eori moxa arch talk 3

Characterization of the Madison Formation, SW Wyoming: First

steps in creating a model for carbon sequestration

Geoffrey Thyne, Mark Tomasso, David Budd, Sharon Bywater-Reyes and Brian Reyes

Why do we need to put it anywhere?Carbon (Dioxide) Emissions

• Current increase comes from burning Fossil Fuels

• Increase in atmosphere is “linked” to climate changes

• There will be a Tax on carbon emissions

Carbon Capture and SequestrationTechnology Options for Stabilization

The Stabilisation Wedge

Emission trajectory to achieve 500ppm

Emission trajectory BAU

1 GtC Slices of the Stabilisation Wedge

Where will we need to put it?

North American CO2

Sources

Why do we need a model?

• Carbon sequestration models will be required for screening, scoping and permitting of storage facilities.

• The model requires a geologic framework including stratigraphic, structural and petrophysical data.

• The primary source for this data are petroleum wellsin the study area.

• Information comes from public sources (WYOGCC, UGSS, WYGS, scientific literature.

Madison Formation Lithostratigraphy

• Platform carbonate

• Early dolomitization

• six 3rd order sequences

• Paleozoic section contains CO2, CH4, H2S and He

Study Area

• Moxa Arch mile long north-south trending anticline

• Late Cretaceous uplift to create structure

• Extensive trap for hydrocarbons (mostly gas) in Mesozoic and Cenozoic sections

• Paleozoic section contains CO2, CH4, H2S and He

• Zone of interest is Mississippian Madison Formation

Madison Formation Lithostratigraphy

• Literature Review

• Facies controlled petrophysics with dolomite having better reservoir quality

• Most good reservoir quality confined to lower two sequences

Core information

• Detailed descriptions of the four cores.

• Recognize several facies (karst breccia, micrite, wackestone, packstone, grainstone).

• Samples for thin sections, XRD, whole rock chemistry and stable isotopes.

• Correlate with petrophysical data.

Core information

• Primarily dolomitic rock.

• Thin limestone layers near top of sequences.

• Fractures in limestones and dolomite with multiple orientations.

• Karstic only at top of Madison.

Primary petrophysical information

• There are 95 wells that penetrate the Madison Formation in the study area

• Most have geophysical logs• Four wells have core• Ten wells have standard

petrophysical data, 1 foot intervals (porosity, Kx, Ky and Kz, oil and water saturation, density and mineralogy)

• Two wells have both petrophysical data and core

Facies dependent por-perm• Core-based data show:

– Facies control on porosity and permeability

Petrophysical data from core

Well information

• Core-based data show:

– Porosity with normal distribution skewed toward lower end

– Permeability with lognormal distribution

• This is the primary data to populate model grid

Petrophysical data from cores

Well information

• Core-based data show:

– Porosity – log Permeability relationship for samples with higher porosity

– Porosity-permeability relationship is very poor with wide range of permeability for lower porosity samples

• Core descriptions and thin sections show fracture-controlled permeability in lower porosity rocks

Potential bias in the petrophysical data

• Petrophysical data comes form core.

• Represents small portion of total formation.

• Limited sampling of facies.

• Ten wells have petrophysical data, but from only two locations.

• One location is in gas cap, one is not.

Dolomite petrophysics by location

CO2 versus no CO2

Average porosity = 1.5

Average porosity in gas cap = 12.6

Average porosity outside gas cap = 7.33

Calcite petrophysics by location

CO2 versus no CO2

Average porosity in gas cap = 2.3

Average porosity outside gas cap = 7.2

Potential bias in the petrophysical data

• Petrophysical data comes form core.

• Represents small portion of total formation.

• Limited sampling of facies.

• Ten wells have petrophysical data, but from only two locations.

• One location is in gas cap, one is not.

Well with core-based petrophysical data have limited sampling of facies

Note positionof cores with petrophysical data

Extending the data

• Extend petrophysical database (n = 2,671) by calculating log-based porosity for nine wells (n = 11,959).

• Use sonic logs to reconstruct bulk density where necessary (Gardner et al. 1974), and bulk density logs to calculate porosity (Schlumberger 1972).

• Calibrate relationship by comparing log-based and measured posoity

• Used calculated values every 0.5 feet to generate porosity for wells with logs.

Core based porosity

Petrophysical data core versus log

Petrophysical data log-based shows no location bias

Modified log-based petrophysical basis

• Good reservoir intervals are in third and fourth sequences.

• Still facies controlled.

• High porosity zones extend over 10’s miles.

• Trimodal porosity distribution?

• Three classes in formation.

• <4%, Mean = 2%, 45% of all porosity values.

• 4 – 12%, Mean = 7.5%, 16.5% of all porosity values.

• >12%, Mean = 14%, 37.7% of all porosity values.

Permeability Isotrophy

Use core based

samples with Kh

and Kv

N = 985

<4%, n = 470,

mean log Kh/Kv =

0.88

4 – 12%, n = 362

mean log Kh/Kv =

0.77

>12%, n = 153,

Mean log Kh/Kv =

0.52

Conclusions

• Core based petrophysical data is available to characterize the Madison in the study area.

• Core-based petrophysical data is somewhat biased by location of cores.

• Not representative of target entire formation.

• Can extend the data using geophysical logs.

• Extended data shows different stratigraphic distribution of reservoir quality than previous work in eastern Wyoming.

• Petrophysical properties in model will have to be modified.

Questions