Importance of regional geological and petrophysical analysis: Bakken/Three Forks formations of the Williston basin North Dakota and Montana Presenter: Stefani Brakenhoff The Discovery Group Team: Bob Cluff, Daniel Hallau, Catherine Murphy, Nikki Oliver, Andrea Simenson
The Highs of the Bakken From Institute for Energy Research 2011, 468 rigs were operating in ND 25% of the US rigs at the time. March 2012 ND produced ~18 million barrels of oil Surpassed Alaska and California, second only to Texas. http://instituteforenergyresearch.org/wp-content/uploads/2012/08/bakken-fact- Sheet.pdf 10/1/2016 www.discovery-group.com 2
What do operating companies focus on in the busy times? In our experience, assembly line approach for: Drilling teams and programs Drilling out leases and acreage Completion and production programs 10/1/2016 steppe.cr.usgs.gov www.discovery-group.com 3
Geological and Petrophysical Variation Looking at variation on a regional scale: Long range planning Lease acquisition High-grading available opportunities 10/1/2016 www.discovery-group.com 4
The Method Collect data Clean up data, Select type wells, Normalize Create stratigraphic framework Create models Type wells --> Bulk application Create maps & cross sections 10/1/2016 www.discovery-group.com 5
1. Collection of Data Data Sources: Publicly available state website data Digitized paper logs Data services (IHS, MJLogs, Mesa Data, etc.) Subset of type wells Log curves needed for models Quality of data Preferred service companies Supplemental data available Areal distribution Well normalized Core Lab reporting differences Type of core data 10/1/2016 www.discovery-group.com 6
Williston Basin Data Example Non-proprietary data All data into Petra From NDGS 494 digital logs ND 692 raster logs ND From MJlogs 211 rasters for MT 32 w/core analyses 10/1/2016 www.discovery-group.com 7
2. Clean up and Normalization Clean up Nulling tool pick-up & invalid data Merge & splice files and curves Potential problems: Multiple runs or missing runs Sources of data Multiple versions of curves Digitization verification Normalization Histograms and cross-sections 10/1/2016 www.discovery-group.com 8
POST PRE Normalization Example 10/1/2016 www.discovery-group.com 9
3. Stratigraphic Framework Create a regional grid of wells: Digital data Full logging suite (GR, NPHI, RHOB, ILD) Cored wells 10/1/2016 www.discovery-group.com 10
Formation Tops in Williston Basin Scallion Fm. Upper Bakken Middle Bakken *E-facies *D-facies *C-facies *B-facies *A-facies Lower Bakken Lower Bakken Silt Pronghorn Limestone Upper Pronghorn Lower Pronghorn Top Three Forks Middle Three Forks Lower Three Forks L. Three Forks Marker Birdbear Duperow Tops picked for 940 wells *Facies picks based on Sonnenberg (2011), Search and Discovery Article #50449, AAPG Annual Convention 10/1/2016 11
4. Setting up the Models 1. Lithology model 2. Total organic carbon (TOC) model 3. Grain density model 4. Porosity model 5. Water saturation model Do you have reliable and significant core data, NMR data, ECS or ELAN to calibrate/compare all models? 10/1/2016 www.discovery-group.com 12
4.1 Lithology Modeling Multi-variate mineral regression Requires accurate core/log depth matching Linear & non-linear options XLSTAT Excel extension Upper & Lower Bakken, Pronghorn, Three Forks Bakken Shales 10/1/2016 www.discovery-group.com 13
Apparent Matrix Volumetric Photoelectric Factor vs. Apparent Matrix Grain Density (RHOMAA-UMAA) WHY in Middle Bakken?? Thinly bedded 6 facies Core statistical significance RHOB, NPHI and PE Doveton* matrix inversion 3 lithology system plus VCLAY from GR PE availability and accuracy NPHI s need to be normalized Blue Triangle Default SLB values Red Triangle End points that match XRD data Pink shading from light to dark pink is increasing volume of dolomite *Doveton, 1994, Computer Applications in Geology No. 2 Geologic Log Analysis Using Computer Methods 10/1/2016 14
TOC 4.2 TOC Modeling Simple log/core relationship 4.5 4 3.5 RHOZ vs. TOC y = -7.9936x + 22.398 R² = 0.6183 3 2.5 2 1.5 1 0.5 0 2.3 2.4 2.5 2.6 2.7 2.8 RHOZ Probabilistic modeling All minerals, TOC, porosity, water saturation 10/1/2016 www.discovery-group.com 15
DeltaLogR Method by Q.R. Passey logr DEN = log 10 R R baseline 2.5 ρ b ρ baseline TOC = logr DEN 10 2.297 0.1688 LOM 10/1/2016 www.discovery-group.com 16 Passey, Q.R., 1990, AAPG V74, No.12
4.3 Grain Density Modeling Variable grain density TOC corrected, weight% minerals to calculate grain density (RHOG) RHOG = 1 TOCcorrected weight% of each mineral σ grain density of each mineral Total clay and TOC grain densities solved for using Excel Solver 10/1/2016 www.discovery-group.com 17
4.4 Porosity Modeling What porosity to use? DPHI, NPHI, SPHI, NDPHI, NMR porosity Quality of data Normalized Deterministic method Var PHI = RHOG RHOB RHOG RHOFL PHIE = Var PHI Vclay gd_clay 10/1/2016 18
4.5 Water Saturation Modeling Archie water saturation (SW) Arps equation 150,000 ppm Corrected with formation temp. ( F) m & n determined by zone using Pickett plots and published values a m n Upper Bakken Shale 1 2.00 2.00 Middle Bakken 1 1.50 1.50 Lower Bakken Shale 1 2.00 2.00 Pronghorn 1 1.55 1.55 Three Forks 1 1.55 1.55 19
5. Type Wells Bulk Application Multiple Software Packages developing: Excel XLSTAT Multi-variate regression for lithology Core data organization LESA Normalization RHOMAA-UMAA lithology Petrophysical modeling IP Probabilistic modeling TOC Passey model PETRA QC for normalization Application of models by user programs QC of models Consolidated to Petra: No LAS import/export Change model easily 10/1/2016 www.discovery-group.com 20
6. Maps and Cross-sections Attribute maps Geological thickness Lithology maps Petrophysical property maps Cross-sections Stratigraphic correlation with grids Petrophysical properties with stratigraphic reference and relevance 10/1/2016 www.discovery-group.com 21
Middle Bakken Thickness 10/1/2016 www.discovery-group.com 22
Middle Bakken median Volume of Calcite 10/1/2016 www.discovery-group.com 23
Middle Bakken Variable Density Porosity 10/1/2016 www.discovery-group.com 24
A+ 5 6 7 8 9 10 11 A+ Petrophysical Index Map B+ C+ D+ E+ F+ G+ H+ I+ J+ K+ L+ 5 6 7 8 9 10 11 B+ C+ D+ E+ F+ G+ H+ I+ J+ K+ L+ 10/1/2016 25
Petrophysical Log Model Cross-Section (H-H ) Upper Bakken Middle Bakken Lower Bakken Pronghorn Three Forks 10/1/2016 www.discovery-group.com 26
Summary Regional geological and petrophysical databases help companies plan Method for large amounts of data 1. Collect log and core data 2. Clean up data, choose type wells, normalize data 3. Create stratigraphic frame-work of consistent formation tops 4. Create and calibrate models 5. Apply models to bulk database 6. Produce maps and cross-sections 10/1/2016 www.discovery-group.com 27
Thank You! Any questions? Stefani Brakenhoff stefanibrakenhoff@discovery-group.com 303-831-1515 x203 The Discovery Group Inc. 6795 E. Tennessee Ave. #500 Denver, CO 80 10/1/2016 www.discovery-group.com 28