Table of Contents

PETROPHYSICAL MODELING......................................................................................................... 2
Important icons used in the process steps:................................................................................2
Exercise Overview:.................................................................................................................... 2
A) DETERMINISTIC MODELING........................................................................................................ 3
Exercise Steps........................................................................................................................... 3
Comments................................................................................................................................. 4
B) DETERMINISTIC MODELING USING TREND MAPS......................................................................4
Exercise Steps........................................................................................................................... 4
Comments................................................................................................................................. 5
C) STOCHASTIC MODELING............................................................................................................ 5
Exercise steps........................................................................................................................... 5
D) STOCHASTIC MODELING - CONDITIONING TO FACIES..................................................................6
Exercise steps........................................................................................................................... 6
Comments................................................................................................................................. 7
E) PROPERTY CALCULATOR.......................................................................................................... 7
Exercise Steps........................................................................................................................... 7
F) HISTOGRAM AND FILTER FUNCTIONALITY....................................................................................9
Exercise Steps........................................................................................................................... 9
Comments................................................................................................................................. 9
G) OPTIONAL EXERCISE................................................................................................................ 9
Calculation of Sw....................................................................................................................... 9

Petrel Introduction Course

P2002 (v.7/02)

Page 1

Petrophysical modeling
Both stochastic and deterministic methods are available for modeling the distribution of
petrophysical properties in a reservoir model. Scaled up well logs and/or trend data is used as input
and various user settings are available. When performing petrophysical modeling all cells for that
property will be given a value. The well and/or trend values are distributed in the volume defined by
the 3D grid.
Before modeling, the user should normally have gone through a detailed data analysis, identifying
trends, correlation between other properties, defining the variogram and viewing the histogram. The
data analysis will not however, be covered in the introduction course.

Important icons used in the process steps:

Show property filter
Toggle Simbox View
Reset settings for all zones to default
Leave Zone unchanged

Exercise Overview:
A)

Deterministic modeling

B)

Deterministic modeling Using trend maps

C)

Stochastic modeling

D)

Stochastic modeling - Conditioning to Facies

E)

Property Calculator

F)

Histogram and filter functionality

G)

Optional Exercise:

Sw Calculation using the Calculator

Petrel Introduction Course

P2002 (v.7/02)

Page 2

A) Deterministic modeling
When the well logs have been scaled up to the resolution of the cells in the 3D grid, the values for
each cell along the well trajectory can be interpolated between the wells in the 3D grid. The result is
a grid with property values for each cell.
There are several deterministic methods available in Petrel, examples are kriging and moving
average. The deterministic methods will produce smooth results. The kriging method can include
information about the variogram, hence producing an anisotrop model that has captured the
geostatistical dependencies between points in the 3D model. The deterministic approaches will not
however, produce local variation, and if you run 100 realizations the outputs will be identical for
each run.

Exercise Steps
1.

Activate the 3DGrid (DC) under the GeoModel project.

2.

Open the Petrophysical Modeling process.

3.

Select Use Existing Property and select the Porosity property as the property to be modeled
from the drop down menu.

4.

Select Tarbert 1 from the pull down menu and de-select the Leave Zone Unchanged
for that zone to create a realization.

5.

Select the Moving average as the Method, leave all other settings as default.

6.

Click OK to create the property model and display the model in the 3D Window. Use the Zone
Filter to display the appropriate zone.

7.

Bring up a Histogram window from the Window menu. Display the Porosity property and use
the Zone Filter to look at the modeled zone. Compare the histogram distribution for the well
logs, the up-scaled cells and the whole property

8.

Double-click on the Porosity property in the Properties folder and check the statistics in the
Statistics tab.

icon

Comments

After modeling it is important to compare the output (3D property) with

the input (up-scaled logs). The histogram is the most obvious. But it is also important to use
geological knowledge and see if the model represent the expected result.

Petrel Introduction Course

P2002 (v.7/02)

Page 3

B) Deterministic modeling Using trend

maps
Trends can be used to constrain the modeling. The trends can represent vertical trends, horizontal
trends or 3D trends represented by the property. In general, the trends should be smooth and the
should not be in conflict with the up-scaled well logs to be able to capture them in the model.
Trends are defined with values between 0 and 1, and if outside this range then they will be scaled.
Horizontal trends, defined by a 2D Grid, will be used in this exercise.

Exercise Steps
1.

Activate a 3D window and click on the Open Dialog for Active Process
Function bar.

icon on the

2.

Select Use Existing Property, select Porosity to be used from the pull down menu and select
Tarbert 1 as the active zone.

3.

Go to the Trends tab. Toggle on the Horizontal trend surface option and insert the Tarbert-1
porosity map located under the Properties/Porosity/ folder in the Petrel Explorer Input tab.

4.

Click OK to create the property model and observe the changes.

5.

Bring up a 2D Window from the Window menu and display the generated property model.
Use the Zone Filter to display only the appropriate zone.

6.

Go to the Petrel Explorer Input tab and double click on the porosity trend map used in step 2.
Go to the Style tab and at the bottom set the Transparency to 70%. Also , change the contour
increment to 0.005.

7.

Display the trend map and click the Map View Position
icon on top. Compare the
contours with the coloring of the property model underneath to verify the trends.

Comments
A trend map is a 2D grid. Such a trend map can be generated by the following procedure:

Petrel Introduction Course

P2002 (v.7/02)

Page 4

Digitize contours that represent the trend map in the Make/Edit

Polygons process. This can be done in a 2D window on top of a horizon.

Give each contour line a value by using the Z-value selector (available
from the Function bar when the Make/Edit Polygons process is active)

Use the Make Surface process to grid these contour lines into a 2D

grid

C) Stochastic modeling
Petrel can generate stochastic petrophysical property models based on the Sequential Gaussian
Simulation method. This is the most straightforward algorithm for generating realizations of a
multivariable Gaussian field. The method can produce local variation and reproduce input
histograms. This means that if you run 100 different realizations (using a different seed 1) based on
the same inputs, then you will get 100 different outputs. All of them will match the input, but since
the input is given by a distribution, then the value given to each cell can vary based on the range
of this distribution. If a large number of realizations is run (~50-100) then the variation between
them reflect the uncertainty in the model. Note however, that this uncertainty is only the true
uncertainty if the model itself is correct! If you have used inputs that are wrong, then all output will
of course be wrong as well.

Exercise steps
1.

2.

Creating first model:

a.
b.

Activate the 3D Grid (DC) under the GeoModel project.

c.
d.
e.

Activate the lock icon for the Tarbert 1 zone and de-select the lock for Ness 2.

f.

Click OK to create the property model.

Select Sequential Gaussian Simulation as the method to use.

In the Variogram tab, select Exponential Variogram type, 3500 as Major Range, 1500
as Minor Range, 10 as Vertical Range and 25 degrees as Azimuth.

Viewing the result:

a.
b.
3.

Open the Petrophysical Modeling process, select Use Existing Property and select
Porosity from the drop-down list.

Bring up a Histogram window from the Windows menu and select Tile Vertical.
Use the Zone Filter to display the 3D property model and the Histogram distribution for
the well logs, the up-scaled cells and the whole property for zone Ness 2.

Changing the model:

a.

Click on the 3D window to make it active and click on the Open Process Dialog
icon on the Function bar.

b.

Go to the Settings tab in the Petrophysical Modeling process window. Under Output
data range click on Estimate (this will estimate the porosity range from the up-scaled

Seed: used in stochastic modeling to create different realizations. If the seed value is the same for
each run, then the output will be identical as well (given that the input is the same). When using
different seeds, then each output realization will be different, but still based on the same input. The
seed value just tells the algorithm which value should be the first values to be drawn from the
distribution. It is a random number unless specified.
1

Petrel Introduction Course

P2002 (v.7/02)

Page 5

cells within the zone). Specify an output data range (e.g. 0.1-0.25).

c.

Click OK and observe the changes in the model.

D) Stochastic modeling - Conditioning to

Facies
Normally, a petrophysical model will be created by using a facies model as input as well. This will
constrain the petrophysical values representing one facies to that facies only. It is obvious that in a
geological environment such as a fluvial system in a shale background, it is critical to distribute the
high-porosity values representing channel in the channel facies only and not average all the
porosity values in the entire zone. That would produce a porosity model that did not represent any
of the input facies.

Exercise steps
1.

Open the Petrophysical Modeling process.

2.

Select the Porosity property as the property to model.

3.

Select Ness 2 as the Zone to model and click the Reset settings of current zone to default
icon to reset the setting for this zone.

4.

Select Sequential Gaussian Simulation as method.

5.

Click on the Facies button and select the Facies Object model from the drop-down list. Select
Sand facies from the facies drop-down list and specify the variogram settings for this facies
according to the table below

6.

Specify the variogram settings for the other facies as well according to the table below.

7.

Go to the Settings tab in the Petrophysical Modeling process window. Under Output data
range click on Estimate to define the minimum and maximum porosity values of the resulting
model. Do this for all the four facies.

8.

Click OK to create the model.

9.

Display the generated Porosity model and compare with the Facies Object model used as
input. Use the Zone filter to view only Ness 2.

Orientation

Major Range

Minor Range

Vertical Range

Clay

500

500

10

Sand

25

3500

1500

10

Silt

25

1000

500

10

Fine Silt

25

850

500

10

Comments

Conditioning to a facies property is time consuming since you must

define the settings for each facies. Since the facies have been defined based on their
difference in petrophysical properties, it is important to preserve this distinction, hence the

Petrel Introduction Course

P2002 (v.7/02)

Page 6

possibility of defining the settings for each facies is very important. It is however possible to
copy settings between facies and between zones to make this easier.

The settings such as variogram ranges, are often difficult to define.

Very often geology (geological knowledge, analogue field, outcrop studies) is used to come
up with some numbers that are reasonable, especially for the horizontal ranges. The vertical
ranges are often possible to calculate based on a variogram analysis.

E) Property Calculator
As for the well logs there is a similar calculator associated with the properties. This calculator can
be used to create new 3D properties and to do operations between properties. There is a whole
range of predefined functions associated with the calculator, such as logical and mathematical
functions. It is possible to write nested statements and there is no limit for the length and the
complexity of such statements. If a calculation is getting complex then it is possible (and useful) to
create macros containing the calculation. The macro can be read from file to do several operations
in one go.

Exercise Steps
1.

2.

Creating a new property model:

a.

Click right MB on the Properties folder in the 3D Grid (DC) and select Calculator from
the pull down menu.

b.

Change the Properties Type to Porosity and type in the white formula field:

test=Porosity [Press Enter]

test=test*0.8 [Press Enter]

c.

See the statistical result by selecting Result in the upper right-hand corner.

d.

Display the new test property model in a 3D Window.

Calculating values:

a.

The Calculator may be used as a normal calculator or for returns of single values using
properties and/or logs.

b.

Type in Sum(Volume()*Porosity) and press enter on the keyboard. The Calculator will
return the pore volume of the entire field.

Petrel Introduction Course

P2002 (v.7/02)

Page 7

F) Histogram and filter functionality

Histograms shows the distribution of values for the selected well logs or for a selected individual
property. The histograms are used for checking and comparing input- and model-data. The
histogram panels are found under Settings for individual properties as well as for well logs, for
individual wells or for all wells. The histogram displays original log values, blocked well values and
property values. The histograms are filter sensitive.

Exercise Steps
1.

Under the Models tab open the Settings window for an individual property (e.g. Porosity) by
double-clicking on it. Within the Settings window, open the Histogram tab.

2.

Select the zone you want to study.

3.

Toggle on/off the boxes for displaying property and well logs (original and up-scaled).

4.

Play around with the option to change the number of columns and check the interval values
in the Legend.

5.

Toggle on the Use filter option and open the Settings window for the Properties folder and
choose the Filter tab. The Filter gives the option to show only parts of a property model.

6.

Choose a filter option, e.g. the Value filter. Observe the changes.

Comments
Remember that the X-axis of the histogram is NOT the values of the property but the number of
columns
Note that for Property histogram, the original well logs are only showing original log values from
cells defined by the up-scaled wells. To view the entire well log, use Well histogram (found under
each well log under the Global well logs folder or under each individual well).

G) Optional Exercise
Calculation of Sw
The water saturation in the hydrocarbon zone is a function of height above the fluid contact, the
permeability and the porosity. The calculator can be used to make a water saturation property that
takes these parameters into account.
In this case we have water saturation in the oil zone that can be
expressed by the formulae to the right:
Where:
h = height above the OWC
K = permeability
=
porosity
a = 0.5, b = 0.2
The contact is given by: OWC = -2010 meters for all zones and all segments.

Petrel Introduction Course

P2002 (v.7/02)

Page 8

A permeability property is needed to generate the Sw property. A linear expression has been found
between porosity and log permeability which can be expressed by: LogPerm=10.9*Porosity-0.48.
A macro has been written to simplify this calculation and for easy reproduction in case of an
update. The macro has been written in a text editor program and saved as a text file with the
ending .mac. All the functions in the macro are listed below each other, just as they would be
written as separate functions using the Property Calculator.

Exercise Steps
1.

Activate the 3DGrid (DC) under the GeoModel project.

2.

Toggle on From file and select the macro file to be used. The macro is called 'Sw_formula'
and it is stored under the directory called Other Data, which is found together with the other
course data.

3.

Press the Run button. Notice that all the formulas will appear in the history window at the top.

4.

Four new properties will be created, called LogPerm, Perm, Temp and Sw. They will all be
attached to the template that was selected next to Type in the property calculator (Seismic
(default) in the picture above).

5.

Change the templates for the Sw and Perm properties by opening the settings window for
each of them, go to the Info tab, and change to another template from the pull-down list.

Petrel Introduction Course

P2002 (v.7/02)

Page 9