Вы находитесь на странице: 1из 9

Advanced Gas Material Balance

in Simplified Format
S. Moghadam, O. Jeje, and L. Mattar, Fekete Associates Inc.

Abstract troduced a new plotting function that keeps the material balance as
Material balance has long been used in reservoir engineering prac- a straight line. Therefore, the modified material balance equation
tice as a simple yet powerful tool to determine the original gas in can be used for overpressured reservoirs. Later, Rahman et al.(2)
place (G). The conventional format of the gas material balance introduced a rigorous form of material balance equation that con-
equation is the simple straight line plot of p/Z vs. cumulative gas siders the effect of the formation and residual fluid expansion.
production (Gp), which can be extrapolated to zero p/Z to obtain The attempt to find a material balance equation for unconven-
G. The graphical simplicity of this method makes it popular. The tional gas reservoirs started when these resources became more
method was developed for a “volumetric” gas reservoir. It assumes popular. Jensen and Smith(3) proposed a simplified material bal-
a constant pore volume (PV) of gas and accounts for the energy ance equation for unconventional gas reservoirs by assuming that
of gas expansion, but it ignores other sources of energy, such as the stored free gas is negligible and consequently omitted the ef-
the effects of formation compressibility, residual fluids expansion fect of water saturation completely. However, King(4) derived a
and aquifer support. It also does not include other sources of gas comprehensive material balance equation for unconventional gas
storage, such as connected reservoirs or adsorption in coal/shale. reservoirs that accounts for the free and adsorbed gas, water en-
In the past, researchers have introduced modified gas material bal- croachment/production and water and formation compressibility.
ance equations to account for these other sources of energy. How- Seidle(5) suggested that the water saturation change does not have
ever, the simplicity of the p/Z straight line is lost in the resulting a significant effect on material balance and substituted constant
complexity of these equations. water saturation in King’s material balance.
In this paper, a new format of the gas material balance equation is This study presents an advanced, rigorous gas material balance
presented, which recaptures the simplicity of the straight line while equation and its plotting function that unifies all the previously
accounting for all the drive mechanisms. This new method uses a mentioned modifications in one equation. The new gas material
p/Z** instead of p/Z. The effect of each of the previously mentioned balance equation has the same format as traditional material bal-
drive mechanisms appears as an effective compressibility term in ance and can be plotted as a straight line with pi/Zi as y-intercept
the new gas material balance equation. Also, the physical meaning and G as x-intercept. A significant advantage of this material bal-
of the effective compressibilities are explained and compared with ance equation is that it can be used to define the total compress-
the concept of drive indices. Furthermore, the gas material balance ibility of the system; therefore, the pseudotime calculated with this
is used to derive a generalized rigorous total compressibility in the total compressibility honours material balance in all situations.
presence of all the previously mentioned drive mechanisms, which
is important in calculating the pseudotime used in rate transient Volumetric Reservoir
analysis of production data. The conventional gas material balance was derived based on the
fact that the remaining gas in the reservoir at any pressure expands
Introduction to fill the reservoir volume, which was initially occupied by G at
It has been of great interest to find G by using material balance. the initial pressure [Fig. 2(a)]. In other words, the reservoir volume
The conventional gas material balance equation was developed occupied by gas stays constant. In this situation, gas compress-
for a “volumetric” gas reservoir. Therefore, the p/Z vs. cumula- ibility is the only production mechanism.
tive gas production plot may give misleading results in some situa-
tions [e.g., when the formation compressibility is of the same order ( )
GBgi = G − Gp Bg , ........................................................................ (1)
of magnitude as gas compressibility (overpressured reservoirs)
or where desorption plays a role (coalbed methane/shale)]. Fig. 1 where Bgi is initial gas formation volume factor, Bg is gas forma-
shows p/Z vs. Gp for several scenarios with the same G. It can be tion volume factor at pressure p, G is original-gas-in-place and Gp
seen from this figure that except for the volumetric reservoir, the is cumulative gas produced.
plot is not a straight line because gas expansion is not the only drive Substituting for Bg from the real gas law, at constant tempera-
mechanism. In fact, water encroachment in water-drive reservoirs, ture, results in:
formation and residual fluid expansion in overpressured reservoirs
and gas desorption in coalbed methane (CBM) or shale reservoirs
p pi  Gp 
can have a significant role as a driving force in these cases. In these = 1-  .......................................................................... (2)
situations, where the gas expansion is not the dominant driving Z Z i  G 
force, modified material balance equations have been developed by
several researchers. Among them, Ramagost and Farshad(1) modi- The previously described equation is the well-known conven-
fied the conventional material balance equation to account for PV tional gas material balance equation.
shrinkage caused by formation and residual fluid expansion and in-
Generalized Equation
This paper (2009-149) was accepted for presentation at the 10th Canadian International
In this paper, we derive the advanced gas material equation to ac-
Petroleum Conference (the 60th Annual Technical Meeting of the Petroleum Society), count for water encroachment in waterdrive reservoirs, expansion
Calgary, 16-18 June, 2009, and revised for publication. Original manuscript received for of formation and of residual liquids in overpressured reservoirs
review 23 March 2009. Revised paper received for review 26 April 2010. Paper peer ap-
proved 28 April 2010 as SPE Paper 139428. and gas desorption in CBM and shale gas reservoirs in the same

90 Journal of Canadian Petroleum Technology

Overpressured Reservoir Water Influx and Production. In a waterdrive reservoir, the aqui-
pi /Z i fer provides pressure support for the reservoir by encroachment of
Waterdrive Reservoir
Volumetric Reservoir water into the gas reservoir. The encroached water (We) decreases
CBM Reservoir the PV available for the remaining gas [Fig. 2(b)]. The reservoir
volume changes because the net encroached water (DVwip) can be
calculated from(6):
p /Z

( )
∆Vwip = 5.615 We − W p Bw , ............................................................. (4)

where We is the water encroachment into the gas reservoir, Wp is

water produced at surface and Bw is the water formation volume
factor (5.615 is a constant used only in oilfield units).
The encroached water may be determined by using the aquifer
models in the literature, such as Schilthuis (steady state)(6), Fet-
0 kovich (pseudosteady state)(7), Carter and Tracy(8) and Van
0 Gp Everdingen and Hurst (unsteady-state)(9). Each of these has its own
assumptions and applications.
Fig. 1—Conventional plot p/Z vs. cumulative gas production.
Overpressured Reservoir. Formation and residual fluid compress-
ibility are usually small in comparison with gas compressibility.
simple format of Equation (2). However, the modification needs to Therefore, in general, ignoring the formation and the residual fluid
be started from Equation (1). expansion does not affect the gas material balance significantly.
Each of the previously mentioned effects can be added to the However, at high pressures the gas compressibility is of the same
right side of Equation (1) as a volume change term. order of magnitude as that of the formation and residual liquids.
Overpressured reservoirs are the most common example of this
GBgi = (G − G p ) Bg + ∆Vwip + ∆Vep + ∆Vd  .......................................... (3) situation, in which ignoring the effect of formation and residual
fluid expansion may result in serious overprediction of G. In over-
The explanation of each of the volume change terms is provided pressured reservoirs, the p/Z vs. Gp plot yields two distinct slopes.
in the following sections. The first slope (shallow) is in the pressure range where formation

Reservoir @ p i Reservoir @ p Produced

volume @ p


GBgi = (G−Gp )Bg


GBgi = (We−WpBw)+ (G−Gp )Bg


GBgi = (GBgi /Sgi)(c f +c w Swi+coSoi )(p i−p)
+ (G−Gp )Bg


G f Bgi = ρBVBVL(p i /(p L+p i )−p /(p L+p ))Bg
+(Gf −Gp)Bg


G f Bgi = (G f Bgi /Sgi)(c f +cw Swi+coSoi )(p i−p)

+(We−WpBw )
+ρBVBVL(p i /(p L+p i )−p /(p L+p ))Bg
+(Gf −Gp)Bg

Fig. 2—Schematic of reservoir volume at initial pressure and a lower pressure.

January 2011 91
and residual fluid expansion play a significant role, while the sec- (Ga). The red-dashed box in Fig. 2(d) shows the volume of “de-
ond slope (steep) reflects the region where gas expansion is the sorbed” gas at reservoir pressure p, which is added to the “free”
dominant production mechanism(1). Ramagost and Farshad(1) con- gas. The desorbed gas volume, which needs to be added to the right
sidered the effect of formation and residual fluid expansion by a side of Equation (1), can be calculated from (for Sgi>0):
volume change equal to
G f Bgi VL pi V p
∆Vd = ρB Bg ( − L ) ......................................... (8)
S giφ pL + pi pL + p
GBgi S wcw + c f ) ( p − p) .
1 − Sw
Advanced Material Balance Equation
Later, Rahman et al.(2) introduced a rigorous form of this volume The advanced material balance equation, with consideration of
change by integrating the compressibility equation for any sub- water encroachment/production, formation and residual fluids ex-
stance in the reservoir. The total effect of formation and the residual pansions and gas desorption, can be derived by substituting DVwip,
fluids compressibility can be added together as(2): DVep, and DVd from Equations (4), (7) and (8) into Equation (3):

 
p p p
G f Bgi = (G f − G p ) Bg + (We − W p Bw )
i i i
− ∫ c f dp ∫ cw dp ∫ co dp
Bgi G    ................... (5)
∆Vep = (1 − e p
)+S (e p
− 1)+S (e p
− 1)
S gi   G f Bgi
wi oi
  + (c f + cw S wi + co Soi )( pi − p)  ................................ (9)
S gi
When matrix shrinkage occurs during CBM production, the G f Bgi VL pi V p
(fracture) porosity containing the free gas increases. In that situa- + ρ B Bg ( − L )
tion, cf has a negative value and is a complex function of pressure. S giφ pL + pi pL + p
If cf, cw and co are constant values, a simplified form of Equa-
tion (5) can be written as: If both sides of the previously described equation are divided
Bgi G G f Bgi
∆Vep = (1 − e − c f ( pi − p ) ) + S wi (ec w ( pi − p ) − 1) + by (reservoir PV), it can be reduced to:
S gi   .......................... (6) S gi

Soi (eco ( pi − p ) − 1) 
The approximate form of Equation (6) can be found considering p p Gp
( S gi − cwip − cep − cd ) = i (1 − ) S gi , ....................................... (10)
e x≈1+x as: Z Zi Gf
∆Vep = (c f + S wi cw + Soi co )( pi − p) ................................................. (7) where cwip, cep and cd are defined as cwip, the change in PV caused
S gi by the water encroachment/production relative to the reservoir PV:

Equation (7) is the format used by Ramagost and Farshad(1).

∆Vwip 5.615(We − W p Bw )
Because of its simplicity, it is also the format that is used in this cwip = =  ................................... (10A)
paper, but for a more rigorous calculation, DVep from Equation (5) G f Bgi S gi G f Bgi S gi
should be used. The effect of formation and residual liquids expan-
sions, DVep, is depicted in Fig. 2(c). cep is the relative change of the PV caused by the formation and
residual fluid expansion (approximate form):
CBM/Shale Gas Desorption. The gas storage mechanism in a
CBM (or shale gas) reservoir is unlike that of a conventional gas
reservoir. In a typical gas reservoir, gas is stored in the pores by cep = = (c f + cw S wi + co Soi )( pi − p ) ........................ (10B)
compression. In a CBM/shale reservoir, in addition to the free gas G f Bgi S gi
(Gf) stored in the fracture network, gas is stored within the coal/
shale matrix by adsorption. As the reservoir pressure is reduced, cd is the relative change of the PV caused by gas desorption:
gas is desorbed from the surface of the matrix. The amount of
gas stored by adsorption can exceed the gas stored by compres-
∆Vd ρB Bg VL pi V p
sion. Desorption of gas is commonly described by the Langmuir cd = = ( − L )  ............................ (10C)
Isotherm as specific gas content G f Bgi S gi φ pL + pi pL + p
VL p
= , Note that the variables cwip, cep and cd are not compressibilities
pL + p (as implied by their symbol), but they represent the relative change
in the PV caused by the specific mechanism.
VL is the Langmuir volume parameter and pL is the Langmuir pres-
sure parameter. Plotting Function of Advanced
Specific gas content is the volume of gas per unit mass of coal. Gas Material Balance
Therefore, the total amount of gas adsorbed can be calculated from: Equation (10) is an easy formulation for a general material balance
equation and can be plotted as (p/Z)(Sgi-cwip-cep-cd) vs. Gp to
VL p
Adsorbed gas=Ga = ρBVB , give a straight line. However, it is derived based on the PV of the
pL + p free gas. Therefore, the straight line crosses the abscissa at Gf (free
gas volume), not G. This is an inconvenience, and is a disadvantage
where rB and VB are the density and volume of the coal, respec- of this plotting format (see Fig. 3) when compared to the conven-
tively, and VL is on a “dry, ash-free” basis. tional material balance (Fig. 1), where the abscissa is G. It is worth
The material balance equation is based on the reservoir volume mentioning that G can be found easily if Gf is known,
that the free gas occupies at the initial pressure. For CBM, this is
G f Bgi ρB VL pi
equal to Gf Bgi. In a conventional gas reservoir, G=Gf , but for a G = Gf + .
CBM reservoir, the total G includes the Gf and the adsorbed gas S gi φ pL + pi

92 Journal of Canadian Petroleum Technology

p i /Z i
p i /Z i

p /Z *


p i /Z i
p /Z

G f not G 0
(p/Z )(Sgi−cwip−cep−cd)
0 Gp
0 Gp
Fig. 4­—King’s p/Z* and p/Z plots.
Fig. 3­—(p/Z)(1-cwip-cep-cd ) and p/Z plots.
p  Gp  p
Note also that Equation (10) must be solved iteratively in the case = 1 −  i  .............................................................. (12)
of water encroachment/production because Gf appears in the cwip Z **  G  Z i **
In his work explaining CBM material balance, King(4) intro- The advantage of the Z** format is that the p/Z** values are
duced Z* as: similar in magnitude to the conventional p/Z values. As shown in
Fig. 5, p/Z** vs. Gp is a straight line that starts from the conven-
Z tional pi/Zi and extrapolates to G. This formulation and presenta-
Z* =
W −W B ρ B φ tion has simplified the applicability of the general material balance
S gi − c f + cw S wi ) ( pi − p ) − Ge G pS w + V p B pg + p
( L ) equation. The definition of Z** was derived from Equations (10)
f gi gi L
and (12) as:
and reformatted Equation (1) as:
Z ** =
 1 p p  G   G  ............ (13)
p  G p  pi  ................................................................ (11)  ( S gi − cwip − cep − cd ) + i  − 1  f
= 1−  S gi Z Zi  G f   G
Z *  G  Z i *

This equation has the same format as the conventional gas ma- Also, Z** is related to King’s Z* by the following relationship:
terial balance equation, and can be plotted as a straight line of
p/Z* vs. Gp, which extrapolates to G, as can be seen in Fig. 4. This  Z 
format has a clear advantage over that of Fig. 3 in that it extrapo- Z ** = Z *  i  .
lates to the greater practical value of G rather than Gf. Whereas this  Z *i 
format is theoretically applicable to gas reservoirs other than CBM,
the fact that the p/Z* values bear little resemblance to the conven- Equation (12) is the general material balance equation for all gas
tional p/Z values detracts from its utility. reservoirs (conventional, overpressured and CBM/shale). When
In an effort to generalize the gas material balance equation for plotted as p/Z** vs. Gp it yields a straight line, which, similar to
all reservoirs (conventional, overpressured and CBM/shale), we the conventional p/Z plot, starts from the conventional pi/Zi and ex-
have developed a Z** variable to replace King’s Z* and have re- trapolates to G.
written the gas material balance equation, Equation (10), as:
Analysis Procedure
To interpret field data, when multiple drive mechanisms exist it is
good engineering practice to select for modification the most im-
p i /Z i portant of these mechanisms and to specify constant values of the
other mechanisms.
In the following example, the rock compressibility was speci-
fied as constant and except for gas expansion, waterdrive is consid-
ered to be the important mechanism to be accounted for. A sample
procedure is presented:
p /Z

p/Z p/Z **
1. Specify cf and pi. Select the aquifer model (in this example
we chose Fetkovich).
2. For each data point, calculate cep [Equation (10B)].
3. Assume cwip = 0.
G 4. Calculate Z** using Equation (13) (note: when there is no ad-
sorption, as in this example, G/Gf is equal to 1. If adsorption is present
G Bgi ρ B VL pi
0 = 1+ ).
0 Gp Gf S gi φ pL + pi

Fig. 5—p/Z** and p/Z plots. 5. Plot p/Z** vs. Gp (Fig. 6).

January 2011 93
Curved data
3600agd u
ag e
ag dk
ag a
3400 d
ke bur
dx b
ag kaje x
First estimate of OGIP=30.5 Bscf
3200 ag
ag aedr
e a
d u
ae xac u w
3000 dk
ag l n
ae ac
xz u g
s u
r ttsse
2800 kg ac z u
e syssabsy
ab z
e qvv
k l u u uurs
ah ag
ae x
g l l abs
yabysyx y
x zyg
x xafsx
ah g
w xfac
h xx
aj x
aj x x x
ab x
af afgzy q ch
laf iif
ah agd ab
tykk aa
aa aeaag
daa g
ag jjad
agaf af
e dag
yeag slaa
aaj afabajag a
d g
akd faf
jjjj af
ke jad
aa aj
ah aj kqcjqai
2600 zaah
x ajfd
aae aj
af eaf
aj kaj k
ah e
ag ad adadcj j
d d
ab ag grae
ah efe ae
kq ai
m x
e ae ocm

p /Z **, psia
f ae
2400 z

2000 yq



1200 ah


800 ai


400 ae

0 100000 300000 500000 700000 900000 1100000 1300000 1500000 1700000 1900000 2100000 2300000 2500000 2700000 2900000 3100000

Net Cumulative Gas Production, MMscf

Fig. 6—p/Z** vs. Gp (cwip is assumed to be 0 in Z** calculation).

6. Choose/adjust aquifer parameters; calculate We. gas reservoir, GDI can be fairly small (or even negligible) in com-
7. Draw best straight line though p/Z** vs. Gp data (this data parison with the desorption drive index, DDI.
may be curved) and extrapolate to G (on the x-axis).
8. Calculate cwip [Equation (10A)]. Total Compressibility and Pseudotime
9. Return to Step 4. In fluid-flow and pressure transient analysis of gas reservoirs, pseu-
10. Repeat until acceptable straight line is obtained (Fig. 7). dovariables (pseudotime and pseudopressure) are used to linearize
the diffusivity equation. Pseudotime is defined as:
Drive Indices
Drive indices were defined to indicate the relative contribution of
t dt ,
ta = ∫
different drive mechanisms in oil recovery(9). The same concept can 0 µ ct
also be applied for gas reservoirs(10). The drive indices are defined
relative to the produced gas volume: where m is the viscosity of the gas and ct is the total compressibility
of the system. The conventional definition of ct is(11):
Gas (Compressibility) Drive Index
ct = c f + Soi co + S wi cw + S gi cg .
G Bg − Bgi )= ∆VG
The problem with the traditional definition of the total com-
G p Bg G p Bg pressibility is that it does not always honour the material bal-
ance equation. Therefore, the computed pseudotime may contain
Formation and Residual Liquids Saturation Compressibility a considerable error. Rahman et al.(12) introduced a rigorous
Drive Index pseudotime definition, which is defined by manipulating the
material balance equation. Their major assumption is that gas
∆Vep was the only mobile phase in a conventional gas reservoir. In
CDI = this paper, a more generalized form of the gas material bal-
G p Bg ance equation is used, which considers water production and is
not limited to conventional gas reservoirs. Therefore, it can be
Waterdrive Index used for waterdrive and also unconventional gas reservoirs (e.g.,
CBM/shale gas reservoirs). The detailed derivation of this pseu-
∆Vwip dotime is given in the Appendix. The total compressibility is de-
G p Bg fined as:

Desorption drive index can be added to the previously men-

 ∂cwip ∂cep ∂cd 
tioned indices as: ct =  cg  S gi − cwip − cep − cd  − − − ,
 ∂p ∂p ∂p 
G p Bg where cwip, cep, cd and their derivatives are defined in the Table 1.

Theoretically, the sum of the drive indices should equal 1. This Conclusions
is identical to Gp Bg=DVG+DVwip+DVep+DVd, as is shown sche- An advanced gas material balance equation has been presented and
matically in the right side of Fig. 2(e). the corresponding plotting function introduced; therefore, the ma-
The variables cwip, cep and cd defined in Equation (10) are re- terial balance equation can be plotted as a straight line with pi/Zi as
lated to the drive indices. The denominator in those variables is the y-intercept and G as x-intercept.
reservoir PV, whereas the denominator in the drive indices is the The similarity of the recommended plotting procedure, pi/Z**
produced gas volume at pressure p, Gp Bg. vs. Gp, to the more commonly used p/Z format is a great practical
It is worth noting that in a conventional gas reservoir, the gas advantage. It allows the use of a rigorous material balance formula-
compressibility drive, GDI, is the dominant drive mechanism, tion for complex and unconventional gas reservoirs, while retaining
whereas in an unconventional gas reservoir such as a CBM or shale the simplicity and familiarity of the commonly used p/Z format.

94 Journal of Canadian Petroleum Technology

3600agd u
ag d
ag ek
Straightened data
ag a
OGIP=19.6 Bscf
3400 b
ke br
u a
ag d ex x
kaj b
3200 u
k b
ag aeag
e u
3000 d u
ajk u
2800 er l n
agrs u
e u
2600 ah
ae ac
s lz u
ae agd l
2400 ye l u
e dag
aaj slt sruu
abd ac srysrsu w

p /Z **, psia
d jab
jg xy ss
f agah
ajaf klab
daa jg xy
2200 ae faj
k xy xyssst g
ae ae eaf
af g ab
jaa xzg
g sq
eaj aa
ah af
ag jg
jjj xab
af xe
fg zx
ah yjjfllh
ah g
kah agaf
2000 zae aeae ad
ag aj
adad af
q afaf
ae q ad
q jhfi
oad aijg
cm j

y ah
y x




600 ai

0 100000 300000 500000 700000 900000 1100000 1300000 1500000 1700000 1900000 2100000 2300000 2500000 2700000 2900000 3100000

Net Cumulative Gas Production, MMscf

Fig. 7—p/Z** vs. Gp.

The advanced gas material balance equation is used to derive a Swi = initial water saturation, %
rigorous definition for total compressibility that can be used for an- t = time, hours, s
alyzing fluid-flow in unconventional gas reservoirs, or when gas is ta = pseudotime, hours, s
not the only mobile phase. T = reservoir temperature, °F, K
Tsc = standard conditions temperature, °F, K
Acknowledgements VB = bulk volume, ft3, m3
The authors acknowledge the contribution of N.M. Anisur Rahman Vi = initial volume, ft3, m3
to this development. JCPT VL = Langmuir volume, scf/ton, m3/kg
We = Water encroachment into formation, bbl, m3
Nomenclature Wp = cumulative water produced, Bbl, m3
Bg = gas formation volume factor at time, t, ft3/scf, m3/m3 Z = Gas compressibility factor, no units
Bgi = initial gas formation volume factor, ft3/scf, m3/m3 Zi = initial gas compressibility factor, no units
Bw = water formation volume factor, bbl/stb, m3/m3 5.615 = Conversion constant in oilfield units, ft3/bbl 3
c = compressibility, 1/psia, 1/Pa DV = change in volume, scf, m3
c = in the Appendix, the summation of cwip, cep and cd DVd = change in volume caused by CBM gas desorption,
cd = relative volume change caused by CBM gas ft3, m3
∆Vd DVep = change in volume caused by formation and the
desorption, S gi residual fluids expansion, ft3, m3
GBgi DVwip = change in volume caused by water encroachment/
cep = relative volume change caused by residual fluid and production, bbl, m3
∆Vep m = viscosity, cp, Pa.s
formation, S gi rB = bulk density, lb/ft3, kg/m3
GBgi f = porosity, %
cf = formation compressibility, 1/psia, 1/Pa y = pseudopressure, psia2/cp, Pa/s
cg = gas compressibility, 1/psia, 1/Pa
co = oil compressibility, 1/psia, 1/Pa Subscripts
cs = CBM sorption compressibility (Appendix) a = adsorbed
ct = total compressibility, 1/psia, 1/Pa B = bulk
cw = water compressibility, 1/psia, 1/Pa d = desorption
cwip = relative volume change caused by water influx and e = encroachment
∆Vwip ep = expansion
production, S gi f = free
GBgi g = gas
G = original-gas-in-place, Bcf, m3 i = initial
Ga = adsorbed-gas-in-place, Bcf, m3 L = Langmuir
Gf = free-gas-in-place, Bcf, m3 o = oil
Gp = cumulative gas produced to time t, Bcf, m3 p = production
k = permeability, md, m2 r = relative
kr = permeability, md, m2 s = sorption
p = pressure, psia, Pa sc = standard conditions
psc = standard conditions reservoir pressure, psia, Pa t = total
PL = Langmuir pressure, psia, Pa w = water
q = flow rate, MMscfd, m3/s wip = water influx and production
S = gas saturation, %
Sgi = initial gas saturation, % SI Metric Conversion Factors
So = oil saturation, % bbl × 1.589 873 E-01 = m3
Soi = initial oil saturation, % cp × 1.0 * E-03 = Pa·s
Sw = water saturation, % ft × 3.048* E-01 = m

January 2011 95
Definition Comment

5.615(We − W p Bw )
cwip =
G f Bgi S gi
Waterdrive reservoir
∂cwip 5.615 ∂We
∂p G f Bgi S gi ∂p

cep = (c f + cw S wi + co Soi )( pi − p ) = ce ( pi − p )
co, cw and c f are
( ) (
= − cf 1 − cf ( pi − p ) + S wi cw 1 + cw ( pi − p ) + Soi co 1 + co ( ) ( ))
pi − p  cep is approximate form.

− c f (pi − p ) c w (pi − p )
cep = (1 − e )+S wi (e − 1)+Soi (eco (pi − p ) − 1) co, cw and c f are
= − Soi co eco ( pi − p ) + S wi cwecw ( pi − p ) + c f e
− c f ( pi − p )
) ) cep is simple form.

pi pi pi

− c f dp ∫ cw dp ∫ co dp
cep = (1 − e p
)+S wi (e p − 1)+Soi (e p − 1) co, cw and c f are
 − c f dp 
pi pi pi
without matrix
∂cep ∫ ∫ cw dp ∫ co dp shrinkage.
= −  c f e p +S wi cwe p +Soi co e p 
∂p   cep is rigorous form.
 

pi pi

∫ cwdp ∫ co dp
cep = (1 − Fφ (p ))+S wi (e p − 1)+Soi (e p − 1) co, cw and c f are
pressure-sensitive, with
 
pi pi
matrix shrinkage.
∂cep dFφ ( p ) ∫ cwdp ∫ co dp
= − − +S wi cwe p +Soi co e p  cep is rigorous form.
∂p  dp 
 

ρB Bg V p V p
cd = (1 − a − m) ( L i − L )
φ pL + pi pL + p
CBM/shale gas
∂cd ρB BgVL   p p   1 p 
= (1 − a − m)  −cg  i −  + − + 2

∂p φ   pL + pi pL + p   pL + p ( pL + p )  

hp × 7.460 43 E-01 = kW 4. King, G.R. 1993. Material-Balance Techniques for Coal-Seam and
°F (°F-32)/1.8 = °C Devonian Shale Gas Reservoirs with Limited Water Influx. SPE
lb × 4.535 924 E-01 = kilogram (kg) Res Eng 8 (1): 67–72; Trans., AIME, 295. SPE-20730-PA. doi:
ton × 9.071 847 E-01 = Mg 10.2118/20730-PA.
5. Seidle, J.P. 1993. Long-Term Gas Deliverability of a Dewatered
*Conversion factor is exact. Coalbed. J Pet Technol 45 (6): 564–569; Trans., AIME, 295. SPE-
21488-PA. doi: 10.2118/21488-PA.
6. Schilthuis, R.J. 1936. Active Oil and Reservoir Energy. SPE-
References 936033-G. Trans., AIME, 118: 33–52.
1. Ramagost, B.P. and Farshad, F.F. 1981. P/Z Abnormally Pressured Gas 7. Fetkovich, M.J. 1980. Decline Curve Analysis Using Type Curve. J  Pet
Reservoirs. Paper SPE 10125 presented at the SPE Annual Technical Technol 32 (6): 1065–1077. SPE-4629-PA. doi: 10.2118/4629-PA.
Conference and Exhibition, San Antonio, Texas, USA, 4–7 October. 8. Carter, R.D. and Tracy, G.W. 1960. An Improved Method for Calcu-
doi: 10.2118/10125-MS. lating Water Influx. SPE-1626-G. Trans., AIME, 219: 415–417.
2. Rahman, N.M.A., Anderson, D.M., and Mattar, L. 2006. New, Rig- 9. Van Everdingen, A.F. and Hurst, W. 1949. The Application of the La-
orous Material Balance Equation for Gas Flow in a Compressible place Transformation to Flow Problems in Reservoirs. SPE-949305-G.
Formation. Paper SPE 100563 presented at the SPE Symposium Trans., AIME, 186: 305–324.
on Gas Technology, Calgary, 15–17 May. doi: 10.2118/100563- 10. Pletcher, J.L. 2000. Improvements to Reservoir Material Balance
MS. Methods. Paper SPE 62882 presented at the SPE Annual Technical Con-
3. Jensen, D. and Smith, L.K. 1997. A Practical Approach to Coalbed ference and Exhibition, Dallas, 1–4 October. doi: 10.2118/62882-MS.
Methane Reserve Prediction Using A Modified Material Bal- 11. Ramey, H.J. Jr. 1964. Rapid Methods for Estimating Reservoir Com-
ance Technique. Paper 9765 presented at the International Coalbed pressibilities. J Pet Technol 16 (4): 447–454; Trans., AIME, 231.
Methane Symposium, Tuscaloosa, Alabama, USA, 12–17 May. SPE-772-PA. doi: 10.2118/772-PA.

96 Journal of Canadian Petroleum Technology

  ∂c  
12. Rahman, N.M.A., Mattar, L., and Zaoral, K. 2006. A New Method for  p Z iG f  ∂p   ∂p −q .
Computing Pseudo-Time for Real Gas Flow Using the Material Bal-   cg −  =
ance Equation. J Can Pet Technol 45 (10): 36–44. JCPT Paper No.  Z pi S gi   S gi − c    ∂t  S gi − c 
06-10-03. doi: 10.2118/06-10-03.    
13. Bumb, A.C. and McKee, C.R. 1988. Gas Well Testing in the Presence ∂p
of Desorption for Coalbed Methane and Devonian Shale. SPE Form The previously described equation can be solved for :
Eval 3 (1): 179–185. SPE-15227-PA. doi: 10.2118/15227-PA. ∂t
qS gi pi Z
Appendix: Total Compressibility and Pseudotime −
∂p G f Zi p
The purpose of this Appendix is to define ct for any reservoir type =  ....................................................... (6A)
in a general form that honours material balance. The gas material ∂t  ∂c 
cg  S gi − c  −  
balance in Equation (10) can be written as:  ∂p 
p 
k p
p p G f − Gp ( ) The next step is involving pseudopressure, ψ = 2 ∫  r  dp.

( )
S gi − c = i
Zi Gf
S gi , ................................................ (1A) Therefore, po  µz 

where p is reservoir pressure at time, t, Z is the compressibility ∂ψ 2kr p  .............................................................................. (7A)

factor, Sgi is the initial gas saturation, c is the summation of cwip, ∂p µz
cep and cd, pi is the initial reservoir pressure, Zi is the initial com-
pressibility factor, Gf is the original-free-gas-in-place and Gp is the Combining Equations (6A) and (7A) together results in
gas produced to time, t.
Equation (1A) is rearranged to: − qS gi pi Z
∂ψ ∂ ψ ∂P 2 k r p G f Zi p
= = ,
p p (G f − G p )  ................................................................. (2A) ∂t ∂p ∂ t µ Z c  S − c  − ∂c
= i S gi g  gi
Z Z i G f  S gi − c   ∂p

Next, Equation (2A) is differentiated with respect to time con- which expands to:

( )
G p = q and
∂c ∂c ∂p , where q represents the rate at
∂t ∂p ∂t 2kr  − qS gi pi  2kr S gi pi
µ 
G Z 
µG f Z i
(− q )
time, t. This results in the following equation: ∂ψ ∂ ψ ∂P  f i  
= = =
∂t ∂ p ∂t c  S − c  − ∂c c  S − c  − ∂c
g  gi  ∂p g  gi  ∂p
  ∂c ∂p  
(G f − G p )  −                                                    ............................................... (8A)
∂  p  pi S gi  − q  ∂p ∂t    .................. (3A)
  =  − 2

∂t Z Z iG   S gi − c   S gi − c   There is a relationship between pseudopressure and pseudo-
  time(12) as:
 
The chain rule can be applied to the right side of Equation (3A) as:
∂ψ 2 pi qkr S gi
=−  ..................................................................... (9A)
∂t a GZ i
∂  p  ∂  p  ∂p  1 p ∂ Z  ∂p .
 =   = −
∂t  Z  ∂p  Z  ∂t  Z Z 2 ∂p  ∂t Therefore, Equations (8A) and (9A) together results in:

1 1 ∂Z 2kr S gi pi
Also, remembering that cg = − , and then multiplying ( − q) 
p Z ∂p ∂ t a ∂t a ∂ψ µG f Z i G f Zi 
p = = − 
by , results in: ∂t ∂ψ ∂t c  S − c  − ∂ c  2 pi qkr S gi 
Z g  gi  ∂p

∂t a 1
∂  p  p ∂p  .................................................................... (4A) =
  = cg ∂t  ∂c   ..................................................... (10A)
∂t  Z  Z ∂ t µ  cg  S gi − c  − 
 ∂p 
Substitute Equation (4A) into (3A): t dt
Also, pseudotime is defined as ta = ∫ , which gives,
 ∂c ∂p  0 µct
(G f − G p )  − c ∂t a 1
p Z i G f ∂p −q  ∂p ∂t   ................. (5A) = . Therefore,
cg = − 2 ∂t µct
Z pi S gi ∂t  S gi − c   S gi − c 

 ∂c 
(G f − G p ) p ZiG ct =  cg  S gi − c  −   ......................................................... (11A)
Equation (1A) can be rearranged as  = .  ∂p 
 S gi − c  Z pi S gi
Depending on what kinds of assumptions are used for reservoir
type, the previously described equation changes to match the reser-
Therefore, Equation (5A) becomes voir. The most general form of the total compressibility equation is:

January 2011 97
 ∂cwip ∂cep ∂cd 
ct =  cg  S gi − cwip − cep − cd  − − −
 ∂p ∂p ∂p 
For CBM reservoirs, cg ( − cd ) − can be further simplified to:
∂cd ρB BgVL  pL .
cg ( − cd ) − =
∂p φ  ( pL + p )2 

The previously described formulation removes the dependence

of the CBM terms on initial pressure. It is defined as the sorption
compressibility, cs(13).
This changes the form of the total compressibility equation to
the following:
 ∂cwip ∂cep 
ct =  cg  S gi − cwip − cep  − − + cs  .
 ∂p ∂p 

Summary of Equations

p p  Gp 
( )
S gi − cwip − cep − cd = i 1 −  S gi
Zi  G f 

p p  Gp 
= i 1− 
Z ** Z i **  G

Z ** =
 1 p p  G G
 ( S gi − cwip − cep − cd ) + i  − 1  f
 S gi Z Zi  G f   G

 ∂cwip ∂cep ∂cd 

ct =  cg  S gi − cwip − cep − cd  − − −
 ∂p ∂p ∂p 


Samane Moghadam is a reservoir engineer

with Fekete Associates Inc., where she is
working on software development-related
R&D projects. She holds B.Sc. degrees in
mechanical engineering and petroleum engi-
neering from Sharif University of Technology
in Iran and an M.A.Sc. degree in petroleum
systems engineering from the University of

Oluyemisi Jeje has been with Fekete and As-

sociates Inc. for over 8 years as a member of
the R&D department. He holds a B.Sc. de-
gree in chemical engineering from the Uni-
versity of Calgary.

Louis Mattar is the president of Fekete As-

sociates Inc. He specializes in the analysis of
production data and well tests. He has au-
thored over 60 technical publications. He has
received the Society of Petroleum Engineers’
Distinguished Author Award and the Out-
standing Service Award. In 2003, Mattar was
the SPE International Distinguished Lecturer
in well testing.

98 Journal of Canadian Petroleum Technology