Filetype PDF | Posted on 29 Jan 2023 | 3 years ago
Authentication
digital resources
255x Filetype PDF File size 0.77 MB Source: www.ige.unicamp.br
Suslick, S.B., Schiozer, D., Rodriguez, M.R. TERRÆ 6(1):30-41, 2009
THEMATIC CONTRIBUTION
Uncertainty and Risk Analysis in Petroleum
Exploration and Production
Saul B. Suslick
UNICAMP, Institute of Geosciences and Center of
Petroleum Studies
Denis Schiozer
UNICAMP, Department of Petroleum Engineering
(FEM) and Center for Petroleum Studies – denis@
dep.fem.unicamp.br
Monica Rebelo Rodriguez
PETROBRAS, Science and Petroleum Engineering
Graduate Program – FEM/IG
Abstract During the past decades, there have cally viable reservoirs, technology and oil price.
been some significant improvements in uncertainty and Even at the development and production stage
risk analysis applied to petroleum exploration and pro- the engineering parameters embody a high level of
duction. This paper presents a brief overview of the main uncertainties in relation to their critical variables
contributions made in the exploration and production (infrastructure, production schedule, quality of
stages, followed by a summary of the main trends in the oil, operational costs, reservoir characteristics etc.).
context of an exhaustible resource. Decisions related to pe- These uncertainties originated from geological
troleum exploration and production are still very complex models and coupled with economic and engineer-
because of the high number of issues involved in the pro- ing models involve high-risk decision scenarios,
cess. However, uncertainty and risk analysis are becoming with no guarantee of successfully discovering and
more popular as new hardware and software advances developing hydrocarbons resources.
appear, contributing in an important manner to clarify Corporate managers continuously face impor-
the range and the impacts of new discoveries as well as tant decisions regarding the allocation of scarce
development and production assets. resources among investments that are character-
ized by substantial geological and financial risk.
For instance, in the petroleum industry, managers
Keywords uncertainty, risk analysis, decision are increasingly using decision analysis techniques
analysis, portfolio. to aid in making these decisions. In this sense, the
petroleum industry is a classic case of uncertainty
Introduction in decision-making; it provides an ideal setting
for the investigation of corporate risk behavior
and its effects on the firm’s performance. The
Exploration and production of hydrocarbons1 wildcat drilling decision has long been a typical
is a high-risk venture. Geological concepts are un- example of the application of decision analysis in
certain with respect to structure, reservoir seal, and classical textbooks.
hydrocarbon charge. On the other hand, economic Future trends in oil resource availability will
evaluations have uncertainties related to costs, depend largely on the balance between the out-
probability of finding and producing economi- come of the cost-increasing effects of depletion
1 Exploration and production of hydrocarbons in this paper encompass all and the cost-reducing effects of new technology.
the activities, such as: basin and play analysis, leads, prospect evaluation, Based upon that scenario, new forms of reservoir
development stages, facilities, logistics, management, etc.
30
TERRÆ 6(1):30-41, 2009 Suslick, S.B., Schiozer, D., Rodriguez, M.R.
exploitation and management will appear where pricing and resource allocation in large monopolis-
the contributions of risk and decision models are tic enterprises. Allais’ work was a useful means or
one of the important ingredients. This trend can preview to demonstrate Monte Carlo methods of
be seen in the last two decades. The new inter- computer simulation and how they might be used
nationally focused exploration and production to perform complex probability analysis, instead of
strategies were driven in part by rapidly evolving simplifications of risk estimation of large areas.
new technologies. Technological advances allowed During this period, there were several attempts
exploration in well-established basins as well as to define resource level probabilities at various stag-
in new frontier zones such as ultra-deep water. es of exploration in a basin using resource distribu-
Those technology-driven international explora- tion and risk analysis (Kaufman, 1963; Krumbein
tion and production strategies combined with new and Graybill, 1965; Drew, 1967; Harbaugh et al.,
and unique strategic elements where risk analysis 1977; Harris, 1984; Harbaugh, 1984, Harris 1990).
and decision models represent important compo- At that time governmental agencies (U.S. Geologi-
nents of a series of investment decisions. cal Survey, Institut Français du Pétrole, etc.) were
This paper covers a brief review of previous also beginning to employ risk analysis in periodic
applications involving the following topics: (1) appraisals of oil and gas resources (Figure 1).
Risk and Decision Analysis in Petroleum Explo- During the 1980’s and 1990’s, new statistical
ration; (2) Field Appraisal and Development, and methods were applied using several risk estimation
Uncertainty in Production Forecasts, (3) the De- techniques such as: (1) lognormal risk resource
cision Making Process and Value of Information distribution (Attanasi and Drew, 1985), (2) Pareto
and (4) Portfolio Management and Valuation Op- distribution applied to petroleum field-size data
tion Approach. This paper describes some of the in a play (Crovelli, 1995) and (3) fractal normal
main trends and challenges and presents a discus- percentage (Crovelli et al., 1997). Recently, USGS
sion of methodologies that affect the present level has developed several mathematical models for
of risk applied to the petroleum industry aimed at undiscovered petroleum resource assessment (Ahl-
improving the decision-making process. brandt and Klett, 2005) and forecast reserve growth
of fields both in the United States (U.S.) and the
world (Klett, 2005).
Risk Analysis: Exploration Throughout 1960’s, the concepts of risk analy-
sis methods were more restricted to academia and
The historical origins of decision analysis can were quite new to the petroleum industry when
be partially traced to mathematical studies of prob- contributions appeared from Grayson (1960), Arps
th th
abilities in the 17 and 18 centuries by Pascal, and Arps (1974), Newendorp (1975, edited as Ne-
Laplace, and Bernoulli. However, the applications wendorp and Schuyler, 2000) and Megill (1977).
of these concepts in business and general man- Newendorp (op.cit.) emphasized that decision
agement appeared only after the Second World analysis does not eliminate or reduce risk and will
War (Covello and Mumpower, 1985; Bernstein, not replace professional judgment of geoscientists,
1996). The problem involving decision-making engineers, and managers. Thus, one objective of
when there are conditions of risk and uncertainty decision analysis methods, as will be discussed later
has been notorious since the beginnings of the in this paper, is to provide a strategy to minimize
oil industry. Early attempts to define risk were the exposure of petroleum projects to risk and un-
informal. certainty in petroleum exploration ventures.
The study by Allais (1956) on the economic The assessment to risk model preferences of
feasibility of exploring the Algerian Sahara is a clas- decision makers can be achieved using a utility
sic example because it is the first study in which the function provided by Utility Theory. If companies
economics and risk of exploration were formally make their decisions rationally and consistently,
analyzed through the use of the probability theory then their implied risk behaviors can be described
and an explicit modeling of the sequential stages by the parameters of a utility function. Despite
of exploration. Allais was a French economist who Bernoulli’s attempt in the 18th century to quantify
was awarded the Nobel Prize in Economics in 1988 an individual’s financial preferences, the param-
for his development of principles to guide efficient eters of the utility function were formalized only
31
Suslick, S.B., Schiozer, D., Rodriguez, M.R. TERRÆ 6(1):30-41, 2009
LowRisk
roject Status
P
Production roduction
On P
Reserves Under Development
Proved Proved+ ending
Commercial Proved+Probable Probable+Possible
etroleum Planed for Development
-in-Place Development P roject Maturity
P
Contigent
Resources Development on Hold
Discovered P IniatiallyLow High
Estimate Best Estimate
Estimate Development not viable
rospect HighRisk
Sub-commercial P
etroleum - Iniatially in Place Prospective Lead
otal P -in- Low Resources High Play
T Estimate Estimate
Best
etroleum Place Estimate
Undiscovered P Iniatially
Unrecoverable
Range of Uncertainty
Figure 1 – Petroleum Resource Classification Scheme (modified from Ross, 2004 and SPE/WPC/AAPG, 2000)
300 hundred years later by von Neumann and jectives and risk policy into the investment choices
Morgenstern (1953) in modern Utility Theory. was made by Walls (1995) for oil and gas compa-
This seminal work resulted in a theory specifying nies using the multi-attribute utility methodology
how rational individuals should make decisions (MAUT). Walls and Dyer (1996) employed the
in uncertain conditions. The theory includes a set MAUT approach to investigate changes in corpo-
of axioms of rationality that form the theoretical rate risk propensity with respect to changes in firm
o
basis of decision analysis. Descriptions of this full size in the petroleum industry. Nepomuceno F et
set of axioms and detailed explanations of decision al. (1999) and Suslick and Furtado (2001) applied
theory are found in Savage (1954), Pratt (1964), and the MAUT models to measure technological prog-
Schailfer (1969). Cozzolino (1977) used an expo- ress, environmental constraints as well as financial
nential utility function in petroleum exploration to performance associated with exploration and pro-
express the certainty equivalent that is equal to the duction projects located in offshore deep waters.
expected value minus a risk discount, known as the More recently, several contributions devise
risk premium. Acceptance of the exponential form petroleum exploration consisting of a series of
of risk aversion leads to the characterization of risk investment decisions on whether to acquire addi-
preference (risk aversion coefficient), which mea- tional technical data or additional petroleum assets
sures the curvature of the utility function. Lerche (Rose, 1987). Based upon these premises explora-
and MacKay (1999) showed a more comprehen- tion could be seen as a series of investment deci-
sible form of risk tolerance that could intuitively sions made under decreasing uncertainty where
be seen as the threshold value, whose anticipated every exploration decision involves considerations
loss is unacceptable to the decision maker or to of both risk and uncertainty (Rose, 1992). These
the corporation. aspects lead to a substantial variation in what is
An important contribution that provides rich meant by risk and uncertainty. Some authors such
insight into the effects of integrating corporate ob- as Knight (1921) make a distinction between risk
32
TERRÆ 6(1):30-41, 2009 Suslick, S.B., Schiozer, D., Rodriguez, M.R.
(where probabilities are known) and uncertainty without significant precision loss. Simplifications
(where one is unable to assign probabilities) fo- are possible, for instance, in the modeling tool,
cusing their analysis on uncertainty. Meanwhile, treatment of attributes and in the way several types
Megil (1977) considered risk an opportunity for of uncertainties are integrated.
loss. Risk considerations involve size of investment One of the simplest approaches is to work with
with regard to budget, potential gain or loss, and the recovery factor (RF) that can be obtained from
probability of outcome. Uncertainty refers to the analytical procedures, empirical correlations or pre-
range of probabilities in which some conditions vious simulation runs, as presented by Salomão and
may exist or occur. Grell (2001). When higher precision is necessary, or
Rose (2001) pointed out that each decision when the rate of recovery significantly affects the
should allow a progressively clearer perception of economic evaluation of the field, using only the
project risk and exploration performance that can expected recovery factor may not be sufficient.
be improved through a constructive analysis of geo- Techniques such as experimental design, re-
technical predictions, review of exploration tactics sponse surface methods and proxy models have
versus declared strategy, and year-to-year compari- been used by several authors (Damsleth et al., 1991;
son of exploration performance parameters. These Dejean, 1999; Ligero et al., 2007) in order to accel-
findings showed the importance of assessing the erate the process. Another possible approach is to
risk behavior of firms and managerial risk attitudes. use faster models such as a streamline simulation or
Continued monitoring of the firm’s level of risk a fast coarse grid simulation as proposed by Hast-
aversion is necessary due to the changing corporate ings et al. (2001), Ballin et al. (1993), Subbey and
and industry environment as well as the enormous Christie (2003), and Ligero et al. (2003).
contribution generated by technological develop- The integration of risk analysis into the defini-
ment in E&P. Over any given budgetary period, tion of production strategy can also be very time
utilization of an established risk aversion level will consuming because several alternatives are possible
result in consistent and improved decision making and restrictions have to be considered. Alternatives
with respect to risk. may vary significantly according to the possible sce-
narios. Schiozer et al. (2004) proposed an approach
Risk Analysis: Field Appraisal and Development to integrate geological and economic uncertainties
with production strategy using geological represen-
tative models to avoid large computational effort.
During the exploration phase, major uncertain- Integration is necessary in order to (1) quan-
ties are related to volumes in place and economics. tify the impact of decisions on the risk of the
As the level of information increases, these uncer- projects, (2) calculate the value of information, as
tainties are mitigated and, consequently, the im- proposed by Demirmen (2001) and (3) quantify
portance of the uncertainties related to technology the value of flexibility (Begg and Bratvold, 2002;
and recovery factor increases. The situation is more Hayashi et al., 2007). The understanding of these
critical in offshore fields and for heavy-oil reser- concepts is important to correctly investigate the
voirs, where investments are higher and there is a best way to perform risk mitigation and to add
lower operational flexibility (Pinto et al., 2001). value to E&P projects.
In the preparation of development plans, field Therefore, risk analysis applied to the ap-
management decisions are complex issues because praisal and development phase is a complex is-
of (1) the number and type of decisions, (2) the sue and it is no longer sufficient to quantify risk.
great effort required to predict production with Techniques today are pointing to: (1) quantifica-
the necessary accuracy and (3) the dependency of tion of value of information and flexibility, (2)
production strategy definition on several types of optimization of production under uncertainty,
uncertainty with significant impact on risk quan- (3) mitigation of risk and (4) treatment of risk
tification. as an opportunity. All these issues are becoming
In order to avoid excessive computation effort, possible due to hardware and software advances,
some simplifications are always necessary. The key allowing an increasing number of simulation
point is to define the simplifications and assump- runs of reservoir models with higher complexity
tions that can be made to improve performance (Gorell and Basset, 2001).
33
no reviews yet
Please Login to review.
