Application of Decision Theory in Assessing Marginal Oilfield Risks

7/30/2019 Application of Decision Theory in Assessing Marginal Oilfield Risks

1/7

Journal of Emerging Trends in Engineering and Applied Sciences (JETEAS) 3(4) 594-600 (ISSN: 2141-7016)

594

Application of Decision Theory in Assessing Marginal OilfieldRisks: Niger Delta Hub Example

Alaneme, Charles Ezemonye and Igboanugo, Anthony Clement

Department of Production Engineering

University of Benin, Benin City, Nigeria.

Corresponding Author: Alaneme, Charles Ezemonye

___________________________________________________________________________Abstract

Decision theory valuation methodology has been identified as an effective tool in the analysis and managementof risks for decision making in marginal oilfield exploitation. Numerous conventional methods in the employ of

most international oil companies seem an aberration in marginal oilfield operation due to the uniqueness of the

risks and the high cost of implementation. This short coming has resulted in the inability to optimally unlockthe economic potentials of marginal oilfields that has remained untapped to replenish fast declining oilfields on

account of their low economies of scale. A decision theory approach was successfully deployed in theoreticallyanalyzing decision alternatives with Isiekenesi Oilfield, one of 251 remotely located marginal oilfields in the

Nigeria Niger Delta. The study yielded corresponding payoff values for different reserve expectations of low,medium, and high cases in barrels of crude oil. Despite its limitations in not aptly defining the risks in crispnumbers, it successfully predicted the risk ratios of fundamental decision alternatives guided by basic

assumptions on state of nature. This approach provides a cost effective first-pass appraisal mechanism needfulfor decision making process open to investment capitalists engaged in marginal oilfield exploitation.

__________________________________________________________________________________________

Keywords: marginal oilfield, reserves, risks, decision theory, Isiekenesi, Niger Delta.__________________________________________________________________________________________INTRODUCTIONCurrent risks assessment and management practices

by International Oil Companies (IOCs) have not beenvery effective in marginal oilfields operation due to

the small size and remoteness of the oilfields, whichin most cases, are very far from existing processing

facilities coupled with the complexities of theoperation among others. Whereas IOCs operate

many oilfields to which they could afford to spread or

absorb the risks, local operators or venture capitalistsare constrained to just one or few marginal oilfields.

In the Nigerias Niger Delta for instance, circa 251identified marginal oilfields are operated by nearly

equal number of indigenous companies who could

pass as small scale oilfield operators. This caliber ofoperators lack basic technological know how and

managerial skills essential for handling thecomplexities off marginal oilfields exploitation.

Marginal oilfields show great potential tosubstantially contribute to the national economic

fortune of a producing country and could at the sametime ruin investors, if the associated risks are not

properly identified and managed. Further, marginal

oilfields contribute significantly to the national oilreserve and in so facto deserve due attention.

Unfortunately, establishing risk management process

and database similar to IOCs, require many years ofbuilt up experience and deployment of discipline

professionals, which can be quite strenuous andexpensive. This proposed approach takes into

cognizance feasible alternative courses of action in

conjunction with the prevailing state of nature and indoing this; appropriate statistical techniques are

employed as decision support in arriving at the rightcourse of action.

There has been considerable interest in the

development and application of models to riskmanagement. Use of Simulation in risk analysis is

common in the risk analysis literature. Some

representative works on risk management usingSimulation approach include Chapman (1983),

Vinnen (1983), Cooper,MaccDonald, and Chapman(1985), Pugh (1986), and Hall (1986). Following

these seminal research in the 1980s, further

application of Monte Carlo appeared in the 1990s.The works Higgins (1993), Kostetsky (1994),

Savvides (1994) as well as Van Groenendaal andKleijnen (1997) are typical. Interest in Monte Carlo

simulation application to risk management has beensustained in recent contemporary period. Coelho, et

al. (2005) compared Monte Carlo simulation withNeural Network techniques and observed that the two

approaches complement each other. Further Chinbat

and Takakuwa (2009) focused on the development ofa simulation method for managing risk in Mongolian

mining project.

Besides Simulation technique, a variety of

approaches had also appeared in the risk managementliterature. Thus in an effort to further establish the

use of risk analysis and, moreover, widen the scope

Journal of Emerging Trends in Engineering and Applied Sciences (JETEAS) 3 (4): 594-600 Scholarlink Research Institute Journals, 2012 (ISSN: 2141-7016)

jeteas.scholarlinkresearch.org


2/7


595

of applicable model, McCray, et al. (2002) as well as

Trumper and Virine (2011) introduced the use ofEvent Chain. These studies demonstrated that the

method is quantitative and also mitigates cognitiveand motivational biases inherent in the project risk

analysis. Again, heuristic method of making

judgment about risk under uncertainty has been

studied by Tversky and Kahneman (1974). Theauthors noted that though the approach is highlyeconomical and effective but it leads to systematic

and predictable errors. Also, Kaiser (2010) offers a

lively modeling of inventory of assets committed tomarginal oilfield production and sketched a

perspective of the producing assets.

In particular, the studies Schuyler (2001) andFlyvbjerg (2006) provide interesting treatments of

risk analysis of projects using decision theory. Inaddition, the book Rose (2012) provides detailedtreatment on risk assessment and the economic

implications for explorationists, managers, andoilfield investors. Yet, a novel approach to risk

management known as unknown unknown ispresented by Raydugui (2011). The study noted that

people risk identification requires thinking outside

the box. Also, Bastos and Barton (2004) appliedPortfolio Theory that is based on Probabilistic

methodology. This was applied to Brazilianelectrical system involving development of standalone hydropower station. Last, Kraft (1982)

discussed risk in policy research and noted that riskanalysis is a veritable tool for managing risks in

projects. Again, works; Ward and Chapman (1991),Harbaugh, et al. (1995), Undram and Takakuwa

(2009), and Andersen and Mostue (2011)emphasizedthe need for the use of risk analysis in managing risks

in projects.

In sum, then, the foregoing sample literature review

provides palpable supportive evidence to the fact thatvery little study appears to have been documented inregards to application of Decision Theory to risk

management in marginal oilfields. This papertherefore seeks to breach this frontier. Thus, the aim

of this paper is to apply Decision Theory in analyzingrisks inherent in the marginal oilfields located in the

Nigeria Niger Delta.

METHODSThis analytical and case study research design is

based on data obtained from three exploratory wellsdrilled in Isiekenesi, a marginal oilfield located in theNigeria flank of Niger Delta. More specifically, the

data relates to wells drilled in the early 1910s with a

2-D seismic survey acquired sixty years later in the

early 70s. The field is a partially appraised, non-

concessionary onshore acreage located approximately63 and 85 Kilometers North East of Izombe and

Egbema fields respectively in the Niger Delta.Figure 1 shows the Oil Mining Lease (OML) map of

the Niger Delta and Benue Basin with relative

location of the Isiekenesi Field.

IsiekenesiAcreage

Figure 1: Location Map of Nigeria Oil Mining Leases

The first well was drilled to a depth of 8,400 feet(2,560 meters) and encountered 271 feet (87 meters)

of net oil in four sands. Also, Figure 2 shows the

cross-sectional map of the only three exploratory

wells. Snowball, non-random sampling techniquewas followed.

NW SE

T.D.L P.

10800 ss10800 ss

O WC 1092 2 ss

ISHII -A3ISHII-A1IHSII-A2

ISHII A3.000P x-section

Figure 2: Cross-sectional Map

The estimated expected reserves were taken from

preliminary evaluations conducted at the early stage.

Further, a 20-year production forecast for three casescenarios; low case, medium case, and high case

representing proved, probable, and possible reserveswere estimated based on data from the three

exploratory wells. Again, another data obtained fromthe oilfield relate to the initial estimated reserves and

are shown in Table 1.


3/7


596

Table1: Reserve Expectation ScenariosS/N Sensitivity Case STOIIP

(MMSTB)

Reserves

(MMSTB)

1 P90 25.3 10.12 P50 36.9 14.2

3 P10 53.5 22.6

For simplicity, we considered these three alternatives

open to prospective venture capitalists for operating amarginal oilfield:i) Direct Labour (Direct Execution)ii) Partnership (Equity Sharing)iii) Outsourcing (with 10% returns)

Expected Value Method (EVM)

This theoretical approach enables the analyst to

determine the following:i) Expected Value E(x) of the reservoir

under uncertaintyii) The associated risk in the management

of the reservoir uncertainties.

Let x be a random discrete variable known as the

pay-off matrix. Consider a probability or densityfunction

( ) . , ( ) 1K

i i

f x p f x

, then

22

1

( ) . ( ),

, ( ) (1)K

i

E x x f x and

Variance Risk x f x

Thus, the higher the variance, the higher the

associated risk.From (1):

2 2 2

2 2

2 2

2 2

( 2 ) ( )

( ) 2 ( ) ( )

( ) 2 . .1

( ) (2)

x x f x

x f x xf x f x

x f x

x f x

For this oilfield, the state of nature in Table 2 has

been estimated as follows:

Table 2: Assumed State of NatureS/N Alternative Investment

Capital

Income

(Reserve)

1 Direct Labour 100% 100%2 Partnership 60% 60%

3 Outsourcing 0% 10%

S/N Sensitivity Reserve

Expectation

Historical

Probability

1 P90 Low 0.25

2 P50 Medium 0.703 P10 High 0.2

Case Analysis

The confidence limits are

0

0

0

/2

/2

/2

/2

, ;

90%, 1.67

50%, 0.67

10%, 0.12

, 3

ConfidenceLimits y Z and fromstatisticalTablen

For Z Z

For Z Z For Z Z

Number of wells n

(3)

And, for spatial contiguity of the wells, being thatthey are in the same region, the standard deviation

among the well voluminosity should be25%

A) Proved Case (90% Confidence level), K=1,where n is number of exploratory wells.

State of nature = 25/75.90%

1.67

for confidence level

UCL yn

(4)

0.25(10.1)10 .1 1.67

3

12.535MMSTB

1.67LCL yn

(5)

State of Nature

Decision

Alternatives

Operating Factor S1 (0.25) S2

(0.75)

1 1.0 12.535 7.666

2 0.6

(Equity share

factor)

7.521 4.600

3 0.1

(Expected Royalty

Factor)

1.254 0.767

(6)

12.535 7.6660.25

7.521 4.6000.75

1.253 0.7666

12.535 0.25 7.666 0.75

7.52 1 0.2 5 4.600 0.75

1.253 0.75 0.7666 0.75

1

8.883

( ) 5.330

0.888

kE x

2 2 2( )Risk x f x (7)

228.88312.535 7.666

0.255.3307.521 4.600

0.750.8881.253 0.767

0.25(10.1)10.1 1 .67

3

7.666MMSTB

( ) . ( )E x x f x


4/7


597

4.444

1.600

0.045

99.648

0.045 35.883

1

100

36

1

B) Probable Case (50% Confidence level),

K=2, where n is number of exploratorywells. State of nature = 70/30

0UCL y Z

n

(8)

0.25(14.2)1 4.2 0 .1 2

3

15.573MMSTB

0

LCL y Zn

(9)

0.25(14.2)1 4. 2 0 .1 2

3

12.827MMSTB

State of Nature

Decision

Alternatives

Operating

Factor

S1 (0.70) S2 (0.30)

1 1.0 15.573 12.827

2 0.6

Equity share

factor

9.344 7.696

3 0.1

Expected Royalty

Factor

1.557 1.283

2( ) . ( )

kE x x f x (10)

15.573 12.8270.7

9.344 7.6960.3

1.557 1.283

(15.573 0.7) (12.827 0.3)

( 9.344 0.7 ) (7 .696 0. 3)

(1 .5 57 0 .7 ) (1 .2 83 0 .3)

1

2

3

14.749

8.850

1.489

d

d

d

2 2 2, ( )Hence the risk x f x (11)

2214.74915.573 12.827

0.7

8.8509. 34 4 7 .69 6 0.31.4891 .5 57 1 .2 83

21 9.122 217.533

78.885 78.323

2.235 2.214

1.589

0.562

0.021

2

75

, ( ) 0.021 27

1

Risk

C) Possible Case (10% Confidence level),

K=3, where n is number of exploratorywells. State of nature = 20/80

0UCL y Z n

(12)

0.25(22.6)22.6 0.12

3

22.992MMSTB

0LCL y Z

n

(13)

0.25(22.6)22.6 0.12

3

22.208MMSTB

State of Nature

Decision

Alternatives

Operating

Factor

S1 (0.20) S2 (0.80)

1 1.0 22.992 22.208

2 0.6

Equity share

factor

13.795 13.325

3 0.1Expected

Royalty Factor

2.299 2.221

3( ) . ( )

kE x x f x (14)

22.992 22.2080.2

13.795 13.3250.8

2.299 2.221

(22.992 0.2) (22.208 0.8)

(13.795 0.2) (13.325 0.8)

(2.299 0.2) (2.221 0.8)

1

2

3

22.365

13.419

2.237

d

d

d

2 2 2, ( )Hence t he risk x f x (15)

2222.36522.992 22.208

0.2 13.41913.795 13.3250.8

2.23662.299 2.221

500.282 500.193

180.105 180.070

5.0033 5.0024

0.089

0.035

0.001

2

89

, ( ) 0.001 35

1

Risk

Assumptions:

1. Three decision alternatives were considered in thisstudy namely:

a. direct in-house execution by Venture Capitalistusing own resources

b. partnership with other operators through equityparticipation

c. direct or complete outsourcing and with expectedreturns through royalty

2. Sixty/forty equity sharing between the investorand participating partners is assumed

3. A royalty of 10% is assumed to be the return foroutright lease or full outsourcing of the oilfield

78.913(157.126 0.25) (58.768 0.75)

28.411(56.565 0.25) (21.160 0.75)

0.7898(1.571 0.25) (0.588 0.75)


5/7


598

4. on account of the spatial contiguity of the three oilwells voluminosity, a fair assumption of 25%variability standard deviation, that is, = 25% is

assumed5. State of nature: A fair probability of achieving the

projected expected reserve cases are assumed as

follows:

a. 25% for Proved Caseb. 70% for the Probable Casec. 10% for the Possible Case6. Number of wells, n = 3, based on the number of

sampled exploratory wells.

RESULTS

The results of this study are presented in the

following sequence:3.1 Resistivity log3.2 Projected cross-sectional map3.3 20-year estimated production schedule3.4 Results of statistical computations of risks

associated with three decision optionsThe foregoing outline is taken seriatim.

Resistivity log Result

Figure 3 presents the reservoir resistivity log showing

the thick accumulation of hydrocarbon distributionwith a three layer yield.

Figure 3: Ishii-3 Resistivity Log

Projected Cross-Sectional Map

Figure 4.2 provides the cross-sectional map of thethree exploratory wells indicating areas of continuity.

I S H I I A 3 0 0 0 N

O W C = 9 8 8 7 f t s s , N o . o f W e l l p e n e t r a t i o n s = 3

I S H I I

Figure 4.2: Cross-sectional Map

The map provides a relative location of theexploratory wells drilled presently and provides the

contiguity pattern in terms of the oil-water contactsfor the reservoirs at 9887 feet of subsurface.

Twenty Year Production Schedule

Figure 4.3 shows the profiles of the estimated 20-year

exploratory production forecast for the three expected

cases based on 2-D seismic data.

Low CaseExpectation

MediumCase

Expectation

High CaseExpectation

0

5001000

1500

20002500

30003500

4000

4500

1 3 5 7 9 11 13 15 17 19 21

Volume(MSTB)

Production Year

Figure 4.3: Oil Production Forecast

The plots represented the three possible scenarios oflow, medium, and high case probabilities.

Expectedly, the plot approximates to a normaldistribution over the 20-year period considered with apeak duration lasting more than five years.

Results of Statistical Computations

Proved Case Scenario

This presents the low case expectation with highestlevel of confidence. The expected pay-off in millions

of barrels for the three decision alternatives is asfollows:

1

2

3

8.883

5.330

0.888

d

d

d

While, the associated relative risk is expressed as:

2

100

( ) 0.045 36

1

Risk


6/7


599

Implying that, the ratio of risks associated with the

three decision alternatives: direct execution;partnership; outsourcing is 100:36:1. This

quantitative risk analysis is an operators guide toaction.

Probable Case Scenario

Again, this presents the medium case expectationwith average level of confidence. The expected pay-off in millions of barrels for the three decision

alternatives is as follows:1

2

3

14.749

8.850

1.489

d

d

d

MSTB

And the relative associated risk is given by:

2

75

, ( ) 0.021 27

1

Risk

Which shows a reduced ratio from the previous

Proved case scenario, thus, for probable case, the risk

ratio is estimated to be 75:27:1.

Possible Case Scenario

This presents the highest case expectation, however,with the lowest level of confidence. In this category,the expected pay-off in millions of barrels for the

three decision alternatives is expressed as:1

2

3

22.365

13.419

2.237

d

d

d

MSTB.

And, the associated risk is given by:

2

89

, ( ) 0.001 35

1

Risk

In this version, the risk proportion is 89:35:1;

showing a significant variation from the previousreserve expectation scenarios.

DISCUSSIONIncreasingly, interest in marginal oilfields is rapidly

emerging owing mainly to the economic benefits.Thus, oil producing countries concerned about the

fast depleting reserves without commensurate newdiscoveries, are now striving to harness the huge

potentials from marginal oilfields which hitherto had

been neglected. Despite this interest, the exploitation

of these marginal oilfields has remained nightmarish

due to numerous risks and uncertainties associatedwith the exploitation as expounded earlier.Apparently, extrapolating the cost of conventional

practices by IOCs to managing marginal oilfield riskshas not been 100% effective due to several reasons.

First, IOCs could absorb or spread some of theserisks among their portfolios which other local

operators or venture capitalists might not be disposedto handle. Second, the practical approach might be

too expensive for the local operators. The approachin question is an accumulation of experience over a

long period which is not readily available to the

venture capitalists. Hence, a less costly theoreticalrisk assessment method becomes a veritable tool

which this study has proposed.

Evidently, the approach adopted in this study,

although a statistical evaluation, but in some way, is aballpark of simulation methodology in the sense that

three different decision alternatives were brought intoperspective and analyzed based on certain stated

assumptions. The payoffs and the associated risks

were weighed up in order to guide the process of

taking well informed decisions on investmentalternatives.

Objectively, this study has successfully unearthed

quantitatively, the relative risks associated with likelydecision alternatives when faced with varied expected

reserve scenarios under uncertain states of nature.

CONCLUSION

Despite limited field data obtained from the marginal

oilfield, the study has been successful in usingdecision tool methodology to provide quantitativeinsight into the nature and distribution of risks

inherent in the marginal oilfield under study. Sureenough, this outcome of the study has demonstrated

that a quantitatively guided decision could be reachedwith sparse data which will not be easy with

conventional approach. All the same, by extending

the spectrum or bandwidth of the stated state ofnature, a sensitivity analysis could be called into play

in order to achieve optimization.

REFERENCES

Andersen, S. and Mostue, B.A., 2011. Risk analysisand risk management approaches applied to the

petroleum industry and their applicability to IOconcepts. Safety Science. 50(10), pp.20102019.

Bastos, P.R.M. and Bortoni, E.C., 2004. Portfolio

analysis applied to small hydroelectric plant

investment. IEE Journal, 8th International

Conference on Probabilistic Methods Applied to

Power Systems, Iowa, September 12-16..

Chapman, C.B. and Cooper, D.F., 1983. Risk

Analysis: Testing Some Prejudices. European Journalof Operational Research, 14(3), pp.238-247, Elsevier

Publishing.

Chinbat, U. and Takakuwa, S., 2008. Using operationprocess simulation for a Six Sigma project of mining

and iron production factory. Institute of Electrical andElectronics Engineers publications, Winter

Simulation Conference Proceedings, New Jersey,pp.24312438.

Coelho, D.K. and Jacinto, C.M C., 2005. Risk

assessment of drilling and completion operations in

petroleum wells using a Monte Carlo and a neural

network approach. Institute of Electrical andElectronics Engineers publications, Winter

Simulation Conference Proceedings, New Jersey,pp.18921897.


7/7


600

Cooper, D.F., MacDonald, D.H., and Chapman, C.B.,

1985. Risk Analysis of a Construction Cost Estimate.International Journal of Project Management, vol. 3,

issue 3, pp.141-149, Elsevier Publishing.

Flyvbjerg, B., Holm, M. K. S., and Buhl, S. L., 2005.

How inaccurate are demand forecasts in public works

projects? Journal of the American PlanningAssociation, 78(2), pp. 131-146.

Groenendaal, W.J.H. and Kleijnen, J.P.C., 1997. On

the Assessment of Economic Risk: Factorial DesinVersus Monte Carlo Method. Reliability Engineering

and Safety System 57, pp. 91-102, Elsevier Science

Ltd., PII:S0951-8320(97) 00019-7.

Hall, J.N., 1986. Use of Risk Analysis in North Sea

Projects. International Journal of ProjectManagement, vol. 4, Issue 4, pp.217-222, ElsevierPublishing

Harbaugh, J., Davis, J., and Wendebourg, J., 1996.

Computing risk for oil prospects: Principles andprograms. Elsevier Science, 13(7), pp.852

Higgins, J.G., 1993. Planning for Risk andUncertainty in oil exploration. Long Range Planning,

vol. 26, issue 1, pp. 111-122, Elsevier Publishers.

Kaiser, M.J., 2010. Marginal production in the Gulf

of Mexico I. Historical statistics & modelframework. Applied Energy Journal: Elsevier,

87(8):2535-2550.

Kostetsky, O., 1994. A Facilitated, Graphics &Monte Carlo Based Predictive Project Process.

Predictive Technologies, IEEE Conference Journal,

pp.177-184.

Kraft, M.E., 1982. The Use of Risk Analysis inFederal Regulatory Agencies: An Exploration.Review of Policy Research, May, DOI: 10.1111,

J.1541-1338, Wiley Online library, Published: PolicyStudies Organization 1(4), pp.666-675.

McCray, G.E., Purvis, R.L., and McCray, C.G., 2002.

Project Management Under Uncertainties: TheImpact of Heuristics and Biases. Project Management

Journal, 33(1), pp.49-57.

Pugh, L.A. and Soden, R.G., 1986, Use of RiskAnalysis Techniques in Assessing the Confidence ofProject Cost Estimates and Schedules. International

Journal of Project Management, 4(3), pp.158-162,

Elsevier Publishing.

Raydugin, Y., 2011. Unknown Unknown in ProjectProbabilistic Cost and Schedule Risk Models.

Palisade White Papers.

Rose, P., 2012. Risk Analysis and Management of

Petroleum Exploration Ventures. AAPG Methods inExploration Series, No. 12. American Association of

Peytroleum Geologists, Tulsa USA.

Savvides, S.C., 1994. Risk Analysis in Investment

Appraisal. Project Appraisal, 9(1), pp.3-18, Beech

Tree Publishing.

Schuyler, J.R., 2001. Risk and decision analysis in

projects. 2nd

ed. Pennsylvania: Project Management

Institute.

Trumper, M. and Virine, L., 2011. Eevent Chain

Methodology in Project Management. European

Journal for the Informatics Professional CepisUpgrade vol. xii, no. 5, Novatica,

http://Cepis.org/upgrade, ISSN 1684-5285.

Tversky, A. and Kahneman, D., 1974. Judgment

Under Uncertainty: Heuristics and biases. Science,185, pp.1124-1130.

Undram C., Takakuwa S., 2009. Using Simulation

Analysis for Mining Risk Management Project.

Proceedings of the 2009 Winter SimulationConference, 978-1-4244-5771-7, IEEE Journal.

Vinnem, J.E., 1983. Quantitative Risk Analysis in thedesign of Offshore Installations. Reliability

Engineering, vol. 6, issue 1, pp.1-12, ElsevierPublishing.

Ward, S.C. and Chapman, C.B., 1991. Extending the

use of risk analysis in project management.International Journal of Project Management, 9(2),

pp.117123.

Documents

Application of Decision Theory in Assessing Marginal Oilfield Risks