SPE 108831

Building the Field Development Plan for a new gas field, located in Algeria, Reggane
Trend
Martin Vilela, Pedro Marin, Maria Rodrigo, Ana Limeres, Angel Medina, Felix Mata RepsolYPF, Madrid; Farid
Djettou, Sonatrach

Copyright 2007, Society of Petroleum Engineers

we will focus on Devonian formation, Azrafil and
This paper was prepared for presentation at the 2007 SPE Asia Pacific Oil & Gas Conference
and Exhibition held in Jakarta, Indonesia, 30 October–1 November 2007.
Reggane fields, which has been tested in 5 wells
This paper was selected for presentation by an SPE Program Committee following review of
and shown remarkable results.
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any The methodology uses an Integrated Asset
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of Management (IAM) approach, linking geological
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is with reservoir information to model hydrocarbon
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 conspicuous
flow going from the reservoir to wells, wellbore,
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.
through the pipeline network and the compressor
system up to the point of delivery based on defined
Abstract HIP, HRP and plateau; uncertainties associated
One of the most critical stages after a field with these parameters were identified and ranked
discovery is to define the Field Development Plan to be included in the FDP. As a conclusion, the
(FDP) for the new field; usually with limited FDP shows that around 30 wells must be drilled
information, managers must define the during the life cycle with a plateau of about 8.5
marketability of a prospect by taking as input the MM3/D of production gas for more than 10 years.
Hydrocarbon In Place (HIP), Hydrocarbons Rate of
Production (HRP), Number of Wells (NW) and Introduction
Drainage Location (DL), Plateau, Well Completion, During the life cycle of a Hydrocarbon Asset, when
Surfaces Facilities (SF) including pipelines and going from the Exploration Phase to the
separation/compression systems; all these Development Phase, there is a critical period where
elements have to be economically evaluated and important decisions for going or not ahead must be
finally a decision must be taken based on most taken: The Appraisal phase.
probable production scenario and economic This paper is related to the difficulties usually found
optimization. at that stage.
The available set of initial data and worked out
In the last five years the importance of figures, does not normally correspond from a
hydrocarbons gas has increased due to the qualitative and quantitative point of view, with our
increase in oil prices, the limited number of new big wishes of a very solid background with no
oil fields discoveries, the environmental uncertainties that will allow us to take solid
contamination due to oil combustion and the decisions and so acquiring strong commitments for
development of new technologies to use gas as a the future development.
competitive energy source. This situation becomes worse in those cases where
commitment for development, meaning investment,
In this paper we implement a methodology to build should be taken prior to what is commonly named
an FDP for a gas field, Reggane, located in the Basic Engineer or FEED period.
South-West of Algeria. Reggane concession is a For overcoming or at least mitigate as much as
structural trend including Azrafil, Reggane, Sali, possible the time constraints (12 months for ending
Tiouliline and Tif-Taf fields. Some of these fields the overall project), the lack of both internal human
have been tested in Devonian, Ordovician and resources and lack of fast track contractors due to
Carboniferous formations with mixed results. Here market situation, during the launching of the
2 SPE 108031

Reggane Azrafil Appraisal Project its own Project RAPORT DÉFINITIF de DECOUVERTE
MASTER DEVELOPMENT PLAN
Execution Plan was intensively discussed in order CONCEPTUAL
CONCEPTUAL
ENGINEERING
PROJECT EXECUTION PLAN
“RESERVES”
to establish the adequate work flow holding at the ENGINEERING
COST ESTIMATE COST ESTIMATES
FACILITIES
same time the quality required for being able of
taking the Development decision.
Essentially the key points so far maintained along GAP MODEL
COST ESTIMATE
the project were: PRODUCTION
ENGINEER
WELL
• Using of fast and simplified models for GATHERING COMPLETION
COST ESTIMATE
going ahead with the rest of chained
activities ECLIPSE DYN.
MODEL

• Working in parallel with more RESERVOIR

CORROSION DRILLING
sophisticated/accurate models establishing ENGINEER /MATERIALS
REPORT
ENGINEER

cross check milestones for verifying the

adequacy of the preliminary assumptions,
mainly in what is related to volumetric
Figure 3. The third branch of IAM Methodology
calculations
This methodology has proven its efficiency along
During the Appraisal phase a detailed analysis of
the project allowing us to work fast track with
all the data generated during the exploration phase
negligible lost of accuracy.
(seismic, petrophysics, well test, PVT, core, etc.)
The work flow is shown in the flow charts here in
must be performed in order to achieve three main
below and the software used is mentioned on the
objectives: define reservoir deliverability (well type,
paragraphs corresponding to each speciality.
well production, well completion, reservoir plateau,
STRUCTURAL MAPS
well location, production strategy, etc), define
TIME/DEPTH surface facilities (pipeline and trunk lines, plan
GEOPHYSICS location and capacity, etc.) and perform an
PROPERTY MAPS economic evaluation and marketing estimation
LKG, SPILL POINT
(OPEX and CAPEX, Hydrocarbon cost, etc.).
ELAN
PROBABILISTIC
GEOLOGIST ANALYSIS
P50 OGIP
Concerning this appraisal phase, the first activity
was to gather all the data acquired during the
exploration phase, which includes seismic, six (6)
wells drilled and partially tested, PVT samples,
PETROPHYSICS
PETREL FACIES
MODEL
RESERVOIR
ENGINEER
conventional and special core analysis from some
(2)
PETREL SAND/SHALE
of the wells, well test performed in three (3) of the
MODEL
wells and then prepare an integrated analysis of the
reservoir.
Figure 1. The first branch of IAM Methodology Based on the previous information the idea was to
re-build a geophysical model of the structure and fill
TYPICAL WELL
INTEGRATED MODEL that model with properties generated from an
PLATEAU
DEFINITION updated and integrated petrophysical (from the
previously generated petrophysical model built
PRODUCTION
ENGINEER.
CONCEPTUAL
during exploration phase) petrophysics model
ENGINEERING
AWARD
(using the same criteria in all the wells evaluation
and a further interpolation in the entire reservoir).
PRELIMINARY
FACILITIES ECONOMICS
AND
DRILLING In here two lines of work were performed in parallel:
ENGINEER.
from one side a well type definition-material
balance model-gathering model was performed so
PRELIMINARY COST ECLIPSE MODEL
ESTIMATES BUILD UP

allowing to define the number of wells, the

(1)
RESERVOIR
ENGINEER (3) production plateau, the pipelines and trunk lines
Figure 2. The second branch of IAM Methodology length and diameter required, the well head
pressure, etc. On the other side, a complete
dynamic reservoir simulation model was built to
SPE 108831
define wells location, forecast wells performance
and define reservoir exploitation strategy.
Using the results from the first line of work (Material
Balance), the plant definition and its dimensioning
was performed. Finally a preliminary economic
evaluation was done in order to define the viability
of the project.
Reggane Nord field is located 1250 km south-
South-West of Algiers, in Algeria. The Reggane
and Azrafil structures are situated in the Reggane
basin, to the southwest of the Sahara platform (Fig.
4). They belong to the Blocks 351c and 352c. This
appraisal project is focused on the primary
objective of gas in the Lower Devonian sands:
Geddinian, Sieginian and Emsian reservoirs.
Figure 4. Map of Algeria showing location of the
Reggane Nord block and the approximate location
of the Reggane and Azrafil Appraisal Project study
area.
Geophysics
New structural depth maps of the reservoir were
elaborated as starting point of the Appraisal
Project.
The target of a new structural interpretation on the
area was to include recently acquired data (2D
lines over Azrafil and a new well drilled in Reggane)
with the existing previous data on the area, making
a depth map of every sand layer in the Lower
Devonian reservoir.
Information Gathering
The seismic interpretation has been done under
Geoframe software. The area of interest covering
3
the two structures of Reggane and Azrafil was
about 3000 km2.
Seismic data included both 2D and 3D lines. The
3D seismic was acquired by REPSOL YPF in 2003,
covering the area of Reggane structure. The 2D
seismic laid mainly over Azrafil and it is composed
of different 2D seismic surveys ranging in age from
1972 to 2006. The data quality and appearance
among the 2D surveys are very variable. The area
benefits from part of a recent 2006 2D seismic
survey with very good quality.
There are 6 exploratory wells reaching the Lower
Devonian Formation.
Two more wells located outside the area of interest
have been used in order to tie with the seismic, and
also in all computations related to the depth
conversion to help the map definition.
All time-to-depth correspondences coming from
check shots or VSP surveys from the wells were
loaded in the project with the proper depth and time
references. The matching with the seismic was very
good in all the wells.
Seismic Interpretation
The reservoir of the project is composed of Lower
Devonian sands, which are five. From bottom to
top: Geddinian A, Geddinian B, Siegenian C,
Siegenian D and Emsian.
Three horizons were globally interpreted:
• Hercynian Unconformity (minimum
amplitude)
• Top Tournaisian (minimum amplitude)
• Top Siegenian D (maximum amplitude)
And the other two were partially interpreted:
• Top Lower Devonian (maximum
amplitude)
• Top Ordovician (minimum amplitude)
The seismic character of all Lower Devonian levels
is very weak. Top Siegenian D had stronger
reflection than Lower Devonian although the
continuity is not always present. It is associated
with a positive contrast of acoustic impedance,
which means maximum amplitude in the seismic.
4 SPE 108031
Figure 5. 3D Random Line crossing three wells in
Reggane Area (Top Siegenian D is the blue line)
There is a contrast in the data density between the
northern part, covered by the 3D seismic, and the
southern part, which has very few 2D data. The
lack of seismic in the southern part (Azrafil) also
caused troubles when interpreting faults. The fault
correlation was somewhat complicated due to the
absence of seismic lines.
Lower Devonian Depth Maps
A cell size of 100mx100m was chosen for making
the structural maps.
Among several depth conversion methods, it was
decided to use pseudo-well velocities. This method
works right when having wells at the apex and the
flanks of the structure, which is the present case.
A velocity map from the SRD (Seismic Reference
Datum) to the Top of Siegenian was produced. This
velocity map came from convergent interpolation of
the pseudo-velocities calculated at the wells, using
the stratigraphic markers (well tops) and the back
interpolated seismic times. With the velocity and
the time grids, the depth map came straightforward.
Figure 6. Top Siegenian D depth map
From the sight of the map, the following
characteristics can be shown:
• The main fault trend is compressive, though
there are also some indications of transfer
faults present in the area.
• The main fault orientation is 305º in the area
of Reggane, rotating 345º to the south, in
Azrafil.
• Two clear culminations can be seen: the
elongated ridge of Reggane, drilled by five
wells (RG) and the more rounded Azrafil
structure, by the moment only drilled with
one well at the centre, AZSE-1.
From the well stratigraphic markers for all the top of
the sands in the Lower Devonian, a table was built
in order to calculate the differences in thickness.
The well tops showed that the thicknesses of the
Lower Devonian sands were very similar. Being the
sands geometrically sub parallel, it was decided to
make the maps by adding the corresponding
average difference in meters, and then calibrating
the resulting map with the wells to honor the actual
markers.
Five more depth maps were obtained this way: Top
Emsian, Top Siegenian C, Top Geddinian B, Top
Geddinian A and Top Silurian.
The six depth maps were ready to be implemented
in the geological model.
SPE 108831
Geology
Location & Structural Style
The Reggane basin is an asymmetrical structure
with an axis oriented NW-SE and deepening north-
eastwards towards the strongly structured north-
eastern border of the basin formed by the
Infracambrian massifs of Touat and Bled El Mass
and by the Ougarta folds – anticline zone of Kahal
Tabelbala. To the east, between Bled El Mass and
Ouallene the depression is limited by the dipping of
Azzel-Matti, shown on the geological map by an
extensive zone of Devonian rock outcrops. The
southern and western border of the depression is
masked by an unconformable Meso-Cenozoic
drape (Erg-Chech-Tanezrouft). The Paleozoic
sediments constitute a series of folds with major
closures parallel to the Ougarta.
The Reggane field is located in the north-eastern
part of the basin and comprises two structures (Fig.
7). The main structure is the Reggane anticline,
which is an important asymmetrical fold with an
independent closure of 500m of vertical relief in
which the Carboniferous beds dip at 40º-60º on the
southwest flank and 20º on the northeast flank.
This structure is fault-controlled on the south-
western flank with a throw of 1300m. Towards the
south of the Reggane anticline the axis of the basin
dips considerably and takes on a meridional
direction.
Reggane
Azrafil
Figure 7. Structural depth map (3D) of Top
Siegenian_C, showing the Azrafil and Reggane
structures and the wells used for this study
A total of six wells reached the Lower Devonian
reservoirs in this zone: RG-102, RG-3, RG-4, RG-5,
RG-6 and AZSE-1.
5
The second structure in the Reggane field is the
Azrafil anticline. Sourcing is from Silurian and
Frasnian shale, the Silurian shale represents an
organic rich source rock (1-5% average TOC) and
the potential reservoirs lie above and below this
source. Good sealing of the Lower Devonian is
assured by the coverage of thick clays from Middle
and Late Devonian and the Carboniferous.
Lower Devonian reservoir characterization
In the Reggane field, the Lower Devonian series
comprise approx. 300m of predominantly clastic
sedimentary rocks, and unconformably overlies silty
shale of the Silurian. Reservoir intervals are made
up by fluvial, tidal or shoreface coarse-grained
clastic deposits, which represent the upper part of
upward coarsening and shallowing sequences. The
lower parts of these sequences are non-reservoir
intervals that are represented by offshore facies
associations and occasionally carbonates. The
Lower Devonian comprises the Geddinian,
Siegenian and Emsian stages that have been
further subdivided into informal units (Geddinian_A
and B, Siegenian_C and D, Emsian E and F) based
on sequence stratigraphic interpretations (Fig. 8).
The producing sands are the Geddinian_A,
Geddinian_B and Siegenian_C. The most important
of those are the two Geddinian sands, which
account for most of the production in the tested
wells.
6 SPE 108031

1 2 3

Swcutoff
Vclcutoff
kcutoff

Фcutoff Фcutoff Фcutoff

Figure 9. Schematic cross-plot for the synergic

identification of dynamically-conditioned cut-offs in
porosity, clay volume fraction and water saturation
from a reference limiting permeability (Kcutoff).
Ellipses represent clouds of data. Modified from
Worthington and Cosentino (2003)

It has been attempted to follow this approach in the

analysis of the cutoffs. Major difficulties arise,
however, from the lack of dynamic data (allowing
definition of a permeability cutoff) and lack of clear
relationships between PHIE and K and between
PHIE and VCL. These cross-plots together with the
sensitivity analysis and comparison to the dynamic
data, lead to determine the cut-offs applied: Phi ≥
7%, Sw ≤ 60% and Vcl ≤ 30%. Our analysis shows
that it is possible to argue for different Phi-cutoffs
ranging from 5% and 13%, depending on the
approach. Given this uncertainty and the high
sensitivity of the Hydrocarbon Pore Volume Height
in this range of Phi-cutoffs, it has been decided to
use a cutoff of 7% to the logs.
Figure 8. Correlation panel showing gamma-ray
For GIIP calculations, a 7% Phi-cutoff represents a
and zone correlations (3rd level: sands). From top to
reasonable compromise between the historical
bottom the zones represent the following sands:
cutoff of 5% and the 9% log-porosity cutoff (the
Emsian F (blue), Siegenian_D (dark red),
conservative value to predict low flow-rates),
Siegenian_C (pink), Gedinian B (yellow, orange
Fig.10.
and cream) and Geddinian_A (light pink and bright
The final step consisted of calculating the GIIP
red). The different colors in Geddinian_A and B
(Gas in Place) and it was attempted in two different
represent from top to bottom B3, B2, B1 and A2,
ways, first a deterministic way by making property
A1.
maps and multiplying grids, the second was
stochastically by working on a 3D model approach
and making 50 realizations, including facies
Five out of six wells have conventional core
modelling. Both attempts were compared and the
analysis, thus porosity and permeability
difference in volume was around 10% on average.
measurements were used to calibrate the log
curves and after the quantitative log evaluation was
made, the next step was one of the most difficult
steps in this study: establishing the cut-offs. A
generally accepted way of establishing cutoffs in an
integrated, dynamically-conditioned way is
illustrated in Figure 9 (Worthington and Cosentino,
2003).
SPE 108831 7

• Detection of wells or areas with possibilities

a) Equivalent Pore-Throat Radius (Dp) for Gedinian A and B
of production increment.
Rock Type I: Rock Type II: • Detection of bottlenecks in increments of
meso/macro porethroats
micro porethroats
production.
50
Lines of constant
45 Pore-throat size

40
dry/low flow-rates high flow-rates
• Detection of obstructions in lines and
35 damage in the productive formation.
D p(=sqrt (k/phi))

30 10 um
25
Working Methodology
20

15

10 2.0 um
5 0.5 um
0
0.2 um
0 5 10 15 20 25 30
phi (%)

b) Equivalent Pore-Throat Radius (Dp) for Sieginian C AZSE-1

RG-3
50
RG-4

RG-5

producer AZSE-1
40
producer SL-1
Dp (=sqrt(k/phi))

30 10 um
Figure 11. Implemented Methodology
20

Fig.11 shows a diagram summarizing the

10
0
0 5
Figure 10. Plot of porosity (phi) against the
equivalent pore-throat radius (Dp, =√ (k/phi)). Note
the change of dominantly low Dp values below, and
high Dp values above ~13% porosity in the
Geddinian reservoirs (a). This trend is not observed
in the Siegenian reservoir (b). Dashed horizontal
lines represent lines of constant pore-throat size.
See text for discussion.
Production and Reservoir Engineering
In this stage a methodology was implemented in
order to carry out the production forecasts for the
various blocks that make up the Reggane / Azrafil
Area.
This methodology will make possible an updated
maintenance of the following tasks:
• Verification of in-situ and remaining
volumes.
• Coordinated pressure and production
histories.
• Production potentials by well, block and
field.
2.0 um
implemented methodology and the relation
0.5 um
between the various system components.
10 15 20 25
0.2 um
30 The right margin shows NOBALPRO components,
phi (%) data base, basic information importing routines,
calculation models, preliminary simulations, and
exporting routines to the module components of the
Petroleum Experts IPM and Schlumberger Pipesim
packages.
The central part shows the BACO application
where more than 45 scenarios were evaluated with
Compositional Balance. In addition, it shows the
figures for the balance module of the selected
package, MBAL, and the PROSPER nodal analysis
module.
The left side shows the module that makes possible
the integral simulation of the production system
(GAP). The link between the latter and the balance
model provides the production forecasts as a
function of time for the various development
alternatives of each block.
Performed tasks
In this stage, the following tasks were performed:
• Generation of Data Base
• Reservoir fluid characterization
• Revision and re-interpretation of all the well
tests carry out in the existing exploratory
wells.
• Nodal Analysis
8 SPE 108031

• Matter Balance. Preliminary Forecasts. data, production controls can be consulted and
• Definition of possible scenarios. Evaluation pressure and temperature gradient calculations can
of each case. Selection of the scenario to be made for both the producing and the injecting
be developed. wells.
• Study and analysis of the surface network
design.
• Optimization of the selected case.
• Preliminary economic evaluation.

Analysis of Information – Generation of Data Base

In evaluations where working with Material Balance

is required, and even more when this methodology
is used as a production forecasting tool, it is
necessary to have information about the reservoir
pressures and fluid production arranged and
coordinated for the Balance and in addition, all the
information relating to the mechanical condition of
the well and the physical data of pipelines. In order
to characterize the performance of wells, it is also Figure 13. Nodal Analysis.
required to have the data relating to production
tests and controls.
For this project, NOBALPRO was used as a tool for Fig. 13 shows a case of nodal analysis that was
data base, nodal analysis calculation, pre- carried out for the REG 5 (Ged A) well. The upper
processing, generation of vpf tables and right corner of the window shows the icon which
exportation to the IPM (MBAL-PROSPER-GAP) allows the generation of vfp tables for Eclipse and
package selected by REPSOL-YPF for developing Mbal format, while the upper left corner shows the
the forecast model for the Reggane-Azrafil area. icons for exportation to PROSPER and PIPESIM.
Well information was broken down by producing The material balance module also includes the
formation, with data access being provided by well, option for exportation to MBAL.
producing formation, groups of wells or by selecting The data base is in mdb format (ACCESS) so
the whole system. consultations of all kinds can be made directly from
ACCESS.

Characterization of Reservoir Fluid

In order to carry out the production forecasts

through the compositional material balance, it was
used a composition of the reservoir fluid that was
calculated from the chromatography data for the
separating gas obtained from the well test and
recombined with a liquid sample calculated from an
iterative process for calculation of balance
constants. The fluid thus obtained is a wet gas
showing a very low condensation.

The chromatography used for this purpose was that

of the Geddinian_A formation. The results for
Figure 12. Nodal Analysis
production forecasts would have no sensitivity if the
Geddinian_B composition or a gas-condensate
Fig. 12 shows a view of NOBALPRO. The left side
relationship should have been taken in order to
shows the data structure tree where information
recombine from double of half the adopted value.
about wells, tubing data, casing, flowlines, fluid
There would only be small differences in the
SPE 108831
condensate volumes to be obtained, but such
volumes have not been included in the economic
evaluations.
The definition of the treatment process required to
make the output gas of the plant comply with the
specifications is where the composition to be
adopted as representative of the field fluids
acquires a prevailing value.
A study was made of the performance of the
various fluids that may be considered as
representative of the field and are detailed below:
a) Gas composition without considering the
liquid in balance
b) Gas composition considering the liquid in
balance
Revision and reinterpretation of Well Tests
A reinterpretation was made of the tests carried out
in the REG 5 (Ged A), REG 5 (Ged B) y REG 6
(Ged A) wells, in addition to tests carried out in a
Kahlouche well.
Figure. 14. REG 5 (Ged A)
From such interpretations, we could draw very
important conclusions when defining the type of
well and the required number of wells to be drilled
in order to comply with the selected development
scenario.
Nodal Analysis
In order to carry out the model representing the
fluid flow from the punched holes to the field
collectors and treatment plant, it is necessary to
characterize the performance of the wells to be
coupled to the tanks representative of each block
9
and/or any hydraulically-independent sectors which
are defined.
For each well, its IPR curve must be calculated, its
point of operation must be verified and the fvp file
intended to characterize the OUTFLOW curve from
the separation unit to the punched holes must be
prepared.
Definition of IPR Curve
In the “Nodal Analysis” of this project, the IPR
curves were calculated for all the exploratory wells
that were carried out, with potentials by layer and
total potentials being calculated in order to
dimension the completion to be used.
This IPR curve takes into account the flow provided
by each one of the layers opened in each well, see
Fig. 15. Anyhow, this multilayer analysis was only
made for the nodal analysis of wells. The Material
Balance was carried out with “a single tank” which
comprised the volumes of each producing layer in
accordance with the probabilistic analysis that was
carried out in COUGAR.
Figure 15. IPR curve for REG 5 well
Typical wells were defined by considering the
Productivity Index per layer and thus define a total
production per well.
10 SPE 108031

Study of Well-Pipeline System Performance
Figs. 16, 17,18 and 19 show the compositional
model defined for carrying out the thermal
performance analysis both from the punched holes
to the wellhead and from the wellhead to the field
collector and plant. Sensitivity studies were carried
out for various ambient temperatures and gas
flows, considering the production profile during the
life of the project. This analysis allowed the
adjustment and calibration of the wellhead
temperature in GAP.

Figs. 20 and 21 show the analysis of the pipeline

Figure 16. Compositional Model of Well
diameter sensitivity at various reservoir pressures
and various capture pressures in plant for both the
high pressure and the medium pressure scenarios
of a typical Reggane well.
It can be rapidly concluded that there is no
appreciable difference in using a 6”- and an 8”-
flowline, which is relevant when estimating the
gathering costs.

Figure 17. Elevation vs. Temperature: Ambient

Temperature 40C

Figure 20. Outflow Sensitivity: Flowline ID

Figure 18. Temperature vs. Total Distance:

Ambient Temperature 40C

Figure 21. Outflow Sensitivity: Flowline ID

(Separator Pressure: 28bar)
Figure 19. Wellhead Temperature Calibrations for
GAP
SPE 108831
Definition of multiphase flow performance in tubing
For the definition of the multiphase flow
performance in tubing, VFP (Vertical Flow
Performance) files were prepared for each one of
the wells included in the model.
The flow range used for this purpose was defined
from the IPR curve, with the AOF of each well
being the maximum flow point. A wide range of
pressures and gas-water relationships was
considered for preparing each VFP.
All of the VFP’s were prepared using NOBALPRO
and PROSPER. Various sets of files were made
using the GID(1) correlation and Petroleum Experts
2. In addition, two different types of completions
were considered, one for crest wells and the other
for flank wells.
Flow in Pipeline
The GAP model was loaded with the whole
pipelines and trunk lines between wellheads,
collectors and plant. The correlation used for the
horizontal flow was the Dukler-Eaton correlation
with Flanigan inclination correction. Anyhow these
correlations must be adjusted when the project is
put into production and real field measurements are
made available in order to adjust the performance
of the surface network.
Material Balance – Preliminary Forecasts
For this project, the Material Balance was only used
as a tool for making forecasts from the volume
defined for each tank in COUGAR, the initial
pressure of the reservoir, the fluid sampled in each
well, and the VFP files that were prepared.
BACO 6.0 Tank Model
A tank model for both Reggane and Azrafil was
developed by using the BACO 6.0 compositional
model for the following reasons:
• Easy to handle and permitting very fast
calculations, the latter being extremely
necessary when evaluating a large number
of scenarios in short periods of time.
• BACO 6.0 is fully compositional and
provides information about the composition
and flows of all the streams from the
11
reservoir to the selling gas. It is strict in the
fluid separation process and calculates the
remaining flow in the reservoir and the
condensed liquid in the field taking into
account the reinjected gas (if any).
• Since the great majority of the wells are
directly connected to the star-shaped
collectors, it is not strictly necessary to use
a surface network model if pipelines are
considered in the VFP files.
• It works with VFP files which are compatible
with eclipse 100 and 300.
• It allows the calculation of the number of
wells required in order to comply with the
defined flow control from the IPR curve of
the wells and the VFP files. This is very
useful in the beginning when starting to
outline the development plan.
• Calculation of compression power required
for the produced or injected gas streams or
per well.
Development Plan – Preliminary Forecasts
Due to the various volumes of gas involved in
Reggane and Azrafil, the distance between both
fields, the different productivity of the wells but with
similar costs and after several conceptual analyses
and studies, it was decided to use a production
strategy that involved various dates for beginning
the development for each field.
Considering that the initial pressures in both fields
are similar and close to 6000 psia, that they are two
separate blocks without a hydraulic connection, and
that the treatment plan will be located north of
Reggane, it was defined as Development Plan the
beginning of the drilling campaign and building of
facilities in Reggane, and then to begin the
development of Azrafil under high pressure, when
Reggane begins to operate at low pressure.
Obviously, the recovery factors for both fields will
be different.
Figure 22. Compositional Analysis of Gas
12 SPE 108031

Fig. 25 shows the well schedule for a case of an 8

MMsm3/d sales gas plateau. It shows the entering
dates for each well and the pressure separation
change in plant.

After this step, the simulations in the strict sense

were carried out in order to define the plateau
duration, compression scenario duration, recovery
factor, evolution of composition and required
compression power, see Figs. 26 and 27.

Figure 23. Reggane typical well. Initial conditions

Figure 25. Well entering schedule and beginning of

compression project

Figure 24. Reggane typical well. Beginning of

compression stage

Fig. 22 shows the compositional analysis that was

used in order to generate the forecasts.
In order to develop the forecasts, typical wells were
used for each field, and in addition, different VFP
files were used in order to make different
sensitivities to tubing diameter (4½ and 5½ inches)
and well depth (crest and flank), see Figs. 23 and
24.

Nearly 40 cases were evaluated considering Figure 26. Example of Reggane and Azrafil
various plateaus, calculating the number of wells development scenario
for each case, various compression scenarios were
evaluated and the entering into production of Azrafil
was synchronized in terms of time and flow.

It is very important to emphasize that all the

production forecasts were accompanied by
preliminary economic evaluations intended to
discard possible development scenarios.
SPE 108831
Figure 27. Estimation of required compression
power
1- Flowlines Analysis, Azrafil Gas Pipeline –
Reggane. GAP Model
The starting points that were adopted for preparing
the MBAL – GAP forecasting model were as
follows:
Forecast starting dates: 01/10/2011
OGIP: of the study in COUGAR
Number of blocks: 2 without hydraulic connection
Number of wells under production: 25
Number of field collectors: 3
IPR curve for Reggane and Azrafil wells
For each one of the cases, these IPR curves take
into account the flow provided by all the layers
opened in each well. Both MBAL and GAP require
the C coefficients, and n if the back pressure
equation is used to define the IPR curve for each
well.
These coefficients were loaded in the MBAL-GAP
model.
Figure 28. Mbal-Gap model as seen on the screen
of the PC
13
Flowlines data
In both the production network that was generated
in GAP and the analyses that were made in
Pipesim, lengths defined from the location of wells,
collectors and treatment plant were used.
The wells were located in the Simulation Model
(Eclipse) under the criterion for the higher GOIS
areas interconnected with good transmissibility
values. These locations were used to define the
flowline lengths and possible location on the ground
of the field manifolds and the treatment plant.
Study of Azrafil – Reggane Gas Pipeline
A study of the gas pipeline that will carry the Azrafil
production to Reggane was carried out. The said
study consisted in defining the trunkline diameter
and studying the type of flow.
The figures shown below summarize the analysis
for the initial condition of the project under high
pressure:
Fig. 29. Outflow Sensitivity: Pressure vs Total
Distance
Fig. 30. Outflow Sensitivity: PI-SS vs Total
Distance
14 SPE 108031

9 Can we estimate loss revenue from

un-optimized well?
9 Can we back compute NPV
smartness would have brought
• Intervention are always successful
9 Ability to repair / isolate zone 100%
9 Mis-diagnosis is never factored in

Cons
• Increase in well completion cost

Fig. 31. Outflow Sensitivity: Mean Slug Length

vs. Total Distance
REPSOL YPF – Preliminary SmartWell®
Completion Schematic for Independent Control of 3
The same analysis was carried out for the medium
zones thru HVC Multi-Step Interval Control Valves.
pressure scenario.
Note: Reference is made to the Reggane Well
(similar considerations apply to Azrafil)
In accordance with available information and the
studies that were carried out, no formation of slugs
is expected. On the other hand, the topography of
the ground shows favorable conditions (there are
no valleys permitting the accumulation of liquids).
Therefore the handling of liquids can be solved with
a conventional two-phase horizontal separator with
appropriate internals and multi-level control in order
to prevent the dragging of liquids to the plant. This
involves a significant decrease in capital
investment.

GAP Forecasts

All the development scenarios analyzed in BACO

6.0 and MBAL were reproduced. The consistency
of results is absolute. Once the model was
completed, sensitivities were carried out at various
Plant locations and separation pressures defined in
the conceptual study of the treatment process to be
implemented.

Well Completion Design

Due to the features of the producing formations and

conception of this project, it was chosen to use an
intelligent completion design. The reasons for this
are as follows:

Intelligent vs. Traditional approach

• Value of data and increased reservoir
understanding is never factored in
• Value of real time optimization vs. deferred
intervention never factored in Figure 32. Intelligent Completion
SPE 108831 15

Drilling Drilling hazard and mitigation recommendation:

Reggane block area wells has been designed in

accordance with experience gained while drilling
the RG-5, SLI-1 and RG-6 wells, combined with
experience acquired while drilling similar wells in
REGGANE 351C and 352C blocks.
Description of project
Basic well data are required in order to perform the
successful drilling campaign in the development
plan of the REGGANE BLOCK. These data are
required in order to perform a successful project
execution.
For development phase and with the previous
experience in the field, the proposal should contain
the following basic information:
Type of Well (Exploration, Delineation, Appraisal)
Development Wells for REGGANE BLOCK have
been designed on the basis of the experiences and
data obtained from the last drilled wells.
There will be different options of design, depending
on the depths, formation lithology and pore
Casing Design
Gas Kick Considerations:
Full evacuation criteria are applied to calculate
the safety factors for burst condition and
REPSOL YPF factor for gas kick is 100 bbls
Safety Factors (Burst, Collapse, Tension, Triaxial
Analysis)
In terms of mechanical design factors for
uniaxial casing design, as stated in REPSOL’s
Casing Design Manual for exploration, appraisal
and development wells, all casings from surface to
bottom should use the values summarized below
when making calculations for Burst, Collapse,
Tension and Triaxial Analysis.
DESIGN FACTOR
CASING/LINER

pressures profiles. SUMMARY

CONDUCTOR SURFACE INTERMEDIATE PRODUCTION

M
BURST 1.1 1.1 1.1 1.1
Twenty (25) development Wells will be drilled I
N
during 2009 and 2015 campaigns, by means of 2 I
M
COLLAPSE 1.0 1.0 1.0 1.0

rigs before first gas what is meaning a rough U

M TENSION
average of 3 wells/year/rig. (Running)
1.6 1.6 1.6 1.6
D
E TENSION
1.4 1.4 1.4 1.4
Essential well design data. S
I
(Post Running)

G COMPRESSION 1.0 1.0 1.0 1.0

N
Predicted Pore Pressure F
The pore pressure data is still not complete and will A
C TRIAXIAL 1.25 1.25 1.25 1.25
be improved while drilling operations go through; so T
O
far we have data from DST’s tests, FIT’s and mud R

weights used in the area.

The pore pressure was based on data from wells
RG-5, RG-6, KL-2 and AZF-1.
The 24” section will be drilled into the top of the
Visean; the 17 ½’’ hole section will be drilled into
the top of the Strunnian; the 12 ¼” section will
penetrate most of the Devonian and the casing is
set at the top of the Frasnian / Givetian.
The primary target (Geddinian_A and B) is
penetrated by the 8 ½” section, with casing set at
the top of the Ordovician to penetrate the Hamra
quartzite with the 6” bit.
The Corrosion Study Report determined:
The selection of materials to be used in the
Downhole Completion Equipment, Well Head and
the Gathering System for the Development of the
Reggane Nord area was determined by an external
consulting study.
According to corrosion study performed by a third-
party company experts, 7’’ liner (production) should
be super 13Cr and 9 5/8’’ production should be “C”
Grade, (Molybdenum 0.25 % weight), according to
formation gas content, pressure and temperature
conditions and expected life of wells.
 16 SPE 108031

The conventional precautions on drilling site should

be taken.

MUD TYPE AND WEIGHT BY SECTION

Interval TVD Mud

Hole (inch) Casing (inch) Mud Type
(m) Weight (sg)

0 to 400 24” 18 5/8” SPUD MUD 1.05 – 1.10

400 to INHIBITED WBM

17 1/2” 13 3/8 ” 1.10 – 1.15
1.400/1900 INVERTED OBM
1.400;1.900 to
INHIBITED WBM
2200;2600; 12 1/4” 9 5/8 ” 1.20 - 1.25
INVERTED OBM
3.400
2200; 2600;
8 1/2” Liner 7 ” INHIBITED WBM 1.32 - 1.34
3.400 to 4.100

8 ½’’ Section (Production Zone)

This section will be drilled efficiently with inhibited

Polymer System. Sodium chloride is added to help
reduce the content of solids and allow the use of
acid-soluble Calcium Carbonate as weighting
agent.
Given the results of shale analysis tests on the
previous wells, highly reactive shale is not expected
and the onus will be on fluids loss and dispersion
control.
Well Type Configuration
Figure 33. Reggane Type 1; Crest (TVD
3250m)
Figure 34. Azrafil Type 1 (TVD 4100m)
Dynamic Modeling
A geocellular model was built in PETREL
Software to describe the reservoir; this geocellular
model uses sand/shale characteristics, due to the
fact that limited information is available (instead of
using a facies model). This model has 1998920
cells. The model was exported into the simulation
software, ECLIPSE. Inside the simulation software,
the grid was edited to remove all the cells that are
located beyond our area of interest. As a result the
grid dimension was reduced to 337215 cells, which
is a model that can be dynamically modeled
directly, without need to do upscaling.
In the next plots, the properties distributions are
presented for different layers, corresponding to the
top of each sand of interest: Siegenian_C,
Geddinian_B y Geddinian_A; also there are
presented two transversal sections corresponding
to north-west to south-east
The model was initialized with a Fetckovich
aquifer, which allows locating the gas down to (for
simulation proposes the gas-water contact) for the
two extreme cases: the Low Known Gas case and
the Spill Point case.
The PVT model used was generated from three
samples taken one in REG-5 at Formation
Geddinian_A and two samples of Geddinian_B,
one taken at REG-5 and the other in REG-6.
Figures 35 to 36 show the more relevant results.
SPE 108831 17

1,5
around 15% connate water saturation as indicated
1,4
by the lab result.
Fig. 37 shows the relative permeability curves
1,3 used for Geddinian_A sand.
Z factor

1,2

1,1

0,9
0 1000 2000 3000 4000 5000 6000 7000

Pressure (psia)

Geddinian_A Relative Permeability Curve

Figure 35. Z factor for Reggane gas 1,00

0,90
0,80
0,70

Rel. Perm
0,60
Gas
0,50
0,40 Water
1,4
0,30
0,20
1,2
0,10
0,00
1 0,00 0,20 0,40 0,60 0,80 1,00
Water Saturation (frac)
B g (cf/scf)

0,8

Figure 37. Relative Permeability Model.

0,6

0,4

According with the material balance model

proposed, at the beginning ten (10) wells will be
0,2

0
0 1000 2000 3000 4000 5000 6000 7000
drilled in Reggane to start production with a plateau
Pressure (psia) of 8.5 MMSCF/D; to keep that plateau six (6)
additional wells will be drilled in the next two years;
Figure 36. Bg factor for Reggane gas Azrafil will start at the end of drilling in Reggane
with 9 wells and compression will be beginning two
years later.
According to the thermodynamics properties and The dynamic model was built according to the
gas composition this is a dry gas with a 4 % of constraint already mention; in the next figures, the
CO2. simulation results are presented, showing gas rate
About the Relative Permeability the only Special production and pressure decline for the life of the
Core Analysis performed en Reggane-Azrafil area, project.
at Devonian level correspond to REG-6 and for this
well only one plug located at Geddinian_B
Formation; for that sample, the only data available
are the end-point saturation, so using correlations,
the corresponding curves were generated. The
biggest uncertainty in this point is the
representative this small plug could be of our
reservoir, more taking into account not only the
different in between one sand but also comparing
one sand with the other: from fluvial amalgamated
sand to upper-shoreface to tidal fluvial the
depositional environment and the rock quality are Figure 38. Porosity at the top of Siegenian_C
too different to use only one model of rock tipe for
this reservoir. So, in our rock model we use three
different curves for each sand, all of them having
18 SPE 108031

Figure 42.. Top of Geddinian_B

Figure 39. Porosity at the top of Geddinian_B

Figure 43. Top of Geddinian_A

Figure 40. Porosity at the top of Geddinian_A

Figure 41. Top of Siegenian_C Figure 44. Initial Pressure at the top of
Siegenian_C
SPE 108831 19

• Conceptual Engineering of Surface

Facilities

The Scope of Work of this contract has

consisted essentially of:
¾ Process scheme selection among the
different alternatives.
¾ Conceptual Design development of the
chosen scheme
¾ cost estimate of the Facilities
Figure 45. Wells in the simulation Model
For CO2 removal an Amine Unit was selected
One way to check all the process construction of as the most adequate option. Regarding water and
the field model is to review that the volumetric HC dew point (50ppmv and -6ºC cricondentherm
calculation at the different stage is approximately respectively) vs. more commonly used technologies
the same. That was done going from the 2D Temperature Swing Adsorption was selected and
geological model to the 3d geological model to the fitted perfectly with the lean gas characteristics of
dynamic model. The variation between each case the wellhead fluid.
was control to be no more than 5 percent. Reasons supporting such a selection are
Finally in Fig. 46 it is shown the predicted gas ¾ Flexibility of the process for feed gas
production rate for the field, according to the compositional changes
dynamic simulation model, built in ECLIPSE. ¾ Reliability and turndown capabilities
covering all the expected profile for the full
life cycle of the project
¾ Reduced Capex and Opex vs. the other
Reggane/Azrafil Nord Field
more conventional approaches (mechanical
10 refrigeration and turbo expansion)
8
FGPR(M3M/D)

6 Conclusions
4 1. The methodology used explain in details,
2
how to perform a reservoir study when the
0
01-11-11 23-04-17 14-10-22 05-04-28 26-09-33 19-03-39 08-09-44 data available is very limited.
Date
2. Although the study has several
uncertainties, those uncertainties can only
Figure 46. Filed Gas Production rate
be reduced with the acquisition of more and
accurate data; in this study, the objective
External Contracts
was to use in the more efficient and
As indicated at Fig. 3 (boxes in yellow) two
profitable way the data available.
mean external contracts have been awarded
• Corrosion and Material Study 3. A complete filed development plan was
Due to the medium CO2 content as well as H2S generated that show how to produce the
presence (traces), corrosion and material selection
became a relevant fact for considering during the field, the rate of production, plateau, the
Project Execution Plan for Development, as well as drilling plan and the facilities required to
for cost estimates purposes at this stage.
Fluid characteristics will oblige using in some reach the production target and specs.
specific cases CRA (well completion) and will allow 4. The scenarios generated with the two
in others the use of CS under strict rules for
corrosion monitoring as well as inhibitor injection different approaches produce the same
20 SPE 108031

results, so stressing our confident in the 5. Bilderbeck, M. and Beck, I.: “Using
Integrated Project Models to Evaluate Field
methodology used.
Development Options”, SPR 93554, 14th
5. In this stage, once solved all the marketing SPE Middle East Oil & Gas Show and
Conference, Manama, Bahrain.
issue, economic evaluation and contract
6. Tianlu, T., and Stein, M.,: “Evaluating
agreement, the development phase can be Operation Strategies via Integrated Asset
Modeling”, SPE 75525, 2002 SPE Gas
started over a relatively solid base.
Technology Symposium, Calgary, Canada.
7. McClay, K., “Advanced Structural Geology
Acknowledgement for Petroleum Exploration”, University of
London, 2003.
Authors wish to express their appreciation to the 8. Carter, M., Siraki, E. and Coelho, D.:
management of RepsolYPF for permission to “Velocity Interpretation and Depth
publish this paper. Thanks are also expressed to Conversion”, IHRDC, 1993, Boston, USA.
Consuelo Moya, Poltak Parulian, Ruben Rieu, 9. Worthington and Cosentino,: “The Role of
Antonio Bencic and Juan Rodriguez for their Cut-Off in Integrated Reservoir Studies”,
collaboration in different phases of this study. Lastly SPE 84387, 2003.
we would like to express our gratitude to our 10. Canel, C, Savoy, E. & Mondoni, L.
partners Sonatrach, RWE and Edison for their “Aplicación de modelos híbridos y desarrollo
support. de una nueva correlación para el cálculo de
gradiente de presión en pozos de gas y
Reference condensado”, 1º Congreso Latinoamericano
de Hidrocarburo, Buenos Aires, Mayo 1974.
1. Guyaguler, B. and Ghorayeb, K.: “Integrated
Optimization of Field Development,
Planning and Operation”, SPE 102557,
2006 Annual Technical Conference and
Exhibition, San Antonio, Texas, USA.
2. Al-Marhoon, J., Clark, G. and Babajan, S.:
“Integration of Surface and Subsurface
Models of Commingled Production of
Different Reservoir Fluids – Case Study:
Gas Condensate Field, Oman”, SPR
101290, 2006 Abu Dhabi International
Petroleum Exhibition and Conference, Abu
Dhabi, U.A.E.
3. Nadar, M., Kulkarni, R et al., :”Development
of an Integrated Production Network Model
for the Heera Field”, SPE 101089, 2006
Asia pacific Oil & Gas Conference and
Exhibition, Adelaide, Australia.
4. Stevenson, B. and O’Shea, C., : “Case
Study: modeling of a Large-Scale Tight-
Gas-Gathering System”, SPE 100433, SPE
Gasechnology Symposium, Calgary,
Canada.