SPE 94093

Integrated Modeling of the El Furrial Field Asset Applying Risk and Uncertainty
Analysis for the Decision Making
L.M. Acosta, J. Jiménez, A. Guedez, E.A. Ledezma, J.A. Bello, A.J. Millán, M. Guzman, and E. Marin, PDVSA;
F.J. Gómez, and I. Herrera, LandMark; and P. Córdoba, Beicip Franlap

Copyright 2005, Society of Petroleum Engineers

This field has two producing formations: Naricual and Los
This paper was prepared for presentation at the SPE Europec/EAGE Annual Conference held Jabillos Formations. One important characteristic of the
in Madrid, Spain, 13-16 June 2005.
reservoirs is that they are volumetric, and produce by rock and
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
fluid expansion.
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to At the beginning, the field produced naturally; however, high-
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the SPE, their officers, or members. Electronic reproduction, distribution, or storage pressure decline observed by the end of 1992 justified a water
of any part of this paper for commercial purposes without the written consent of the Society of injection process to be implanted in Naricual Formation, and
Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract
of not more than 300 words; illustrations may not be copied. The abstract must contain in 1997 in Los Jabillos Formation. Subsequent reservoir
conspicuous acknowledgment of where and by whom the paper was presented. Write
Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435. studies led to identify opportunities to increase the recovery
factor through gas injection. Therefore, in 1998 a miscible gas
injection process was implemented in Naricual. Currently, a
Abstract
combined process of water and miscible gas injection1
The El Furrial field is one most important hydrocarbon giants
supports the exploitation plan of the field, maintaining a
of Venezuela. It’s located at the north of the Monagas State
Production Plateau of 400.000 BOPD.
and it handles approximated 45 % of the production of this
A huge advance in the reservoir characterization has impulsed
area. It was discovered in 1986 and its characteristic makes it
the evaluation of new development scenarios in order to
unique. The strategy of development has been supported by
increase the recovery factor of the field by means of a new
the combined injection of miscible gas and water; this has
development plan.
allowed maintaining the production plateau of 400 MBPD of
Due to the complexity and size of the actors involved in the
28 ° API oil. A transcendental advance in the characterization
implementation of the new exploitation plan, the development
of the field was done with a new dynamic model called V9,
plan was selected following an innovative methodology
what promoted the evaluation of new development scenario.
(Integrated Asset Modeling) in order to reduce decision time
and better take account for associated risks.
This paper presents one of the first integrated development
plan built by PDVSA; focus on the Integrated Modeling of the
Definition and Philosophy of the Process.
Asset, considering risk and uncertainty analysis in the decision
The Integrated Asset Modeling (IAM), is a way of working
taking. The project was developed under a philosophy of
that integrates all the core disciplines of the asset in order to
collaborative work and changing the way of thinking from
generate and/or update the Integrated Exploitation Plan from
deterministic to probabilistic. The Project was executed by a
the business point of view (Figure 2). By means of IAM,
multidisciplinary team integrated by personnel of the core
professionals working in different activities are temporally
areas of the asset, creating a milestone in the traditional way
detached, in order to conform an extended multidisciplinary
of working. This integration allowed evaluation through
work team. The day to day work of IAM is constantly
statistic and expert interview, the uncertainties associated to
enriched by experience exchanges between professionals
the variables involved in the evaluated scenarios. Huge
belonging to different technical and administrative
reduction of capital expenses related with the grow of facilities
departments. This is very helpful when it comes to identify
were obtain due to the estimation of production profiles based
and quantify risk and uncertainties associated to every factor
on risk and uncertainty analysis. As a result of this analysis a
involved.
decision matrix was built, with the recovery factor,
The use of this methodology allowed the reduction of
investment, qualitative risk factor, and economic indexes,
technical analysis time, focusing only on the key factors.
which were evaluated to decide the optimum development
Additionally, as people from different section worked togeter
plan.
in order to select the optimum exploitation plan,
implementation time is considerably reduced.
Introduction
The El Furrial Field is located in the Monagas State, at 25 Km
of Maturin City (Figure 1), at the area corresponding to the
The process is basically divided in three major steps
West Venezuela’s Basin (Sub Basin of Maturin). It was
described as follows.
discovered in February 1986, by the FUL 1X well,
representing at the time, the most important discovery made in
Venezuela and one of the largest worldwide.
2 SPE 94093
Brainstorming and Scenario Hierarchy
As a first step of the methodology a brainstorm is performed,
in order to identify the entire factors and/or activities related to
the exploitation plan. The interdepartmental multidisciplinary
team allows the consideration of a huge set of factors and
variables to be considered.
From the brainstorming, a set of possible exploitation
scenarios is created. These scenarios have to be value adding,
quantifiable and feasible to implement.
Quantification
This consists in the technical evaluation of each one of the
scenarios defined during the brainstorm. During this phase, all
the technical work that could generate data to stochastic
evaluation is performed, including risk and uncertainty in each
parameter involved.
Economic Evaluation
During this phase the profitability of each one of the scenarios
is estimated. Global risk for each scenario is also determined.
IAM Methodology Application for El Furrial Field
In this section the methodology application is described in
detail for El Furrial Field. The interdepartmental
multidisciplinary team was composed by 11 professionals
working togeter for a period of four months.
El Furrial Field Brainstorming and Scenario Hierarchy
At the beginning of the Brainstorming a discussion took place,
in order to identify the most important factors affecting the
exploitation plan choice. As this on shore field contains
significant volumes of oil reserves, it was recognized that
recovery process was the key factor affecting economics.
The analysis was then focused over the technical decision
of studying the secondary and/or EOR process for the El
Furrial Field.
Several business cases were built considering different
injection processes. These cases included the short-term
optimization of the existing injection processes and the
subsequent implementation of a new EOR process. The
following processes were studied:
• Crestal Gas Injection/Flank Water Injection (Present
Process)
• Crestal Gas Injection/Flank Dewatering
• Flank Water Injection / Non-Crestal Injection
• Flank WAG Injection / Crestal WAG Injection
• Flank WAG Injection / Non-Crestal Injection
• Crestal WAG Injection / Water Flank Injection
• Crestal and Flank WAG Injection
Additionally, a reference case was defined for comparison
purposes.
Furrial Field Quantification
Production Profile
In order to create the production profiles, the latest simulator
model was used2. From this model for all the cases the oil
production, water and gas profiles were obtained, as well as
the requirements of injection volumes.
Risk and Uncertainty analysis
The uncertainties related with a variable can be quantified
through a process that covers: variable identification,
uncertainty quantification, and finally uncertainty
management, where they can be reduced by caption of
additional information and studies.
The variables related to the subsurface and subjected to
uncertainty management are:
Original Oil In Place (OOIP), Recovery Factor, oil, gas and
water production rates.
Probabilistic OOIP
The probabilistic OOIP was used to evaluate the reliability of
the simulator model that was used for generation of the
production profile for each scenario. To quantify it the
equation of the volumetric calculation was used (Figure 3).
Once gathered all the required data, an initial OOIP of each
reservoir was calculated, doing afterwards the calculation of
the OOIP for the entire field. The difference between the
OOIP probabilistic (P50) and the one estimated by the
reservoir model was of 2%, meaning that the simulation model
was reliable enough to be used in the study.
Risk Analysis of the Production Profile
Usually, real production differs from simulation model
forecasts. This creates an uncertainty that has a huge impact in
the facilities requirements, translating this into economic
impact (Oversizing, low profitability, etc). In this sense and
with the purpose to reduce the uncertainty of the production
profiles (oil, gas and water), the difference between the
production forecast and the real field production was
calculated. This calculation was used to quantify the deviation
of the forecast for those scenarios including water injection,
water alternating gas (WAG) and dewatering process.
Risk Analysis for the Process of Gas and Water Injection
To make this analysis, field production information was
compared with those obtained from simulator forecast. This
was possible because crestal gas injection combined with
flank water injection is the current recovery process.
Subsequently, statistic was used to evaluate the deviation
between real and simulate profile (Figure 4).
Risk Analysis for the WAG Process
As WAG was never implemented in the field, it was not
possible to measure the deviation between estimated and real
production. Actually there is a WAG pilot project, and the
results are not conclusive by the time being, for this reason it
was necessary to use worldwide experience in order to
estimate the forecast deviation of WAG process3.
The risk analysis was based on the field results of WAG
projects that have been finalized. An analysis of published
SPE 94093
results2 show significant difference between forecast and real
field production.
The statistic analysis applied to the published data made it
possible to quantify roughly the uncertainty related with the
modeling of this kind of process (Figure 5). It was observed that
most of WAG forecasts are significantly optimistic, so that, in
the present study, WAG forecast where affected by a high
degree of uncertainties.
Risk Analysis of Dewatering Process
Simulation results showed that the success of this case
depends strongly on the ability to extract previously injected
water from the flanks, as gas is injected in the crest. Also, the
performance of this dewatering with crestal gas injection,
depends on the ability to maintain the integrity of the gas cap,
reducing gas channeling. Therefore, some extra simulation job
was required in order to evaluate different water extraction
conditions and different crestal injection patterns. The
sensibility shown by simulation results allowed to estimate
statistically the uncertainties associated to this scenario.
Recovery Factor Deviation
The risk analysis allowed associating uncertainties to
deterministic recovery factor forecasts (Figure 6). Results show
that conventional processes (Water or Gas Injection) have a
low risk, and results are often as expected. However, non-
conventional processes such as WAG or Dewatering have a
high risk associated. The analysis shows that real behavior for
non-conventional processes can be not only far from forecasts,
but also it can even be poorer than conventional recovery
processes.
New Wells Location
New wells location were determined based on the following
data:
• Present simulated state of the field (Saturation Map /
Flow Capacity, etc).
• Environmental and legal aspects.
• Basic Cartography4.
• Existing Facilities.
The same set of wells was used for all the scenarios, only
the well planning program was adapted to account for
different fluid movements in each case.
Probabilistic Drilling Cost
The methodology of calculating the drilling cost based on risk
and uncertainty analysis includes all the variables related to
unexpected operational events, engineering improvements,
geologic uncertainties, etc. For this analysis all the variables
that affect the cost of drilling were identified and quantified.
One matrix was then created considering statistical data for
time and cost of drilling by phases and well completion. This
database allowed calculating the cost of each one of these
variables as the total cost of drilling the well5.
3
Workovers
Workovers were considered in the economical evaluation. The
total number of workover was estimated considering:
• Workover requirement for maximum allowed GOR, water
cut, and minimal well head pressure (from reservoir
model).
• Workover requirement for mechanical problem,
estimulations, well cleaning, etc. (from field statistics).
Uncertainties associated to workovers were estimated by
measuring deviation between past planified and real data.
Facilities Evaluation
In order to identify new facilities requirements, production
forecast, existing capacities and associated uncertainties were
analyzed. The statistical analysis over the simulation forecast
showed that simulated results overestimate to some extent
water production. This deviation was considered, as it has a
high impact in capital investments related to new facilities
construction (Figure 7).
As result of the analysis upgrading requirements for
facilities were estimated. Uncertainties associated to facilities
CAPEX requirements were determined by comparing past
planified and real data, taken into account several projects
from the company database.
Economic Evaluation of the El Furrial Field
To execute the economic evaluation, all the information
previously gathered, such as production profiles, capital
investments and costs were included. This information was
ranged using the economic software6 as a triangular function
P10, P50 and P90. In this way, risk was included in the
economic evaluation.
From this information the stochastic economic evaluation
was performed, with 500 realizations. As a result, the
probabilistic economic evaluation was obtained. The P50
value was set as the most probable value, and the team
decided to use the standard deviation as the risk of not
achieving P50, and considering P10 as the minimum value
that could be achieved
It is important to mention that the economic evaluation was
done to all the studied scenarios, with the original production
profiles from the reservoir model, and the values adjusted by
the deviation determined for each case. In this way, technical
uncertainties were translated into economical uncertainties.
In order to identify the impact of the economic risk, an
efficiency frontier graph was generated for all the cases (Figure
8).
Decision Matrix
The most important result from this is the decision matrix that
consolidated all the key performance indicators as a result of a
multidisciplinary team analysis. This matrix helped view and
select the optimum exploitation scenario, and was built by,
Probabilistic NPV, ROR, Risk Factor (standard deviation),
Investment, Recovery Factor and the social development
Factor.
4 SPE 94093

It is important to mention that the dynamics of the

company allows some factors to vary with time. This is why it CARITO
FURRIAL
is so important that in this matrix all those factors are PIRITAL
considered in order to support the final decision, assigning a
MATURIN
weight to each variable depending on the needs of the

0
-1400
-15000

-14000
-14000

-15000

company.
N CARIBBEAN SEA
NORTH
Conclusions AMERICA MONAGAS
• A multidepartamental temporal team showed to be very
efficient in order to define and project the future
development of the field. COLOMBIA
• The multidepartmental team allowed a broad analysis
SOUTH
over the problem, and a very specific treatment of AMERICA
uncertainties. BRAZIL
• The implemented methodology allowed reducing time for
technical analysis, generating a high impact in the
strategy and vision of the business. Figure 1. El Furrial Field Location
• The probabilistic analysis showed that even though the
EOR methods offers a higher recovery factor, this kind of
process has associated a higher level of uncertainty. Reservoir
• The estimation of uncertainty in the production forecast
reduces capital investments, by minimizing facilities
oversizing risks. Economic
• The decision matrix became a support for the hierarchy of Evaluation
the business cases that may be implemented for a better Production
future field development.
Risk
Acknowledgment Analysis
Special thanks are offered to PDVSA INTEVEP for its
valuable collaboration in accomplishment this project. Surface
Facilities

References
1. Echeverria, I., Auxiette, G.: ”Reservoir Management of El Drilling/Completion
Furrial Field-Venezuela under Enhanced Oil Recovery
Process”, SPE 75201, 2002.
2. ATHOS User’Guides, Release 4.3 , 2004, IFP. Figure 2. The Integrated Asset Modeling (IAM) integrates all the
core disciplines of the asset in order to generate the Integrated
3. Christensen, J.R., Stenby, E.H., Skauge, A.:”Review of Exploitation Plan
WAG Field Experience”, SPE 71203, 2001.
4. OpenExplorer, Release 1998.8, 2000, LandMark
Graphics.
5. CRYSTAL BALL 2000 professional Edition, v5.2, Area

Decisioneering. Inc.
6. TERAS Evaluation Module Reference Guide, Volume 1,
2000, Landmark Graphics. Arena Neta

Glossary
CGI/FWI = Crestal Gas Injection/Flank Water Injection OOIP = 7758*Área*ANP*Φ*(1-Sw) Porosidad

CGI/Fdewatering = Crestal Gas Injection/Flank Dewatering Boi

FWI = Flank Water Injection / Non-Crestal Injection OOIP

Sat. Agua
FWAGI/CWAGI = Flank WAG Injection / Crestal WAG
Injection
FWAGI= Flank WAG Injection / Non-Crestal Injection
Factor Volumétrico
CWAGI= Crestal WAG Injection / Flank Water Injection Boi

CWAGI/FWAGI=Crestal and Flank WAG Injection

Figure 3. Equation of the volumetric calculation for Probabilistic
OOIP calculs.
SPE 94093 5

600
Without Deviation With Deviation
500
350 350
More and Early
Qg (MMCFD)

400 Investmant
300 300 Less and Later

Qw (MBWPD)
300 Investmant
250 250

200 200 200

100 150 150

0 100 100
1/22
1/15

1/29

2/19
2/26

3/25

4/15
4/22
2/12

3/11
3/18

4/29
1/8

2/5

3/4

4/8
4/1
1/1

50 50

Time / weeks 0 0
2004 2007 2010 2013 2016 2019 2022 2004 2007 2010 2013 2016 2019 2022
Real Plan
Time / Years Time / Years

Figure 4. Real and Predictive Gas production rate versus time

with estatistical analysis for a same possible exploitation Figure 7. Water production rate versus time for a exploitation
scenario. scenario with and without deviation. Note the huges impact in
investments when the deviation is considered.

CWAGI Original
160 7 5 1000
00
CWAGI Modified
140 Standard Deviation

120 7 0 0750
0

NPV (MM$)
Qo MBPD

100
44% 23% 6 5 0500
0
80

60 23% 6 0 0250
0
40 With Uncertainties
Without Uncertainties
20 5 5 0 00
225
00 2 30
20 2 35
40 240
60 245
80 350
00
0
2004 2008 2012 2016 2020 2024 R is k (s td . D e v .) M M $
FWAGI/CWAGI CWAGI/FWAGI CWAGI FWAGI
Time / Years

Figure 8. Net Present Value (Referential) versus risk for somes

Figure 5. Oil production rate versus time for a WAG scenario with scenarios. Note the huges economic impact to consider the risk.
and without deviation. This procedure was applied of all the
scenarios considered with WAG processs.

7 0

6 0
Recovery Factor (%)

5 0

4 0

3 0

2 0

1 0

0 C a so B a se C a so C a so G a s A G A F la n c o A G A C r e s ta A G A F la n c o F U L L A G A F u ll G a s
BASE CGI
O p tim iz a d o CGI5 7 0 FWAGI
g a s c r e s ta CWAGI FWAGI
A g u a FWAGI CGI/F F u FWI
ll A G U A

F la n c o s
FWI FWI CGI FWI CWAGI dewatering
Deterministic value Probabilistic value

Figure 6. Recovery factor with and without deviation for all

exploitation scenarios. Note that the non-conventional recovery
factors processes can be significally higher, however, the
uncertainties related to the processes create a high risk on them.