Unconventional Petrophysical Unconventional Petrophysical Analysis in Unconventional Analysis in Unconventional

ReservoirsReservoirsPutting the Puzzle Together in Gas Shales

Lee Utley

“Intuitively, it is my belief that this magnitude of money could be better spent on other projects.”

Executive with Mitchell Energy in his recommendation for attempting the

first completion in the Barnett Shale ‘discovery’ well (Slay #1) - 1982

“Why are we spending all this money to find out how much gas is in the Barnett? If we really want to know what will happen in Johnson County, we just need to drill some damn wells!

Engineering executive with Mitchell Energy upon finding out the

magnitude of our planned spending on coring and analysis to reevaluate the gas content of the Barnett - 1999

IntroductionIntroduction

Has this happened to you?Has this happened to you?

Somebody just dumped some stuff in your officeLarge stack of logs

Several CDs/DVDs of digital dataCore reports

Several maps and cross-sections

You are told that your company wants to get into this Barnett Shale play everyone is talking about so

you need to figure this out.

ProblemsProblems

General GoalsGeneral Goals

• Areal extent• Thickness• Type of hydrocarbon• Possible production mechanisms• Barriers to economic production

Evaluate the resource

Specific Goals to Achieve Using Specific Goals to Achieve Using Log AnalysisLog Analysis

• Gas Content • Analysis of ‘conventional’ formations• Maturity• Total Organic Content• Porosity• Water saturation• Lithology• Rock Properties• Fracture types

Why is this so hard to do?Why is this so hard to do?

• Old logs with limited information• Little or no core data• Complex lithologies cause problems with

typical methods• TOC calculation is difficult at best• Porosity determination is complicated by

presence of TOC

Useful Core DataUseful Core Data

• Geochemical analysis (Ro, TOC, etc…)• Porosity• Water saturation• Gas content (including adsorption isotherm

information)• Mechanical properties

Gas ContentGas Content

Gas Storage SitesGas Storage Sites

• Sorption – TOC• Pore space• Open natural fractures

Most gas is stored in the pore space and the TOC. Fracture storage is usually minimal and probably can’t be quantified.

Calculation of Gas ContentCalculation of Gas Content

• For sorption, relate TOC to gas content – usually through Langmuir parameters.• Don’t forget about non-methane adsorption

• For pore space, use conventional gas-in-place equations.

TOC and porosity are two of the biggest keys in looking at gas shales.

‘‘Conventional’ AnalysisConventional’ Analysis

Why look at ‘conventional’ areasWhy look at ‘conventional’ areas

• Production pathways• ‘Unfavorable’ porosity• Stimulation barriers• Uphole ‘bail-out’ zones

MaturityMaturity

Log Indicators of MaturityLog Indicators of Maturity

• Resistivity• Density – Neutron Separation

Use averages of these values in very well defined geologically correlative areas to compare to core

vitrinite reflectance data.

Use resistivity as a predictorUse resistivity as a predictor

(OGJ – Morel – 1999)

Use Old Resistivity Logs TooUse Old Resistivity Logs Too

• Use resistivity inversion modeling to get old ES logs and induction logs up to modern standards – compare apples to apples

1940’s 1980’s Modern

Density – Neutron SeparationDensity – Neutron Separation

Gas Shale Well One Gas Shale Well TwoLower Vitrinite Reflectance Higher Vitrinite Reflectance

TOCTOC

Four main methodsFour main methods

• Use average TOC from published accounts and apply it to every well

• Density log regression• Delta log R

• Passey, et al – AAPG 1990

• Neural Networks

PorosityPorosity

Standard Porosity TransformStandard Porosity Transform

• Core matrix numbers exclude organic material.• Normal log presentations show very high apparent

porosities. These porosities are closer to the volume of pore space and organic material combined.

fluidmatrix

matrix

log

Basic Porosity EquationBasic Porosity Equation

fluidmatrix 1log

Rock co

ntributio

n

Fluid contri

bution

Porosity Equation with TOCPorosity Equation with TOC

TOCTOCfluidTOCmatrix VV 1log

Rock co

ntributio

n

Fluid contri

bution

TOC contri

bution

Solved for PorositySolved for Porosity

fluidmatrix

TOC

matrixmatrix TOC

TOC

1log

ccgmto

ccgmto

ccgmto

fluid

TOC

matrix

/0.14.0

/4.13.1

/68.260.2

Water SaturationWater Saturation

What are the correct parameters?What are the correct parameters?

n

tmw

wR

aS

R?

Pickett PlotPickett Plot

Calculate Water SaturationCalculate Water Saturation

LithologyLithology

Two most common methodsTwo most common methods

• Probabilistic methodology• Integrated neural network solution

Neural Network SolutionNeural Network Solution

Rock PropertiesRock Properties

Standard Rock Mechanic EquationsStandard Rock Mechanic Equations

Use Lithology to Correlate with Use Lithology to Correlate with Rock PropertiesRock Properties

Neural Network of Young’s Modulus in Two Permian Basin wells using a Fort Worth Basin Model

Neural Network Computed Young’s Modulus

Roc

k Pr

oper

ties

Com

pute

d Y

oung

’s M

odul

us

FracturesFractures

Imaging LogsImaging Logs

• Fracture Size• Direction(s)• Complexity• Open/Closed• Induced fracture direction (stress field)

Barnett Shale Case StudyBarnett Shale Case Study

Core Data AcquiredCore Data Acquired

Conventional and pressure cores – Extensive data suite• Porosity• Water Saturation• TOC• XRD• Canister desorption• Adsorption isotherms• Capillary pressures• CEC

Integrate Core DataIntegrate Core Data

Quartz Plagioclase Calcite Dolomite Apatite Pyrite Total Total Organic Porosity Water Bulk Volume Bulk Volume

Clays Carbon Saturation Water Hydrocarbon

34 1 10 6 0 0 35 6 8 54 4 4

37 4 13 3 1 0 29 5 7 46 3 4

32 2 20 3 1 0 29 5 8 50 4 4

23 2 46 4 1 1 22 1 0 76 0 0

13 1 41 20 0 0 18 3 4 37 1 2

12 2 61 17 1 1 4 1 2 32 1 1

23 2 33 4 0 0 30 3 4 56 2 2

10 1 74 10 0 0 3 0 1 67 1 0

30 0 26 8 0 0 33 1 3 89 2 0

16 1 23 13 0 0 40 2 5 75 3 1

31 3 11 3 1 0 42 4 5 70 3 1

34 2 17 12 1 1 23 4 5 74 4 1

15 1 15 41 1 1 24 1 1 79 1 0

35 5 8 4 1 1 39 5 2 100 2 0

Train a Volumetric Neural NetworkTrain a Volumetric Neural Network

Apply integrated solution to all wellsApply integrated solution to all wells

Fort Worth Model Applied to Fort Worth Model Applied to Permian Basin WellPermian Basin Well

ComparisonComparison

ConclusionsConclusions

Gas shales can be effectively Gas shales can be effectively analyzedanalyzed

• Maturity, TOC, and porosity are some of the keys to gas shale analysis and can be determined from logs.

• Even without extensive core data, gas shales can still be analyzed, at least in a relative sense.

• Other gas shales can be evaluated from log data and core data using these techniques. An integrated study is required for full evaluation.

Unconventional Petrophysical Unconventional Petrophysical Analysis in Unconventional Analysis in Unconventional

ReservoirsReservoirsPutting the Puzzle Together in Gas Shales

Lee Utley