Sie sind auf Seite 1von 19

SPE-171519-MS

Optimizing Recovery in Gas Condensate Reservoirs


Y. H. Seah, Schlumberger; A. C. Gringarten, Imperial College London; M. A. Giddins and K. Burton,
Schlumberger

Copyright 2014, Society of Petroleum Engineers


This paper was prepared for presentation at the SPE Asia Pacific Oil & Gas Conference and Exhibition held in Adelaide, Australia, 14 16 October 2014.
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
Gas condensate reservoirs exhibit complex flow behavior below the dewpoint pressure, caused by
compositional changes and the creation and growth of a condensate bank around the wellbore, which
effectively reduces the relative permeability to gas flow. As a result, gas production decreases and liquid
condensate, a valuable resource, is left behind in the reservoir.
Well deliverability impairment resulting from liquid dropout has been a main focus of gas condensate
studies for over 60 years. We have used compositional reservoir simulation, with representative lean- and
rich-condensate fluids and velocity-dependent relative permeability models, to predict condensate dropout
under typical operating conditions. The effectiveness of different production methods and remediation
solutions in minimizing condensate buildup below the dewpoint pressure was quantified, using first a
single-well model and then full-field models of tight and low-permeability gas condensate reservoirs in
which six vertical wells were compared with two horizontal wells.
We found that, although both horizontal wells and vertical well stimulation do improve well productivity, productivity enhancement depends greatly on well and reservoir parameters such as horizontal well
lengths, well placement, reservoir permeabilities, and gas condensate compositions. For gas condensate
production below the dewpoint pressure, it is possible to achieve an optimum balance between gas
production rate and pressure drawdown, thus minimizing condensate dropout effect while producing at a
reasonable rate. In a low-permeability gas condensate reservoir, the six vertical wells perform slightly
better than the two horizontal wells for the same amount of gas production. In a tight gas condensate
reservoir, on the contrary, gas recovery with two horizontal wells is significantly greater than with six
vertical wells. Although, in this model, well stimulation can increase the productivity of the vertical wells
in tight gas condensate reservoir, it is not as effective as using horizontal wells.

Introduction
Continuous developments in drilling and logging technologies have enabled deeper, higher pressure, and
higher temperature gas condensate reservoirs to be discovered and developed. In addition, most mature
gas condensate fields worldwide are at the later stage of their field life, gradually approaching their
dewpoint pressure below which liquid condensate drops out. Globally, the increasing demand for natural
gas liquids on world markets has stimulated interest in optimizing gas condensate resources (BP 2014).

SPE-171519-MS

Gas condensate reservoirs exhibit complex flow behavior as a result of retrograde condensation when
the bottomhole flowing pressure drops below the dewpoint pressure. Gas condensate dropout near the
wellbore must be considered, as changes in fluid saturations affect relative permeabilities. As a result, it
affects the final decision made in the production planning of a gas condensate field.
This has been the subject of many studies for over 60 years (Muskat 1949). It was discovered that three
concentric regions with different liquid saturations emerge around a gas condensate well producing below
the dewpoint pressure (Kniazeff and Naville 1965). Away from the well, an outer region contains the
initial liquid saturation. Next, an intermediate region shows a rapid increase in liquid saturation and a
corresponding reduction in gas relative permeability. Liquid in this region is still immobile because the
critical saturation has not been reached yet. Nearer to the well, an inner region forms where liquid
saturation exceeds the critical saturation and both the reservoir gas and liquid condensate flow into the
well with constant composition.
A fourth region in the immediate vicinity of the well with low interfacial tension resulting from high
flow rate yields lower liquid saturation and higher gas relative permeability (Gringarten et al. 2000). The
existence of this fourth region is important because it counters the productivity loss caused by liquid
dropout. This region is where the capillary number effect is dominant over the inertial resistance effect,
contributing to higher gas recovery.

Background and Theory


Muskat (1949) first suggested that liquid condensate builds up around the wellbore when bottomhole
pressure drops below the dewpoint pressure and explained mathematically the condensate saturation
profile around a producing well in a gas condensate reservoir. Eilerts et al. and Kniazeff and Naville
published papers in 1965, in which they modeled gas condensate well deliverability numerically with
respect to time and radial distance and studied the effects of non-Darcy flow. Two years later, Gondouin
et al. (1967) extended the work by Kniazeff and Naville (1965) and demonstrated the significance of
non-Darcy flow effects and condensate blockage by applying their formulations to the results data of
backpressure tests from Hassi Er RMel field, Algeria. ODell and Miller (1967), on the other hand,
applied Muskats solutions and presented a pseudopressure function to describe the gas condensate
blockage.
Fetkovich (1973) developed a rate- and time-dependent skin based on Muskats results and incorporated it into the standard gas flow equation. Subsequently, Fevang and Whitson (1996) applied the
modified form of the Evinger-Muskat pseudopressure (Evinger and Muskat 1942) to calculate gascondensate well deliverability and provided a simple method for calculating bottomhole flowing pressure
in coarse-grid models. A more recent study by Gringarten et al. (2000) established the use of a
three-region composite model to analyze gas condensate well test data.
Well productivity index (i.e., the relationship between pressure drop and flow rate) is complicated in
gas condensate reservoirs below the dewpoint pressure due to the existence of the fourth region with
lower interfacial tension and higher gas relative permeability. The two velocity-dependent relative
permeability effects contributing to the existence of the near-wellbore region are the capillary number
effect, also referred to as velocity stripping, viscous stripping, or positive coupling effect, and the
non-Darcy flow or inertial resistance effect introduced by Forchheimer (Belhaj et al. 2003).
A compositional reservoir simulator has been used to model the complex flow behavior of gas
condensates, with very fine grids near to the well, as pressure drawdown predominantly occurs within 10
ft of the wellbore (Mott et al. 2000).
Velocity-Dependent Relative Permeability Effects
The methods used in this study to model the two velocity-dependent relative permeability effects are
described below.

SPE-171519-MS

Capillary number effect Previous studies have reported that gas deliverability increases as a result of an
increase in the gas relative permeability at high velocity when the capillary number is high (Danesh et al.
1994; Henderson et al. 1995, 2000; Ali et al. 1997; Gringarten et al. 2000). Danesh et al. (1994) was the
first to address the improvements of relative permeability to gas flow in condensate systems with increases
in velocity or decreases in interfacial tension, through laboratory experimental results. Miscibility between
the flowing phases due to reduced interfacial tension at high velocities, which increases mobility for oil
and gas, can be modeled in compositional simulation using various different options for velocitydependent relative permeability.
Capillary number is the ratio of viscous to capillary forces and can be represented by a single
parameter, Nc. Production data from several fields have revealed that deliverability is higher than the
conventional simulated data because of the positive effect of capillary number (Mott 2002). It is essential
to incorporate the capillary number model into compositional simulations below the dewpoint pressure in
systems that exhibit retrograde condensation. Otherwise, well productivity will be underestimated
whereas pressure drops and condensate dropout will be overestimated.
The capillary number formulations applied in this study are after Henderson et al. (1995), as shown
below. In their model, the capillary number for both phases depends only on the gas properties.
(1)
The base capillary number Ncbp is defined as the lower threshold value below which Ncp has no effect
on phase relative permeabilities. This number is taken from input data and used to calculate a normalized
capillary number Ncnp:
(2)
At higher capillary numbers, the residual saturation is reduced by multiplying the phase residual
saturation Srp by a factor Xp:
(3)
The capillary number is also used to define the interpolation between the immiscible relative
permeability curves defined by input tables and the miscible curves that are constructed internally as
straight lines.
(4)
The immiscible relative permeability is krp, and the miscible relative permeability is calculated as
follows, giving a straight line with endpoints defined by the adjusted phase residual saturation:
(5)
where Xp is defined in Eq. 3. The interpolation factor f is calculated as
(6)
(7)
Non-Darcy Flow Effect Darcys law does not explain gas flow accurately when the flow rate is high,
especially near the wellbore. At high gas rate, pressure drop for flow into the well is no longer proportional

SPE-171519-MS

to the rate. Turbulence, which causes increased pressure drop, can be modeled in the simulator using the
Forchheimer correction to account for inertial effects:
(8)
Non-Darcy flow effect is typically modeled as a rate-dependent skin in most numerical simulators.
can be determined from multirate pressure test analysis or from theoretical or empirical correlations (Li
and Engler 2001). A typical empirical correlation for single-phase flow is:
(9)
Geertsma (1974) proposed an empirical correlation for two-phase flow, relating to the fluid saturation
and relative permeability:
(10)
In the case of two-phase gas condensate flow, inertial effects generate extra condensate dropout
because the inertial resistance in one phase is affected by the presence of the second phase.
Parameters Affecting Productivity of Horizontal Wells
The performance of horizontal wells in comparison with vertical wells has been investigated in previous
studies (Joshi 1988; Hashemi and Gringarten 2005). Results show that the relative productivity of a
horizontal well, compared to a vertical well in the same reservoir, depends on the producing horizontal
length and the ratio of vertical to horizontal permeability, kv/kh.

Methodology
This study was carried out in three stages: first, gas condensate compositional analysis; second, base
model study, and third, the full field model design. The following assumptions were made:
1. The near-well phase equilibrium and fluid flow interactions are accurately represented using a fine
radial grid model and a fine tartan grid, with fluid properties calculated from an equation of state
(EOS) model.
2. There is no limit on the tubing head flowing pressure, and only well bottomhole pressure limits
are considered.
3. The pressure drop due to condensate blockage is significant compared to the pressure drop in the
tubing.
4. No aquifer support was considered in this study.

Gas Condensate Compositional Analysis


Four gas condensate mixtures with different compositions have been studied to distinguish between rich
and lean gas condensates (TABLE 1). A typical lean gas condensate generates a small volume of liquid
(i.e., 30 to 50 stb/MMscf), whereas a rich gas condensate generates a relatively larger volume of liquid
(i.e., 50 to 300 stb/MMscf) (Zheng et al. 2006).
Referring to Table 1, gas condensate B was selected for use in single-well sensitivity studies and for
investigating production strategies in the full field model, because it represents a gas condensate with a
condensate/gas ratio (CGR) that is not too rich and not too lean (Fig. 1). All four gas condensates were
used in the comparison of rich versus lean gas condensate for the full field model study.

SPE-171519-MS

Table 1FLUID PROPERTIES FOR GAS CONDENSATE A, B, C, AND D


Gas condensate

CGR (stb/MMscf)

Dewpoint pressure (psia)

Reservoir temperature (F)

161
87
48
27

4767
4298
4028
3440

340
200
251
251

A
B
C
D

Figure 1Phase diagrams of gas condensate A (left) and B (right). Gas condensate B is leaner than gas condensate A and therefore its phase diagram
appears to be shifted to the left when compared to gas condensate A.

Base Model: Single Well


Compositional gas condensate simulation cases have been constructed to simulate fluid flow behavior and
to predict condensate dropout under typical operating conditions using single-well models.
Single-Well Model Gridding and Description The single-well models consisted of a five-layer homogeneous reservoir of 100-ft thickness with a drainage area of 891 acres, 10% porosity, horizontal
permeability of 1 md, and vertical-to-horizontal-permeability ratio (kv/kh) of 0.1 (typical of sandstone
reservoirs). Initially, the reservoir pressure is slightly above the dewpoint pressure.

Table 2SINGLE-WELL RADIAL GRID MODEL GRID PROPERTIES


Grid properties
Number of active cells
Innermost grid radius
Outermost grid radius
Cell spacing (ft)

Value

Unit

20
0.15
ft
3515
ft
Cell sizes increase geometrically
in radial direction away from the
center where the well is located.

Figure 2Single-well radial grid model.

SPE-171519-MS

Table 3SINGLE-WELL TARTAN GRID MODEL GRID PROPERTIES


Grid properties

Value

Unit

Number of active cells


Innermost grid size
Areal cell spacing
Uniform
Logarithmic, Central
Uniform

28125
2
Number of cells
30
15
30

ft
Average cell size (ft)
101.35
10
101.35

Figure 3Single-well tartan grid model.

A finely gridded 1D radial, fully compositional model was used, with no wellbore storage or
mechanical skin effect. A large outermost grid size of 1,335 ft was used to avoid boundary effects (Fig.
2; TABLE 2).
A tartan grid model was then designed to match the radial grid model, to be used in the full field model
studies at the later stage of the project (Fig. 3; TABLE 3).
Velocity-Dependent Relative Permeability Effect The same capillary number model and Forchheimers
non-Darcy model were incorporated into all the simulation cases for the compositional simulation runs.
The capillary number model introduced by Henderson et al. (1995) has been used to interpret the
difference between the base relative permeability and the miscible relative permeability models. A
multirate drawdown (200 Mscf/D for 1 day, 400 Mscf/D for 2 days, 800 Mscf/D for 3 days, and 1,600
Mscf/D for another 4 days) followed by 10 days of buildup were simulated to observe the positive effect
of the velocity-dependent relative permeability phenomenon.
The high capillary number and non-Darcy flow effects are both rate-dependent functions acting in
opposite directions. Fig. 4 shows a stronger capillary number effect over the inertial effect, resulting in
a positive velocity-dependent relative permeability effect. The pressure drawdown is lower than that
without velocity-dependent relative permeability models at the same designated flow rates.
Effect of Different Gas Production Rate The single-well radial grid model has been used to examine
the condensate dropout effect as production goes below the dewpoint pressure by operating at different
specified rates. Various scenario cases were run for 10 days of drawdown followed by another 10 days
of buildup to achieve an optimum balance between gas production rate and well bottomhole pressure. This
is important to minimize the condensate dropout effect in the reservoir.
The final operational decision was made after analyzing the results obtained from the scenario cases
at high and low rates whilst also taking into account the initial reservoir conditions.
Fig. 5 shows two options for operating this gas condensate model, yielding different amounts of
condensate dropout as the bottomhole pressure drops below the dewpoint pressure. For the well producing
at 2 MMscf/D, oil saturation as high as 37% was observed around the wellbore at the end of the 10-day
drawdown. For a production rate of 1 MMscf/D, about 26% oil saturation was observed around the well.
There is a trade-off between high production rate and small liquid dropout effect. The two operating
scenarios shown in Fig. 5 are both producing at reasonable rates. An operational decision can be made,
either aiming at reducing condensate dropout and therefore a lower production rate or aiming at producing
at a higher rate and accepting a higher liquid dropout around the wellbore.
For the next stage of the study, it was decided that the pressure drawdown will be allowed to go below
the dewpoint pressure, to allow the well to produce at a reasonable rate.

SPE-171519-MS

Figure 4 Effect of velocity-dependent relative permeability models on well bottomhole pressure.

Figure 5Gas production rate (GRAT) control mode at 1 MMscf/D and 2 MMscf/D, respectively, with a limiting bottomhole pressure (BHP) target
or lower limit at 3000 psia.

Full Field Model and Simulation Results


The full field simulation model consisted of a five-layer homogeneous reservoir of 300-ft thickness with
a drainage area of 4,768 acres, 10% porosity, horizontal permeability of 1 md, and kv/kh ratio of 0.1
(typical of sandstone reservoirs). A tartan grid was constructed with logarithmic cell spacing, with fine
grid cells near the wells for enhanced grid definition to capture the fluid flow movement. All the models
have the same initial gas in place. (Fig. 6)
Six vertical producing wells were placed equally apart from each other in the gas condensate field.
Capillary number and non-Darcy flow effects were taken into account, but frictional losses in the
wellbore, wellbore storage, and skin were neglected. The four main sensitivity studies conducted are gas
condensate composition, reservoir permeability, well configuration, and well stimulation. A sensitivity
study on rates was performed, and it was decided to produce at a field production daily contracted quantity
(DCQ) of 12 MMscf/D from the field; with each well producing at 2 MMscf/D to minimize the condensate
dropout effect around the wellbore. The condensate saturation profiles around well 1, well 2, and well 3
were plotted against the radial distance from the wellbores (i.e., along distance A from well 1, distance
B from well 2, and distance C from well 3. (Fig. 7) Different radial directions from the wells gave distinct
condensate saturation plots, but their saturation profiles generally follow the same trend.

SPE-171519-MS

Figure 6 Full field model with six vertical wells producing from the gas condensate reservoir. Tartan grid (right) was used for enhanced grid
definition.

Figure 7Comparison of condensate B saturation profile around the wellbores at different production rates (i.e., 2MMscf/D, 5 MMscf/D, and 10
MMscf/D).

Phase Diagram
As the well bottomhole flowing pressure falls below the dewpoint pressure, lighter components of gas
condensate B are produced as gas phase whereas the heavier components drop out as immobile liquid. The
produced gas becomes leaner as the near-wellbore region becomes richer in liquid condensate. As can be
seen in Fig. 8, the phase diagram of the gas condensate near the wellbore has shifted to the right at the
end of the drawdown, indicating that condensate dropout has occurred around the well.
Fig. 9 further confirms that the produced reservoir gas has become leaner at the end of the drawdown
period as condensate has dropped out in the reservoir. The CGR decreases rapidly from 87 stb/MMscf to
approximately 72.5 stb/MMscf during the first year of production; it then remains relatively constant at
a plateau CGR until the end of the production.
Rich versus Lean Gas Condensate
A first objective of the simulation runs was to compare the condensate saturation profile around the
wellbore for different gas condensate compositions, with all other parameters kept constant. The four gas
condensates shown in Table 1 were used in this study. Compositional simulation runs were made with
capillary number and inertial effects, at a constant gas flow rate of 2 MMscf/D per well.
A plot of condensate saturation profile against distance from well 1 at the end of the 5-year drawdown
is shown in Fig. 10. Gas condensate A (the richest gas condensate) exhibits the highest amount of liquid

SPE-171519-MS

Figure 8 Comparison of phase diagrams at the start and the end of the 5-year production from the full field model using six vertical wells.

Figure 9 Changes of CGR in the gas condensate field with respect to producing time.

dropout near the wellbore whereas gas condensate D (the leanest gas condensate) exhibits the least amount
of liquid dropout around the wellbore.
Gas condensate B, which has been used for all the simulation models, has about 31% of liquid
saturation within 2 ft of the wellbore; this increases gradually up to approximately 35% 10 ft away from
the wellbore and then eventually tails off farther away from the wellbore. Examination of the condensate
saturations and fluid properties near the wellbore indicates the existence of the fourth saturation region
due to capillary number effect (Gringarten et al. 2006). This was not observed in the simulation results
of the other three gas condensate types in this study.
Tight versus High-Permeability Gas Condensate Reservoir
A series of simulation runs was performed to compare the condensate saturation profile and gas and liquid
production rates for different reservoir permeabilities. Four different reservoir permeabilities were
studied, with each representing different types of gas condensate reservoir (i.e., ultratight reservoir (0.001
md), tight reservoir (0.1 md), low-permeability reservoir (1 md), and high-permeability reservoir (1,200
md). All four reservoirs have a kv/kh ratio of 0.1 and a porosity of 10%. Gas condensate B was used in
this study.

10

SPE-171519-MS

Figure 10 Comparison of condensate saturation profiles for different gas condensate compositions (A, B, C, and D).

Figure 11Comparison of condensate B saturation profiles in ultratight, tight, low-permeability, and high-permeability gas condensate reservoirs.

The condensate dropout effect is most significant in the tight reservoir, even more than in the ultratight
reservoir. Gas condensate production is lowest in the ultratight reservoir because the nature of the
reservoir inhibits the flow of the gas and condensates. There is no liquid dropout in the high-permeability
reservoir due to the very good formation connectivity, which enables both the gas and liquid phases to
flow easily into the producing wells and eventually up to the surface. (Fig. 11 and Fig. 12)

Well Configuration
Horizontal wells can enhance reservoir contact and improve well deliverability. The simulation model
used in this study is exactly the same as the full field model with six vertical wells discussed earlier, except
that two horizontal wells are used instead of the six vertical wells (Fig. 13). The two horizontal producing
wells were placed equally apart from each other in the gas condensate reservoir to minimize boundary
effects. Both of the horizontal wells are fully cased and perforated over their entire horizontal length.
Frictional losses in the wellbore, wellbore storage, and skin were neglected.
In this section, sensitivity studies were performed in the low-permeability gas condensate field (1 md)
and the tight gas condensate field (0.1 md). A series of simulation runs were performed to compare the
pressure drawdown and the condensate saturation profile when the flowing bottomhole pressure drops
below the dewpoint pressure. To match the field production DCQ at a rate of 12 MMscf/D, each of the
two horizontal wells was produced at 6 MMscf/D. The full field model was operated for 5-year drawdown
followed by 500 days of buildup period.

SPE-171519-MS

11

Figure 12Comparison of pressure drawdown, well gas cumulative production volume, well gas, and liquid production rates in ultratight, tight,
low-permeability, and high-permeability gas condensate reservoirs.

Figure 13Full field model with two horizontal wells producing from the gas condensate field. Tartan grid (right) was used for better runtime
performance and enhanced grid definition.

Low-Permeability Gas Condensate Field The liquid dropout effect in the low-permeability gas condensate field (1 md) is shown in Fig. 14. A plot of pressure against time is shown in Fig. 15 (left) for the
two horizontal wells with different completion lengths (1,000 ft, 2,000 ft, and 3,000 ft) and for the six
vertical wells (from the previous full field simulation) with 300 ft of completion length. Compositional
runs were made with capillary number and inertial effects at a constant gas flow rate of 6 MMscf/D.

12

SPE-171519-MS

Figure 14 Liquid condensate dropout effect from the first year to the fifth year of production.

Figure 15Comparison of pressure drawdowns (left) and of field gas and liquid cumulative production volume (right) for vertical and horizontal
wells in the low-permeability gas condensate field.

Figure 16 Comparison of condensate saturation profile for vertical and horizontal wells in the low-permeability reservoir.

As can be seen in Fig. 15 (left), the pressure drop in horizontal wells decreases as the horizontal well
length increases, although six vertical wells yield a smaller pressure drop than the two horizontal wells
(producing the same total amount). The horizontal well length of 3,000 ft was selected because it yields
a lower pressure drawdown, thus reducing the condensate dropout around the wellbore. Nonetheless, this
must be justified against the cost of using a longer horizontal well length. Fig. 15 (right) shows that for
the same field gas cumulative production, the six vertical wells produce approximately 10% more liquid
condensate compared with the two horizontal wells at the end of the drawdown.

SPE-171519-MS

13

Figure 17Comparison of pressure buildups (left) and of field gas and liquid cumulative production volume (right) for vertical and horizontal wells
in the tight gas condensate reservoir.

The condensate saturation profile around the vertical and horizontal wells in the field at the end of the
5-year drawdown is illustrated in Fig. 16. For the two horizontal wells, the condensate saturation is plotted
against the distance from the heel in the x-direction (perpendicular to the well). In this particular reservoir,
condensate saturation is fairly uniform because of the isotropic horizontal permeability throughout the
field. Fig. 16 also indicates that, as the horizontal well length increases from 1,000 ft to 3,000 ft, the
condensate saturation decreases.
Tight Gas Condensate Field To compare well productivities between vertical and horizontal wells and
to quantify the effectiveness of horizontal wells in tight gas condensate field (0.1 md), compositional
simulation runs of 5-year duration were made for different horizontal well lengths (1,000 ft, 2,000 ft and
3,000 ft). As for the low-permeability case, the runs were made with capillary number and inertial effects
at constant gas flow rate of 6 MMscf/D per well.
A plot of pressure against time for a 5-year drawdown followed by a 500-day buildup is shown in Fig.
17 (left) for the two horizontal wells with different completion lengths (1,000 ft, 2,000 ft, and 3,000 ft)
and for the six vertical wells with 300 ft of completion length. Due to the very low permeability of the
reservoir, it is difficult for the gas condensate to flow from the field into the wells. As can be seen in Fig.
17 (left), the bottomhole pressure limit of 400 psia for the vertical and horizontal wells has been reached
once the production was started.
Fig. 17 (right) shows that cumulative gas production from the two horizontal wells is approximately
4.7 Bscf more than production from the six vertical wells. In addition, it was found that the two horizontal
wells produce approximately 85 Mstb more liquid condensate than the six vertical wells.
The difference in well productivity between horizontal wells and vertical wells is very significant in a
tight gas condensate field. The cumulative gas production increases with the horizontal well length and
exceeds the production from the vertical wells for lengths over 2,000 ft (Fig. 18). Fig. 19 compares the
cumulative field liquid production for each horizontal well length and the vertical wells.
The condensate saturation profile around the vertical and horizontal wells in the tight gas condensate
reservoir at the end of the 5-year drawdown is illustrated in Fig. 20. The different horizontal well lengths
yield similar condensate saturation profiles around the wellbore. The vertical and horizontal wells showed
a maximum saturation of about 48%, which eventually tails off as it gets further away from the wellbore.
In the tight gas condensate field, horizontal wells yield higher total gas recovery than vertical wells.
This was not the case in the low-permeability gas condensate field, as discussed above.

14

SPE-171519-MS

Figure 18 Comparison of field gas cumulative production for vertical and horizontal wells in the tight gas condensate reservoir.

Figure 19 Comparison of field liquid cumulative production for vertical and horizontal wells in the tight gas condensate reservoir.

Figure 20 Comparison of condensate saturation profile for vertical and horizontal wells in the tight gas condensate reservoir.

Well Spacing between Horizontal Wells


Well spacing is generally defined as the maximum area of a reservoir that can be efficiently and
economically drained by one well (Keuengoua and Amorin 2011). Sensitivity studies were performed to
study the effects of well spacing between the horizontal wells. In this study, the low-permeability gas
condensate field was used to examine the boundary effects caused by the neighboring well when the two
horizontal wells were placed closer to each other. Fig. 21 shows another full field model having exactly
the same size and fluid properties as the original full field model, but with different well location and

SPE-171519-MS

15

Figure 21Horizontal wells placed closer to each other to observe boundary effects, where each well acts as a no-flow boundary against the other.

Figure 22Comparison of pressure drawdowns for the two horizontal wells being placed closer to each other and the two horizontal wells in the
original full field model.

smaller well spacing. As a result, the two horizontal wells acted as no-flow boundaries against each other,
leading to an increase in the pressure drawdown at a constant flow rate (Fig. 22).
Stimulation of Vertical Wells
Sensitivity analysis of well stimulation was performed to investigate the well productivity/deliverability
enhancement upon stimulating the six vertical wells in the tight gas condensate reservoir. The possible
effect of well stimulation was modeled by a productivity index (PI) multiplier in the simulation. The
resulting cumulative gas production is shown in Fig. 23.
Well (acid) stimulation enhances the productivity index and reduces the pressure drawdown at constant
flow rate, thus reducing the condensate dropout effect in the tight reservoir. However, in this study, as
shown in Fig. 23, localized well stimulation does not have a major impact on the productivity of the
vertical wells, compared to the use of two horizontal wells. Well stimulation yields an incremental field
gas recovery of 600 MMscf compared to the unstimulated vertical wells. However, the two horizontal
wells produce 4.1 Bscf more gas than the six stimulated vertical wells.
Neither hydraulic fracturing nor well stimulation on horizontal wells was investigated in this project
due to time constraints. Further study would be recommended to evaluate these options.

Discussion
In this study, well deliverability impairment due to liquid dropout in gas condensate fields was successfully represented in compositional simulation models. A sensitivity study on gas condensate compositions

16

SPE-171519-MS

Figure 23Productivity index enhancement after well stimulation (left) and the comparison of field gas cumulative production volume in vertical and
horizontal wells (right).

(rich versus lean) has been conducted to understand production behavior for different fluid characteristics.
A richer gas condensate yields higher liquid dropout in the field than a leaner gas condensate. A sensitivity
study on reservoir permeability was also performed to investigate the effects of formation permeability
around the wellbore on gas recovery and liquid dropout. This study assumes a homogeneous reservoir. In
real life, gas condensate reservoirs are complex and highly heterogeneous, resulting in more complicated
near-well flow behavior. Sensitivity studies on gas condensate reservoir properties are essential to confirm
the best production method for individual cases.
Good operational decisions in managing the gas condensate field yield higher recovery efficiency. The
production options considered in this study were production rates, well configuration, and well stimulation. In this study, for a low-permeability gas condensate reservoir, six vertical wells have a slightly better
performance than two horizontal wells (well length 3,000 ft), with both producing at the same DCQ.
On the other hand, the performance of two horizontal wells in a tight gas condensate reservoir is superior
to that of the six vertical wells, obtaining 47% more gas recovery.
Well stimulation (e.g., acid treatment) enhances the productivity of the vertical wells and yields 6%
more gas recovery than the unstimulated vertical wells in the tight gas condensate reservoir. Hydraulic
fracturing was not investigated in this particular work and is recommended for further study. Hashemi and
Gringarten (2005) compared horizontal wells and hydraulically fractured wells in single-well models and
showed that these well configurations are comparable and can improve productivity and reduce pressure
drop. The final choice between a horizontal well and an equivalent vertical well with a fracture can only
be determined from economic evaluation.
In practice, particularly in a gas condensate reservoir where reservoir pressure is close to dewpoint
pressure, there are two extreme options for operating the field. The first is to allow for very quick
depletion to produce most of the gas and condensates out of the field within the next few years of
production, without considering the effect of liquid dropout. Then, shut in the well for pressure buildup
for some time and produce from the field again, assuming the resulting pressure buildup plateau is above
the dewpoint pressure and the condensate bank will revaporize. The second option is to produce the field
at a lower rate by limiting the pressure drawdown above the dewpoint pressure. The optimum operating
decision for a specific field is likely to be between these two extremes, and it will require more simulation
studies to determine the best operating rates.

SPE-171519-MS

17

Conclusions
The results of this study demonstrate the critical importance of liquid dropout effects in a gas condensate
field. The main conclusions are as follows:
1. Well productivity index impairment due to liquid dropout in a gas condensate field will be
overestimated if the compositional simulation model ignores the velocity-dependent relative
permeability effects in the immediate vicinity of the wellbore.
2. For gas condensate production below the dewpoint pressure, simulation studies can be used to
achieve a balance between the gas production rate and the pressure drawdown, to minimize
condensate dropout effect while at the same time producing at a reasonable rate from the field.
3. The optimum well configuration and selection of vertical or horizontal wells for a gas condensate
field depend on the type of reservoir (low permeability versus tight reservoirs).
4. Horizontal well length affects gas production in gas condensate fields. A longer horizontal well
can increase reservoir contact and enable operation at a higher bottomhole pressure, thus reducing
condensate dropout in fields operated below the dewpoint pressure.
5. Well stimulation enhances the productivity of the vertical wells in the tight gas condensate
reservoir; however, in this study it is not as effective when compared to the use of horizontal wells.

Recommendations for Further Study


For this study, reservoir models with fine grids near the wellbore were successful, and the simulation
models ran in reasonable timescale. For a bigger field, it may be necessary to use coarser grids
incorporating methods of upscaling such as the generalized pseudopressure model proposed by Fevang
and Whitson (1996).
This work focuses on operational decisions and reservoir management of gas condensate fields. For a
more comprehensive understanding of the options available for field development, the following additional studies should be considered:
1. More detailed analysis of the costs and economic benefits of using horizontal wells compared to
vertical wells.
2. A study of hydraulic fracturing effects, before and after the formation of condensate bank.
3. Investigation of the effects of acid stimulation and hydraulic fracturing for horizontal wells.
4. Investigation of gas cycling, which can be used to maintain reservoir pressure above the dewpoint
pressure while producing at a reasonably high rate, thus avoiding condensate dropout in the
reservoir.
5. Research into solvent injection treatment to mitigate condensate banking.
6. Studies on the impact of aquifer drive and water encroachment for a gas condensate field with
aquifer support.
Nomenclature
a, b, c, d
Non-Darcy coefficients (experimental)
dp/dL
Pressure drop along distance L
f
Interpolation factor
Subscript g Phase (gas)
k
Absolute permeability (md)
Horizontal permeability (md)
kh
Miscible relative permeability of phase p
krmp
Immiscible relative permeability of phase p
krp
Interpolated (velocity-dependent) relative permeability of phase p
krvp

18

kv
LH
mp, n1p, n2p
Ncbp
Ncnp
Ncp
p
Subscript p
Sp
Srp
Xp

SPE-171519-MS

Vertical permeability (md)


Horizontal well length (ft)
Dimensionless user-defined input parameters
Base capillary number for phase p
Normalized capillary number for phase p
Capillary number for phase p
Pressure (psia)
Phase (oil, gas)
Saturation of phase p
Residual saturation of phase p
Residual saturation multiplying factor for phase p
Forchheimers non-Darcy coefficient
Viscosity (cp)
Interstitial velocity in the reservoir (ft/d)
Density
Interfacial tension
Porosity

References
Ali, J.K., McGauley, P.J., and Wilson, C.J. 1997. The Effects of High-Velocity Flow and PVT
Changes Near the Wellbore on Condensate Well Performance. Paper SPE 38923 presented at the SPE
Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 5 8 October.
Belhaj, H.A., Agha, K.R., Nouri, A.M. et alet al. 2003. Numerical Simulation of Non-Darcy Flow
Utilizing the New Forchheimers Diffusivity Equation. Paper SPE 81499 presented at the SPE Middle
East Oil Show & Conference, Bahrain, 5 8 April.
BP Energy Outlook 2035, January 2014, pp. 2350. http://www.bp.com/en/global/corporate/aboutbp/energy-economics/energy-outlook/outlook-to-2035.html. Accessed 11 July 2014.
Danesh, A., Khazam, M., Henderson, G.D. et alet al. 1994. Gas Condensate Recovery Studies. Paper
presented at the DTI Improved Oil Recovery and Research Dissemination Seminar, London, June.
Eilerts, C.K., Sumner, E.F. and Potts, N.L. 1965. Integration of Partial Differential Equation for
Transient Radial Flow of Gas-Condensate Fluids in Porous Structures. SPE J. 5(2) 141152. SPE 716-PA.
Evinger, H.H. and Muskat, M. 1942. Calculation of Theoretical Productivity Factor. Trans AIME
146(1) 126 139. SPE-942126-G.
Fetkovich, M.J. 1973. The Isochronal Testing of Oil Wells. Paper SPE 4529 presented at the Fall
Meeting of the Society of Petroleum Engineers of AIME, Las Vegas, Nevada, USA, 30 September-3
October.
Fevang, . and Whitson, C. H. 1995. Modeling Gas-Condensate Well Deliverability. Paper SPE
30714 presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, USA, 2225
October.
Geertsma, J. 1974. Estimating the Coefficient of Inertial Resistance in Fluid Flow through Porous
Media. SPE J. 14(5): 445450. SPE 4706-PA.
Gondouin, M., Iffly, R. and Husson, J. 1967. An Attempt to Predict the Time Dependence of Well
Deliverability in Gas Condensate Fields. SPE J. 7(2): 113124. SPE 1478-PA.
Gringarten, A.C., Al-Lamki, A., Daungkaew, S. et alet al. 2000. Well Test Analysis in Gas Condensate
Reservoirs. Paper SPE 62920 presented at the SPE Annual Technical Conference and Exhibition, Dallas,
Texas, USA, 1-4 October.

SPE-171519-MS

19

Gringarten, A.C., Bozorgzadeh, M., Daungkaew, S. and Hashemi, A. 2006. Well Test Analysis in
Lean Gas Condensate Reservoirs: Theory and Practice. Paper SPE 100993 presented at the SPE Russian
Oil and Gas Technical Conference and Exhibition, Moscow, Russia, 3 6 October.
Harisch, R.A., Bachman, R.C., Puchyr, P.J. and Strashok, G.W. 2001. Evaluation of a Horizontal
Gas-Condensate Well Using Numerical Pressure Transient Analysis. Paper SPE 71588 presented at the
SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, USA, 30 September3
October.
Hashemi, A. and Gringarten, A.C. 2005. Comparison of Well Productivity Between Vertical, Horizontal and Hydraulically Fractured Wells in Gas-Condensate Reservoirs. Paper SPE 94178 presented at
the 2005 SPE Europec/EAGE Annual Conference, Madrid, Spain, 1316 June.
Henderson, G.D., Danesh, A., Tehrani, D.H. et alet al. 1995. Measurement and Correlation of Gas
Condensate Relative Permeability by the Steady-State Method. Paper SPE 30777 presented at the SPE
Annual Technical Conference and Exhibition, Dallas, Texas, USA, 2225 October.
Henderson, G.D., Danesh, A., Tehrani, D.H. and Al-Kharusi, B. 2000. The Relative Significance of
Positive Coupling and Inertial Effects on Gas Condensate Relative Permeabilities at High Velocity. Paper
SPE 62933 presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, USA, 1 4
October.
Joshi, S.D. 1988. Production Forecasting Methods for Horizontal Wells. Paper SPE 17580 presented
at the SPE International Meeting on Petroleum Engineering, Tianjin, China, 1 4 November.
Keuengoua, C.D.S. and Amorin, R. 2011. Well Spacing for Horizontal Wells. Research Journal of
Applied Sciences, Engineering and Technology 3(6): 486 493.
Kniazeff, V.J. and Naville, S.A. 1965. Two Phase Flow of Volatile Hydrocarbons. SPE J. 5(1): 3744.
Li, D. and Engler, T.W. 2001. Literature Review on Correlations of the Non-Darcy Coefficient. Paper
SPE 70015 presented at the SPE Permian Basin Oil and Gas Recovery Conference, Midland, Texas, USA,
1516 May.
Mott, R. 2002. Engineering Calculations of Gas-Condensate Well Productivity. Paper SPE 77551
presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 29
September2 October.
Mott, R., Cable, A., and Spearing, M. 2000. Measurements and Simulation of Inertial and High
Capillary Number Flow Phenomena in Gas-Condensate Relative Permeability. Paper SPE 62932 presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, USA, 1 4 October.
Muskat, M. 1949. Condensate Reservoirs. In Physical Principles of Oil Production, Chap. 13, pages
738 800. McGraw-Hill.
ODell, H.G. and Miller, R.N. 1967. Successfully Cycling a Low-Permeability, High-Yield Gas
Condensate Reservoir. J Pet Tech 19(1): 4147. SPE 1495.
Zheng, S.Y., Zhiyenkulov, M. and Yi, T.C. 2006. Productivity Evaluation of Hydraulically Fractured
Gas-Condensate Reservoirs. Journal of Petroleum Geoscience 12(3): 275283.

Das könnte Ihnen auch gefallen