See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/289361913

Guidelines for developing gas fields associated with edge-water drive

Article · January 2012

CITATION READS
1 498

1 author:

Hazim Al-Attar
United Arab Emirates University
42 PUBLICATIONS   273 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Enhanced oil rcovery View project

Modeling of Fractured Oil Reservoirs View project

All content following this page was uploaded by Hazim Al-Attar on 09 April 2016.

The user has requested enhancement of the downloaded file.

Advances in Sustainable Petroleum Engineering Science ISSN 1937-7991
Volume 1, Issue 1 © 2009 Nova Science Publishers, Inc.

GUIDELINES FOR DEVELOPING GAS FIELDS

ASSOCIATED WITH EDGE-WATER DRIVE

Hazim H. Al-Attar
United Arab Emirates University, Chemical & Petroleum Engineering Department, Al-
Ain P.O.BOX 17555, UAE, [email protected]

ABSTRACT
The objective of this work is to investigate the effects of key reservoir and aquifer
parameters on gas recovery factor for finite and infinite aquifers. The first part deals with
performance calculations of cumulative water influx and cumulative gas production at
successive time steps. The effects of aquifer size to reservoir size ratio, rate of pressure
depletion at the gas/water contact, initial pressure at the gas/water contact, initial gas in
place, angle of water encroachment, and volumetric sweep efficiency on the cumulative
water influx and cumulative gas production are investigated. In the second part, a new
analytical approach is proposed to determine abandonment pressure and gas recovery
factor for the case of infinite aquifers. The third part caters for performance calculations
of gas recovery in the case of finite aquifers. For this purpose, another new analytical
approach is suggested for estimating the rate of pressure depletion at the gas/water
contact.
It is concluded that the reservoir and aquifer parameters considered in this study can
significantly affect, to a varying degree, the amounts of water influx and gas production.
For infinite aquifers, higher rates of pressure depletion at the gas/water contact are found
to yield higher gas recovery factors, less number of lbm-moles of remaining gas in the
reservoir, and lower abandonment pressures. For finite aquifers, higher ratios of aquifer
size to reservoir size are shown to yield higher abandonment pressures and larger number
of lbm-moles of remaining gas in the reservoir at abandonment conditions. The gas
recovery factor, however, is found insensitive to the ratio of aquifer size to reservoir size.
The results of this work can be of great value to reservoir and production engineers
dealing with this type of gas reservoirs and should provide them with the necessary
guidelines for planning their production strategies.

Keywords: gas reservoirs, edge-water, drive mechanism, reservoir performance, material

balance, sensitivity analysis, aquifer size, sweep efficiency, water influx, abandonment
pressure, depletion rate.
104 Hazim H. Al-Attar

NOMENCLATURE
Bgabn gas FVF at abandonment pressure pabn, ft3/STB [res m3/std m3]
Bg gas FVF at pressure p, ft3/STB [res m3/std m3]
Bgi gas FVF at pressure pi, ft3/STB [res m3/std m3]
Bw water FVF, RB/STB [res m3/stock-tank m3]
cf pore compressibility, psi-1 [Pa-1]
ct total compressibility, psi-1 [Pa-1]
cw water compressibility, psi-1 [Pa-1]
e rate of reservoir/aquifer boundary pressure depletion, psia/day [Pa/day]
Eg underground gas expansion, ft3/scf [res m3/std m3]
f (encroachment angle)º / 360
F total gas and water production, ft3 [res m3]
G original gas in place, scf [std m3]
Gp cumulative gas produced, scf [std m3]
Gpabn cumulative gas produced at abandonment pressure, scf [std m3]
h net thickness, ft [m]
p pressure, psia [Pa]
pabn abandonment pressure, psia [Pa]
pi initial reservoir pressure, psia [Pa]
pn reservoir pressure at time tn,psia [Pa]
ra radius of aquifer, ft [m]
RD dimensionless radius = ra/rg
Sgr residual gas saturation, fraction less than one
Swc connate water saturation, fraction less than one
Swi initial water saturation in the gas reservoir, fraction less than one
t time variable, hrs
tD dimensionless time (see Eq. 1)
tDj dimensionless time at time step j (see Eq. 2)
TD dimensionless time at time step n (see Eq. 2)
T reservoir temperature, ºF [ºC]
We cumulative water influx, RB [m3]
WD(tD) dimensionless cumulative water influx function giving the dimensionless
influx per unit pressure drop imposed at the reservoir/aquifer boundary from
t = 0 up to TD
Wp cumulative water produced, RB [m3]
Z gas deviation factor
Zabn gas deviation factor at abandonment pressure
Zi gas deviation factor at initial reservoir pressure

Greek Letters

α volumetric sweep efficiency, per cent

Δ difference
 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 105

ε rate of reservoir/aquifer boundary pressure depletion, psia/day [Pa/day]

φ porosity, fraction less than one
γ gas specific gravity (air = 1.0), dimensionless
μ viscosity, cp [Pa.s]
Θ angle of water encroachment, degrees [rad]

Subscripts

a aquifer
D dimensionless
e influx
emin minimum influx
f formation
g gas
gi gas at initial condition
gr gas at residual condition
gabn gas at abandonment condition
i initial condition
j index of loops
n number of time steps
p produced
wc connate water

SI Metric Conversion Factors

cp x 10-3 = Pa.s
degrees x 1.745 329x10-2 = rad
ft x 3.048 x10-1 = m
ft3 x 2.831 685 x10-2 = m3
ºF (ºF - 32)/1.8 = ºC
psi x 6.894 757 = kPa
psi –1 x 1.450 377 x10-1 = kPa –1

1. INTRODUCTION
Predicting the depletion of waterdrive gas reservoirs have been investigated by many
authors. Agarwal et al. (1965) used a material balance model to study the effect of water
influx on gas recovery. They concluded that gas recovery depends on production rate,
residual gas saturation, aquifer strength, aquifer permeability, and the volumetric sweep
efficiency of the encroaching water zone. Bruns et al. (1965) studied the effect of water influx
on the p/z versus cumulative gas production (Gp) curves. They concluded that it is dangerous
to extrapolate the p/z charts on a straight line without considering the possibility of water
106 Hazim H. Al-Attar

influx. Geffen et al. (1952) conducted an experimental study of residual gas saturation under
waterdrive. They concluded that residual gas saturation under waterdrive varies from 15 to
50% pore space, depending on the type of sand. Knapp et al. (1968) developed a two-phase,
two-dimensional model to predict gas recovery from aquifer storage fields. The model was
used to study the effects of heterogeneity, aquifer strength, and gas production rates. From
their results, they concluded that gas recovery is a function of gas production rate, aquifer
strength, and heterogeneity. Their conclusions agree with those of Agarwal et al. (1965) in
terms of gas production rate and aquifer strength. Shagroni (1977) studied the effect of
formation compressibility and edge water on gas field performance. He concluded that it is
incorrect to extrapolate the early part of the p/z vs. cumulative gas production curves as a
straight line to p/z = 0.0, to estimate the initial gas inplace without considering the possibility
of water influx and the effect of formation compressibility, and that the sensitivity of the
performance curve (p/z vs. Gp) to reservoir compressibility increases as the initial reservoir
pressure increases. Pepperdine (1978) used a mathematical model to study the performance of
the Devonian gas fields in northern British Columbia. He concluded that to achieve maximum
gas recovery, the depletion process should be increased as much as possible by production
practices, and that the important factor in the low efficiency of gas recovery was water influx
rather than coning phenomenon in the portion of the Clarke Lake field that was modeled. Al-
Hashim and Bass Jr. (1988) predicted the depletion performance of partial waterdrive gas
reservoirs to study the effect of aquifer size, gas production rate, and initial reservoir pressure
on the rate at which the gas-water-contact advances and on gas recovery. Based on their
results and for constant reservoir permeability of 300 md, they concluded that regardless of
the size of the reservoir, when the ratio of aquifer radius to gas reservoir radius, ra/rg > 2.0,
the pressure in the unsteady-state water influx equation has to be corrected to the original gas-
water-contact. They also concluded that gas recovery appeared to be sensitive to initial
reservoir pressure and the aquifer size (when ra/rg > 2.0), and as ra/rg and the initial reservoir
pressure increase, gas recovery decreases. Their results indicated that gas recovery appears to
be sensitive to gas production rate when ra/rg > 3.
The objective of this study is to investigate the effects of key reservoir and aquifer
parameters that have not been thoroughly investigated before, on gas recovery factor for finite
and infinite aquifers. The outcomes of this study should provide the practicing engineers with
the necessary guidelines mostly needed in the development of gas reservoirs associated with
edge-water drive.

2. THEORETICAL BACKGROUND
The Van Everdingen-Hurst Model; “The Aquifer Fitting Model”

Van Everdingen and Hurst (1949) proposed solutions to the dimensionless diffusivity
equation, and are known as the Constant Terminal Pressure Solution (CTP) and the Constant
Terminal Rate Solution (CTR). Engineers prefer the application of the CTP solution in the
subject of gas reservoir performance under the effect of water influx, because they are more
interested in predicting rates of water influx in terms of a certain pressure drop at the
gas/aquifer boundary. The final form of the CTP solution is written as:
 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 107

We = U ΔP WD(tD) (1)

where, U is aquifer constant for radial geometry ( = 1.119 fφhct rg2) in bbl/psia, We is
cumulative water influx due to a pressure drop Δp (psia) imposed at the reservoir radius rg, at
time t = 0, in bbls, WD(tD) is dimensionless cumulative water influx function giving the
dimensionless influx per unit pressure drop imposed at the reservoir-aquifer boundary at t = 0,
f is (encroachment angle)º/360º, which is used for aquifers which subtend angles of less than
360º at the center of the reservoir-aquifer system, φ is aquifer porosity fraction, ct is total
aquifer compressibility in psia-1, tD is dimensionless time (= 2.309 kt / φμctrg2) and μ is water
viscosity in cp.
The dimensionless water influx WD(tD) is presented in tabular form or as a set of
polynomial expressions giving WD as a function of tD for a range of ratios of the aquifer to
reservoir radius RD = ra/rg, for radial aquifers. In this work the polynomial approach proposed
by Klins et al. (1988) is used and found much easier to deal with than the lookup tables or
charts that may sometimes require interpolations. Polynomial equations are available for
finite and infinite aquifers with absolute errors less than 0.03% and 0.02%, respectively.
When applying the van Everdingen-Hurst aquifer fitting model in history matching, it is
necessary to extend the theory to calculate the cumulative water influx corresponding to a
continuous pressure decline at the reservoir-aquifer boundary. In order to perform such
calculations it is conventional to divide the continuous decline into a series of discrete
pressure steps. For the pressure drop between each step, Δp, the corresponding water influx
can be calculated using Eq (1). Superposition of the separate influxes, with respect to time,
will give the cumulative water influx, and as follows:

n −1
We n (T ) = U ∑ Δp jWD (TD − t Dj ) (2)
j =0

This model is then combined with the material balance equation and the resulting
equation is used to solve for the single unknown which is the average reservoir pressure at
any time step n, (pn), within the pressure decline history.
If it is felt confident that the aquifer fitting model is satisfactory in matching the history,
then the next step is to use it in predicting the future reservoir performance. The aim here is
usually to determine how the reservoir pressure will decline for a given gas offtake rate.
Knowledge of this decline will assist in calculating the recovery factor, consistent with
production engineering and economic constraints. The basic equations are the reservoir
material balance and the water influx equation. These can be solved simultaneously, by an
iteration process, to calculate the reservoir pressure.

Gas Field Volumetric Material Balance

a) Appropriateness in Application
Whether material balance can be applied to a hydrocarbon accumulation as a whole
depends upon how rapidly any pressure disturbance is equilibrated in the reservoir so that it
108 Hazim H. Al-Attar

may be treated as zero dimensional. This, in turn, is dependent on the magnitude of the
hydraulic diffusivity constant, k/φμc; the larger the value of this parametric group, the more
rapidly is pressure equilibrium achieved. It can be shown that, in spite of the high gas
compressibility, its extremely low viscosity dominates in making the diffusivity constant
many times greater than for oil which enhances the prospect for meaningful application of
material balance, even in tight gas reservoirs.

b) Havlena and Odeh (1963) Interpretation

Neglecting water expansion and pore compaction, the material balance equation for gas
reservoirs subjected to water influx can be expressed as:

F/Eg = G + WeBw/Eg (3)

where, F (= GpBg + WpBw) is total gas and water production in rcf, Eg (= Bg – Bgi) is
underground gas expansion in rcf/scf, G is initial gas in place in scf, Gp is cumulative gas
production in scf, We is cumulative water influx in rcf, Wp is cumulative water production in
rcf, Bgi is gas formation volume factor at initial reservoir pressure in rcf/scf, Bg is gas
formation volume factor at current reservoir pressure in rcf/scf, and Bw is water formation
volume factor in rbbl/stb, usually set equal one.
Using the production, pressure and PVT data, the left-hand side of this expression should
be plotted as a function of the cumulative gas production, Gp. This is simply for display
purposes to inspect its variation during depletion. Plotting F/Eg versus production time or
pressure decline, Δp, can be equally illustrative. If the reservoir is of the volumetric depletion
type, We = 0, then the values of F/Eg evaluated , say, at six monthly intervals, should plot as a
straight line parallel to the abscissa, whose ordinate value is the GIIP. Alternatively, if the
reservoir is affected by natural water influx then the plot of F/Eg will usually produce a
concave downward shaped arc whose exact form is dependent upon the aquifer size and
strength and the gas offtake rate. Backward extrapolation of the F/Eg trend to the ordinate
should nevertheless provide an estimate of the GIIP (We ≈ 0) but the plot can be highly non-
linear in this region yielding a rather uncertain result. The main advantage in the F/Eg versus
Gp plot, however, is that it is much more sensitive than other methods in establishing whether
the reservoir is being influenced by natural water influx or not.

c) p/Z-Interpretation Technique
This is by far the most popular method of applying gas material balance. Neglecting
water expansion and pore compaction, the equation is formulated at standard conditions (scf)
as:

p/Z = pi/Zi{[1 – (Gp/G)] / [1 – (WeBwEi/G)]} (4)

The term WeBwEi/G represents the fraction of the hydrocarbon pore volume invaded by
water and consequently, the greater the influx the higher the pressure for a given offtake of
gas. In the event that there is no influx and the reservoir is of the volumetric depletion type,
then the equation may be reduced to the form:
 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 109

p/Z = pi/Zi[1 – (Gp/G)] (5)

Equation (5) is a simple linear relationship between p/Z and the fractional gas recovery. It
gives rise to the popular field technique of plotting the reservoir averaged values of p/Z , in
which the pressures are referred to some common datum level, as a function of the cumulative
gas production Gp. If the reservoir is of the volumetric depletion type, then the plot must
necessarily be linear, and its extrapolation to the abscissa (p/Z = 0) enables the effective GIIP
to be determined as Gp = G.
Alternatively, if there is natural water influx from an adjoining aquifer, the p/Z plot is, in
principle, non-linear. The technique may seem fairly straightforward but this is where the
potential danger lies in application of the p/Z plot: deciding what is and what is not a straight
line. In great many cases the plot for a water drive field will appear to be linear until a very
advanced stage of depletion when, in fact, it is not. Consequently, it is suggested that this plot
is made with an enlarged p/Z scale where the only linear portion of the plot occurs very early
in the lifetime of the field, before the water influx is significant and extrapolation of this early
trend will give a more reliable value of the GIIP, although it is still likely to be too large.

3. METHOD OF CALCULATION
The calculations were divided into three parts and as follows:
Part I: Calculations of cumulative water influx We and cumulative gas production Gp at
successive time steps are made for different ratios of reservoir size /aquifer size, pressure
depletion rates, initial reservoir pressures, initial volumes of gas in place, and angles of water
encroachment. These calculations are accomplished, as described above, by combining two
fundamental mathematical models:

A. The Van Everdingen and Hurst aquifer fitting model (1949) (constant terminal
pressure solution of the diffusivity equation).
B. Dake (1978) expressed the material balance equation which includes water expansion
and pore compaction as
⎡ B gi ⎤
G p = {G ⎢( B g − B g i ) + (c w S w i + c f ) ΔP ⎥ + 5.615W e } / B g (6)
⎣ 1 − S wc ⎦

where cw and cf are water and pore compressibility, psia-1, respectively.

To simplify the calculations, it is assumed that the cumulative water production, Wp, is
negligible and the water formation volume factor, Bw, is equal 1.0.
The gas deviation factor, Z, was estimated with the Hall-Yarborough equation (1996) as
follows.

0.06125Ppr x exp[−1.2(1 − x) 2 ]
Z= (7)
y
110 Hazim H. Al-Attar

To solve for Z, however, another empirical equation is used to solve for y using the MS
Excel solver:

y + y2 + y3 − y4
− 0.06125Ppr x exp[ −1.2(1 − x ) 2 ] +
(1 − y ) 3
− (14.76 x − 9.76 x 2 + 4.58 x 3 ) y 2 + (90.7 x − 24.2 x 2 + 42.4 x 3 ) y ( 2.18+ 2.82 x ) = 0

Part II: Calculations of gas recovery factors for infinite aquifers; ra/rg ≥ 10
The following equations are proposed to run the calculations of this part.

We min = (Original pore volume )α (1 − S gr − S wc ) (8)

where α is the volumetric sweep efficiency, dimensionless.

Dividing both sides by initial volume of gas in place (G):

Wemin Bgi
= α (1 − S gr − S wc ) (9)
G 1 − S wc

The cumulative gas produced at abandonment pressure is calculated as follows;

(G p ) abn = Initial gas in place − [Trapped Re sidual Gas + Bypassed Gas ] (10)

or

⎡ Sgr ⎤
(Gp )abn = G − ⎢αGBgi ( ) + (1−α)GBgi ⎥(Bg )−1abn (11)
⎣ 1− Swc ⎦

The recovery factor is therefore:

(G p ) abn ⎡ We min S gr Bgi ⎤

= F = 1− ⎢ + (1 − α ) ⎥ (12)
G ⎣⎢ (1 − S gr − S wc )GBgabn Bgabn ⎦⎥

Agarwal (2004) presented an empirical equation to calculate the residual gas saturation
for limestone formations as follows

A1 (100φ ) + A2 (logk ) + A3 (100S gi ) + A4

S gr = (13)
100

where: A1 = -0.53482234, A2 = 3.3555165, A3 = 0.15458573, A4 = 14.403977

 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 111

The value of Sgr thus obtained is assumed constant in all calculations.

The amount of remaining gas at abandonment conditions in the reservoir, n, in lbm-mole
is calculated by applying the real-gas law as follows;

Pabn π rg h φα S gr
2

n= (14)
Z abn (T + 460 ) R

In this work, a new graphical procedure is proposed to determine the abandonment

pressure for this case, as shown in Figure 1. Using the results of part I, plots of We/G versus P
are prepared for three different values of pressure depletion rates, e or ε. The value of Wemin/G
then is calculated with Eq (9) and used in the above figure to estimate Pabn.
The recovery factor is calculated with Eq (12) and the amount of gas remaining in the
reservoir is calculated with Eq (14). Values of recovery factor, F, amount of remaining gas, n,
and abandonment pressure, Pabn, thus obtained are plotted versus pressure depletion rates, ε.
This procedure is applied to investigate the effects of initial reservoir pressure and volumetric
sweep efficiency on F, n, and Pabn, respectively.

Figure 1. New graphical technique: Determination of pabn-infinite aquifers.

Part III: Calculations of gas recovery factors for finite aquifers; ra/rg < 10.
The following equations are proposed for the calculations of this part.

θ
We min = πφhct (ra − rg )(Pi − Pabn )
2 2
(15)
360
Since,
112 Hazim H. Al-Attar

⎡ θ ⎤
G = ⎢πφhrg (1 − S wc )
2
/ B gi (16)
⎣ 360 ⎥⎦
Then,

⎡r 2 ⎤
(c f + c w ) ⎢ a 2 − 1⎥ ( Pi − Pabn ) B gi
We min ⎢⎣ rg ⎥⎦
= (17)
G (1 − S wc )

Equating equations (9) and (17) and simplifying, an equation for the abandonment
pressure is obtained.

α (1 − S gr − S wc )
P abn = Pi − (18)
⎡r 2 ⎤
(c f + c w ) ⎢ a 2 − 1⎥
⎢⎣ r g ⎥⎦
In this part, a new graphical approach is designed to estimate the rate of pressure
depletion for each case investigated, as presented in Figure 2.

Figure 2. New graphical technique: Determination of ε-finite aquifers.

The values of Wemin, Pabn, n, and F are calculated with Eqs. (15, 18, 14 and 12),
respectively.
 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 113

4. RESULTS AND DISCUSSION

Due to space limitation, only selected figures are presented in this paper.

Part I: A. Effects of reservoir and aquifer parameters on cumulative water influx (We). A
base case is selected as a reference for comparisons; ε = 2 psia/day, pi = 5,000 psia, G = 1011
scf, and θ = 45º.
A.1) Effect of aquifer size to reservoir size ratio (RD) on We (ε = 1, 2, and 4 psi/day,
respectively): The results are shown in Figures 3-5. Regardless of RD, higher rates of pressure
depletion reduce the amount of water influx, as the later will have less time available to
encroach into the gas reservoir and catch up with the fast moving gas towards the producers.
For RD ≤ 2 (small aquifer), the effect of the aquifer on gas reservoir performance becomes
increasingly insignificant and may even be neglected. For a given value of ε and boundary
pressure, larger aquifers (RD > 2) show higher amount of water influx, but this effect is
diminished as ε is increased to 4 pisa/day. This implies that We appears to be insensitive to RD
as the gas offtake rates become increasingly high.
A.2): Effect of aquifer size to reservoir size ratio (RD) on We (θ = 180º): The results are
shown in Figure 6. The water encroachment angle enters the calculations in the aquifer
constant (U). Consequently, as θ is increased from 45º to 180º, the results show significant
increase in We (approximately by 4 folds).
A.3) Effect of aquifer size to reservoir size ratio (RD) on We (G = 1013 scf): The initial
gas in place is increased to 1013 scf and the results of calculations are shown in Figure 7. At a
given pressure and regardless of RD, larger volume of gas in place yields approximately the
same amount of water influx as the smaller one does. This may be attributed to the fact that
RD is simply a ratio between aquifer size to reservoir size, and increasing G would result in an
increase in the aquifer size in the same proportion.
A.4) Effect of aquifer size to reservoir size ratio (RD) on We (pi = 10,000 psia): To
investigate this effect, the initial reservoir boundary pressure is increased from 5,000 psia to
10,000 psia (doubled), and the results of calculations are presented in Figure 8. For RD = 6,
going from 10,000 psia down to 7,000 psia (reduction of 3,000 psia in pi) results in 17
MMbbl of water influx compared with approximately the same amount of We in Figure 4
when pi goes down from 5,000 psia to 2,000 psia. However, for RD = 4, and for the same pi
reduction, We values are 16 MMbbl and 12.5 MMbbl for pi of 10,000 psia and 5,000 psia,
respectively. Thus, the effect of pi seems to become significant for RD ≤ 4. Effects of reservoir
and aquifer parameters on cumulative gas production (Gp). A base case is selected as a
reference for comparisons, where RD =4, pi = 5,000 psia, G = 1011 scf, and θ =45º; Figure 9.
B.1) Effect of ε on Gp (RD = 10): The results are shown in Figure 10 and they reveal that
low values of ε tend to keep the p/Z vs. Gp curves at high level, while high ε permits drawing
down the gas reservoir pressure before water influx completely floods the reservoir.
Moreover, as the aquifer size (RD) is increased from 4 to 10, the p/Z vs. Gp curves depart
from the straight line relationships for a volumetric gas reservoir and tend to be at higher
levels. In fact, when RD ≤ 4 the effect of the aquifer on gas reservoir performance can be
neglected (Figure 9).
114 Hazim H. Al-Attar

ε = 1 psi/day, Pi = 5000 psi, G = 1011 SCF, θ = 45

6000
5000

P re s s u re , p s i
RD=10
4000
RD=6
3000
RD=4
2000
RD=2
1000
0
0 10 20 30 40
We , MMBBL

Figure 3. Effect of RD on We; ε = 1 psi/day.

11
ε = 2 psi/day, Pi = 5000 psi, G = 10 SCF, θ = 45

6000
5000
P re s s u re , p s i

RD=10
4000
RD=6
3000
RD=4
2000
RD=2
1000
0
0 5 10 15 20
We , MMBBL

Figure 4. Effect of RD on We; base case, ε = 2 psi/day.

11
ε = 4 psi/day, Pi = 5000 psi, G = 10 SCF, θ = 45

6000
5000
P ressu re, p si

RD=10
4000
RD=6
3000
RD=4
2000
RD=2
1000
0
0 2 4 6 8 10 12
We , MMBBL

Figure 5. Effect of RD on We; ε = 4 psi/day.

 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 115

ε = 2 psi/day, Pi = 5000 psi, G = 1011 SCF, θ = 180

6000
5000

P re s s u re , p s i
RD=10
4000
RD=6
3000
RD=4
2000
RD=2
1000
0
0 20 40 60 80 100
We , MMBBL

Figure 6. Effect of RD on We; θ = 180º.

13
ε = 2 psi/day, Pi = 5000 psi, G = 10 SCF, θ = 45

6000
5000
P re s s u re , p s i

RD=10
4000
RD=6
3000
RD=4
2000
RD=2
1000
0
0 2 4 6 8 10 12
We , MMBBL

Figure 7. Effect of RD on We; G = 1013 scf.

ε = 2 psi/day, Pi = 10000 psi, G = 1011 SCF, θ = 45

12000
10000
P re s s u re , p s i

RD=10
8000
RD=6
6000
RD=4
4000
RD=2
2000
0
0 20 40 60 80
We , MMBBL

Figure 8. Effect of RD on We; pi = 10,000 psia.

116 Hazim H. Al-Attar

11
G=10 SCF, Pi=5000 psi , RD=4, θ=45

4000

3500

3000

2500

P/z, psia
ε=4
2000 ε=2
ε=1
1500

1000

500

0
0 20000 40000 60000 80000
G p, MMSCF

Figure 9. Effect of ε on Gp; base case, RD = 4.

11
G=10 SCF, Pi=5000 psi , R D=10, θ=45

4000

3500

3000

2500
P/z, psia

ε=4
2000 ε=2
ε=1
1500

1000

500

0
0 20000 40000 60000 80000
Gp, MMSCF

Figure 10. Effect of ε on Gp; RD = 10.

11
G=10 SCF, Pi=5000 psi , RD=4, θ=180

4000

3500

3000

2500
P/z, psia

ε=4
2000 ε=2
ε=1
1500

1000

500

0
0 20000 40000 60000 80000 100000
Gp, MMSCF

Figure 11. Effect of ε on Gp; θ. = 180º.

 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 117

B.2) Effect of ε on Gp (θ = 180º): The results of performance calculations are presented

in Figure 11. They indicate that increasing the angle of water encroachment from 45º to 180º
tends to keep the p/Z vs. Gp curves at higher levels (higher pressure for the same Gp). This
effect is attributed to the larger amount of water influx into the reservoir, as the angle of water
encroachment increases.
B.3) Effect of ε on Gp (G = 1013): The results are shown in Figure 12. Increasing the
initial gas in place from 1011 scf to 1013 scf is found to further diminish the effect of the
aquifer on gas reservoir performance and the reservoir would behave volumetrically,
regardless of the rate of the reservoir boundary pressure depletion.
B.4) Effect of initial reservoir boundary pressure on Gp: The results of reservoir
performance calculations are presented in Figures 13 and 14 for RD = 4 and RD = 10,
respectively. For a given ε, when pi is doubled the p/Z vs. Gp curves tend to be at higher
levels, resulting in higher abandonment reservoir pressure.

Part II: Sensitivity of gas recovery factor (F), abandonment reservoir pressure (pabn), and
number of lbm-moles of remaining gas in the reservoir (n) to the rate of reservoir boundary
pressure depletion (ε)-case of infinite aquifers (RD ≥ 10).
A base case is selected as a reference for comparisons where G = 1011 scf, pi = 5,000 psia,
volumetric sweep efficiency (α) = 0.6, as shown in Figs. 15 and 16. The results of the base
case reveal that the recovery factor tends to be lower at lower values of ε, which is indicative
of the strong dependency of gas recovery on field gas offtake rate. Higher rates of boundary
reservoir pressure depletion would result in lower abandonment pressures, higher recovery
factors, and less number of lbm-moles of remaining gas in the reservoir.
The effect of increasing the volumetric sweep efficiency from 0.6 to 0.8 on the gas
reservoir performance curves are shown in Figs. 17 and 18, respectively. For a given value of
ε, a higher recovery factor, a lower pabn, and a lower n are obtained compared with those in
the base case. The reason for this improvement in the overall gas performance may be
attributed to less amount of gas being left behind the water front at the higher volumetric
sweep efficiency.
The effect of doubling the initial reservoir boundary pressure is shown in Figs. 19 and 20,
respectively. For a given ε, a lower recovery factor, a higher pabn and a higher n are obtained
compared with the base case. This may be attributed to the larger amount of water influx at
the higher pressure than that at the lower pressure.

Part III: Sensitivity of F, pabn, and n to RD-case of finite aquifers (RD < 10). A base case
is selected here for comparisons; α = 0.6, pi = 5,000 psia, G = 1011 psia. The results of this
base case are presented in Figs. 21 and 22. The recovery factor is very high and practically
insensitive to RD, but pabn and n seem to be significantly affected when RD ≤ 4. This may be
attributed to the fact that when RD ≤ 4, the gas reservoir behaves volumetrically, resulting in
lower pabn.
The effect of increasing α on the above parameters is shown in Figs. 23 and 24. Similar
trends to those of the base case are observed, and as expected, they show better overall
performance characteristics when compared with the base case.
Finally, the effect of increasing the initial gas in place to 1012 scf is shown in Figures 25
and 26. Similar trends and performance characteristics to the base case are observed.
118 Hazim H. Al-Attar

13
G=10 SCF, Pi=5000 psi , RD=4, θ=45

4000

3500

3000

2500

P/z, psia
ε=4
2000 ε=2
ε=1
1500

1000

500

0
0 1000000 2000000 3000000 4000000
Gp, MMSCF

Figure 12. Effect of ε on Gp; G = 1013 scf.

11
G=10 SCF, Pi=10000 psi , RD=4, θ=45

6000

5000

4000
P/z, psia

ε=4
3000 ε=2
ε=1
2000

1000

0
0 20000 40000 60000 80000
Gp, MMSCF

Figure 13. Effect of ε on Gp; pi = 10,000 psia, RD = 4.

11
G=10 SCF, Pi=10000 psi , RD=10, θ=45

6000

5000

4000
ε=4
P/z, psia

3000 ε=2
ε=1
2000

1000

0
0 20000 40000 60000 80000 100000 120000
Gp, MMSCF

Figure 14. Effect of ε on Gp; pi = 10,000 psia,RD = 10.

 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 119

Figure 15. Sensitivity of F & pabn to ε-infinite case.

Figure 16. Sensitivity of pabn & n to ε-infinite case.

120 Hazim H. Al-Attar

Figure 17. Sensitivity of F & pabn to ε; α = 0.8.

Figure 18. Sensitivity of pabn & n to ε; α = 0.8.

 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 121

Figure 19. Sensitivity of F & pabn to ε; pi = 104 psi.

Figure 20. Sensitivity of F & n to ε; pi = 104 psi.

122 Hazim H. Al-Attar

Figure 21. Sensitivity of F & pabn to RD-F.A

Figure 22. Sensitivity of n & pabn to RD-F.A.

 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 123

Figure 23. Sensitivity of F & pabn to RD-α = 0.8.

Figure 24. Sensitivity of n & pabn to RD-α = 0.8.

124 Hazim H. Al-Attar

Figure 25. Sensitivity of F & pabn to RD;G=1012scf.

Figure 26. Sensitivity of n & pabn to RD;G=1012scf.

CONCLUSION
On the basis of the results of this study for a hypothetical reservoir/aquifer case, the
following conclusions are reached. These conclusions should provide the practicing engineer
with some useful guide lines when developing gas reservoirs subjected to edge-water influx.
 Guidelines for Developing Gas Fields Associated with Edge-Water Drive 125

Part I

1. When RD ≤ 2 (small aquifer), the effect of the aquifer on gas reservoir performance
becomes increasingly insignificant and may even be neglected.
2. Water influx is extremely sensitive to the angle of water encroachment.
3. At a given reservoir/aquifer boundary pressure and regardless of RD, the water influx
is found to be insensitive to the initial gas in place.
4. The effect of pi on We seems to become significant when RD ≤ 4.
5. Low rates of gas offtake tend to keep the p/z vs. Gp curves at high level, while high
gas offtake rates permit drawing down the gas reservoir pressure before water influx
completely floods the reservoir. When RD ≤ 4 the effect of the aquifer on gas
reservoir performance may be neglected.
6. Increasing the angle of encroachment from 45º to 180º tends to keep the p/z vs. Gp
curves at higher levels.
7. Regardless of the gas offtake rate (or ε), the effect of the aquifer on gas reservoir
performance is found to be negligible as G increases from 1011 to 1013 scf, and that
the gas reservoir would behave volumetrically.
8. Doubling of pi tends to shift the p/z vs. Gp curves to higher levels, which would
result in higher Pabn

Part II

9. For a given volumetric sweep efficiency, higher rates of boundary reservoir/aquifer

pressure depletion is found to decrease pabn, increase recovery factor, and reduce the
number of lbm-moles of remaining gas in the reservoir.
10. A new graphical approach has been proposed in this work to estimate the
abandonment reservoir pressure for Infinite aquifers.

Part III

11. For a given volumetric sweep efficiency, the recovery factor seems to be very high
and practically insensitive to RD. However, Pabn and n appear to be significantly
affected when RD ≤ 4 .
12. A new graphical approach has been proposed to estimate the pressure depletion at the
reservoir/aquifer boundary in finite aquifers.

REFERENCES
Agarwal, R.G., Al-Hussainy, R., and Ramey, H.J.Jr.:“The importance of Water Influx in Gas
Reservoirs,” JPT (Nov. 1965) 1336-1342, Trans., AIME, 234.
Al-Hashim, H.S., and Bass, D.M. Jr.:“Effect of Aquifer Size on the Performance of Partial
Waterdrive Gas Reservoirs,” SPERE (May 1988), 380-386.
 126 Hazim H. Al-Attar

Bruns, J.R., Fetkovich, M.J., and Meitzen, V.C.:“The Effect of Water Influx on p/z-
Cumulative Gas Production Curves,” JPT (March 1965) 287-295.
Dake, L.P. Fundamentals of Reservoir Engineering, Elsevier, Amsterdam, 1978.
Geffen, T.M. et al.:“Efficiency of Gas Displacement from Porous Media by Liquid
Flooding,” Trans., AIME (1952) 195, 29-35.
Havlena, D. and Odeh, A.S.:“The Material Balance as an Equation of a Straight Line,” JPT
(Aug. 1963) 896-900, Trans., AIME.
John Lee and Robert A. Wattenbarger:“Gas Reservoir Engineering,” SPE Textbook Series
Vol.5, SPE, Third Printing 2004, Appendix J.
Klins, M.A., Bouchard, A.J., and Cable, C.L.:“A Polynomial Approach to the van
Everdingen-Hurst Dimensionless Variables for Water Encroachment,” SPERE (Feb.
1988) 320-326.
Knapp, R.M. et al.:“Calculation of Gas Recovery Upon Ultimate Depletion of Aquifer
Storage,” JPT (Oct. 1968) 1129-1132.
Pepperdine, L.:“The Recognition and Evaluation of Water Drive Gas Reservoirs,” Paper 78-
29-38 presented at the 1978 Petroleum Soc. Of CIM Annual Technical Meeting.
Shagroni, M.A.:“Effect of Formation Compressibility and Edge Water on Gas Field
Performance,” M.Sc. Thesis, Colorado School of Mines, Golden, CO. (1977).
van Everdingen, A.F. and Hurst, W.:“The Application of the Laplace Transportation to Flow
Problems in Reservoirs,” Trans., AIME (1949), 186, 305-324.
William, C. Lyons: Standard Handbook of Petroleum and Natural Gas Engineering, Vol.2,
Gulf Publishing Co., Texas, 1996.

View publication stats