustness on the analysis and helps to calculate a more accurately gas and oil in place using
resistivity data. It also provides a different picture of the transformations occured in the rock during the thermal
maturation process.
Natural Fractures, Fracture Facies, and Their Applications in the Well Completion – Case Studies From the Permian
Wolfcamp Formation, Midland Basin, West Texas, USA
1 2 2 1
B. Li; ; J. Wan; ; P. Lascelles; ; A. Coker; ;
1. Blackriver Geoscience LLC, Houston, TX, United States.
2. EP Energy, Houston, TX, United States.
Objectives/Scope: Natural fractures are present in the Permian Wolfcamp shales and have the potential to impact well
completion operations and hydrocarbon production. There are two main objectives for this paper. First of all, we aimed
to get calibrated natural fracture attributes using latest borehole image measurements and whole core data. Secondly
it is also intended to build better connection from natural fractures to the well completion process so that the fracture
data can be better used.
Methods/Procedures/Process: In this paper, natural fractures are interpreted on various latest borehole image
technologies including both acoustic and micro-resistivity measurements. Whole core data is also made available to
fully calibrated borehole image interpretation for better understanding of the many fracture attributes. Orienting whole
core using image logs was also made possible for the benefits of maximizing core value for the fracture study.
Fracture facies analysis is first introduced in this study using borehole images, conventional open-hole logs and whole
core data for the purpose of better understanding of rock fracability.
Results/Observations/Conclusions: Litho-bounded, calcite-healed or partial-healed fractures are the dominant type
within the formation. Open fractures, even though far less common, are also observed, with most of the being partially
open. The interpretation concluded that the predominant set of fractures strike NE-SW, with a secondary NW-SE set,
in the studied field area. However, fracture density showed considerable variability vertically and spatially among the
studied wells. Three basic types of fracture facies are proposed including fractured shale, non-fractured shale and limy
beds. Natural fractures are observed in all fractured shale facies but not evident on non-fractured shale, or seen
occasionally on limy beds.
Applications/Significance/Novelty: It is the first time to propose the fracture facies study in the unconventional
reservoirs. The fracture facies output has potential applications in understanding relative rock fracability in the studied
Wolfcamp shales. The naturally fractured shales are expected to be the easier rock to hydraulically fracture compared
to non-fractured shale and limy beds, because, in a sense, they were already fractured in geological history. The
existence of abundant planes of weakness within the facies, takes far less energy to reactivate than to create new
fractures. The relative rock fracability data can be used as part of the input for landing zone decisions. The interpreted
natural fracture trends and in-situ stress data helps to understand the hydraulic fracture network complexity.
Microscale Laboratory Studies for Determining Fracture Directionality in Tight Sandstone and Shale During Hydraulic
Fracturing
1 1 1 1
M. A. Ante; ; G. Manjunath; ; B. Jha; ; F. Aminzadeh; ;
1. Petroleum Engineering, University of Southern California, Los Angeles, CA, United States.
Objectives/Scope: Hydraulic fracturing, also known as “fracking”, has become a go-to-solution for producing from low
permeability reservoirs, for improving continuity/connectivity across thin-layered reservoirs, and for bypassing near
wellbore damage. While hydraulic fracturing has undoubtedly increased recovery from many oil and gas fields,
predicting and controlling the initiation and cessation of a fracture still remains a challenge due to compositional
heterogeneity and inelasticity at the grain scale.
Methods/Procedures/Process: In this paper, we propose a method to investigate the fracture initiation and propagation
behaviors by estimating fracture toughness and directionality using lab scale micro scratch tests. Micro scratch tests
provide a means to account for grain-scale heterogeneity and inelasticity during measurement of fracture properties in
different directions. The characterization of micro scratch track using Scanning Electron Microscopy provides a means
for determining the initiation and propagation nature of failures based on the critical points of failures at different
loading points. The nature of failure helps to predict and identify preferred regions of fracturing during hydraulic
stimulation.
Results/Observations/Conclusions: We have selected two different lithologies for our analysis: tight sandstone and
shale. The fracture toughness values along with the directionality behavior of these samples are compared for
understanding the nature of failures. Three consecutive scratch tests will be performed with uniformly spaced
scratches for a representative coverage of the sample surface in terms of the number of grains. We will discuss the
influence of the packing density of matrix, porosity and pore size distribution on the nature of failures.
Applications/Significance/Novelty: Using chemical and mechanical processes we investigate fracture initiation and
growth; The presence of organic matter in different proportions also affects the rock-mechanical properties and
directionality in fracture initiation and propagation processes. The microscale analysis provides high resolution
understanding of fracture behavior and has the great potential of improving hydraulic fracturing design and
implementation.
Overall, understanding the fracturing behavior of a rock at the microscale plays a critical role in predicting the
performance of hydraulic fracturing during and after the “frac job”.
Geomechanical Investigation of Fracture Hits and Its Implications on Well Integrity and Productivity
1
P. Pankaj; ;
Objectives/Scope:
With the ambition to extract the maximum possible recovery from the unconventional reservoirs, operators target to
place their wells as close as possible. With such well spacing pattern, hydraulic fracture hits amongst the horizontal
wellbores have become norm. The study presented here aims at studying the impact of such frac-hits on wellbore
integrity and production performance by numerical simulation and finite element analysis.
Frac-hits has been reported in almost all the unconventional reservoir basins across North America. In some wells,
their impact on production is claimed to be a positive frac-hit while some others claim as negative. To further drill down
the impact of any frac hit, the study in this paper presents some case studies examined though the advanced
modeling workflows. The workflows include complex fracture simulation model, coupled numerical reservoir simulation
and finite element geomechanical modeling to evaluate the rock deformation due to hydraulic fracturing, impact on
wellbore and casing integrity, shear deformation of casing, near wellbore proppant dislodgement, and effect of time
and distance from the offset wells.
From the study, it is proven that the impact due to a frac-hit on a well is not only dependent on the geomechanical
stress alteration arising from stimulation in the offset well and its production, but also determined by the degree of fluid
and proppant dislodged in the near wellbore proximity. The offset well hydraulic fracture propagation, pressures, fluid
and proppant transport are also critical in deciding the future productivity post frac-hit. Case study demonstrates that
the production of the considered well could have improved by 6% over a span of 5 years if the offset well hydraulic
fractures were pumped with higher proppant volumes. The combined two well configuration had over 15% incremental
production over the 5 years’ timeframe.
The modeling approach allows operators to improve their strategy of pad development and look at the impacts of frac-
hits more closely.Well spacing decsions are streamlined better follwoign the recommendations from this study. While
closing-up the well spacing to extract the last drop of oil and gas is important in a pad development mode, it is
required to modify the treatment designs as required to negate the effects of frac-hits.The study allows operators to
realize the true benefit of optimizing treatments multi-well level for developing an engineered pad development
strategy.
Micromechanical Modeling of Hydraulic Fracturing in Kerogen Rich Shales
1 1
Y. Fang; ; Y. Han; ;
1. Houston Research Center, Aramco Services Company, Houston, TX, United States.
Objectives/Scope: Fast production decline of some hydraulically fractured wells in unconventional reservoirs has
puzzled oil industry for years. Our laboratory experiments and modeling work indicate that kerogen can be a key factor
in this process. Based on laboratory test, a source shale mechanics module is developed in a DEM based lattice code
then applied to simulate hydraulic fracturing stimulation in KRS. Simulation results are theoretically analyzed then
compared & calibrated with field observation and data reported in the literature.
Methods/Procedures/Process: The mechanical properties & behaviors of KRS are calibrated from nano- and micro-
scale laboratory experimental data. The lattice code is based on a discrete element method in which the rotation of
particles is prevented to speed up the calculation and the contacts in classical DEM are replaced by elasto-plastic
springs. The reliability and accuracy of the lattice model in capturing fracture initiation and propagation is
demonstrated in simulating initiation and propagation of a single fracture in the classic PKN problem. The influence of
heterogeneity of kerogen distribution on the fracture generation is investigated at micro- and macro-scales. The
multistage hydraulic fracturing is modeled using operational pumping data and results are compared with
microseismic data in the field.
Results/Observations/Conclusions: Single-stage fracturing simulation showed that the fracture initiation pressure,
length and opening profile of a single fracture predicted by the lattice code agreed well with their analytical
counterparts; for the same pumping schedule, hydraulically induced fractures are wider and shorter where the
kerogen is sparser. Multistage fracturing simulation showed that planar fractures are generated if the stage spacing is
large; bowl-shaped fractures are resulted when the injection point is close to an existing fractured spot; due to the
abundant kerogen, the fracture network created by HF treatment in KRS is significantly smaller than the case without
kerogen, which implies quicker production decline and shorter productivity lifetime of stimulated multi-stage well.
Applications/Significance/Novelty: Strain softening/hardening tensile mechanical behaviors of kerogen rich shales,
discovered in recent microscale laboratory experiments, are implemented in a DEM based lattice code. Applications of
the code in multi-fracturing of KRS seem able to reveal the mechanisms behind the poor production performance and
relatively early time production declines, of hydraulically fractured wells in unconventional reservoirs.
Downhole Microseismic Mapping of More Than 400 Fracturing Stages on a Multiwell Pad at the Hydraulic Fracturing
Test Site (HFTS): Discussion of Operational Challenges and Analytic Results
1 1
N. A. Stegent; ; C. Candler; ;
1. Halliburton Energy Services, Houston, TX, United States.
Objectives/Scope: The US Department of Energy (DOE) and the National Energy Technology Laboratory (NETL) co-
funded a joint industry research project (JIP) in the Wolfcamp formation to better understand the fundamental
hydraulic fracturing process in naturally fractured shales. State of the art diagnostics were used to evaluate production
interference between laterals, effectiveness of completion sequences, and hydraulic fracture geometry in
Methods/Procedures/Process: To support these initiatives, downhole microseismic data was acquired on more than
400 fracturing stages to evaluate the generated hydraulic fracture geometries under various pumping conditions.
Eleven horizontal wells were drilled in the upper and middle Wolfcamp formation in the Permian Basin. The 10,000-ft
laterals were drilled in a chevron pattern to help improve interwell fracture communication and maximize hydrocarbon
recovery. The completion sequences were planned to restrain upward fracture height growth into a lower pressure
zone using a combination of zipper frac and simultaneous fracturing operations.
Results/Observations/Conclusions: Downhole microseismic was used to record more than 90% of the planned
fracture stages. Downhole tiltmeters were also used to validate upward fracture height growth. Multiple geophone
arrays, both vertical and horizontal, were used during this completion; this paper discusses coordination between
maintaining completion efficiency and gathering microseismic data. Additionally highlighted are completion challenges
associated with collecting diagnostic data along with results and findings.
Applications/Significance/Novelty: This paper is being submitted as part of the findings from the Hydraulic Frac Test
Site (HFTS) project and would best fit in the special session.
Using Stage Level Microseismic Analysis to Gain Insight Into Fracture Efficiency and Completion Effectiveness
1 1 1 1 2
I. H. Wang; ; R. Fairfield; ; D. Gray; ; J. Courtier; ; S. Lee; ;
Objectives/Scope: Microseismic event density heat maps generated at sequential time intervals during the treatment
of individual completion stages have been developed. Animating heat maps through time provides insight into the
dynamics of the completion process within a single stage, imaging sequences of microseismic event populations and
helping interpret potential genesis. This approach leverages benefits from URTEC paper 2674376, which mitigates
spurious microseismic events and improves interpretation of geologic features, stimulated reservoir volume, cluster
spacing effects, fracture efficiency and stress shadowing. Unambiguous changes in microseismic event character
from far-field pre-ramp up events to near-field events occurring at the onset of pressure are observed with this
method. Applying a stage-by-stage analytical methodology provides robust statistical conclusions pertaining to
reservoir behavior during well completion.
Using Stage Level Microseismic Analysis to Correlate and Ground Truth Cored Hydraulic Fractures
1 1 1 2 1 1
J. Wicker; ; J. Courtier; ; T. Campbell; ; S. Lee; ; R. Fairfield; ; S. Trowbridge; ;
Objectives/Scope: The Hydraulic Fracture Test Site in Reagan County, Texas provides sufficient data to correlate
microseismicity to physical observations of hydraulic fractures taken from a low-angle whole core, acquired adjacent
and subsequent to a hydraulically fractured horizontal well within the Wolfcamp formation. Microseismic event density
heat maps visualized through time, image concentrations of microseismic events in relation to independently
documented hydraulic and naturally occurring fractures. Microseismic activity observed over the duration of two single
completion stages, at the closest proximity to the cored interval, demonstrate conformity and correlation with the
presence of hydraulic fractures. Given caveats often cited with microseismic acquisition, processing and
interpretation, this study provides a ground truth example and insights into the degree of accuracy microseismic can
offer as a technology in discerning the efficacy of modern completions.
Surface Seismic Monitoring of Hydraulic Fracturing Test Site (HFTS) in the Midland Basin, Texas
1 2 1 3
A. Kumar; ; K. Chao; ; R. W. Hammack; ; W. Harbert; ;
1. Department of Energy, National Energy Technology Laboratory, Pittsburgh, PA, United States.
2. Northwestern Institute on Complex Systems, Northwestern University, Evanston, IL, United States.
3. Department of Geology and Environmental Science, University of Pittsburgh, Pittsburgh, PA, United States.
Objectives/Scope: The use of the microseismic event cloud (area of brittle deformation) as a proxy for stimulated
reservoir volume during hydraulic fracturing has been questioned by recent research focused on Long-Period Long-
Duration (LPLD) seismicity, which suggest that slow-slip deformation may also play an important role in the stimulation
of unconventional reservoirs. Other evidence for non-brittle deformation is the imbalance in the energy budget during
hydraulic stimulation- energy input greatly exceeds the radiated and fracture formation energy output as indicated by
microseismicity.
Methods/Procedures/Process: We carried out surface seismic monitoring at the Hydraulic Fracturing Test Site (HFTS)
in Reagan County (RC), TX. The test site consists of 11 wells, targeting the Upper and Middle Wolfcamp Formations
at ~7500 feet depth. The field deployment is a surface network of 48 three-component geophones. The study
objective is to identify the presence of low frequency events (10-60 Hz) of long duration (few seconds) in the seismic
data collected during hydraulic fracturing of horizontal well at a test site in the Midland Basin, Texas.
Results/Observations/Conclusions: Preliminary investigation of waveforms from Reagan County indicates the
presence of low frequency events of relatively longer duration. Each of these events is approximately 5-10 seconds
long, with no clear first wave arrival. The seismic energy of the observed LPLD events is largely visible on the vertical
component suggesting that the LPLD events are dominated by P waves. We also noticed a finite temporal moveout in
the observed arrival time of these events across the seismic network suggesting a local source of deformation as their
origin.
Applications/Significance/Novelty: More in-depth analysis of waveforms may allow us to better constrain the dominant
mechanism of deformation responsible for LPLD events during hydraulic fracturing. We also aim to investigate the
temporal correlation between the frequency of LPLD events and various pumping parameters (pressure, rate,
proppant concentration) at HFTS.
Microseismicity Analysis for HFTS Pad and Correlation With Completion Parameters
1
D. Maity; ;
1. gas technology institute, Des Plaines, IL, United States.
Objectives/Scope: Traditional microseismic data analysis involves looking at hypocentral “event” locations and
interpreting their spread to understand fracturing. This study aims to look at the frequency-magnitude distribution
(FMD) of the seismicity to distinguish between various zones of interest both spatially as well as in time to better
characterize the hydraulic fracturing process. This study also attempts to validate correlations between the measured
microseismicity and other independent data such as fracturing parameters, baseline seismic data and OBMI image
logs.
Methods/Procedures/Process: Power law behavior of FMD as well as spatial distributions is well understood.
Comprehensive FMD analysis of microseismic catalog from the HFTS project is carried out. The seismic “b-value” is
computed using the maximum likelihood approach with the magnitude of completeness (M ) is estimated using the
C
goodness of fit test to minimize residuals. Care is taken to make sure that adequate catalog size is used in both
spatial and temporal analysis of the microseismic data. To determine the 3D spatial distribution, a grid search
th
algorithm is used with a boundary overlap of 90 quantile of the maximum hypocentral errors along relevant axis.
This is done to account for location errors within the microseismic catalog.
Results/Observations/Conclusions: Significant correlation is observed between the b-value distribution across various
fracture stages for the study well pad and the degree of natural fracturing prevalent in the major horizons of interest.
We find that estimated “b-values” are positively correlated with instantaneous shut-in pressures (ISIP). Low “b-values”
correlate with locations where significant lineaments are identified from the baseline 3D seismic data. Higher values
straddle on the sides of these lineaments indicating significant occurrence of shear failure along these lineaments
impacting associated FMD behavior. When the development of seismicity is studied with time, we observe interesting
correlations between operational perturbations during hydraulic fracturing and accompanying changes in seismicity.
Applications/Significance/Novelty: While FMD analysis for microseismic monitoring programs has been reported in the
past, this study successfully validates observations and relationships using significant background data that was
collected as part of the test site study. We identify contrasting correlation between b-values and ISIP’s compared with
other studies and propose possible reasons for said discrepancies. FMD or similar methods can be used in future
hydraulic fracturing programs where microseismic monitoring is performed to provide valuable information about stage
behavior and identify fracturing issues which can be used to improve future hydraulic fracturing treatments. This study
shows examples of such behavior and establishes probable cause that can be used as a guideline for similar studies
in the future.
Environmental Monitoring of the Hydraulic Fracture Test Site (HFTS)
1 1
S. Eisenlord; ; T. Hayes; ;
1. Energy Supply and Conversion, Gas Technology Institute, Des Plaines, IL, United States.
Objectives/Scope: The HFTS experiment tested new extraction and production techniques and technologies to
understand fracture geometry and increase the efficiency of shale hydrocarbon production. Eleven test wells were
drilled in the Upper and Middle Wolfcamp formations, hydraulically fractured, and produced. Air quality, groundwater
quality, and produced water chemistry and microbiology were monitored through the duration of the field experiment
and one year into production.
Methods/Procedures/Process: BTEX, methane, VOCs, SOx, NOx, and PM were measured 1,000 ft up-wind and
10
1,000 ft down-wind of the HFTS experiment before operations began, during hydraulic fracturing, flowback, and
several months into production to evaluate operational impact on local air quality. The underlying Edward-Trinity
Plateau aquifer was monitored using 5 groundwater wells within 2 miles of the HFTS pad; BTEX, methane, petroleum
hydrocarbons, and a comprehensive suite of anions and metals were analyzed from samples taken before operations
began, during hydraulic fracturing, flowback, and one year after production. Water produced from 3 of the 11 HFTS
wells was analyzed for chemical and microbiological properties, 1.5 years into production.
Results/Observations/Conclusions: There was no significant impact to air quality from HFTS operations, although
elevated BTEX was measured 1,000 ft away from the well-pads during the flowback period. We detected no migration
of hydrocarbon or fracture fluid chemical migration into the Edward-Trinity Plateau aquifer. Source waters used for
hydraulic fracturing contained naturally occurring high concentrations (~300ppm) of sulfate. Rapid microbially induced
corrosion was detected in several wells, leading to biocide dosing. Microbiological populations were significantly
altered by biocide treatment, but still continued to increase in population size six months after application
Applications/Significance/Novelty: After hydraulic fracturing, detailed chemical and microbiological analysis provides
evidence for microbiologically-mediated sub-surface transformations of supplied sulfate and organic carbon into
sulfides. This report discusses air quality, water quality, and produced water analytical results for the HFTS
experiment and further implications for the unconventional oil and gas industry.
Review of the Bone Spring Hybrid Play in the Delaware Basin
1 1 1 1
K. Schwartz; ; A. Starr; ; H. Meier; ; N. Stolte; ;
1. Chevron, Houston, TX, United States.
Objectives/Scope: The presentation will review the Bone Spring play within the Delaware Basin. As drilling marched
south landing strategies changed. This presentation will review landing zones with production and completion designs
to provide guidance on the optimal reservoir development. The Bone Spring reservoir commonly has both a stacked
sand and shale component making it a hybrid, or mixed reservoir of both tight sands and organic-rich shales. Decline
curve analysis will show if the sand component is needed to make a better well. New well results will be shown to
demonstrate the ultimate play extent.
Methods/Procedures/Process: Several well logs will be shown demonstrating how the play differs from the north
(sand-prone), central (hybrid), and south (mostly shale). Pay calculations will be reviewed within the mix lithology
reservoir and landing zone strategies will be confirmed from log-based multimineral analysis. Completion data and
production forecasts will be reviewed to validate the reservoir producibility and preferred landing strategy. Finally, 3D
seismic will show the extent and impact of known mass transport deposits and how wellbores that stayed within zone
ultimately had better production. Core data with reservoir properties will be shown which was used to validate the
mutlimineral models. Favorable well bore trajectory will be reviewed and rotates throughout the basin based on FMI
data.
Results/Observations/Conclusions: Ultimately, reservoirs that having a mixed lithology produced better in comparison
to areas with only tight sands or mostly shales. The interbeded marriage of source with carrier ultimate produces the
most economic returns. Mass transport deposits tend to enhance production as wells are linking multiple isloated
tanks.
Applications/Significance/Novelty: The Bone Spring formation saw the bulk of early horizontal development in the
Delaware Basin. Now, the Wolfcamp is seeing the bulk of rig concentrations. The Bone Spring can rival the Wolfcamp
and ultimately more economic.
Predicting Success in the Haynesville Shale: A Geologic, Completion, and Production Analysis
1
B. Johnston; ;
Objectives/Scope: Recent well results from the Haynesville Shale have shown significant improvements in both the
historical core of the play and previously uneconomic areas as well costs have decreased and completion designs
have intensified. This study analyzes geologic, completion, and production data across the Haynesville play to
characterize and identify the key drivers that make this play successful. Recent proppant-intensive completion designs
have unlocked previously uneconomic regions of the play and increased drilling opportunity as economic limits of the
play expand.
Methods/Procedures/Process: This study analyzes digital well logs and production and completion data from
horizontal wells to understand how geologic characteristics and well performance vary between the historical core and
more recently tested, and previously uneconomic, areas of the Haynesville play. Key geologic parameters such as
porosity, organic content, and lithology, are compared with production data from historical and current wells to help
better understand the key parameters that make the play successful. Well completion parameters such as lateral
length, proppant intensity, and proppant loading are also analyzed across the play to understand how different rock
types reacts to varying completion design.
Results/Observations/Conclusions: Over the past few years, well completion designs have shifted towards increased
proppant intensity, with some operators testing as high as 5,000 lbs/ft of proppant. Although geologic parameters vary
across the Haynesville play, the marked increase in completion design in recent years has produced successful wells
in historically poor areas of the play. Whereas historical, low-intensity completion designs were insufficient to produce
economic volumes of gas, new completion designs overcome the poorer rock quality outside of the core region, and
recent well results suggest the economic limits of the Haynesville play are expanding.
Applications/Significance/Novelty: The Haynesville Shale was viewed as one of the most prolific shale-gas resources
when it first emerged in late 2008. The crash of natural gas prices in 2012 and a surge of Appalachian gas production
quickly shifted activity and interest away from the Haynesville. In 2015, despite depressed commodity prices, rig
activity in the Haynesville returned as operators began to drill more cost-effective and proppant-intensive wells.
Recent results have shown significant improvements in both the historical core of the play and previously uneconomic
areas. The shift to a more proppant-intensive completion design has greatly improved well results in the Haynesville
and has increased drilling opportunity as economic limits of the play expand.
The Niobrara Formation in the Southern Powder River Basin, Wyoming: An Emerging Giant Continuous Petroleum
Accumulation
1
S. Sonnenberg; ;
1. Geology, Colorado School of Mines, Golden, CO, United States.
Objectives/Scope: Review the Niobrara Petroleum System in the Powder River Basin.
Discuss thickness variations in the Niobrara A, B, and C marl/chalk units.
Discuss porosity and permeability distributions in marl/chalk units.
Describe geologic factors related to exploration and development success (e.g., source rock quality, geothermal
gradients, overpressure distribution in basin, regional fracture development and trends).
Methods/Procedures/Process: Methods used include a combination of subsurface mapping of reservoir distribution
(Niobrara A, B, and C units), geopressure distribution, geothermal gradient distribution, and analysis of fracture
development and importance.
XRD and XRF trends in the Niobrara are analyzed and discussed.
SEM techniques are used to analyze pore size distributions and porosity types.
Geologic factors related to successful exploration and development include excellent source-rock quality, source-rock
maturity, reservoir thickness, matrix and fracture porosity and permeability development, high geothermal gradients,
overpressure, oil gravity, gas-oil ratios, and regional fracture development.
Results/Observations/Conclusions: The Niobrara Petroleum System (NPS) covers an extensive area across the
Rocky Mountain Region, USA. The Niobrara is a deep-water hemipelagic carbonate mudrock and will be the subject
of this paper.
The Niobrara in the southern PRB is Coniacian to early Campanian in age and approximately 300 to 500 ft thick. The
formation, where productive, has low porosity (< 10%), low permeability (<0.01 md), and pore throat sizes less than
0.1 micron. The immature-mature present-day depth boundary is approximately 8,000 ft. The formation is subdivided
informally into three units in the PRB (A, B, and C). The units consist of cycles of marls and chalks. The main reservoir
target is the B chalk/marl zone.
Applications/Significance/Novelty: The Niobrara Petroleum System is widespread in the Rocky Mountain Region.
Recent success in the Powder River Basin by several operators demonstrates the potential for a giant unconventional
accumulation.
The results of this study may apply elsewhere in the Rocky Mountain Region.
Regional Appraisal of Shale Resource Potential Within the Permian, Anadarko, and Arkoma Basins: How Does the
Alpine High Stack Up?
1 1
A. Bromhead; ; T. Butt; ;
1. Halliburton, Abingdon, United Kingdom.
Objectives/Scope: The discovery of the Alpine High in the Permian Basin, US, is one of the most significant
exploration successes within the North American shale industry over the last few years. The Alpine High is an
emerging unconventional subplay in the southwestern part of the Delaware subbasin where the Woodford, Barnett,
and Wolfcamp shale plays constitute a stacked resource interval. All three shale plays have been successfully
developed elsewhere in Texas and Oklahoma but were untested on the Alpine High until the first shale appraisal wells
were drilled in 2016.
Methods/Procedures/Process: Initial well results and production tests on the Alpine High have been promising.
However, it is not yet apparent how resource concentration and stratigraphic resource distribution on the Alpine High
compares against established areas of proven production within nearby subplays. This study uses diverse public
domain datasets integrated within a sequence stratigraphic framework to derive a wide variety of modelled outputs
with which to assess shale-play prospectivity, including regional maps for the presence of organic-rich facies, organic
content, gross thickness, net thickness, maturity, gas-to-oil ratio, porosity, pressure, and depth, which are used to
screen and derive resource concentration assessments in shale plays within the Permian, Anadarko, and Arkoma
basins.
Results/Observations/Conclusions: The results of the resource assessment indicate a large in-place risked resource
2
concentration across the Alpine High with a mean value of 96 MMBOE/km . Within the stacked resource interval, 30%
of the total shale resource is contained in the Woodford, 15% in the Barnett, and 55% in the Wolfcamp. This combined
total shale resource appears highly significant when compared against the nearby subplays. The Alpine High is
broadly comparable to the Southern Liquids, Middle Hotspot, and Midland Core subplays targeting the Wolfcamp in
the Permian Basin and appears to be superior to the SCOOP and STACK subplays in the Anadarko Basin where the
Woodford is the primary target.
Applications/Significance/Novelty: A stratigraphic comparison of the Woodford play indicates that the depositional
setting, organic content, proportion of net shale and mineralogy is comparable between the Alpine High and the
SCOOP and STACK. However, the Woodford is, on average, more than twice as thick on the Alpine High and, hence,
has a significantly higher in-place risked resource concentration. An integrated regional model that defines shale-play
presence and resource potential, designed using an internally consistent methodology, enables direct comparisons to
be made in different shale subplays. This unique platform allows the Alpine High discovery to be evaluated within its
regional context and is an effective comparison tool to manage an exploration portfolio and to identify new exploration
opportunities.
Diagnosing the Health of Your Well With Rate Transient Analysis
1 1 1
D. Anderson; ; J. M. Thompson; ; H. Behmanesh; ;
1. Anderson Thompson Reservoir Strategies, Calgary, AB, Canada.
Objectives/Scope: The primary objective of this paper is to demonstrate how production data (rates and flowing
pressures) can be used to identify underperforming producing wells with significant uplift potential, along with
recommended remedial measures. This work introduces a detailed classification process, whereby specific problems
are identified and attributed to “surface”, “wellbore” or “subsurface” categories, based on simple diagnostics that can
be performed very quickly.
Methods/Procedures/Process: The analysis method used in this work is Rate Transient Analysis (RTA). RTA is ideal
for establishing a well’s productive capability early in its life (after 3-6 months) and categorizing subsequent well
decline behaviour as either reservoir depletion or productivity loss through time (or a combination of the two). Through
interpretation of diagnostic plots, such as the Specialized Linear Flow, Flowing Material Balance Analysis and Type
Curve Analysis plots, specific productivity issues can be easily identified and classified using simple pattern
recognition. The potential uplift from remediation can then be quantified by production forecasting using calibrated
RTA models.
Results/Observations/Conclusions: Field examples from multiple plays/basins across North America are included in
our study. Deliverability issues such as surface back pressure restrictions, wellbore liquid loading, wellbore blockages,
compaction-related productivity losses and frac conductivity degradation are identified through this work. Results show
how RTA can be used as a practical screening tool, identifying underperforming wells and quantifying the potential
benefits of remedial action. Case studies include the Falher, Marcellus, Eagle Ford, Bakken and Montney formations.
Results from the Falher study, for example, reveal a 2000 psi pressure blockage in or near the wellbore. The Bakken
example indicates a 15% loss in reserves due to compaction effects.
Applications/Significance/Novelty: This paper presents a novel application for RTA – as a screening tool for
diagnosing underperforming wells and estimating their uplift potential. The industry tends to undervalue the potential
benefit of all but the simplest remedial measures for underperforming producing wells for two reasons- 1. poor
performance is often simply accepted as a statistical reality and 2- operators do not even realize that their wells are
underperforming. There is little published literature that addresses this important issue. However, our experience has
taught us that it is pervasive in our industry, spanning most (if not all) unconventional plays (both oil and gas) to a
degree. If the issue(s) can be resolved, production improvements aggregated field-wide could have tremendous
economic implications.
Artificial Lift Selection and Its Applications for Deep Horizontal Wells in the Unconventional Reservoirs
1
P. Pankaj; ;
Objectives/Scope: Several methods have been tried and deployed as artificial lift method in deep long horizontal wells
completed in unconventional settings. Some have been successful where as some others are not. This paper studies
the various lift mechanisms and draws an envelope for their application to such horizontal wellbores using sensitivity
study through a transient wellbore fluid flow model.
Methods/Procedures/Process: A calibrated earth model from the Eagle Ford Shale basin with hydraulic fracture
geometries in the horizontal wellbore is used for a sensitivity study. The wellbore profile was changed in the simulation
model to four different types of profiles: Toe Up, Toe Down, Toe Up with Hold Trap, Toe Down with Hold Trap. Other
factors such as location of the artificial lift equipment in the wellbore, reservoir performance and deliverability has been
considered for the deployment of the artificial lift method. Transient fluid-flow wellbore simulations and numerical
reservoir simulations are used to determine the performance potential and effectiveness of the artificial lift mechanism
for long term productivity.
Results/Observations/Conclusions: Multi-phase fluid flow and transient flow phenomenon are critical modeling
considerations for horizontal wellbores. It was found that the critical flow rate in horizonal wells can vary considerably
when the well profile is considered. As the drilling dog leg severity increases, the chances of wellbore slugging and
liquid hold-up increases. However, if the fluid-flow rate is maintained above the critical rates as determined in this
study, the production issues can be controlled. Therefore, it is clear from this study that the well trajectory and drilling
uncertainty window must also account for the artificial lift method that is planned to be deployed. Adjustments to the
artificial lift method placement in the wellbore would help offset negative impacts if the wells are poorly drilled.
Applications/Significance/Novelty: Recommended practice for drilling, completion and artificial lift can be derived from
this study. The need to integrate the artificial lift selection to the earth model, drilling trajectory and landing, hydraulic
fracture and completion model is paramount to improve the efficiency of artificial lift in the unconventional reservoirs.
Use of Chemical Tracers Reveals Details of Cleaning of a Non-Conventional Gas Well in Vaca Muerta Shale
1 1 1 1
C. D. Ferlaza; ; J. V. Ramirez; ; D. Garcia; ; D. Ceccon; ;
1. Unconventional projects, YPF, Buenos Aires, CABA, Argentina.
Objectives/Scope: In a multi-fractured horizontal well that navigates an unconventional formation such as the shale,
whether in the oil or gas window, the following questions arise: do all fracture stages produce? Which produce more
and why? Is injected water or formation water produced? These and many others are the questions that have to be
answered. This requires the introduction of new technologies, such as the use of chemical tracers, to help us improve
the understanding of the reservoir-fracture-well system
Methods/Procedures/Process: One challenge that is posed, apart from the selection of horizons within the Vaca
Muerta Formation, is the cleaning process in the shale gas wells to guarantee the best performance of the well during
its life cycle, therefore it is of vital importance to know how this process is carried out along the entire horizontal
section.
The cleaning process of a well during flowback, and the timing in the opening of chokes in a gas well were studied
using specific tracers designed to accompany the gas phase.
Results/Observations/Conclusions: This paper presents the case study of an unconventional gas well drilled in the
Vaca Muerta formation through two producing horizons. The well was drilled directional (80 °) to penetrate both
horizons, evaluate the cleaning process and compare the productivity of both in order to define the most productive
navigation horizon for development.
In this project, it was possible to evaluate the cleanup of the well, which proved to be a gradual process, correlating
the entry into production of all the stimulated stages with time and verifying their behavior before the changes of
production orifice (choke). Other correlations were possible, such as the difference in lithology and TVD depth of the
different zones.
With production by zone, It was possible to estimate the pressure profile and pressure loss along the horizontal
section
The tracers used both for the gas and water phases, made it possible to identify the gradual process of entry into
production of all the well zones, helping to understand how the different operational variables contribute to the
cleaning dynamics of the different stages.
Multifunctional Surfactant Provides Superior Post Frac Production by Enhancing Polymer and Load Fluid Recovery
1 2 1 1 3
R. C. Plasier; ; J. Delorey; ; K. Cooney; ; J. Leguizamon; ; C. Thomson; ;
1. Engineering, BJ Services , Calgary, AB, Canada.
2. Delorey Consulting Inc, Calgary, AB, Canada.
3. Completions, Paramount Resources, Calgary, AB, Canada.
Objectives/Scope: Initial and sustained production are realized using a multifunctional surfactant in an under
pressurized tight sandstone reservoir. In the field trials the multifunctional surfactant was used either as a foamer or a
flow back aid. Field trials and laboratory tests presented demonstrate the multifunctional surfactant has superior
performance over conventional surfactants or foamers in terms of proppant pack clean up and long term production in
these depleted sandstone liquids rich gas wells.
Methods/Procedures/Process: Foamed fracturing fluids are often used to stimulate low pressure reservoirs by
providing the necessary viscosity for excellent proppant transport and distribution and supplying energy for enhanced
flow back. Long term production data of wells from a pressure depleted area of the Glauconite were analyzed. The
wells chosen all had a similar vintage, to mitigate varying completion type and reservoir pressure, and were mainly
energized or foamed treatments. Lab data characterizing the multifunctional surfactant in terms of adsorption,
interfacial tension, and core flow data will be presented. Flow back analysis detailing polymer recovery and surface
tension as a percentage of flow back from wells stimulated using a conventional foamer versus the multifunctional
surfactant will be compared.
Results/Observations/Conclusions: The multifunctional surfactant was used as a foamer or a flow back aid in the
treatments examined in the production analysis. Initial and prolonged production results are typically higher when
compared to previous completions in the Glauconite formation if the multifunctional surfactant was present in the
fracturing fluid. Flow back analysis using the same synthetic polymer but different foamers indicates the wells
stimulated with the multifunctional additive had better polymer recovery, lower interfacial and surface tensions and
better compatibility with the formation fluids over a larger portion of the flow back volume than the conventional
foamer. Core flow testing of the multifunctional additive shows an increase in the relative permeability to oil.
Applications/Significance/Novelty: Using foamed or energized systems for hydraulic fracturing provides excellent
proppant transport and distribution, reduced water requirements and aids in load fluid recovery. One consideration,
often over looked, is foamer/surfactant selection to ensure compatibility with the formation fluids and to improve
polymer and load fluid clean up. The multifunctional surfactant provides lower interfacial tension, sustained lower
surface tension and better compatibility with the formation fluids over a larger volume of the flow back than the
conventional foamer did. More efficient polymer removal and load fluid recovery would aid in higher initial and
sustained production. Initial and prolonged production results are typically higher for wells using the multifunctional
surfactant.
Fluid Heterogeneity for Tight Unconventionals on a Well-Box Scale
1 1 1
C. H. Whitson; ; F. Alqahtani; ; E. Chuparova; ;
1. NTNU, Trondheim, Norway.
Objectives/Scope: This paper presents the supposition that significant fluid heterogeneity (GOR, or composition) may
exist within the drainage volume of a tight unconventional well. The important question to be answered is whether
"well-box" fluid heterogeneity can have a significant impact on how produced GOR changes with time, i.e. ultimate oil
and gas recovery performance. Our proposal is to use Eagle Ford as the field of study.
Methods/Procedures/Process: Compositional reservoir simulation of a detailed well-box drainage volume, combined
with geochemical and petrophysical modeling of insitu fluid and rock-property distributions will provide the basis of this
study, using Monte Carlo methods. Simulation results will be compared with actual GOR performance. Both correlated
and uncorrelated spatial variations of key variables will be used: molar composition (GOR) and heavy-ends
description (API), and petrophysical properties including permeability, porosity, and water saturation. A detailed EOS
fluid model and comprehensive compositional library of relevant reservoir fluids has been developed for the Eagle
Ford field. Correlation of geochemical maturity index to GOR and composition is similar to a correlation develoeped for
the Barnett shale.
Results/Observations/Conclusions: This paper will present the detailed quantitative modeling methods that we are
using, and preliminary results of our study. The main purpose of our paper is to prompt a discussion about the
potential impact of insitu spatial fluid heterogeneity on producing GOR performance and ultimate EURs of oil,
condensate, and gas. Given the little (if any) existing studies of this subject, we hope that this paper will forge a
multidisciplinary alliance of geochemists, geologists, petrophysicists, and reservoir modelers together with solicitation
of industrial partners who can share production performance data illustrating GOR(time) behavior that has not been
readily understood with conventional assumptions of uniform fluid within the well-box drainage volume.
Applications/Significance/Novelty: This study will provide the ability for industry to understand how "local" fluid
heterogeneities of individual well drainage volumes may influence ultimate EURs and GOR behavior. The
investigation of fluid heterogeneity on a well-box scale has not, to our knowledge, been studied previously in terms of
quantitative impact on production performance. If the impact is found to be significant, and substantiated by actual well
performance behavior, then the ultimate application of this understanding would necessarily be in how wells are drilled
and acerage is acquired to maximum liquids production. The study may also reveal how such fluid heterogeneity is
best quantified and used in design of optimal well completion.
Production of Migrated Oil From Horizontal Wells Drilled Into the Eagle Ford Formation on the San Marcos Arch
1
A. S. Kornacki; ;
1. Weatherford Laboratories Inc., Houston, TX, United States.
Objectives/Scope: High-maturity Eagle Ford (EF) source-rock (SR) beds generated the mobile oil produced from the
Austin Chalk in south Texas. Because EF SRs generate more viscous oil at lower thermal maturity, the oil generated
and retained by Lower Eagle Ford marl SR beds at VR <0.80 may not readily flow from that tight reservoir. I
determined the origin of oils produced from Austin Chalk, EF and Buda reservoirs on the San Marcos Arch by
comparing their composition to oil extracted from core plugs.
Methods/Procedures/Process: HRGC analyses were performed on 7 produced oil samples and on extracts obtained
from 20 core plugs selected in the Austin Chalk, Upper EF (UEF), and LEF marl or clay-shale intervals. The LEF core
plugs contain oil-prone kerogen at VR ≈0.75. Leaner Austin Chalk and UEF core plugs contain gas-prone or inert
kerogen. Oil source and maturity parameters were calculated using HRGC data. 17 HRGC peak-height ratios were
used to classify the oil samples and extracts via a Hierarchical Cluster Analysis (HCA). End-member samples and
mixing lines were identified by performing a Principal Components Analysis. Commingled oil samples were allocated
using HC fingerprinting software.
Results/Observations/Conclusions: ≈29-31°API oil samples produced from the Austin Chalk were generated at higher
temperatures than ≈23-27°API EF oil samples. Pristane/phytane (Pr/Ph) ratios are lower in the produced oils and in
UEF core-plug extracts that contain migrated oil than in LEF marl extracts obtained from the same well. Pr/Ph ratios
are even higher in Austin Chalk, UEF, and LEF core-plug extracts obtained at a downdip well. HCA results show the
produced oils – and the migrated oil extracted from Austin Chalk and UEF core plugs – are significantly different than
LEF core-plug extracts. None of the produced oil samples are commingled mixtures of extracts obtained from LEF
marl or clay-shale core plugs.
Applications/Significance/Novelty: These results help define the level of maturity (VR ≤0.75) at which oil generated by
LEF kerogen does not readily flow. Allocation results indicate the migrated oil in UEF core plugs contains ≈65-92 wt%
of the type of oil produced from two wells and ≈8-35 wt% of the type of oil extracted from LEF marl core plugs. The
presence of non-producible bitumen in LEF core-plug extracts does not invalidate these results because bitumen
contains a small amount of the C -C compounds used to classify and allocate the samples. The migrated oil
10 25
present in the LEF reservoir may occur in native microfractures that drain rapidly after stimulating the reservoir.
Permian Basin Petroleum Systems—Geochemical Insight Into Hydrocarbon Generation, Migration, and Well
Performance
1 1
J. B. Curtis; ; J. E. Zumberge; ;
1. GeoMark Research Ltd., Englewood, CO, United States.
Objectives/Scope: Objectives: 1) Geochemically assess Permian Basin crude oils produced from both conventional
and ‘unconventional’ reservoirs using biomarkers and stable carbon isotopes, 2) Determine the number of different
petroleum systems using biomarkers, 3) Predict the nature of the source rock for each petroleum system: depositional
environment, lithofacies, geologic age, 4) Determine the thermal maturity of each oil sample using biomarkers, and 5)
Map the oils from each petroleum system to establish the geographic extent & maturity trends to locate areas for
optimal ‘unconventional’ plays.
Methods/Procedures/Process: Methods: Over 600 Permian Basin crude oils were grouped into 21 oil families that
share common organic matter sources using genetic-specific biomarkers (geochemical fossils) and stable carbon
isotope compositions, by means of hierarchical cluster and principal component analyses. Mapping of the geographic
and stratigraphic distribution of oil families throughout the greater Permian Basin provided insight into timing of
hydrocarbon generation, migration and fluid mixing histories.
Biomarker maturity sensitive ratios track the level of thermal maturity of oils from the same or similar source rocks.
Selected biomarker ratios were subjected to further principal component analysis with the resulting principal factor
converted to estimated vitrinite reflectance equivalent values, %VRE.
Results/Observations/Conclusions: Results: Geologically significant differences in oil occurrences were noted. For
example, Delaware Basin Wolfcamp-sourced oils (Family 8) are different from Midland Basin Wolfcamp-sourced oils
(Family 11). The Delaware Basin Wolfcamp oils are from more of a marl source lithofacies, while the Midland Basin
oils were generated from a distal shale lithofacies. Wolfcamp-sourced oils are relatively isotopically positive with high
C27 and low C29 steranes, the reverse of UD Woodford-sourced oils. Oils sourced from Wolfcamp-Bone Spring
transitional facies (Families 9 & 10) appear to have been derived from organic-rich horizons stratigraphically between
the L. Wolfcamp shales and the Leonardian Upper Bone Spring/Avalon marls/carbonates. Family 14 oils are of ‘mixed’
Woodford and Wolfcamp origin.
Applications/Significance/Novelty: Application: A Delaware Basin case study of 15 oils from Wolfcamp A & B laterals
in Reeves & Pecos Counties, TX illustrates an approach to geochemically determine if hydraulic fracturing of multiple
zones results in co-mingling. Study oils from the WC A and WC B Shale intervals compositionally group separately,
i.e., at least two different Wolfcamp organofacies are evident. The genetic differences between the produced A & B
Shale oils strongly suggest that there is no/little co-mingling between A & B Shale fracturing treatments. Furthermore,
the calculated thermal maturity of the oils varies between 0.79 and 0.88% VRE. The difference in maturity between
the closely spaced A Shale and B Shale oils again attests to the lack of oil co-mingling between A & B Shale
completions.
Insights From Stable Isotope Geochemistry Surveillance in the Unconventional Horn River Basin Play
1
G. Norville; ;
1. Department of Earth and Atmospheric Sciences, Univeristy of Alberta , Toronto, ON, Canada.
Objectives/Scope: We use chemical and stable isotope compositions of gases from the Horn River Group in the Dilly
Creek area (NTS 94/O8) of NE British Columbia, Canada to gain insight into the origin and evolution of gas in a highly
mature shale system. The Horn River Basin is an unconventional shale gas play with estimated GIP reserves of
448TCF hosted within organic-rich Devonian Horn River Group (HRG) shales.
Methods/Procedures/Process: Production gases from the HRG were collected in duplicate at the well head (from 10
wells for time periods up to 1250 days) and analyzed for chemical composition and stable isotope ratio. Two and a
half kilometer long, horizontal isotope gas mud logs were also measured for formations in the HRG.
Results/Observations/Conclusions: Chemical compositions indicate the HRG shale gases are dry with methane mole
fraction values ranging from 0.8 to 0.9. The carbon isotope compositions of the gases (δ13C1, δ13C2 and δ13C3)
were found to be 13C rich and indicative of very highly mature gas. The horizontal mud logs reveal that the gas may
be compartmentalized in some formations. Interestingly, in contrast to most conventional systems, these gases
exhibited temporal changes during gas production and appeared to show some positive correlation with the
corresponding gas production decline curves. Several of the gases analysed were isotopically reversed gases;
however these reversals were more commonly observed in later stages of production. In some cases for periods in
which the well was ‘shut in’ 13C1 enrichment occurred.
Applications/Significance/Novelty: It is evident from the stable carbon isotope values of gas components that the shale
gas system is complex and the subtle isotopic variations observed along the mud log and during production suggest
new models are required to fully describe these unconventional gas systems. To improve our understanding of these
systems it is essential to integrate geochemical data with other geological and geophysical data sets for unravelling
associated reservoir complexity.
Rapid Reservoir Modeling With Automated Tops Correlation
1 1 1
C. Grant; ; W. M. Bashore; ; S. Compton; ;
1. Drillinginfo, Littleton, CO, United States.
Objectives/Scope: In unconventional resource plays, constructing a sound geological model that ties various well
information is mandatory for properly extracting and integrating well and seismic information for predictive analytics
workflows. Unlike conventional plays, unconventional plays that span basins have potentially tens of thousands of
wells. Constructing geological models to include all of these wells and updating them as additional ones become
available can be a daunting task.
When constructing large cross sections, regional stratigraphic patterns are easily discernable visually. Converting
these geologic events and spatial patterns to digital information using the power of the computer and new machine
learning techniques is becoming more important than ever as geoscientists attempt to “keep up” with all this
information. This talk will cover a modern technology toward that end.
Methods/Procedures/Process: We introduce a technology and workflow that uses the power of the computer and
novel machine learning techniques to capture and model well log patterns for correlating geologic events across many
wells. Using one or more well logs at a source well, a signature ‘thumbprint’ is correlated over many target wells to
optimally find stratigraphic intervals for well pick estimation. Our patent-pending method uses a variant of the Dynamic
Time Warping algorithm known as Subsequence Dynamic Time Warping to perform all of the stratigraphic matching.
As candidate well picks propagate from a source location, computed quality and confidence measures are propagated
to the target wells until a threshold is exceeded, and the correlation stops. The propagation paths are carried out along
a Dijkstra path constraint, and all possible paths are investigated in a 3D model. Local statistical modeling is
performed when many estimated picks for a given well occur.
Any previously interpreted geologic events are used as stratigraphic constraints to winnow out non-viable solutions.
The method then improves in both accuracy and speed as it is further stratigraphically constrained.
Results/Observations/Conclusions: In this study, we correlate 7 geologic events in the Wolfberry section of the
Midland Basin over a 240 square-mile seismic survey just west of the town of Midland, TX. Using the process outlined
below, two geoscientists working together picked 7 stratigraphic horizons, two in less than a day, consisting of over
10,000 well picks in a week's time. This represents a dramatic time savings not possible in conventional manual
picking.
The stratigraphic interval is defined by a strong well log signatures: the top at the base of the San Andres and the
base by the Strawn. These signatures are easily discernable and easily picked by the automated approach. A
‘dynamic’ west-to-east geologic cross section is constructed across the study area to include as many vertical wells as
possible. Once the automated picks are deemed to be accurate they are changed to ‘manual status’. North-to-south
depositional strike sections are correlated from the ‘seed picks’ of the depositional dip section. The process is
continued until the geologic event is correlated over the study area.
The basal San Andres and Strawn geologic events are mapped in this fashion with almost 95% of all the well picks
accurately picked using this approach. The intervening horizons are not as well defined on the log signature but still
can be picked with about a 50% manual to automated ratio.
Applications/Significance/Novelty: A new technology that utilizes the power of modern computers with increasingly
powerful and accurate machine learning algorithms provide the geoscientist with a new toolkit to correlate obvious
geologic events rapidly over thousands of wells.
The combination of dynamic transects, geologic cross sections that can support hundreds of wells, automated well top
picking, and stratigraphic flattening provides a new workflow to identify quickly correlative stratigraphic patterns that
initially seem erratic or uncorrelated.
Because of the speed of picking events, local ‘phantom events’ can be picked and then flattened to investigate
bounding stratigraphy.
Once an initial model is built, new interstitial data can be added and interpreted nearly automatically.
Used by multidisciplinary teams for fast scouting of plays, it is a technique that can get the majority of the answer in a
fraction of the time.
Stimulation Performance Indicators and Machine-Learning-based Analytics in the Utica shale: Case Study and
Lessons Learnt
1, 2
S. Perrier; ; A. Delpeint; 1, 2; A. Shrestha; 2; Z. Shawuti; 2;
1. UTICA Joint Venture, TOTAL, Oklahoma City, OK, United States.
2. UTICA Joint Venture, Chesapeake Energy, Oklahoma City, OK, United States.
Objectives/Scope: The Utica shale play exhibits a significant variability of petrophysical and fluid properties, primarily
due to structure dip and differences in burial between sectors.
This paper, based on the case study of the Utica shale, discusses several aspects associated to the implementation of
machine learning in predicting the performance of gas shale wells, adressing many issues associated to the choice of
the indicator, the data treatments of predictor/attributes, and proposes a method to deal with plays with strong
geologically related performance trends.
Methods/Procedures/Process: The methods presented in ths paper involve:
A comparison of different machine learning prediction approaches applied to the Utica shale, and discussion
A presentation of an original hybrid machine learning model, applied to the Utica shale, that enhances the distinction
between the geological and stimulation component
The value of modern machine learning versus conventional multi-variate analysis is also reviewed
The choice of the cost function to be predicted and optimized is also discussed, starting from classical cumulative
production indicators (Cum360, Cum360 per foot) to more sophisticated rate-pressure indicators (EUR, AsquarerootK,
and their normalized pairs, etc…), to provide an insight on the trade off robustness/simplicity of each indicator.
Results/Observations/Conclusions: The key results presented in this paper are:
The value of the methodology that combines a de-trending of the field leading trends based on geology with an
appropriate level of sophistication of machine learning models. Indeed, on-the-shelf ML solutions need to be adjusted
to the specificities of gas shale data paradigm (small datasets, high complexity of parameters, non linear geological
and spatial trends).
The choice of the indicator used for predictive models matters, as the level of intrinsic noise can vary much between
indicators. Although conclusions based on most indicators are consistent, the choice of the appropriate indicator is key
to take informed economic decisions when it comes to completion design.
Applications/Significance/Novelty: The paper will not disclose the detailed practical conclusions extracted through the
use of these multiple modles in term of comparison of the completion strategies, but will show with real data numerous
technical solutions reducing the complexity of the deconvolution of the impact of multiple parameters, to define the
best stimulation design.
Unlocking Reservoir Potential With a Multilayer Inversion Technique From a Directional Resistivity Tool
1 1 2 2 2
T. Rathmann; ; P. Lemay; ; A. Nandlal; ; J. Gremillion; ; M. Flowers; ;
1. US Exploration, Crescent Point Energy, Denver, CO, United States.
2. NAM Well Placement, Schlumberger Ltd., Denver, CO, United States.
Objectives/Scope: The objective of the study was to examine the effectiveness and results of geosteering horizontal
wells in the Green River-Wasatch Formation systems of the Uinta basin in northeast Utah using a deep directional
resistivity tool. Using the real-time Logging While Drilling (LWD) measurements allowed maximizing exposure in open
and marginal lacustrine sands and at the same time provided imaging information that would help describe the
horizontal target zone thickness along with the presence of any faults without drilling additional vertical delineation
wells.
Methods/Procedures/Process: Steering within the lacustrine sands of the Uinta Basin is often difficult because the
primary target sands can be laterally discontinuous and below seismic resolution where seismic data is
available.Changes in channel direction or thickness can create log signatures that make it difficult to determine if the
target sand has pinched out or the trajectory exited the target zone due to a local change in formation dip or faulting.
Using a deep directional resistivity tool that can image the resistivity contrasts at the boundaries of the target zones
allowed staying in the better resistivity reservoir rock with fewer directional changes. Inversion information was useful
to measure the top and base of the target zone without drilling up or down which mitigated a risk of exiting the target
zone.
Results/Observations/Conclusions: By using azimuthal resistivity information while drilling, the horizontal wells drilled
were able to be geosteered more effectively in the highest prospective reservoir rock leading to longer lateral lengths
(one mile or greater). In addition, an increase in the lateral length can potentially support drilling multiple wells from
the same location which would lead to a smaller surface footprint in a development program.
The real-time LWD data allowed the trajectories to be steered in zone for longer and the data was used post well to
create new estimations of sand thickness in areas with sparse offset well data. Using the inversion technique,
formation thicknesses were interpreted to be thicker than seen on the offset well data and faults were more easily
identified.
Applications/Significance/Novelty: Using an advanced inversion technique in real-time enables the mapping of the
reservoir thickness while drilling, and because the inversion isn’t dependent on a prior model, a more accurate
representation of the stratigraphy can be mapped with confidence. Based on the information provided by the
inversion, the target thickness was able to be mapped as the well drilled away from the pilot well. After encountering a
fault, the deep reading inversion allowed for a smooth re-landing back in the target zone. The data was easily
visualized and shared over the internet in real-time allowing decisions to be made quickly without slowing the drilling
operation.
Novel Diversion Case Study for Improved Near-Wellbore Connection Between Wellbore and Hydraulic Fracture
1 1 1 2
W. P. Scanlan; ; K. Pierskalla; ; D. W. Sobernheim; ; R. Boehringer; ;
1. Engineering, Keane Group, Denver, CO, United States.
2. Engineering, Imerys, Denver, CO, United States.
Objectives/Scope: Recent industry advancements have utilized 100% degradable polymeric diverting agents as a
means of enhancing stimulation efficiency on new wells via intrastage diversion, as well as improved coverage in
refracturing of existing wells. These diverting agents redirect treatment fluid and proppant into previously under-
stimulated perforation clusters by bridging and packing of the initial flow paths. Obstructing the current treatment flow
paths of active clusters encourages the propagation of new hydraulic fractures, thereby expanding the stimulated rock
volume (SRV).
Methods/Procedures/Process: As the degradable particles within the diverting agent hydrolyze, the space previously
filled with the particulate closes, reducing conductivity to the fracture network adjacent to the wellbore. Therefore,
despite an improved hydraulic fracture placement across various clusters, the ultimate productivity can be impaired
due to this lack of a near-wellbore conductivity. To address this problem, a novel diversion material blend was created
to mitigate fracture throat closure and dramatically enhance near-wellbore conductivity. Through the combination of
conventional 100% degradable material of varying form-factors and a high-conductivity, rod-shaped ceramic proppant,
both effective diversion and high near-wellbore conductivity can be achieved during new well operations and
refracturing.
Results/Observations/Conclusions: The rod-like ceramic material also impedes flowback into the wellbore of any
proppant on the same level as a fully curable resin sand.This paper chronicles the evaluation of various geometric
bridging efficiencies, retained conductivities, temperature range suitability, and placement performance for application
in the field
Applications/Significance/Novelty: A case study in the Williston Basin is included describing the practical execution of
treatments using this formulation and the results obtained from the application.
Simulation of Proppant Transport in Foam Fracturing Fluid Based on Experimental Results
1 1 2 1
S. Tong; ; K. Mohanty; ; M. Gu; ; R. Singh; ;
2. Petroleum Engineering, West Virginia University, Morgantown, WV, United States.
Objectives/Scope: Slick water fracturing is used commonly because it produces long skinny fractures which are
essential for production from shales. The proppants however settle near the wellbore region in slick water fracturing
leaving a significant part of the fractures unpropped. Gel fracturing can carry the proppants better, but can damage the
fracture faces because the pores in shales are submicron in size. Foam fracturing can improve the proppant transport
and at the same time reduce water usage. The goal of this study is to understand and model transport of concentrated
proppants in foams.
Methods/Procedures/Process: A transparent slot, 750mm x 150mm x 2mm, was constructed to mimic hydraulic
fractures. Foam slurry was prepared with a blender and injected into the slot with a peristaltic pump. Sand and
ceramic proppants were used; ceramic proppants were easier for visualization. “Tracker” software was used to track
the proppant trajectory. The effects of several factors including proppant loading, foam quality and injection rate were
investigated. Correlations were developed between the key factors and the proppant velocity vector. These
correlations were incorporated into an in-house finite-difference numerical foam-fracturing simulator. The simulator
incorporates a geomechanical model, a planar fracture propagation model, a proppant transport/settling model, and a
foam rheological model.
Foam significantly decreased gravity settling of proppants compared to that of water. At high foam quality, proppants
followed the trajectory of the foam. At low foam quality, the drainage of the liquid phase due to gravity leads to two
effects: 1) a decreasing foam quality from the top to the bottom of the fracture; 2) a drag force on proppants
depending on the vertical position. A new settling velocity correlation was developed based on the two effects
observed in the lab test, and then incorporated into a field-scale fracture simulator. Different fracture conductivity
distributions along fracture height and length in the field scale were computed from the simulator as a function of
proppant mesh sizes, proppant loadings, foam qualities, and pumping schedules.
Applications/Significance/Novelty: The proppant transport in foam within a fracture is visualized and quantified. The
settling velocity of proppant depends on the vertical position of the proppant due to foam drainage process. A new
correlation is developed for the proppant settling velocity in foam as a function of key foam and proppant parameters.
The new correlation is incorporated into a field scale fracture simulator that estimates fracture conductivity distribution
along the fractures.
Multi-Physics Pore-Scale Modeling of Particle Plugging Due to Fluid Invasion During Hydraulic Fracturing
1 1 1 1
Y. Zapata; ; X. Dong; ; T. N. Phan; ; Z. A. Reza; ;
Objectives/Scope: The productivity from hydraulically fractured wells is contingent to different mechanisms of
formation damage by the fracturing fluid, such as fluid leakoff, proppant embedment, and fines generation from
proppant crushing. Particle plugging during fluid invasion in hydraulic fracturing is considered to have a significant
detrimental effect on production. In this work, we developed a multi-physics pore-scale plugging simulator integrated
with a fracture simulator to investigate the impact of particle plugging and quantify the formation damage of the
invaded region near hydraulic fractures.
Methods/Procedures/Process: Our multi-physics simulator bridges pore-scale phenomena with those occurring at the
reservoir scale. The pore-scale particle-plugging simulator is capable of using realistic 3-D pore-network models. We
construct these models based on pore-body and pore-throat characteristics derived from scanning electron
tomographic (FIB-SEM) images. Based on filtration theory we compute dispersion, retention, and permeability
changes due to fracturing-fluid invasion. These parameters are interfaced with a fracture simulator that provides
pressure, leak-off rate, proppant size and concentration to the pore-network model. These parameters are
representative of typical multi-stage hydraulic-fracturing operations. The fracture simulator also provides fracture-
geometry and fracture-conductivity.
Results/Observations/Conclusions: Our methodology enables a novel and practical way to understand the particle-
plugging process in the matrix-fracture interface during hydraulic-fracturing operations. The coupled model considers
heterogeneity at pore scale to account for the role of pore structure on particle transport, concurrently studying its
behavior on the fracture. We used FIB-SEM images obtained from Eagle Ford shales in this study. From the
sensitivity analysis performed for different hydraulic-fracturing operational constraints and fluid properties like proppant
size and concentration, we are able to determine the effect of these variables on matrix-fracture permeability
impairment at the pore-scale. The multi-scale integration allowed us to quantify the influence of filtration on the
fracturing operations.
Applications/Significance/Novelty: This novel integrated particle-plugging simulator aims to improve the design and
modeling of hydraulic-fracturing operations in unconventional shale reservoirs. Using this modeling workflow in
conjunction with other hydraulic-fracturing modeling techniques, one can improve the decision-making process for
fracturing fluid and proppant selection. Matrix-fracture permeability alteration during production phase can also be
addressed using our proposed modeling framework. Furthermore, our workflow can be applied in similar scenarios
that include fines migration or particle deposition due to fluid injection or production in water-disposal operations or
asphaltene-deposition processes, when the pore-scale model is coupled with a reservoir-scale simulator incorporating
appropriate physics.
Understanding the Mechanism of Interwell Fracturing Interference Based on Reservoir-Geomechanics-Fracturing
Modeling in Eagle Ford Shale
1 1 1 1
X. Guo; ; K. Wu; ; J. Killough; ; J. Tang; ;
1. Petroleum engineering, Texas A&M University, College Station, TX, United States.
Objectives/Scope: Tightly spaced horizontal wells are widely used in the development of unconventional resources.
The effectiveness of this strategy is largely affected by interwell fracturing interference indicated by interwell fracture
geometry and frac hits, as interwell interference affects both parent and infill well productions. This work proposes a
reservoir-geomechanics-fracturing modeling workflow for understanding the interference mechanism and quantifying
effects of parent well fracture geometry, differential stress, and the design of infill well completion on interwell
fracturing interference.
Methods/Procedures/Process: Reservoir models are constructed for the analysis of various Eagle Ford scenarios. The
numerical workflow involves a finite element model that fully couples reservoir flow and geomechanics and a complex
multi-fracture propagation model coupling rock mechanics and fluid flow in wellbore and fractures. The workflow
characterizes the temporal-spatial evolution of pressure and stress caused by legacy parent well production. The
fracture model is employed to simulate the complex fracture geometry created by infill well completion based on
updated heterogeneous reservoir stress state. The resulting fracture geometry quality is quantified by occurrence of
frac hits, relative growth of fractures on longitudinal and transverse directions, and depletion potential of the fractured
reservoir volumes.
Results/Observations/Conclusions: Non-uniform fracture geometries lead to more complex stress change induced by
depletion than uniform fracture geometries in parent wells. A smaller in-situ differential stress results in stronger stress
reorientation caused by parent well depletion, which induces longitudinal fractures in infill wells and greatly reduces
stimulated reservoir volume and initial well performance of infill wells. A larger in-situ differential stress induces less
stress reorientation and is more likely to lead the fractures propagating toward pre-existing fractures, generate frac
hits, and affect production of parent wells. The quantification study in the sensitivity analysis indicates that differential
stress and the infill well completion design have the most significant influences on interwell interference.
Applications/Significance/Novelty: Based on numerical modeling using field data, the study suggests optimum infill
well completion designs for several Eagle Ford scenarios. The study also provides insights for infill well completion
design in unconventional reservoirs developed by tightly spaced horizontal wells in terms of how to adjust field
operational schedules to avoid frac hits and change the complexity of the interwell fracture networks.
Modeling of Fluid Injection in Depleted Parent Wells to Minimize Damage Due to Frac-Hits
1 1 1
D. P. Gala; ; R. Manchanda; ; M. Sharma; ;
Objectives/Scope: Preventing negative “frac-hits” on parent wells during infill-well fracturing remains a significant
challenge for the shale operators. Some of the available options include (a) changing the frac-job designs of the child
wells (b) pumping fluid into the parent well to re-pressurize the near wellbore area (c) refracturing the existing parent
well before completing the child wells. In this work, we further explore the option of pumping fluid (water or gas) into
the parent well to form a protective shield by increasing the pressure (stresses) and ward off incoming fractures from
child well.
Methods/Procedures/Process: In this work, we developed and used a simulator which solves component mass
balance, pressure equations which are coupled with rock deformation. The model calculates stress changes due to
both poroelastic (pressure changes) and mechanical (fracture opening) effects. The phase behavior of gas injection
with a specified composition which is different form the in-situ reservoir fluid composition is accounted using phase
stability and flash calculation algorithms. Steps are a) Parent well is produced for few years and pressure,
composition, saturation and stresses are calculated b) Injection into parent well is simulated by specifying the injection
rate, composition and maximum allowable pressure c) The stresses and pressure around the parent well are recorded
both during and after injection.
Results/Observations/Conclusions: We performed several simulations on a representative dataset from an
unconventional reservoir. We are able to observe the following trends using our simulation studies: (a) Stresses travel
much farther away from the parent well as compared to pressure during injection (b) water and gas injection
simulations show considerable differences in the observed pressures and stresses due to differences in
compressibility, relative permeability and phase behavior of these fluids (c) the practice of shutting in the parent well
does not lead to substantial increase in pressure and stresses as compared to fluid injection in parent wells (d) In
some cases, the required injection gas amounts are very large in order to observe any substantial increase in
pressures or stresses around the parent well.
Applications/Significance/Novelty: We developed a fully coupled geomechanical compositional reservoir simulator in
this work to model fluid injection (water and gas) for achieving parent well protection. This work will be able to provide
any shale operator with significant insights on how to design the fluid injection process in a parent well. Specifically,
several design questions can be answered such as (a) injection volumes (b) injection rates (c) composition of the
injected gas (d) how far from the parent well do the pressure and stresses increase and (e) time interval before the
increased pressure and stresses start diffusing. Also, this work will further improve understanding of when gas
injection works and when it should be preferred over water injection option.
The EGS Collab Project: A Field Stimulation Study in Crystalline Rock to Validate Models
2 1
T. Kneafsey; ; D. A. Blankenship; ;
1. Sandia National Laboratories, Albuquerque, NM, United States.
2. Lawrence Berkeley National Laboratory, Berkeley, CA, United States.
Completion Engineer for a Day: How Geology and Geomechanics Can Influence Completion Designs in
Unconventionals
1
N. Nagel; ;
1. OilField Geomechanics, Richmond, TX, United States.
Modeling of Hydraulic Fracture Height Growth Through Weak Interfaces
1
X. Weng; ;
The Formation and Properties of Complex Fracture Networks in Shales
1
M. Sharma; ;
1. Department of Petroleum, Geosystems and Chemical Engineering, University of Texas, Austin, TX, United States.
Quantitative Mineralogy of Vaca Muerta and Alum Shales From Core Chips and Drill Cuttings by Calibrated SEM-EDS
Mineralogical Mapping
1 1 1 1 1 1 1 1
C. I. Fialips; ; B. Labeyrie; ; V. Burg; ; V. Maziere; ; Y. Muneral; ; H. Haurie; ; I. Jolivet; ; R. Lasnel; ;
1 1 1
J. Laurent; ; L. Lambert; ; L. Jacquelin-Vallee; ;
1. TOTAL S.A., PAU, France.
Objectives/Scope: Well logs are often the only continuous acquisitions across targeted reservoir intervals. In shale
reservoirs, the most important uncertainties for quantitative log interpretations are related to the clay and organic
matter end points but difficulties also arise from complex mineral mixtures without a well differentiated lithology.
Accurate quantitative mineralogical analyses of reservoir samples are essential for calibrating well log interpretations.
Our new quantitative mineralogy workflow will be presented, with two case studies: Vaca Muerta Shales (Argentina)
and Alum Shales (Denmark).
Methods/Procedures/Process: Our reference quantitative mineralogy method integrates X-Ray Diffraction analyses
with a series of other measurements performed on crushed samples. It is very accurate but labor-intensive and not
well adapted when applied on contaminated drill cuttings. Our newly developed workflow is based on the
mineralogical mapping of thin-sections, core chips and resin-impregnated drill cuttings using Scanning Electron
Microscopy equipped with Energy Dispersive X-Ray Spectroscopy (SEM-EDS). The 2D SEM-EDS elemental maps
are post-treated using a dedicated software allowing the deconvolution of each EDS spectrum using up to three
mineral spectra. The SEM-EDS data are calibrated with a limited number of bulk quantitative mineralogical analyses
performed on crushed samples.
Results/Observations/Conclusions: Non-calibrated mineralogical data of Vaca Muerta and Alum Shale samples are
not directly equivalent to volume% because microporosity is not resolved at the resolution of EDS acquisitions. The
data in area% can however be converted into fairly accurate mass% based on apparent densities through careful
calibration with quantitative mineralogy data obtained on crushed samples. For drill cuttings, drilling mud
contaminations and fall-out materials can be numerically removed to only analyze the mineralogical composition of the
targeted formation intervals. One significant limitation of the SEM-EDS mapping is that it does not allow correctly
quantifying organic matter. For source rocks, insoluble organic matter contents measured on crushed samples are
therefore integrated in our workflow.
Applications/Significance/Novelty: Results obtained for Vaca Muerta Shale samples from several wells have shown
that the calibration of SEM-EDS data is similar throughout the Neuquen Basin but is different from that required for
Alum Shale samples or other reservoir rocks. The calibration of SEM-EDS data is therefore basin-dependent and
should be carefully completed for each new play in order to produce reliable quantitative mineralogy data. One great
advantage of the calibrated SEM-EDS workflow is that it produces both a set of bulk mineralogical composition data
for calibrating well log interpretations but also a set of 2D maps of the samples showing the micrometer-scale
distribution and size of mineral grains that can be used to identify the main lithofacies.
Development of a Mixed Polymer Fracturing Fluid for High Temperature Applications
1 2 2 2 1
T. Almubarak; ; J. Ng; ; K. Sokhanvarian; ; H. Nasr-El-Din; ; M. Khaldi; ;
1. Saudi Aramco EXPEC ARC , Bryan, TX, United States.
Objectives/Scope: Recent interest in deep low permeability formations demands growth in high-temperature hydraulic
fracturing technologies. Conventional guar systems break down above 300°F and require higher polymer loadings to
maintain thermal stability. Synthetic polymers designed for high-temperature applications utilize specialty monomers
that are expensive. Individually, both these polymer are lacking in either performance or price. However, when these
polymers are combined, they exhibit a synergistic interaction and produce higher viscosity and thermal stability than
the individual polymer systems.
Methods/Procedures/Process: The mixed polymer solution was prepared at a concentration of 20, 30 and 40 lb/1,000
gal at a ratio of 1:1 (CMHPG: Polyacrylamide). Hydrated polymer solutions were crosslinked with a metallic crosslinker
and broken with an oxidizer at 300°F. Testing was carried out in a HP/HT rheometer to evaluate viscosity, thermal
stability, salt influence, and breaking. HP/HT see through cell and aging cell were used for visual assessment of
proppant settling, and final broken polymer residue, respectively.
Results/Observations/Conclusions: Results show that the 20, 30 and 40 lb/1,000 gal mixed polymer fracturing fluid
-1
generates a stable crosslinked viscosity at 300°F and 100 s . Results also indicate that the synthetic polymer
enhances thermal stability while CMHPG increases the peak viscosity of the mix. Therefore, the mixed polymer
system produces a fracturing fluid that can handle much harsher conditions than the individual polymers. Moreover,
the polymer mixture fracturing fluid exhibits good tunable delayed crosslinking, is able to carry proppant effectively at
high temperature and shows a controlled clean break with an oxidizer.
Applications/Significance/Novelty: Extensive experiments were pursued to evaluate the new high temperature mixed
polymer system for the first time. This mixture exhibits a positive interaction between polysaccharide and
polyacrylamide families and generates excellent thermal stability. Mixed polymer systems are a viable solution to
combat the challenges currently faced in the industry such as increasing high temperature stability, high salt tolerance,
and reducing the polymer loading used. Polymer mixture technology can potentially reduce material cost, simplify field
operation and reduce damage to the proppant pack and formation.
Enhancing Friction Reducer Performance in High Salt Conditions
1 1 1
B. Seymour; ; A. Sanders; ; D. Friesen; ;
1. Oilfied, Stepan Company, Houston, TX, United States.
Objectives/Scope: This study covers an ongoing investigation on the discovery and understanding of how nonionic
surfactants enhances the performance of friction reducer polymers in increased amount of salinity. Initially, surfactants
were screened with a freshwater friction reducer in seawater. Surfactants that showed potential were then tested with
single cations (potassium, calcium, sodium, magnesium, and iron) to understand role of each ion. Lastly, surfactant
blends were evaluated in brines simulating produced water conditions from 23,000 to 174,000 mg/l total dissolved
solids.
Methods/Procedures/Process: Friction reduction performance was measured using a friction flow loop with a 20 foot
segment of ½ inch outer diameter tubing at a flowrate of 8 gal/min. Initial screening used a brine that simulated
seawater with a freshwater friction reducer. Surfactants that showed positive results were then tested with single
cations at increasing concentrations (500 ppm increments for divalent cations and 5000 ppm for monovalent cations).
Tests were then conducted with at least six different brines that simulated field produced water. Further testing looked
at the optimized concentration of the surfactants were looked at, as well as, testing with various different friction
reducer polymers.
Results/Observations/Conclusions: Initial screening found two chemistries that not only reduced the inversion time,
but also increased the maximum percent friction reduction by 20% and maintain stability over time. The single cation
testing showed a known trend that polymers are more sensitive to divalent cations over monovalent cations, however
surfactants still enhanced overall performance. The optimized concentration of the surfactant blend was between 0.02
to 0.08 gpt depending on the salt conditions. The tests with the simulated produced water indicated that surfactant
boosted system could withstand higher concentrations of divalent cations in the presence of other ions, compared to
the single cation testing. Surfactant blends enhance a variety of FRs including anionic and cationic types.
Applications/Significance/Novelty: The use of these nonionic surfactant chemistries in stimulation applications is novel.
There has been one field study completed, and currently undergoing a second. The use of these surfactant blends
could potentially reduce the amount of polymer needed to obtain performance, it increases the range of performance
of less functionalized polymers, and allows friction reducers systems to perform much better in produced water.
Can Friction Reducers Transport Sand During Fracturing Treatment?
1 1 1 1
L. Shen; ; L. Vigderman; ; D. Heller; ; D. Fu; ;
1. BJ Services, Tomball, TX, United States.
Objectives/Scope:
The use of friction reducers (FRs) has increased rapidly, not only in traditional slickwater treatments, but also as a
replacement for linear and even crosslinked guar based fluids for proppant transport at higher loadings. However, the
current understanding of proppant carrying capability of FRs is limited since most of the literature is focused on simple
viscosity measurement at high shear rate. We report herein the study of full rheological characterization of FR based
fracturing fluids.
The viscosity and elasticity of three FRs were measured across a wide range of concentrations under varying shear
conditions using advanced rotational rheometer. Proppant carrying capability was assessed using the static proppant
settling test as well as via slot flow testing designed to mimic the fracturing fluid traveling through a fracture. For
comparison purpose, the traditional linear gel and crosslinked guar gel system were tested with the same apparatus.
Field data were also collected to validate key conclusions from the study.
Unlike traditional fracturing fluids, there are different mechanisms that lead to enhanced proppant suspension for FR
fluids. The authors correlated the results of viscosity measurements over a wide range of shear rates, elasticity
measurements, static proppant settling tests, and dynamic proppant settling tests, and proposed mechanisms that are
dependent on different conditions. This helped to explain why some products showed poor proppant transport
capability while having relatively high viscosity. The industry accepted testing standards appear not suitable in
measuring FR proppant carrying property. The selection of FRs should follow different testing protocols.
This paper is the first to provide a comprehensive study of FRs as linear or crosslinked guar system replacement for
transporting proppant. An in-depth understanding of the key factors controlling proppant carrying capability of FRs is
critical to improved fracturing fluid design and can provide guidance for future FR development.

URTeC 2018 Julio

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

URTeC 2018 Julio

Uploaded by

Copyright:

Available Formats

Program Schedule and Abstract Book

Time Session or Event Info

8:30-8:35 AM Moderator Welcome

8:35-8:45 AM Opening Remarks S. Winberg

Re-Invent, Re-Tool, Re-Imagine: Finding Success in the Resources

9:00-9:15 AM Enigmatic Shale S.W. Tinker

What the Investor Community Wants From the Unconventional

9:30-10:00 AM Moderated Discussion

10:45-10:50 AM Introductory Remarks

10:50-11:00 AM Reservoir Engineer, Shell R. Cao

11:00-11:10 AM Vice President of Development, Paramount Resources G. Stotts

11:10-11:20 AM Reservoir Development Manager, Alta Mesa S. Geetan

11:20-11:45 AM Moderated Discussion

11:45-12:05 PM Audience Q&A

10:45-10:50 AM Introductory Remarks

Novel Pore Pressure Prediction Technique for Unconventional

Using Traditional Methods to Predict Pore Pressure in Devonian

10:45-10:50 AM Introductory Remarks

Montney Key Drivers: An Integration of Multidisciplinary Data

Porosity and Organic Content Analysis, Bone Spring and Wolfcamp

A Water Saturation Interpretation Model for Organic-Rich Shale

10:45-10:50 AM Introductory Remarks

Spider Bots: Database Enhancing and Indexing Scripts to Efficiently

Integrated Workflow for the Definition of a Type Well Using

Toward a Management Science for Unconventional Wells: A

10:45-10:50 AM Introductory Remarks

Emerging Shale Oil Plays in Hypersaline Lacustrine Qianjiang

Unconventional Reservoir Development in Egypt’s Western Desert:

10:45-10:50 AM Introductory Remarks

Mechanism Study of Casing Deformation in Multistage Hydraulic

Optimization and Drilling of Horizontal Wells using a Bayesian

TST3D: Automated Structural Interpretation in Horizontal Wellbores

10:45-10:50 AM Introductory Remarks

10:50-10:58 AM Oak Ridge National Laboratory Y. Polsky

10:58-11:06 AM Los Alamos National Laboratory R. Pawar

11:06-11:14 AM Lawrence Berkley National Laboratory T. Kneafsey

11:14-11:22 AM Senior Fellow, Geologic and Environmental Systems G.S. Bromhal

11:22-11:45 AM Moderated Discussion

11:45-12:05 PM Audience Q&A

10:45-10:50 AM Introductory Remarks

Reliable Solid Organic Matter Thermal Maturity Assessment Using

Integrated Inorganic and Organic Geochemistry Approach in the

Geochemical Characterization of the Eagle Ford Formation in

10:30-10:35 AM Introductory Remarks

Using Drone Magnetic and LiDAR Surveys to Locate Unmarked,

Advanced Characterization and Novel Waste Management for Drill

11:25-11:50 AM Estimating Carbon Intensity of Unconventional Plays D.N. Meehan

10:30-10:35 AM Introductory Remarks

Mapping the Barnett Shale Gas With Probabilistic Physics-Based

Hindcasting Production Forecasts in Four Shale Gas Basins Using a

Montney Versus North America – Completions Comparison K. Ogilvy;

10:30-10:35 AM Introductory Remarks

10:35-11:00 AM Wolfcamp Geologic Reservoir Modeling Challenges B.J. Casey

Additional Applications of Optimal Artificial Lift Strategies in the

Additional Applications on Determining Optimal Lateral Lengths and

1:45-1:50 PM Introductory Remarks

An Integrated View of the Petrology, Sedimentology, and Sequence

Outcrop and Subsurface Geology Applied to Drilling, Sweet Spot and

Geological Facies Prediction Using Computed Tomography in a

3:05-3:55 PM Refreshment Break

Recent–Pleistocene Immature Mudrocks of Anoxic Basins of

Integration of Core Fracture and Lithofacies Descriptions in the