Log Analysis Using Microsoft Excel® Focus on the Marcellus Tim Carr West Virginia University Advanced” log analysis techniques using commonly available modern logs and a Microsoft® Excel spreadsheet that is applicable to the Marcellus Shale

My Observations http://www.snopes.com/photos/natural/canyonleap.asp

Background Costs Are Becoming More Significant High Land Costs More Moderate Commodity Price High Capital Costs Horizontal Wells & Large Multi-Stage Fracture Stimulations Key Reservoir Parameters Thickness Unit Definitions (Formation  Bed) Lithology Thermal Maturity Total Organic Carbon (TOC) Gas Fraction (Adsorbed and Free) Permeability More Energy at Lower Real Cost Resources are Adequate Increased Access to Energy Slow Incremental Changes Fossil Energy will be the Major Component of our Energy System for this Century Need for Investment

AVERAGE WELL HEAD PRICE $2.95 per MMBtu 2002 $6.25 per MMBtu 2007 $7.96 per MMBtu 2008 $3.71 per MMBtu 2009 $4.33 per MMBtu 2010 $4.04 per MM Btu on 11/16/2010 EIA (http://www.eia.gov )

Recent Growth in Natural Gas Production, Lower 48 States, Attributed Largely to Unconventional Gas What caused the recent growth in domestic gas production? Next slide shows the unconventional gas production accounts for the 2007 growth; unconventional gas has been increasing while production from other types of gas has decreased. (EIA, 2010)

Natural Gas Supply by source, 1990-2030 (trillion cubic feet) History Projection Unconventional Alaska Net imports Non-associated offshore Associated-dissolved Non-associated conventional Source: Energy Information Administration, Annual Energy Outlook 2009

Marcellus Shale Resource Marcellus Resource U.S. Resources1 2,080 Tcf U.S. Proved Reserves2 244 Tcf Marcellus Shale Resource3 256 Tcf Annual U.S. Consumption 23 Tcf 1 Potential Gas Committee, June 18, 2009 2 U.S. Energy Information Administration 3 Marcellus Proved Reserves < 1 Tcf

Marcellus Shale Production Forecasts Sources: “An Emerging Giant: Prospects and Economic Impacts of Developing the Marcellus Shale Natural Gas Play.” T. Considine, R. Watson, R. Entler, J. Sparks, The Pennsylvania State University, College of Earth & Mineral Sciences, Department of Energy and Mineral Engineering. July 24, 2009. Integrated Resource Plan for Connecticut. The Brattle Group. January 1, 2010. (Wood Mackenzie)

Marcellus Shale Production Outlook Source: Williams Partners L.P.

Unconventional Resource Production Technology, Economies of Scale, Integration Successful pursuit of shale gas requires: Extensive acreage position leading to large scale developments Strong integration of engineering and geoscience disciplines Innovative engineering and tolerance for experimentation Embedded process for continued integration of technology advances and cost reduction Cheaper to work underground Maximize recovery of a cube of shale Completion technology and horizontal drilling have emerged as the shale gas enablers to extract as much gas as possible from the resource Cost in developed shale gas plays has been driven down through innovative use of technology, economies of scale, and effective integration of all the disciplines

Unconventional Resource Production Technology, Economies of Scale, Integration Successful pursuit of shale gas requires: Extensive acreage position leading to large scale developments Strong integration of engineering and geoscience disciplines Innovative engineering and tolerance for experimentation Embedded process for continued integration of technology advances and cost reduction Cheaper to work underground Maximize recovery of a cube of shale Completion technology and horizontal drilling have emerged as the shale gas enablers to extract as much gas as possible from the resource Cost in developed shale gas plays has been driven down through innovative use of technology, economies of scale, and effective integration of all the disciplines

Gas Shale Characteristics Very High Gamma Ray Activity (Kerogen Content) High Uranium High Resistivity – Low Water Saturation Relatively Low Clay Content Smectite to Illite Transition Low Bulk Density (Kerogen Content) Kerogen - Petrophysical Characteristics Bulk Density 1.0 to 1.2 g/cm3 U 0.18 to 0.24 Neutron Porosity 50 to 65 p.u. Gamma Ray Activity 500 to 4000 API Sonic Slowness 160 µs/ft

Three Approaches Logs to be used Bulk Density g/cm3 Density Porosity Percent or Decimal Neutron Porosity Percent or Decimal Photo-Electric Barns Gamma Ray API Units Clay Typing – Related to Deposition & Diagensis Spectral Gamma Ray Logs Uranium (PPM), Thorium (PPM) and Potassium (Percent) Compositional Lithology Logs Rhomaa-Umaa Computational Analysis (Linear)

Spreadsheets Ubiquitous and Low Cost Provide Some Hands-On Understanding of the Process Allow Easy Export to Higher End Packages Use Basic Logs Clay Typing Estimate Uranium Content from Full Spectrum Gamma-Ray Logs Compositional Lithology Logs Rhomaa-Umaa Computational Analysis (Linear) Organic Content (Next Time) Saturation (Next Time) Heavily Modified Archie

Gamma-Ray Log Analysis 4/21/2017 Gamma-Ray Log Analysis U Th K

Gamma-Ray Spectrum Uranium Thorium

Gamma-Ray Spectrum Schlumberger Log Interpretation Principles 1989, Page 3-7

Geochemists’ concept of typical shale and black shale North American Shale Composite (NASC) Gromet et al. (1984) Th 12.3 ppm, U 2.66 ppm, K 3.2% GR = 121.7 API units Black Shale Composite (BSC) Quinby-Hunt et al. (1989) Th 11.6 ppm, U 15.2 ppm, K 2.99% GR = 215.8 API units API unit multipliers: Th ppm 4 : U ppm 8 : K% 16

Typical Spectral Gamma-Ray Log Presentation Format

Potassium-Thorium Crossplot with Generalized Mineral Fields (after Schlumberger)

Potassium-Thorium Crossplot with Generalized Mineral Fields (after Schlumberger) Increasing Th/K Ratio

Thorium and Uranium Concentration and Redox Potential Adams and Weaver (1958)

Gamma-Ray and Spectral Ratio Logs Permian – Cretaceous Central Kansas

Photo-Electric and Spectral Gamma Ray Schlumberger, Log Interpretation Principles 1989, Page 6-4

Photo-Electric and Spectral Gamma Ray Schlumberger, Log Interpretation Principles 1989, Page 6-4

Idealized Kansas Pennsylvanian Cyclothem

Spectral Gamma-Ray Log Lansing Group, Wabaunsee County, Kansas

Chestnut Drive Section Spectral Gamma Ray Response

Devonian Shale Analysis Harrell Tully Mahantango Marcellus Onondaga

Devonian Shale: Oxidizing and Reducing Conditions Reducing Vs. Oxidizing conditions determined by Th/U Oxidizing

Devonian Shale: Clay Type Clay type can be determined from Th/K Illite-Pink Smectite-Green Illite can increase porosity by 4%

Wells 1 & 3

Wells 1 & 3

Well 2

Project 1 Make sure you open an LAS File with Notepad http://www.geo.wvu.edu/~tcarr/PTTC_11_2010 Make sure you open an LAS File with Notepad Import a LAS File to EXCEL Well 3.LAS Open Spectral Gamma Ray Template Well 1.LAS Marcellus (7375’-7562’) Well 2.LAS Marcellus (7359’-7501’) Create & Examine Plots What is the difference in the two wells

Open with Notepad

Importing a LAS File to EXCEL

Importing a LAS File to EXCEL

Importing a LAS File to EXCEL

Introduction to Porosity Logs Porosity Logs DO NOT Directly Measure Porosity Acoustic (Sonic) Logs Measure Wave Travel Time Density Logs Measure Formation Bulk Density Neutron Logs Measure Formation Hydrogen Content

Neutron Log Applications Porosity Lithology with Density and/or Sonic Gas Indicator Clay Content Correlation Cased Hole

Neutron Tool Background Outgrowth of Work by Italian Physicists (1935) Slowing down and stopping of neutrons by a hydrogen rich material (e.g., water). Radioactive Source of High Energy Neutrons Americium and Beryllium Fairly Shallow Zone of Investigation ~ 6 inches (Flushed Zone (Rxo) in most cases) Neutrons lose energy each time they collide with nuclei as they travel through the formation Greatest loss in energy when neutrons collide with nuclei of a similar mass Hydrogen atoms As the neutrons slow they can be captured and emit a gamma ray. Reduction in Neutron Flux (Increased Gamma Rays) is largely controlled by concentration of hydrogen in the formation. Water (Oil) Filled Porosity in Flushed Zone of Clean Units Clays Lithology Effect Hydrocarbon Gas Effect Depress apparent neutron porosity

The Neutron Porosity Tool

Historical Development of Neutron Logging Common Curve Mnemonics ΦN, PHIN, NPHI Usually Tracks 2 or 3 and dashed line. Units Counts %, Decimal Fraction

Neutron Energy Loses

Density Log Applications Porosity Lithology with PE, Neutron and/or Sonic Gas Indicator Synthetic Seismograms with Sonic Rock Properties with Sonic Poisson’s Ratio, Young’s Modulus Clay Content Borehole Conditions (Size and Rugosity)

Density Tool Background Source of High Energy Gamma Rays Cesium 137 Shallow Zone of Investigation <2 inches Gamma rays interact with the electron clouds of the atoms they encounter, with a reduction in the gamma ray flux, which is measured by both a near and far detector. Higher Energy Range Affected by Compton Scattering Reduction is a function of the electron density of the formation Number of Electrons Matched by the Number of Protons In Most Cases Z/A = 0.5 Z - Atomic Number A – Atomic Mass Two Density Values Bulk Density (RhoB or ρb) – Measured by Logging Tool – Solid + Fluid DEN, ZDEN Matrix Density (ρma) – Density of the Rock that has no Porosity Hydrocarbon Gas Effect Enhances apparent density porosity For any element, the number of electrons is matched by the number of protons, which is the atomic number, Z. The atomic mass is contained in the atomic nucleus, is effectively the sum of the number of protons and neutrons, and is given by the atomic number, A. In general, the number of protons is approximately the same as the number of neutrons, so that the Z/A ratio of many elements is close to 0.5, particularly for elements that are in the lower part of the periodic table. Using this simple atomic theory and a Z/A ratio of 0.5, the actual measurement of electron density can be converted to an apparent density, measured in units of grams per cubic centimeter, which is usually close to the real density of common rock types. The primary use of the density log is as a measure of porosity, using a simple mass balance relationship, and interpolating between the matrix mineral density (2.65 quartz; 2.71 calcite ; 2.87 dolomite) and mud filtrate, since most of the density reading is from the flushed zone. The difference in density between that of any residual oil and the mud filtrate is usually insufficient to affect porosity estimations. However, the markedly low density of any residual gas will strongly influence the density reading to suggest a higher apparent porosity.

The Formation Density Tool

Density Porosity ΦD = (ρma – ρb) / (ρma – ρfluid) DPHI, PHID, DPOR Sandstone 2.644 gm/cm3 Limestone 2.710 gm/cm3 Dolomite 2.877 gm/cm3 Anhydrite 2.960 gm/cm3 Halite 2.040 gm/cm3 Freshwater 1.0 gm/cm3 Saltwater ~1.15 gm/cm3

Question Why does ΦN read much higher Than ΦD in the red boxed area? What are the general lithologies in this well?

Photo Electric Pe Tool Lithology with Density, Neutron and/or Sonic Supplementary Measurement of the Density Tool 1970’s Onward Lower Energy Range Gamma Rays Affected by Photoelectric Effect Logged Value is a function of Z - Atomic Number Pe = (Z/10)3.6 Barns per electron Only mild affect of Pore Volume and Fluid/Gas Content Quartz = 1.81 Barns Dolomite = 3.14 Barns Calcite = 5.08 Barns Pe, PE, PEF For any element, the number of electrons is matched by the number of protons, which is the atomic number, Z. The atomic mass is contained in the atomic nucleus, is effectively the sum of the number of protons and neutrons, and is given by the atomic number, A. In general, the number of protons is approximately the same as the number of neutrons, so that the Z/A ratio of many elements is close to 0.5, particularly for elements that are in the lower part of the periodic table. Using this simple atomic theory and a Z/A ratio of 0.5, the actual measurement of electron density can be converted to an apparent density, measured in units of grams per cubic centimeter, which is usually close to the real density of common rock types. The primary use of the density log is as a measure of porosity, using a simple mass balance relationship, and interpolating between the matrix mineral density (2.65 quartz; 2.71 calcite ; 2.87 dolomite) and mud filtrate, since most of the density reading is from the flushed zone. The difference in density between that of any residual oil and the mud filtrate is usually insufficient to affect porosity estimations. However, the markedly low density of any residual gas will strongly influence the density reading to suggest a higher apparent porosity.

Photoelectric factor log

Compositional Analysis Combing More Than Two Logs

Compositional Analysis Determine Lithology Graphic Plots Computation Identification and Semi-Quantitative Estimates

Porosity Log Combinations Single Porosity Measurement Lithology is Specified for Correct Porosity Choice of Matrix Value Two Porosity Measurements Two Lithologies can be Predicted along with Porosity Three Porosity Measurements Three Lithologies can be Predicted along with Porosity Greater the number of Measurements the Greater the Complexity of the Lithology that can be Estimated

2 Logs 2 Minerals

Dolomitic-Limestone System

Three-Measurement Cross-Plot Three Mineral Matrix Can Be Determined Usually Reduce From 3-D to 2-D Collapse the 3 measurements to two axes with common denominator M-N Plots Axis 1 – Sonic and Density Axis 2 – Neutron and Density Problem of Density and Sonic being Correlated Addition of Pe in Newer Methods

M-N Cross Plot

M – N Crossplot Remove the effect of pore fluid 4/21/2017 M – N Crossplot Remove the effect of pore fluid Usually drilling fluid Combine Sonic and Density Logs (M) M = (∆tfluid – ∆tmatrix) / (ρmatrix – ρfluid) Combine Neutron and Density N = (Φnfluid – Φn matrix) / (ρmatrix – ρfluid)

M-N Cross Plot

RHOmaa – Umaa Crossplot Mineral Identification (MID) Plots Apparent Matrix Density RHOmaa Density and Neutron Apparent Matrix Photoelectric Cross Section Umaa Density, Neutron and Photoelectric Effect

Apparent Matrix Density RHOmaa          

Photoelectric (PE) Factor

Volumetric Photoelectric Absorption U/cm3 The photoelectric absorption index (Pe) is measured in units of barns per electron. In order to linearize its relation with composition, the variable must be converted to a volumetric photoelectric absorption index (U) with units of barns per cc and is approximated by:    

Volumetric Photoelectric Absorption U of the matrix This is the volumetric photoelectric absorption coefficient of the zone (matrix plus fluid). The hypothetical volumetric photoelectric absorption coefficient of the matrix is UMAA.   or approximated by  

Umaa Values (Apparent 𝜙)

RHOmaa Umaa Plot Pyrite

Shale Characterization

Computational Analysis 2 Logs 2 Minerals

Computational Analysis C - matrix of the log responses of the components V - vector of the component proportions L - vector of the log readings To Solve for V need the inverse of the component matrix CV=L V = C-1L

Log response equations: Rewritten as matrices:

We are Saved - Easily computed in Excel The compositional solution vector is then given by pre-multiplying the log response vector by the inverse of the coefficient matrix We are Saved - Easily computed in Excel

Compositional Analysis

Project 2 http://www.geo.wvu.edu/~tcarr/PTTC_11_2010 Use Parameters From Appendix B Open Compositional Template Load in Separate Template Well 1.LAS Marcellus (7375’-7562’) Onondaga (7562.5’ 7578’) Why are data points outside the Rhomaa-Umaa Triangle Load in Separate Template Well 2.LAS Marcellus (7359’-7501’) Onondaga (7501.5’ 7516’) Create Computational Plots What is the difference in the two wells

My Observations http://www.snopes.com/photos/natural/canyonleap.asp

http://www.snopes.com/photos/natural/canyonleap.asp

Email: tim.carr@mail.wvu.edu Earth at Night Credit: C. Mayhew & R. Simmon (NASA/GSFC), NOAA/ NGDC, DMSP Digital Archive Tim Carr Phone: 304.293.9660 Email: tim.carr@mail.wvu.edu