An introduction to artificial intelligence applications in petroleum exploration and production

vinegarclothAI and Robotics

Jul 17, 2012 (4 years and 11 months ago)


An introduction to artificial intelligence applications in petroleum
exploration and production
1.Artificial intelligence
During the last two decades,the petroleum industry
all over the world has experienced a rapid increase in
the number of artificial intelligence (AI) applications.
This upsurge in the number of applications of AI is due
to the greater availability of human experts and
publication of a larger number of case studies.
Artificial Intelligence (AI) is the science and
engineering of making intelligent machines.AI is
devoted to designing ways to make computers perform
tasks that were previously thought to require human
intelligence.AI studies are divided into two main
categories;studies that try to mimic the operations of
human brains and studies that understand and apply
thinking methodologies.The first is the Artificial
Neural Networks (ANNs) and the second is the
classical Artificial Intelligence.Since AI techniques
became aligned with conventional computer hardware
architecture in the middle 1980s,their applications to
petroleum exploration and production have become
available.Artificial neural networks,fuzzy logic
systems,and expert systems are three AI technologies
having a major impact in the petroleum industry.
Artificial neural networks (Fig.1),a biologically
inspired computing methodology,have the ability to
learn by imitating the learning method used in human
brain.It is an interconnected assembly of simple
processing elements,units,or neurons,whose func-
tionality is loosely based on the brain neuron.The
processing ability of the network is stored in the inter-
unit connection strengths,or weights,obtained by a
process of adaptation to,or learning from,a set of
training patterns.Neural networks are well suited to
complex problems.They generally have large degrees
of freedom,thus they can capture the non-linearity of
the process being studied better than conventional
regression methods.ANNs are relatively insensitive to
data noise,as they have the ability to determine the
underlying relationship between model inputs and
outputs,resulting in good generalization ability.A
neural network model can be subjected to additional
training in order to adapt itself to new situations at
which its input–output performance is inadequate.
Fuzzy logic (Fig.2),invented in 1964,is an
approach to reasoning where the rules of inference are
approximate rather than exact.It is useful for manip-
ulating information that is incomplete,imprecise,or
unreliable.Traditional set theory defines set member-
ship as a Boolean predicate (e.g.btallQ means being
greater than some specific height,and either you are tall
or you are not).bFuzzyQ set theory represents set
membership as a possibility distribution (the greater the
numeric value assigned to your height,the more likely
you are to be tall).Once set membership has been
redefined in this way,you can define a reasoning
system based on techniques for combining distribu-
tions.Fuzzy logic has applications in control theory.
When you are programming things to function in a
0920-4105/$ - see front matter D 2005 Published by Elsevier B.V.
Fig.1.A typical neural network.
Journal of Petroleum Science and Engineering 49 (2005) 93–96
complex environment,fuzzy rules may be easier to
derive and faster to use than explicit formulae.Since
fuzzy logic is used mainly for efficiency,some people
think that it is doomed by the emergence of massively
parallel computing.
Expert systems (Fig.3),also known as Knowledge-
Based Systems (KBS),are programs that contains a
knowledge base and a set of algorithms or rules that
infer new facts from knowledge and from incoming
data.An expert system uses the knowledge base of
human expertise to provide expert advice and aid in
solving problems.The degree of problem solving is
based on the quality of the data and rules obtained from
the human expert.Expert systems allow the human
expertise to be accumulated and stored in a computer.
Once stored,the expertise can be retrieved at any time
and used to solve problems in a specialized area.Expert
systems,once successfully developed,provide a per-
manent knowledge base.This base can be easily
transferred in a concise and economical way.Expert
systems can also be used to ameliorate the performance
of individuals because of incorporation of knowledge
from other experts in the field.
An expert system program contains three parts:a
knowledge-base;an inference engine and user interface.
The knowledge base is the first part of an expert system
which contains all the knowledge or expertise in the
form of rules and facts.This step requires the
acquisition,or gathering of available knowledge,and
then storage of this knowledge using a knowledge
representation language in a form recognizable by a
computer.The expert system develops answers by
running the knowledge base through an inference
engine (a software program that interacts with the user
and processes the results from the rules and data in the
knowledge base).Thus,the inference engine provides
the path that directs one toward the solution.The user
interface is the part that establishes the communication
between the user and the expert program.It requests
input from the user and presents the results obtained
fromthe expert systemto the user.Some expert systems
require running large external programs,which is the
case in this study.If this is the case,an expert system
can be directly linked to such programs so that their
results can be used in the inference process.
Expert systems can be developed using either an
artificial intelligence language,or an expert systemshell.
Artificial intelligence languages are more flexible
compared to expert systemshells,but they require much
more programming.These languages are different from
the conventional programming languages.Each of the
AI languages works on a different paradigm and thus
offers different features.The expert systemshells contain
a built-in inference engine which has a knowledge
representation language and pre-specified control stra-
tegies.Several commercial expert system shells,each
with different features,are available in the market.
Expert systemshells are more convenient to use than AI
languages;however,expert systemshells offer limited or
no capabilities for modification of the inference engine.
One of the early AI applications in the petroleum
industry is DIPMETER ADVISOR which was deve-
loped to performwell log analysis.Since then,a number
of other AI applications were developed in various
disciplines of petroleum engineering.Such applications
include interpreting logs,diagnosing and prescribing
remedies for stuck drill pipe,locating mineral deposits,
configuring seismic processing runs,selecting the
optimal drilling mud;problem diagnoses,identifying
the cause of a chemical spill and recommending action,
selection and design of EORprocesses,well stimulation,
testing and logging,and prediction of fluid properties.
2.About this special issue
Given the large number of publications,it was
appropriate to showcase the research efforts in one
Fig.2.Geometry of fuzzy logic.
Fig.3.General structure of an expert system.
special issue of the Journal of Petroleum Science and
Engineering.Efforts for the special issue started in
February 2004.Out of the papers submitted,a total of
11 papers were accepted for publications in this special
issue.The included papers in this issue came from
different researchers working in various research
centers and universities around the world.Titles and
authors of abstracts are given below,where corres-
ponding authors are specified by *:
1.bOptimization of formation analysis and evalua-
tion protocols using neuro-simulationQ,T.Erte-
kin* and N.Silpngarmlers,Pennsylvania State
2.bAI applied to evaluate waterflood response,gas
behind pipe,and imbibition stimulation
treatmentsQ,William bBillQ Weiss*,Jason Weiss,
Visveswaran bVishuQ Subramaniam and Xina
Xie,Correlation Company,USA.
3.bApplication of artificial intelligence tools to
characterize naturally fractured reservoirs in Hassi
Messaoud Oil Field,Algeria:a case studyQ,
Abdelkader Kouider El Ouahed and Djebbar
Tiab*,University of Oklahoma,OK,USA.
4.bA process-knowledge management approach for
assessment and mitigation of drilling risksQ W.F.
*,J.M.Peden,and K.W.Wong
University of Technology,Perth Australia,
nyang Technological University,Singapore.
5.bAweb-based expert system for the planning and
completion of multilateral wellsQ,Haitham M.S.
Lababidi and Ali Garrouch*,Kuwait University,
6.bReservoir properties determination using fuzzy
logic and neural networksQ,Jong-Se Lim*,Korea
Maritime University,Republic of Korea.
7.bEstimating the fracture gradient of Middle East
reservoirs using artificial neural networksQ,Adel
Malallah* and Ibrahim Sami Nashawi,Kuwait
8.bApplication of artificial neural networks for
reservoir characterization with limited dataQ,K.
Aminian* and S.Ameri,West Virginia Universi-
9.bLeak detection in petroleum pipelines using a
fuzzy systemQ,Henrique Ventura da Silva
Kazuyuki Morooka
*,Ivan Rizzo Guilherme
and Jose Ricardo Pelaquim Mendes
State University of Campinas,
Paulista State
10.bA new and novel methodology for the identifi-
cation of best practices in the oil and gas industry,
using intelligent systemsQ,Shahab D.Moha-
ghegh*,West Virginia University,USA.
11.bApplication of an expert system to optimize
reservoir performanceQ,Ridha Gharbi*,Kuwait
The 1st paper by Ertekin and Silpngarmlers of
Pennsylvania State University proposes neuro-simula-
tion methodology that involves the use of conventional
reservoir engineering analysis tools to generate some
sound data base and teaching this data base to a neural
network which is designed as an expert system for the
same class of problems.The data may include
laboratory reservoir engineering data,well tests con-
ducted in the field with the relevant analysis protocols
followed and any other reservoir engineering analysis
studies conducted with the help of numerical and
analytical models.Once the model is trained,the
network is then used in a predictive mode for new
systems.The paper includes examples showing the
application of the proposed method.
The 2nd paper by Weiss fromCorrelations Company
used artificial intelligence technology to predict (1) the
secondary to primary ratio of a water flood candidate
using public domain information,(2) the potential gas
producing rate of a behind pipe interval given only
gamma ray and density logs,and (3) the performance of
single well chemical imbibition treatments.A technique
based on conventional statistical parameters was
developed to numerically describe the patterns ob-
served in log cross-plots.These numerical descriptions
were then prioritized and used as neural network inputs
to be correlated with known production response.
Application of AI in naturally fractured reservoirs is
the subject of the 3rd paper by El-Ouahed and Tiab of
the University of Oklahoma.In this study,a two-
dimensional fracture intensity map and fracture network
map in a large block from Hassi Messaoud field in
Algeria have been developed using Artificial Neural
Network and Fuzzy Logic.The paper discusses the
methodology used to map the fracture network.
In the 4th paper by Prassl et al.of Curtin University
of Technology,a Process-Knowledge Management
System (P-KMS) was developed.The P-KMS system
was designed to investigate drilling in gas hydrate
environments,identify potential well risks (well con-
trol,borehole stability and/or well integrity),and assess
mitigation of them due to alteration of drilling para-
meters and/or strategies.
Aweb-based fuzzy expert system is presented in the
5th paper by Lababidi and Garrouch of Kuwait
University.The expert system was developed in an
Editorial 95
integrated process for planning the style and completion
of multilateral wells.The reasoning process in this
expert system is based on a systematic planning
approach for screening and selecting multilateral well
candidates,lateral-section completion types,and the
junction levels of complexity.
In the 6th paper by Limof Korea Maritime University,
an intelligent technique that uses fuzzylogic and artificial
neural network is suggested in order to determine
reservoir properties fromwell logs.Fuzzy curve analysis
basedon fuzzylogics is used for selecting the best-related
well logs with core porosity and permeability data.
Artificial neural network is used as a nonlinear regression
method to develop transformation between the selected
well logs and core analysis data.The technique is
validated by a case study from offshore Korea.
Estimating the fracture gradient of Middle Eastern
reservoirs using artificial neural network is the subject
of the 7th paper by Malallah and Nashawi of Kuwait
University.The neural network model is able to predict
the fracture gradient as a function of pore pressure,
depth and rock density.A detailed comparison between
the results predicted by this method and those predicted
by other techniques are presented.
The 8th paper authored by Aminian and Ameri from
West Virginia University discusses the application of
artificial neural network for reservoir characterization
with limited data.The method combines statistical and
artificial intelligence techniques to predict the missing
information.A systematic and synergistic approach was
then employed to integrate and interpret various
geological and engineering data that are obtained at
different scales to characterize a complex oil reservoir.
The 9th paper by da Silva et al.fromState University
of Campinas discusses a fuzzy system that is developed
for detecting leaks in petroleumpipelines.The procedure
on howthe systemwas developed,and the evaluation of
such system,is presented.A fault detection module
evaluates the inlet–outlet flow rate deviation in order to
detect a leak or an abnormal operation condition.
In the 10th paper by Mohaghegh of West Virginia
University,a methodology for the identification of best
practices in the oil and gas industry using intelligent
system is presented.The methodology is named
bIntelligent Best Practices AnalysisQ and includes
artificial neural networks,genetic algorithms and fuzzy
logic.The methodology is applied to a database of
stimulation procedures in the Golden Trend fields of
Oklahoma to clearly demonstrate its use and benefits.
The 11th paper authored by Ridha Gharbi from
Petroleum Engineering Department of Kuwait Univer-
sity presents an optimization methodology combined
with an economic model,which is implemented into an
expert system to optimize the net present value of full
field development with EOR processes.In this paper,
the details of the proposed expert system and the effect
of several design parameters on the project profitability
of the studied EOR processes are reported.
We would like to express our sincere thanks to the
authors for their contribution to this special issue.We
also would like to express our appreciation to the
reviewers of papers submitted to this issue.Despite
their heavy schedules,the reviewers have enthusiasti-
cally participated in the review process.Their valuable
suggestions and criticisms greatly enhanced the quality
of this special issue.Our appreciation goes to Julius
Langlinais (Louisiana State University),Abdel Zellou
(Prism Seismic),Patrick Wong (Veritas Geophysical
Corp),Turgay Ertekin (Pennsylvania State University),
Tarek Darwich (SIPETROL),Luis Gomez (University
of Tulsa),Ali Garrouch (Kuwait University),Adwait
Chawathe (ChevronTexaco),Iraj Ershaghi (University
of Southern California),Shedid Elgaghah (United Arab
Emirates University),Mahmut Sengul (Schlumberger
Oilfield Services),Tao Zhu (University of Alaska),
Adel Malallah (Kuwait University),Jonathan Kwan
(University of Oklahoma),and Tongjun Ruan (New
Mexico Tech).We also would like to thank Ms.Tirza
van Daalen and Ms.Tonny Smit of Elsevier Science for
her help to complete this special issue.
Ridha B.C.Gharbi
Department of Petroleum Engineering,
Kuwait University,PO Box 5969,
Safat 13060,Kuwait
Corresponding author.
G.Ali Mansoori
Departments of Chemical and Bio Engineering,
University of Illinois at Chicago,Chicago,
IL 60610,USA
13 September 2005