Sunteți pe pagina 1din 18

SPE 131582

Condensate Banking Phenomenon Evaluation in Heterogeneous Low


Permeability Reservoirs
D. Giamminonni, G. Fanello, M. Kfoury, I. Colombo, A. Bonzani, eni e&p.

Copyright 2010, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE EUROPEC/EAGE Annual Conference and Exhibition held in Barcelona, Spain, 1417 June 2010.

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 have not been reviewed
by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or
members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of 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 conspicuous acknowledgment of SPE copyright.

Abstract

In low and tight gas formations, condensate banking will form in shortly time after production start-up due to pressure drop below
the saturation pressure. Mobility reduction near wellbore area will affect well productivity. The prediction of gas condensate wells
production will strongly depend on oil banking evaluation and modeling.

A benchmark radial fine well model has been built using constant petrophysical properties per each layer. Several coarse Cartesian
grids have been considered to evaluate discrepancies in terms of production and flowing pressure with respect to the benchmark
grid. For a coarser Cartesian grid, it has been deduced that Generalized Pseudo-Pressure (GPP) is a key parameter to avoid well
performance over-estimation. An alternative solution consists in defining a local grid refinement (LGR) near wellbore to honour
the benchmark solution without using GPP.

In this case study a LGR technique has been used to incorporate future hydraulic fractures for the wells development. A real
application case has been considered to extend lessons learned from benchmark to field scale. A proper geological model has been
built using a sedimentological model as driver for petrophysical properties distribution.

Two DST have been considered to analyze condensate banking phenomena evaluation in a low and medium permeability matrix.
To this purpose, three analytical models have been considered. Thus, to validate a representative analytical model a numeric
simulation has been performed. Based on the obtained results, it can be affirmed that the radial composite is the most appropriate
analytical model reproducing the phenomenon of gas mobility reduction in the nearest wellbore region.

Introduction

Condensate banking is the build up of liquid around the production wellbore occurring when the pressure falls below the dew point
and it is a critical issue in low permeability scenarios.
This paper focuses on the use of 3D dynamic model to physically describe this phenomenon and to better understand the real
deliverabilty in terms of gas and condensate production.
Our study has been applied on real reservoir layers which are gas condensate bearing and whose permeability varies in a range
between 0.01 and 50 mD. The overall workflow of the study started setting-up a benchmark synthetic radial model, that, due to
cells refinement, allows to correctly describe the physics of the banking build-up around the wellbore and to asses the impact on
production performance.
Since reservoirs are generally modeled using a Cartesian geometry, several coarse Cartesian grids have been then considered to
evaluate the discrepancies in terms of production and flowing pressure with respect to the benchmark grid.
As a central matter of this study we will show how the Generalized Pseudo-Pressure algorithm or an appropriate local grid
refinement (LGR) near the wellbore are mandatory in order to reproduce the benchmark production profile and to avoid well
performance over-estimation.
2 SPE 131582

Moreover three different analytical models have been used to match production data from DST on two different wells, well-2 and
well-4. A proper numerical simulation has been performed to select the best analitycal model, which better reproduces the gas
mobility reduction in the near wellbore region.
As it will be explained later on, radial composite is the most suitable analytical model to face this issue.

Condensate banking phenomenon Generalized Pseudo-Pressure

Gas-condensate is a gaseous hydrocarbon mixture from which liquid condenses in the reservoir when the pressure falls below the
dew point. Nevertheless gas-condensate production consists predominately of gas from which more or less liquid drops out at the
surface separators, hence the name gas-condensate. Gas-condensate reservoirs generally produce gas with associated light-
coloured stock tank liquid (gravity above 45API) at Gas Oil Ratio (GOR) in the range of 3000 to 100000 SCF/bbl.
It is worthwhile noticing that gas condensate is intermediate between volatile oil and dry gas. Its typical composition is
characterized by a content of C7+ between 0.5 to 12% [1].
A physical characteristic of the condensate is that usually it is not a liquid at reservoir conditions. In most cases the hydrocarbon
mixture exists at reservoir conditions as a single gas phase or two phases (gas plus condensate, if the pressure is below dew-point).
Such a phase transformation may take place within the reservoir due to pressure depletion, through a process of isothermal
retrograde condensation. The liquid developed within the reservoir generally remains entrapped and provides a negligible
contribution to the liquid hydrocarbon stream produced at surface. When the pressure of a producing reservoir falls below the dew
point, CGR starts reducing (a percentage of liquid remains in the reservoir as liquid phase) and condensate starts to be formed
within the reservoir (dropping out from gas) until a maximum value. The maximum saturation of liquid condensing in the
reservoir, when pressure falls below the dew point, is generally small and frequently it is below the critical value (minimum
saturation above which the liquid becomes movable). Therefore the liquid hydrocarbon deposited within the reservoir (retrograde
condensate) is not produced. This represents a huge loss in hydrocarbon recovery.
It may be imagined that a further pressure drop would lead to re-vaporization of the liquid. This hardly occurs because once the
reservoir pressure falls below dew point, the overall molar weight of the hydrocarbons mixture increases, since some of the
heaviest paraffins remain in the reservoir as retrograde condensate (whilst the lighter components are produced as gas). Therefore,
the phase envelope for the reservoir fluids changes strongly, inhibiting re-vaporization.
The liquid build-up around production wellbore occurring when the pressure falls below dew point is known as condensate
banking (CB) which is usually a remarkable issue in gas condensate reservoirs with low to moderate permeability (<10-50 mD).
Wells with high KH (permeability x thickness > 1.53,000 mD*m) are typically not affected by condensate blockage since they
produce at small reservoir pressure drop and deliverability is constrained almost entirely by tubing. As far as condensate banking is
concerned, three regions can be distinguished around the well focusing on gas-condensate flow and saturation distribution:

Region 1: Near the well (0-10 meters during reservoir life). Here gas and condensate simultaneously flow at different
velocities. The single phase gas entering Region 1 has the same composition as the producing well mixture.
Region 2: Quite near the well (5-100 meters during reservoir life). It defines a region of condensate accumulation. Only
gas flows in this region because oil saturation is below the critical value. Region 2 shifts from well during time.
Region 3: Far from the well. Hydrocarbon is one-phase only. Gas preserves its original composition because pressure is
above dew point [2,3,4].

The CB may appreciably reduce gas well deliverability, though the severity depends on reservoir and well parameters. Condensate
blockage can be important if the pressure drop between the reservoir and the wellbore is a significant percentage of the total
pressure drop from reservoir to delivery point (e.g. a surface separator). Its effect can reduce the well potential between 0% and
50%[5].
In terms of well performance, the near-well flow behaviour, determined by the near-well relative permeability functions, is the
dominant factor and Region 1 is the main source of deliverability loss in a gas-condensate well. Gas relative permeability is
reduced owing to condensate build-up. Region 2 has somewhat reduced gas relative permeability, but this generally has a second-
order effect on well performance (Figure 1).
The Generalized Pseudopressure (GPP) approach allows an accurate description of the well inflow in a coarse grid compared to
the traditional approach (pseudopressure) and it is useful to model the condensate banking phenomenon. Numerical simulators
approximate the pressure in each grid block with an average value (pressure at Peaceman radius). In a coarse grid block containing
a producer well, the average block pressure could be higher than the dew point pressure even if the well flowing pressure is below
the dew point pressure. In this case the numerical model will predict no condensate drop out in that specific grid block
SPE 131582 3

underestimating the formation of condensate bank in the reservoir. As a result the surface condensate production can be largely
overestimated. The GPP approach is schematically presented in Figure 2.
For a detailed theoretical explanation you should see the references [6]. Alternative approach is the use of a LGR in the near
wellbore cell.

Synthetical model approach

The condensate blockage is clearly an issue limited to the well scale. Its extension rarely reaches hundred of meters from the
wellbore. Pressure and liquid accumulation strongly change near the well during time [7]. For this reason radial fine grids can
describe correctly the pressure and the fluid saturation profiles around the wellbore while a full field scale Cartesian grid is not
aimed to accurately reproduce it.
At this purpose a fine radial model has been built with the compositional reservoir simulator Eclipse 300 in order to describe the
condensate blockage and to study its effect on well productivity. This is our benchmark case. The radial grid was built using 30
gridblocks (N) with regular spacing in ln(r), according to Eq.1
1
ri r N
= e Eq.1
ri 1 rw

in which re = 563.8 m and rw = 0.108 m. The inner block has a width of 0.06 m. The petrophysical properties are always kept
constant areally, with a vertical distribution only. The average permeability, considered in this sector study, is around 3 mD.
Since we have generally to deal with Cartesian geometry, especially in full field model simulation, it is necessary to built an
appropriate Cartesian grid and compare it to the radial model in terms of productivity and gas/oil cumulative production, after
checking that the two sector models have the same OOIP and OGIP (Figure 3).
In particular, in our study, the use of a Cartesian grid is justified by the necessity to introduce hydraulically fractured wells in the
full field development plan. In this case the flow pattern in the near fracture/well bore region is essentially perpendicular to the
fracture plane (linear flow), so the Cartesian geometry best fits these flow regimes. As explained before the pseudopressure model
is used in well calculations for the coarse-grid models with relatively big grid dimensions (>50 m), capturing near-well condensate
blockage [3]. In this case the sector Cartesian model has been built using 90 m*90 m grid blocks, since the full field model that
we have to deal with has the same space discretization.
The effect of the GPP option is shown in Figure 4, Figure 5 and Figure 6, in which gas and oil production profile are plotted. We
can notice that the cases without GPP option show a drastic early production in terms of gas and condensate. So it is clear that
GPP option is mandatory, otherwise the well production performance is too optimistic and completely unreliable.
The use of the GPP could be supplied by a proper radial or Cartesian local grid refinement (LGR) in the near wellbore region.
A fine space discretization obviously allows to accurately describing the change of fluid saturation distribution and the wellbore
pressure profile. The cost to pay is the increase in CPU time of the simulation, due to the huge increase in the number of cells in
the dynamic model. In our experience the elapsed time could triplicate with respect to the use of GPP, depending on the specific
refinement.

Real case application

Field A is a gas condensate and oil field with formation hydrocarbon bearing belonging to Cretaceous age (Neocomian-
Barremian). In this formation 8 layers can be distinguished from top to bottom. We focused on A, B, C layers, the upper ones,
which are mainly gas condensate bearing intervals.
Five explorative wells have been drilled in the field A. The fluid samples collected during several DSTs and the pressure gradient
from MDT measurements have confirmed the presence of gas condensate phase. The MDT acquisition has been performed on two
wells and has shown a pressure gradient about 0.03 bar/m.The reservoir pressure is equal to 361.2 bar.
Several DSTs have been done but we have retained just two of them for the purpouse of this paper. The observed gas condensate
ratio (GCR) during the well testing has been almost higher than 2300 sm3/sm3.
Wells deliverability has shown a degradation of the petrophysical properties from South to the northern part of the field. For
instance, well-4 drilled in the southern part of field A has shown an effective permeability to gas around 50 mD, while well-2,
situated 2 km to the North, has pointed out an effective permeability of 2.3 mD.
4 SPE 131582

A PVT study has been performed on the fluid samples collected from well-2. The dew point pressure obtained from fluid
recombination has been 19 bar higher than the reservoir pressure. It has been decided to correct the dew point and to reduce the
reservoir pressure. This assumption is reasonable because the condensate volume produced from 380 to 361.2 bar is relatively
small Figure 7. The gas condensate phase has been characterized with 10 components fluid model using Peng-Robinson EOS.
No SCAL analyses have been available during this study. We have considered experimental data, obtained from an analog field in
terms of fluid and petrophysical properties, to tune the gas and condensate relative permeability curves. The applied methodology
has followed the Fevang and Whitson approach [8, 9].

Geological setting

The structural and stratigraphic setting of Field A is related to a Rift system basin. The reservoir object of this study is the
formation A, a lacustrine turbiditic complex that is partially heteropic with shaly-marly marginal lacustrine sediments. This
formation is generally represented by an alternation of thinly laminated sandstones/siltstones and shales and it presents well
developed turbiditic facies locally affected in the depocenter area by slumping phenomena (Cretaceous age Neocomian-
Barremian).
The presence of amorphous organic matter, which generally increases toward the top of the formation, suggests that the deposition
happened in a subaqueous environment.
The formation hydrocarbon bearing was discovered through the exploration Well-1 and Well-2. The additional exploration Well3,
drilled in the northern part of the field, found the formation mainly shale due to a facies variation. Then Well 4 and Well5 have
been recently drilled. The first one confirmed the gas condensate mineralization and assessed oil mineralization in the lower part of
the reservoir while the second one confirmed the expected closure by facies change from sand to shale in the northern part of the
reservoir.
On the basis of lithological vertical variation at field scale, 8 levels have been distinguished from Top to Bottom: A, B, C, D, E, F,
G & H. Levels G & H were excluded and not considered mineralized (shaly layers). Thin shaly interlayers have been recognized
between the different levels, neverthless they have not been considered as vertical barrier but the entire reservoir have been
considered in vertical communication. As said before we focused on levels A, B & C which are mainly gas condensate bearing.
Gas-Oil-Contact (GOC) was estimated at depth **65 m ssl from MDT analysis performed at Well4. Basing on several DSTs
performed at Well4 an Oil-Down-To (ODT) has been estimated at the bottom of E level (**34 m ssl).
Using software Petrel, a 3D geological model has been performed taking into account well data and seismic interpretation (seismic
horizons, faults, CPI). A fine geological grid has been created in order to reproduce the reservoir geometry. The dimensions of the
cells were equal to 90 m * 90 m, 6 zones have been defined inside the model; inside each zone, an internal layering at fine scale
(0.5 meter) has been defined, in order to up-scale correctly the logs inside the 3D grid.
A simple conceptual sedimentological model has been idealized to characterize the reservoir in terms of petrophysical properties.
In agreement with it the depositional system (delta or delta-fan) develops in the central-southern area (Wells 1, 2 & 4).
The clastic source is mainly from W-SW and the reservoir sequence tends to shale out towards Well3, which is in distal/marginal
position (Figure 8). The best reservoir petrophysical properties represented by coarser sediments in Wells 2 & 4 have been
progressively substituted by finer sediments towards northwest (as confirmed by Wells 3 & 5).
Some geological boundaries have been chosen in order to honour the geological and structural constraints (Figure 9).
Usually the petrophysical modeling is based on the facies modeling. This model is affected by the lack of a facies model because
no cores have been available to define a qualitative sedimentological model.
For this reason some assumptions have been made (Figure 10):
porosity has been used as main driver to distribute all the petrophysical parameters (Water saturation, Permeability & Net
to Gross);
porosity CPI curves have been imported into Petrel and then up-scaled to the 3D grid cells at the well locations and
extended all over the model using a geostatistical Sequential Gaussian Simulation algorithm. In this phase the conceptual
sedimentological model has been used to constrain the porosity distribution;
for each level an empirical relation Sw vs has been calibrated and used to predict water Saturation in all the grid cells;
permeability has been distributed all over the model using a linear relationship between and Sw ;
NTG has been determined using a and Sw cut-off.

The fine geological grid has been up-scaled in order to obtain a dynamic coarse grid; cell dimensions have been maintained equal
to the geological model (90 m*90 m) in X and Y directions in order to honour the geological heterogenety. The layers have been
grouped only at vertical scale, in order to reduce the cells number performing a coarse geostatistical internal vertical layering
spacing variable from 1.5 m to 14 m.
SPE 131582 5

DST interpretation

Two DST have been performed in the gas condensate bearing formation at Well-2 and Well-4. The main purpose has been to
evaluate the well deliverability from B and C layers as from two different areas in the field. This section is divided in two main
parts: the analytical well test interpretation succeeded by the numerical validation. For the first one the package InterpretTM by
PARADIGM [11] has been used while the numerical validation has been performed by using Eclipse software [6].

Analytical interpretation

Well-2 is located in the central part of the field. The well has been perforated in B and C layers. The total interval perforated is
about 30 meters. The gross thickness of B and C layers is 80 meters while the net thickness is about 61 meters. The ratio of total
perforated interval versus net thickness is equal to 49 %, which implicates a spherical flow at early time during DST. Reservoir
pressure is about 360.4 bar at the top of the perforation. Four flowing periods and one main build-up have been considered during
the test. The measured gas rate at surface is ranging from 120 to 256 ksm3/d. The GOR is also variable from 2100 to 2500
Sm3/Sm3.
A partial penetration/radial composite analytical model has been considered to correctly describe both the spherical flow and the
oil banking phenomenon observed near the wellbore area. The effective horizontal permeability to gas is projected at 2.3 mD
which means a low permeability formation. D factor, representing non Darcy or turbulence flow, is equal to 2.0E-05 [D/Sm3].
Main interpretation results are reported in Figure 11 and Table 1.
Well-4 is located in southern part of the field A where there are better petrophysical properties. B layer has been the primary
target. The gross thickness of B level is 23 meters at Well-4 and the perforated interval is the lower five meters. This partial
penetration will be rendered by a spherical flow at short time. From the DST pressure analysis, the reservoir pressure at the top of
the perforation is about 360.4 bar. Six flowing periods and two build-ups have been considered during the well-testing operation.
The measured gas rate at stock tank condition has been ranging from 150 to 600 Ksm3/d. The GOR has been also variable from
2600 to 4000 Sm3/Sm3.
The pressure derivative plot did not give a single answer. In addition, two radial stabilizations can be observed in Figure 12. The
second one offers the opportunity to deduce K*H of the formation while the first one is affected by geometrical, non-Darcy flow
and oil banking effects. For this reason, three well/reservoir analytical models have been taken into account for the interpretations
of these build ups, which are in increasing order of complexity:
Vertical well with a radial composite reservoir (RC),
Partial penetration well in a homogeneous reservoir (PP),
Partial penetration well with a radial composite reservoir (RC+PP).

The three interpretation approaches have confirmed a good K*H ranging between 942 and 1162 mD*m (2nd stabilization) with a
permeability of about 50 mD and a mechanical/turbulence skin ranging between 4 and 11. The well damage or the skin factor can
be split in several components:
Mechanical skin in mud filtrate zone;
Geometrical skin due to partial penetration geometry, which can be easily removed with the perforation extension;
Condensate banking skin, related to liquid accumulation in the near wellbore area;
Turbulence skin due to non Darcy flow effects which is proportional to the gas rate.

We also observed that the mechanical skin is higher during the flowing periods before the first build up due to overbalance
perforation and exposure of the perforation to the drilling mud for four days. The value of each skin depends on the well/reservoir
interpretation model adopted and main results are reported in Table 2 and Figure 13 turbulence or non-Darcy flow is represented
by a D-factor which is varying between 4E-06 and 7E-06 d/Sm3 (Figure 14).

Numerical validation

Following the analytical interpretation done at Well-2, a local grid refinement (LGR) has been defined in the wellbore area in
order to numerically reproduce the DST, using the results recovered from analytical interpretation. A double cross section of 3D
dynamic model at Well-2 is illustrated in Figure 15. The LGR integration into full field model will preserve both the pressure
relaxation and the fluid flow types that can be produced in gas condensate situation. For the Well-4, we will present another
equivalent approach that consists to build a fine radial sector model without using LGR. We deduced that the simulated bottom
6 SPE 131582

hole flowing pressure is matching the measured pressure from the gauge as shown in Figure 16. The partial penetration/radial
composite analytical model is a representative model to use for gas condensate well-testing interpretation.
The analytical interpretation of the DST done at well-4 was further deepened throught a reservoir simulation, building a 3D
compositional model at a sector scale. The purpose of the sector study was to validate the interpretation results of the analytical
approach, regarding interpretation model, permeability and the skin values. The skin validation (mechanical, turbulence,
geometrical and related to condensate banking) was considered crucial for the full field model development plan, as their values
affect the production performance of a gas condensate well. The idea behind the 3D sector model is that it correctly describes the
condensate banking phenomenon, while a fine layering can properly define the wellbore geometry and partial penetration effects.
The radial model has an inner radius of 0.11 meters and an outer radius of 600 meters Figure 17. The grid is divided into 30 layers
in radial direction and 10 in vertical one. The petrophysical properties are reported from CPI study and used to populate the grid by
an average value (Table 3).
For rock relative permeability, we consider an analogue from a similar low permeability gas condensate formation bearing. Gas
Condensate Velocity Dependent Relative Permeability (VDKR) has been considered to tune relative permeability in cells having a
good permeability (high condensate velocity). Two key parameters n and control VDKR option in ECLIPSE as introduced
by Fevang-Whitson model and described in Eq. 2 and Eq. 3:

k rg = f I k rgI + (1 f I )k rgM Eq. 2

1
fI =
( N )cg
n
+1
Eq. 3

Where, fI is the immiscible contribution to the gas relative permeabiltity curves.


Relative permeability depends on miscibility between gas and condensates [2,8,10]. High values for n and mean good
gas/condensate miscibility and consequently better relative permeability values (Figure 18).
For the validation phase, we considered the three previous analytical interpretations:
Case 1 or Radial Composite (RC);
Case 2 or Partial Penetration + Radial Composite (RC);
Case 3 or Partial Penetration + Homogeneous.

As a matter of fact, well-4 has investigated good matrix permeability and those parameters (n and alpha) should be adapted to the
formation type. The flowing pressure match is obtained using n equal to 0.7(default value in Eclipse) and alpha equal to 15000
(default value in Eclipse is 4000). The analytical models, relative to case 2 and case 3, are ensuring numerical match of well
bottom hole flowing pressure. Case 1 analytical model is not honouring partial penetration aspect which has implicated a high well
skin due to spherical flow.
This model does not allow the flowing pressure match (Figure 19). In addition and in order to validate the case honouring all
results obtained from analytical model, we calculated the mobility ratio between inner and outer radius (kh/)(1)/(kh/)(2) for each
of those three cases at the production test start-up (Figure 20) and immediately before the first build-up (Figure 21).We observed
that the ratio calculated due to oil banking is higher than 0.9. Moreover, the inner radius is around 5 to 6 meters.
We deduced also that the case 2, partial penetration & radial composite (Table 2), honours the (kh/)(1)/(kh/)(2) highlighted by the
sector model. As a consequence, the radial composite & partial penetration model is the one honouring the observed flowing
pressure, condensate accumulation in the inner radius, mobility reduction and hydrocarbon production rate, using velocity
dependent relative permeability (n=0.7; =15000).

Conclusions

The main conclusions we can drive from this study are summarized as follows:
In gas condensate fields, the condensate banking phenomenon needs to be evaluated at a sector model scale and
properly described at the full field model scale by using Generalized Pseudopressure (GPP) option.
The GPP approach allows an accurate description of the well inflow in a coarse grid compared with the traditional
approach (pseudopressure) and it is useful to model the condensate banking phenomenon at the field scale. Without
using the GPP option, in corser grid, the numerical model will not catch the condensate accumulation in the wellbore
region and this limitation will impact well performance. As a result the surface condensate production can be largely
overestimated.
SPE 131582 7

An alternative approach is to performe a LGR in the near wellbore region. This can be useful in low permeability
scenarios in order to dynamically simulate the hydraulic fracturing.
Based on the results obtained, it can be affirmed that, in this case, the partial penetration (PP) plus radial composite
(RC) is the most appropriate analytic model reproducing the phenomenon of gas mobility reduction in the nearest
wellbore region and spherical flow. Moreover Gas Condensate Velocity Dependent Relative Permeability and non
Darcy flow options are mandatory in order to match the well BHP profile, by tuning relative permeability in cells with
good petrophysical properties.

Acknowledgements
Authors wish to thank eni e&p and all colleagues which gave a support to the realization of this study.

Nomenclature

H= net thickness
OGIP=original gas in place
OOIP=original oil in place
keff= effective permeability
krg= gas relative permeability
krgI= immiscible gas relative permeability
krgM= miscible gas relative permeability
Ncg = gas phase capillary number
fI = immiscibile contribution to gas relative permeability
Sw = water saturation
= non-Darcy flow coefficient
rw= wellbore radius
re = outer radial sector model radius
s= skin
= porosity
= fluid viscosity

References

1. Petroleum Reservoir Engineering, B.C.Craft and M.F.Hawkins.


2. . Fevang, SPE, and C.H. Whitson, SPE, U. Trondheim, NTH, Modeling Gas Condensate Well Deliverability, paper SPE - 30714
3. Kameshwar Singh, SPE, PERA, and Curtis H.Whitson, SPE, NTNU/PERA: Gas Condensate Pseudopressure in Layered Reservoirs, paper
SPE 117930.
4. A.C. Gringarten, A Al-Lamki, S.Daungkaew, Centre for Petroleum Studies, Imperial College of Science, Technology & Medicine, London,
UK; R.Mott, AEA Technology; T.M. Whittle, Baker Hughes, Well Test Analysis in Gas Condensate Reservoirs, paper SPE 62920.
5. Deddy Afidick, N.J. Kaczorowski, and Srivanes Bette, Mobil Oil Indonesia Inc,Production Performance of a Retrograde Gas Reservoir: A
Case Study of the Arun Field, paper SPE - 28749
6. Eclipse, Ver. 2007, software, Schlumberger, 2008.
7. Ahmed H.El-Banbi, and W.D.McCain, Jr., SPE, Schlumberger Holditch-Reservoir Technologies, and M.E. Semmelbeck, SPE, Battlecat Oil
& Gas, : Investigation of Well Productivity in Gas-Condesate Reservoirs, paper SPE 59773
8. Curtis H. Whitson/NTNU nad Pera, ivind Fevang/PERA, and Aud Svareid/ResLab,Gas Condensate Relative Permeability for Well
Calculations, paper SPE 56476.
9. Robert Mott, SPE, Andrew Cable and Mike Spearing, SPE, AEA Technology, A New Method of Measuring Relative Permeabilities for
Calculating Gas-Condensate Well Deliverability, paper SPE -56484.
10. M.Jamiolahmady, A.Danesh, G.Henderson, and G.D.Tehrani, Variations of Gas-Condensate Relative Permeability with Production Rate at
Near Wellbore Conditions: A General Correlation, paper SPE 83960.
11. InterpretTM, version 3.4.005, Paradigm.
8 SPE 131582

Table 1 Well-2 analytical interpretation results.

Table 2 Well-2 analytical interpretation results.

Petrophysical Properties Average Value


Porosity [%] 16.1
Effective Horizontal Permeability [mD] 50
Effective Vertical Permeability [mD] 0.50
NTG [%] 98.9

Table 3 Radial grid petrophysical properties.


SPE 131582 9

Figure 1 - Evolution of the condensate ring around the wellbore [2].

Figure 2 - Comparison between cases with and without the GPP approach.

Figure 3 Radial and Cartesian sector mode.


10 SPE 131582

SM3/DAY SM3/DAY

Figure 4 - Sector gas production rate and cumulative imposing a gas plateau of 350,000 sm3/day: comparison between benchmark radial
model (red), Cartesian model with GPP option (blue) and without GPP option (green).

Figure 5 - Sector gas (left) and condensate (right) cumulative imposing a minimum BHP of 50 bar: comparison between benchmark radial
model (red), Cartesian model with GPP option (blue) and without GPP option (green).

Figure 6 - Sector gas (left) and condensate (right) production rates imposing a minimum BHP of 50 bar: comparison between benchmark
radial model (red), Cartesian model with GPP option (blue) and without GPP option (green).
SPE 131582 11

Constant Volume Depletion


Liquid Saturation
9.00
8.50
8.00
7.50
7.00
6.50
6.00
Liquid Saturation (%)

Psat Observed
5.50 Pres observed
5.00
4.50
4.00
3.50
3.00
2.50
2.00
1.50
Psat observed
1.00
P reservoir
0.50
0.00
0 25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400
Pressure (bar)

Figure 7 - CVD experiment (Liquid saturation vs Pressure).

Figure 8 - Conceptual sedimentological model Three schematic bodies are reported in the figure. The orange body is characterized by
the best petrophysical properties of the area; the yellow one represents the transition from sand to shale. These two main bodies were
considered as the reservoir area (blue line). The green one is the marginal body which closes the depositional area (red line) of the
sedimentation and has been considered completely shaly out.
12 SPE 131582

Figure 9 - Reservoir area boundaries (yellow line): faults closure and stratigraphical closure.

Figure 10 - Comparison among Porosity, Water saturation and Permeability 3D spatial distribution.
SPE 131582 13

TM
Figure 11 Well-2 analytical interpretation using Interpret .

Figure 12 Well-4 first and second derivative pressure stabilization phases.


14 SPE 131582

Figure 13 Well-4 Analytic interpretation: radial composite, partial penetration & radial composite and partial penetration models.

Skin vs. Qg Skin vs. Qg


Radial Composite Interpretation Radial Composite & PP Interpretation

12.00 5.0
y = 7E-06x + 6.8524 4.5
10.00 y = 4E-06x + 1.8237
4.0
8.00 3.5
3.0
6.00 2.5
S

2.0
4.00
y = 7E-06x + 2.3382 1.5
y = 4E-06x - 0.6052
2.00 1.0
0.5
0.00 0.0
0.0E+00 2.0E+05 4.0E+05 6.0E+05 0.0E+00 2.0E+05 4.0E+05 6.0E+05
Qg [Sm3/d] Qg [Sm3/d]

Skin vs. Qg
PP Interpretation & Homogeneous

5
4.5
y = 4E-06x + 2.1196
4
3.5
3
2.5
S

2
1.5
y = 4E-06x - 0.8131
1
0.5
0
0.0E+00 2.0E+05 4.0E+05 6.0E+05
Qg [Sm3/d]

Figure 14 Well-4 skin versus gas rate (non Darcy flow) for three different analytical models.
SPE 131582 15

Figure 15 Well-2 double cross section and LGR refinement.

Figure 16 Well-2 simulated and observed BHP.


16 SPE 131582

Figure 17 Well-4 radial sector model.

Figure 18 Real Relative permeability are weighted average between miscible (straifght) lines and rock curves. Increasing capillary
number, increase relative permeabilities.

Figure 19 Well-4 simulated BHP using petrophysical properties computed from analytical interpretation.
SPE 131582 17

Figure 20 - Well-4 fluid mobility reduction relative to the three analytical models at start of first draw down.
18 SPE 131582

Figure 21 - Well-4 fluid mobility reduction relative to the three analytical models immediately before the first build-up.

S-ar putea să vă placă și