Education

- MSc. Petroleum Geosciences Engr.(2012) Stavanger, Norway

- BSc. Geological Sciences (2006) UNIZIK, Nigeria

Work

- Now: Geoscientist (Consultant) working at Statoil, Norway

- Previously:
Geomodeler (Intern) with Total E&P, Norway
Graduate Assistant University of Stavanger, Norway
Reservoir Geoscientist / Geomodeler with RDSL, Nigeria
Geologist with Geo-Fluids PLC, Nigeria
 Terminology
Introduction to 3D geo-modelling
What is a 3D geo-model
Dataset
Modelling Tools
Model Purposes
Modelling Steps (Workflow)
Uncertainty
Conclusion
 Geological Modeling This is a sub discipline of geology that
integrates structural geology, sedimentology,
stratigraphy, paleoclimatology, and
Static Model diagenesis

Reservoir Model The same

meaning
Geo-Model

3D Geo-Model

Geo-cellular Model
A muti-disciplinary approach

Geophysicist
Geologist
Petrophysicist
Geomodeler
Reservoir Engineer
 Problems? Definition

Reservoir Model
A set of parameters that describe a
representation of the sub-surface,
especially oil and gas fields

A 3D grid or group of grids based on

the same fault structure and boundaries
suitable for geological modeling and flow
simulation.

Deterministic & Stochastic Models

Deterministic model produces only
single outcome
Stochastic model produces multiple
equi-probable realisations
 Why is Reservoir Modeling Important?

Itallows us to understand, assess, and predict a reservoir by

incorporating all available data/information into a quantitative
digital representation
 What should a 3D Model Represent?

A realistic description of the sub-surface

Realistic geological concepts

Input data to its own level of accuracy

Uncertainty
 Described using grid-cells
Represent the structure
(shape) of a reservoir
Faults modelled as
discrete structures
Zones identify geological
units
Layers used to represent
fine scale detail
Structural Model:
describes the shape of a reservoir
 Components
surface structure fault offset modelled
faults modelled
accurately/realistically
as discrete
surfaces

stratigraphic cells defined

zones and sub- by 8 corners
zones

includes
reservoir
cells assigned uncertainty
multiple properties

includes
cell-faces assigned reservoir
Facies Dependent Porosity multiple properties heterogeneity
 Properties

facies property
porosity property

Property Model: assign number to

cells to represent geological and
reservoir characteristics

saturation property
 Dataset
Well data
Seismic Core

Analogues/Concept
2main widely used industry standard commercial
modelling tools today:
Petrel, owned by Schlumberger (www.slb.com)
RMS (Irap RMS), owned by Roxar (www.roxar.com)
Both tools originate in Norway, and much of the
development still takes place in Norway.
Othertools include Gocad/SKUA(Paradigm), Earth
vision (Dynamic Graphics) Jewel Suites (Baker
Hughes)
 Reservoir Simulation
good geological models form a strong basis for reservoir simulation input helps
to get it right first time
Reservoir Volumes
3D models are good for calculating volumes, in general, form a better basis than
map based methods
Well Planning & Operations
help engineers of all disciplines understand the geology to be encountered during
drilling
extremely useful for monitoring the progress of a well during drilling, including
updating whilst drilling
Visualisation
helps multi-disciplinary teams work together to understand many different kinds of
geological and reservoir problems
Decommissioning
demonstrate to authorities that there are no economics volumes left to extract

Maximize/Optimize Production and Assess Uncertainty

this will help in field development planning
 Project Scope & Design
Input Data (Seismic Interp.)
Well Correlation Stratigraphic
Facies Interpretation Modeling
Fault Modelling
Grid Building
Structural
Horizon Modelling
Modeling
Zone Modelling
Layering
Contact Modelling
Upscale Well Logs
Property
Facies Modelling Modeling
Petrophysical Modelling
Volumetrics
Upscaling & Simulation Reservoir Engineering
Documentation & Housekeeping
Petrel Workflow Tools
 Facies Interpretation

Gas Zone Pay Zone

Water Zone

Gas Zone

Pay Zone

Gamma Ray (GR) Log:depicts reservoir/non reservoir unit (rock type)

Resistivity Log (RDEP):determines hydrocarbon bearing zone (oil/gas)
Density+ Neutron Log: determines the presence of oil/gas/water
Water saturation (Sw): shows amount of hydrocarbon saturation
Porosity (t): indicates the available pore spaces
Top & Base
Reservoir
 seismic sub-division seismic plus geological sub-division

Top

Base
Zones

Zones divided by layers

Zones: Reservoir sub-division of well based
correlatable units with similar petrophysical
properties

Layers: Sub-division of reservoir zones which is

appropriate for the required heterogeneity
modelling
 Stochastic
Deterministic

- Kriging - Simulation
- Interpolation/estimation technique - Random technique
- Single deterministic result - Multiple equiprobable result
- Used when many data are available -Used when limited data are
(many wells + seismic) available (few wells)
- Produces smooth image (smooth - Produces fuzzy image (same
away from wells) variability every where)
 STOIIP = GRV*NTG*Porosity* So
Bo
GIIP = GRV*NTG*Porosity* Sg
Bg
Recoverable oil = STOIIP * Rfo
Recoverable gas = GIIP * Rfg

STOIIP: Stock tank Oil Initial In Place

GIIP: Gas Initial In place
GRV: Gross Rock Volume
NTG: Net to Gross Ratio
Sw: water Saturation
So: Oil Saturation
Sg: Gas saturation
Bo: Oil Formation Volume factor
Bg: Gas Formation Volume Factor
Rfo: Recoverable Factor for Oil
Rfg: Recoverable Factor for Gas

Note: So/Sg = (1- Sw)

Constant: Bo;Bg;Rfo;Rfg (varies from field to field)

Geo model are often created at high levels
of resolution in order to capture reservoir
heterogeneities; however, this high level of
details incorporated into reservoir model
tend to exceeds the capabilities of standard
reservoir simulators

Therefore, a procedure commonly referred

to as Upscaling or Scale up technique is
often used to coarsen (up scale) fine scale
Geocellular model in a Simulation model
 Uncertainty is everywhere
Uncertainty includes Knowns and Unknowns

We know, there are known knowns; there are things we know we

know. We also know there are known unknowns; that is to say we
know there are some things we do not know. But there are also
unknown unknowns - the ones we don't know we don't know."
 An uncertainty analysis will generally involve the study of
different scenarios and multiple realisations of those
scenarios.
Create high & low case
Scenario - represents an hypothesis, an alternative scenario
may usually involve the creation of a new deterministic object
e.g. different geological concepts or development strategies
Realisation involves variation of parameters within a
hypothesis, e.g. stochastic variation of facies / petrophysical
model input
 Focus on key uncertain parameters like:
Isochore thickness
Contact
Velocity model
Facies model (channel direction, sand/ shale
ratio)
Petrophysical model ( variogram azimuth,
according to channel direction)
 In the sub-surface there are basically 3 types of
uncertainty

Measurement errors
Processing / Interpretation uncertainty
Predicting what cannot be directly observed
E.g. estimate Porosity from Density or Neutron Log
 Layer thickness for heterogeneity

0.1 0.5 1 2

28
 3D view
Plan view
(a)
(c)
(b)

Understand data distribution

Half distance between all samples = 10 km (a)= Range

Minimum distance between two samples = 600m (b)= Lag Distance

Average cell thickness = 1m (c)= Vertical Resolution

29
Spherical

Exponential

Petrel
Gaussian

30
 x +
Error Map Random Variable Map Base Case Structure Map

Example showing how

structural maps are
manipulated, other data are
handled in a similar manner.
=
Then the 3D grid is re-built
for each realisation.
Realisation Structure Map

Realisation = Base Case + (Error x Random Component)

 traditional method for
accounting for uncertainty
in 3D models

allows for uncertainty at a

given location based on
same statistics, different seed fixed input data
 modelling with uncertainty

Replace Model Input

Manipulate for
Uncertainty
Input Data

Objects

GOC = 2235m GOC= 2235m

OWC = 2356m OWC= 2356m

facies, facies,
NTG, f NTG, f
P10=94
P50=96
P90=98 Realizations:200 Base case=97

1 20

200 realisations
GIIP
0.9 18

0.8 16
Cumulative Frequency

0.7 14

0.6 12

Frequency
0.5 10

0.4 8

0.3 6

0.2 4

0.1 2

0 0
 A reservoir model should be a realistic description of the sub-surface (model
should look like concept with uncertainty quantified)

Before you start understand the geology (make a sketch, design a workflow)

There are known's and a lot of unknowns about our reservoirs (its very easy to
make a simple model of our reservoir with a small amount of information & feel
confident about it)

Always cross check modelling results with conceptual model (importance of

qualitative & quantitative geological modelling)

Agree with reservoir engineer beforehand the required level of complexity in

the model

Multi-disciplinary teams work together to understand many different kinds of

geological and reservoir problems

Note: desire in the oil industry is to build fit-for-purpose models

MANY THANKS...NEXT.....

QUESTION???...

THERE IS NO ONE POSSIBLE MODEL ANYWHERE;

ALL MODELS ARE GOOD BUT NOT ALL MODELS ARE
USEFUL
 Practical Example: Modelling
Facies Modelling Petrophysical Modelling
Practical Example: Flow Simulation
Static Model + Dynamic Model