A COMPARATIVE ANALYSIS OF WATER INFLUX MODELS FOR RESERVE

ESTIMATION USING MBAL ON NIGER DELTA OIL FIELDS

Deigh Yeminifie Wilfred1, Gbonhinbor Jeffery2

Department of Petroleum Engineering, Niger Delta University, Wilberforce Island-Amassoma,

Bayelsa state, Nigeria

E-Mail: [email protected]

ABSTRACT

In petroleum exploration and development, one thing of particular interest to the reservoir
engineer is accurate reserves estimates for use in the financial reporting to the Securities and
Exchange Commission (SEC) (for the USA), Department of Petroleum Resources (DPR) (for
Nigeria) and other corporate bodies. An accurate description of the volume of fluid present is
very important in quantifying the resources and selection of production techniques, rates and
overall management of the reservoir. The information obtained is also the basis for resource
development and plan making hence the need for cross examination to mitigate inherent problem
resulting from overestimation or under estimation due to the use of inappropriate aquifer models
and other inherent source of errors. In this work, oil originally in place and aquifer volumes of
reservoirs E-1000X and G-9200Y in the Niger Delta, were estimated using Hurst-van
Everdingen-Modified, Fetkovich steady state, Fetkovich semi-steady state and the Carter-Tracy
aquifer models by performing a non linear regression of average pressure against cumulative oil
production, possible causes for variance in the models also, has been highlighted. What has been
presented in this study is not really the first globally, but is, on Niger Delta oil fields; at
providing the reservoir engineers and production managers a means of determining an
appropriate aquifer model to use when calculating water influx and performing reserve
estimation and as such; cannot be said to be conclusive. Consequently, the study recommends
that the Hurst-van Everdingen-Modified model should be used and that the model should be
compared with the Carter-Tracy model.

Keywords: water influx, reservoir, reserve, STOIIP, aquifer.

INTRODUCTION very important part of reserve exploration
Accurately estimating hydrocarbon and development, adequate management of
reserves is most important in oil and gas the reservoir, and serves also as a guide for
exploration and production because it affects the selection of production techniques and
every phase of the oil and gas business. rates schedule control. There is therefore,
Quantifying the uncertainties in estimating the need for careful and adequate study of
the Original Hydrocarbon in Place (OHIP) water influx models as it is seemingly one of
can support development and investment the most neglected source of uncertainty,
decisions for individual reservoirs. The usually considered in reserve estimation. No
petroleum engineer is often faced with doubt, several research work have been
the challenges of accurately determining carried out with high degree of success, on
this volume of hydrocarbon and what is input data in MBAL reserve estimation
economically recoverable. An accurate ranging from PVT data to geological data
description of the volume of fluids present and so on; but little focus has been placed on
in the reservoir is very important, as this water influx models. Water influx is an
serves as the basis upon which resources are important parameter in determining the
quantified and the selection of production optimum production level in rate sensitive
techniques. The information obtained is also reservoirs and contributes the highest
the basis for resource development and plan primary drive mechanism to production and
making, production rates, enhance adequate also serves as the highest percentage of
schedule control and as well, for the overall primary pressure maintenance in the
management of the reservoir throughout its reservoir. Thus, the analysis of water influx
life. models is extremely an important step. The
Several uncertainties and complexities purpose of this research work is to carry out
characterize the accurate description of the a comparative study of the effects of water
volume of fluid present in a reservoir. Water influx models in reserve estimation, using
influx is one of the sources of uncertainties the material balance tool in MBAL to
influencing the common problem of perform reserves estimation, recommend a
underestimation and/or overestimation of reliable and more appropriate model that
hydrocarbon reserves. The accurate would adequately resolve it as it relates to
description or estimation of reserves forms a Niger Delta oil and gas fields. . The
reservoir management is often influenced by Brian E. Ausburu et al (1998) identified that
the type of petroleum reservoir being differences in tectonic, depositional
evaluated. For oil reservoirs with strong environment, rock type and quality,
aquifer drive mechanism, water influx into reservoir fluid types, well drainage area and
oil zone remains a primary concern. This has recovery per well are factors that has
led to water influx models being developed influence in the selection of the best method
by several authors (Ahmed T, 2000). The to estimating reserve.
models are commonly incorporated with the
Petris W.J.E in his journal, stated that for
material balance equation to promote better
general studies and large scale planning
reserve estimation (Dake L.P, 1994).
purposes statistical methods may be used to
Water-bearing rocks called aquifers project reserve within an acceptable range of
surround nearly all hydrocarbon reservoirs. uncertainty, but specific projects beyond the
These aquifers may be substantially larger earliest exploratory phase require at least a
than the oil or gas reservoir they adjoin as to minimum of physical data. The material
appear infinite in size, and/or they may be as balance method entails the application of a
small in size as to be negligible in their computer simulation in the description of
effect on reservoir performance (Ahmed T, fluid flow in a reservoir.
2000).
Fair .Jr. et al (1994) presented statistical
Reserves estimation is one of the most approaches to various material balance
essential tasks in the petroleum industry. It methods in a bid to obtain comparison for
is the process by which the economically optimum MBE analysis. In a work done by
recoverable hydrocarbons in a field, area, or Eugene et al (2004), the material balance
region are evaluated quantitatively (Demiren method improved the quality of history
2007). matched results obtained during 3D
reservoir simulation studies of a Niger-Delta
Under the SPE/WPC definitions, “Reserves
field through established connected
are those quantities of petroleum which are
hydrocarbon in-place volumes, reservoir
anticipated to be commercially recovered
compartmentalization, possible
from known accumulations from a given
hydrocarbon-water contact and most
date forward” -Petroleum Reserves
uniquely historical reservoir drive
Definitions (1997).
mechanisms and proof of an undetected with water influx than when reservoirs are
aquifer support. volumetric. They also obtained similar
results for reservoirs with initial large gas
Rafael (2011) also stated that wrong
caps.
estimate using the material balance can be
complicated by the drive mechanism Material balance
prevailing in the reservoir or combination of
Schilthus (1936) first presented a method of
drive mechanisms; solution gas drive, gas
reservoir estimation using the material
cap expansion or pushing water.
balance. Material balance is a volumetric
In a work by Walsh (1999), pressure balance which states that since the volume
uncertainty effects were studied on different of a reservoir (as defined by its initial limits)
material balance methods. He found out that is constant, the cumulative observed
the Havlena-Odeh graphical methods for production, expressed as an underground
reservoirs with initial gas caps had no withdrawal, must equal the expansion of
tolerance to pressure uncertainty. fluids in the reservoir resulting from finite
pressure drop.
Baker et al (2003) studied PVT uncertainties
Nnaemeka .E. (2011) stated that the
and their effects and they discovered that
description of the equation that is developed
such effects could be very significant on
as a material balance equation is in a strict
material balance estimates if the pressure
sense not accurate. Actually, it is a
decline is small or if a PVT correlation is
volumetric balance based on reservoir
used with no adjustment to field data.
barrels instead of a material balance based
Mc Ewen (1961) studied pressure on mass, He opinionated.
uncertainties for water influx reservoirs and
Latest work has proven the material balance
his results showed that material balance
is a valuable tool in evaluating and
method is not reliable even when there is
analyzing complex hydrocarbon reservoir
small pressure uncertainty.
behavior and compartmentalization by
Wang et al (1997) studied the influence of means of single and multi-tank models (M.
drive mechanisms on the original oil in A. E. Ela 2007).
place (OOIP) estimates. They found that it is
more difficult to estimate OOIP for reservoir
Derivation of the Material Balance balance as the equation of a straight line, the
Equation first paper describing the technique and the
The equation is derived as a volume balance
second illustrating the application to
which equates the cumulative observed
reservoir case histories. (Dake L. P.1994).
production, expressed as an underground
withdrawal, to the expansion of the fluids in Using the nomenclature of Havlena and
the reservoir resulting from a finite pressure Odeh, the above equation can be written in
drop. the following form

That is, E o=( Bo−Boi ) + ( R si −R s ) B g (rb/stb)

Underground withdrawal (rb) = Expansion represents the expansion of oil and

of oil (rb) + originally dissolved gas cap (rb) originally dissolved gas
+ change in HC pore volume (HCPV) due to
Bg
connate water Expansion (rb) E g=Boi ( −1) (rb/stb)
B gi

N p [ Bo + ( R p −Rs ) Bg ]=N B oi
[ B oi (
( B o−Boi ) + ( R si −Rs ) Bg represents
B
+m g −1 the
B gi ) C S +C
expansion
+ ( 1+m ) { w wcof thef }∆
gascap
P + ( gas
1−S wc ]
W e −W p ) B w

¿ ( 1+m ) B oi
( 1−S wc )
C w S wc+ C f
(rb/stb)

The Material Balance Expressed as a

Straight Line represents expansion of connate water and
Havlena and Odeh (1963) presented two of reduction in the pore volume
the most interesting papers ever published
on the subject of applying the material [ ]
F=N p Bo + ( ( R p−R s ) ) B g +W p B w

balance and interpreting the results. Since (rb)

the advent of sophisticated numerical
represents underground withdrawal
reservoir simulation techniques, the
Schilthus material balance equation has been Using these terms the material balance
regarded by many engineers as being of equation can be written as;
historical only; a technique used back in the
F=N ( Eo + m E g+ E f w ) +W e Bw
nineteen forties and fifties when people still
used slide rules. Their papers, described the
technique of interpreting the material
For an oil reservoir producing above the can give OHIP values that are
bubble point pressure, (undersaturated erroneously high.
reservoirs without water influx), the c) The reservoir drive indices are a
mobility ratio is zero and R p = R s = R si = 0, useful tool for determining the
mEg = 0, E f w= 0, and W e B w = 0. correctness of the material balance
solution because they sum to unity.
Hence, the material balance equation can be
The drive indices should never be
expressed as
normalized to sum to unity because
F=N EO this obscures their usefulness and
leads to a false sense of security.
For oil reservoirs with water influx, the
material balance equation is given as Dake (1994) presented an excellent
discussion of the strength and weaknesses of
F=N EO +W e
the MBE as a straight line. He point out that
the plot of F / Eg versus Gp will have a
Pletcher J. (2002), in his work, gave
shape depicting the presence, size and
the following observations with results
strength of the aquifer. The main advantage
obtained from his analysis that:
of the F / Egvs. Gp plot is that it is more

a) Failure to account for a weak water sensitive than other methods in establishing

drive can result in significant whether the reservoir is being influenced by

material balance errors. natural water influx or not. Nonlinear plot

b) The assertion of previous authors will result if the aquifer is improperly

that weak water drive exhibits a characterized.

negative slope on the Cole (for gas) He further highlighted that material balance

and Campbell (for oil) plots has is excellent at history matching production

been confirmed. A weak water drive performances, but has considerable

is much more unambiguous on these disadvantages when it comes to prediction,

plots than the commonly used plots, which happens to be the domain of

meaning that depletion drive plots numerical simulation modelling.

p Dake (1978) pointed out from the material
such as are ambiguous in the
z balance equation,
presence of weak water drives and
 F=N ( Eo + m E g+ E f w ) +W e Bw water influx. The errors are found to be
more significant for a strong aquifer.
(2.11)
Mattar L. and Anderson (2005) pointed out
That, the term Efw, can frequently be that the principles underlying this method
neglected in water drive reservoirs. are best illustrated using constant rate
production. When the flow becomes
According to him, this is not only for the
dominated by boundaries i.e. stabilized or
usual reason that the water and pore
pseudo-steady-state conditions are achieved,
compressibility is small but also because a
the pressure at every point in the reservoir
water influx helps to maintain the reservoir
decline at the same rate.
pressure and therefore the pressure
differential appearing in the Efw term is They went further to show that the material
reduced to balance calculation could be done without
shutting in our wells. Their findings utilizes
F=N ( Eo + m E g ) +W e B w
a method called the “Flowing Material
(2.14) Balance” which entails converting the

Dake (1978) further highlighted that in an measured flowing bottom hole pressures

attempt to use the equations (2.13) and (FBHP) to average pressures at a particular

(2.14) in water drive reservoirs to match the time with the accurate knowledge of the

production and pressure history of a flow rates at that time. However, while the

reservoir, the greatest uncertainty is always method has proven to be very good, it is

the determination of the water influx, We. limited to a constant flow rate and fails
when the flow rate varies. Unfortunately,
Carlos et al. (2007) observed the combined most oil reservoirs follow a varying flow
effects of pressure and PVT uncertainties in rate pattern (i.e. Pseudo-steady state and
material balance calculations and their unsteady state flow patterns/conditions).
analysis revealed that the oil originally in
place uncertainty increased for reservoirs Matter L. and Anderson (2005), went further

with the following specifications and the on to obtain an extension of the Flowing

error increases with increasing gas cap and Material Balance method. It was termed
“Dynamic Material Balance” and it is
applicable to both constant and variable flow
rates and is based on the same initial 3. Fetkovich semi-steady state model
principle. 4. Carter-Tracy model

Water Influx Models Van Everdingen and Hurst (1949) provides

A good number of models have been the exact solution to the radial diffusivity
developed for estimating water influx that is equation and therefore is considered the
based on assumptions that describe the correct technique for calculating water
characteristics of the aquifer. Due to the influx for practically all flow regimes,
inherent uncertainties in the aquifer provided that the flow geometry is actually
characteristics, all of the proposed models radial(Ahmed, T. 2000). Van Everdingen
require historical reservoir performance data and Hurst solutions are for both the
to evaluate constants representing aquifer constant-terminal-rate case and the constant-
property parameters since these are rarely terminal-pressure case of finite and infinite
known from exploration-development edge-water aquifers.
drilling with sufficient accuracy for direct When an oil well is brought on production at
application. The material balance equation a constant flow rate after a shut-in period,
can be used to determine historical water the pressure behavior is essentially
influx provided original oil-in-place is controlled by the transient (unsteady-state)
known from pore volume estimates. This flowing condition (Ahmed T, and
permits evaluation of the constants in the McKinney, P. 2005).
influx equations so that future water influx
rate can be forecasted. Fetkovich M.J (1971) proposed a model that
uses a pseudo-steady-state aquifer PI and an
For the sake of this study, the water influx aquifer material balance to represent the
models analyzed are limited to those system compressibility. Like the Carter-
commonly used and found in the MBAL Tracy method, Fetkovich's model eliminates
software. the use of superposition and therefore is

1. Hurst-van Everdingen-Modified much simpler than van the Everdingen-

model Hurst method.

2. Fetkovich steady state model However, because Fetkovich neglects the

early transient time period in these
calculations, the calculated water influx will The Material Balance Tool in MBAL
always be less than the values predicted by The material balance tool in MBAL consists
the previous two models (Fetkovich M.J.: of three main sections; the input section, the
1971). history matching and the production
Joao B. M and Osvair V. T (2007), alerted prediction section.
on the inadequacy of the use of the The input section is where the reservoir
Fetkovich aquifer model when the ratio engineer or analyst would key in the data
between the radius of the aquifer and the that describes the reservoir properties:
radius of the reservoir (outer/inner radius known and estimated reservoir parameters,
ratio) is below 12. known and estimated aquifer type and
properties, pore volume fraction versus
The Carter-Tracy method is quite similar to
depth (if available), relative permeability
the Hurst by adopting the assumption of
parameters, transmissibility parameters (if
constant water influx rate for finite time
applicable) and production and injection
periods. The principal difference between
history on a well to well basis or total tank
this method and the Hurst-van Everdingen
production. The software also has provision
models is as follows. The Hurst-van
for validation of input data.
Everdingen model assumes a constant
The history match section is where a
pressure over a time interval and thus use
graphical method ( either Havlena-Odeh,
the constant terminal pressure solution of the
Campbell, etc) is used to quantify the
diffusivity equation with the principle of
missing reservoir and aquifer properties, an
superposition to find the water influx
iterative non-linear regression is used to
function (Dake L. P., 1994).
automatically find the best mathematical fit
Theory of MBAL for a given model, and a simulation of
Schilthus’ material balance equation which production, can be run to check the validity
equates the cumulative observed production of the results of the above two techniques.
(expressed as underground withdrawal) to Production prediction section, where
the expansion of the fluids in the reservoir reservoir performance can be simulated
resulting from finite pressure drop is the assuming production and constraint
governing principle of the MBAL software. schedules, well performance definition, well
schedule or drilling program, and relative
permeability.

DISCUSSION

Brief description of field E-1000X

Reservoir E-1000X is an oil producing

reservoir located in the Niger Delta. The
reservoir is producing currently from four Figure 1-Energy plot. Drive mechanism for
(4) wells; the first well which was field E-1000X
exploratory was initiated and first produced
Fig. 3 shows a pictorial view of the
in March 1987 with oil production of
fractional contributions of energy useful in
0.356222MMSTB, and gas cumulative
hydrocarbon recovery of reservoir E-1000X
production of 0.178111Bscf without water
after the aquifer fitting has been performed. .
production. The reservoir started producing
As shown in fig. 3, it is seen that three
water after thirty (30) months of production.
drives affects the recovery of oil from the
The reservoir STOOIP was estimated at
reservoir; pore volume compressibility
206MMSTB from geological and petro
(green), fluid expansion (blue) and water
physical data.
influx (red), with water influx being the
Table 1- Geological and Petro physical Data
dominant drive.
GEOLOGICAL DATA
RES RA THIC POR SATUR OII
ERV DI KNE OSI ARATI P
OIR US SS TY ON (M
(fee (ft) (Sw) MS
t) TB)
E- 250 250 0.23 0.25 206
1000 0
Figure 2-Analytical method results for E-
X
1000X.
G- 920 100 0.25 0.15 350
Fig.4. above shows the analytical history
9200 0
matched model after regression analysis has
Y
been carried out. Regression analysis was
carried out to improve the quality of the reservoir was initiated with an exploratory
match. The red line shows the simulated well in 1987 and two additional wells were
model without aquifer strength while the later drilled. It was opened for production at
blue line shows the simulated model with an initial pressure of 2725.3Psi which was
aquifer strength after regression. The black also recorded as the bubble point pressure. It
zigzag line represents the history (measured produced a cumulative oil of
data points), and it depicts a uniform 1.25191MMSTB from the first well drilled
pressure decline. As can be seen, the blue and a high gas cumulative production of
line clearly matches with the measured data 809.484MMSCF.
points after regression, thus depicting
success in creating a zero-dimensional
model for reservoir E-1000X. It can also be
seen from fig. 4, that the reservoir pressure
is supported by an aquifer, and is seen to be
significant in maintaining the reservoir
energy.

Figure 4- Drive mechanism for field G-

9200Y

Fig. 5 shows the energy plot of field G-

9200Y. A pictorial view highlighting the
fractional contributions of energy useful in
Figure 3-Graphical plot for E-1000X after hydrocarbon recovery of G-9200Y after the
regression aquifer fitting was performed. Fig.5 shows
four energy systems sustaining the recovery

Brief description of field G-9200Y of oil and gas from the reservoir. The three

G-9200Y is a saturated oil reservoir also drives dominant in the reservoir are water

located in the Niger Delta, currently under influx (red), pore volume compressibility

production from a total of four wells. The (green) and the fluid expansion (blue). It is
seen that the water influx is the dominant
drive with high contributions from fluid syste
expansion. m
Outer 4.7961 4.2524 4.4381 4.7821
/inner 9 3 2 5
radius
Calc. 39722. 37895 41486. 39194.
aquife 2MMft MMft3 8MMft 7MMf
3 3
r t3
volu
me
Encro 160.00 187.27 171.11 160.93
Figure 5-Analytic method result for G- achm 4 3 3 4
9200Y ent
angle
Table 2A- comparison of estimated OIP and (degr
other reservoir parameters by aquifer models ees)
for field E-1000X Aquif 9.0152 10.812 6.0486 9.0389
Para E-1000X er 8md 2md 4md 4md
meter perm.
s Domi Water Water Water Water
Hurst- Fetkov Fetkov Carter nant influx influx influx influx
van ich ich -Tracy Drive
Everdi Steady Semi-
ngen State Steady
Modifi
ed
OIP 195.21 239.87 239.94 193.81
8MMS 8MMS 8MMS MMS
TB TB TB TB
Aquif Radial Radial Radial Radial
er
Table 3B- comparison of estimated OIP and er md 1md 8md
other reservoir parameters by aquifer models perm.
for field G-9200Y Domi Water Fluid Fluid Water
Para G-9200Y nant influx expans expans influx
meter Drive ion ion
s
Hurst- Fetkov Fetkov Carter-
van ich ich Tracy
Discussion and comparison of Aquifer
Everdi Steady Semi-
Models for fields E-1000X and G-9200Y
ngen State Steady
From table 1, it is observed that reservoir E-
Modifi
1000X has a geological estimated STOIIP of
ed
206MMSTB. From table 2A, after the
OIP 364.10 450.37 444.56 373.83
modeling and history matching; it is shown
6MMS 1MMS 1MMS 5MMS
that the Hurst-van Everdingen-Modified
TB TB TB TB
aquifer model gave a STOIIP of
Radial Radial Radial Radial
195.218MMSTB; the Fetkovich steady-state
Outer 8 4.3019 7.9663 7.9782
model gave 239.878MMSTB; Fetkovich
/inner 7 2 5
semi-steady 239.948MMSTB, and the
radius
Carter-Tracy aquifer model an estimated
Calc. 20987 72138. 25406 21242
STOIIP of 193.81MMSTB.
aquife 3MMft 7MMft 0 92MM
r 3 3
ft3 It is observed also, for reservoir G-9200Y
volu from table 1 that the geological estimated
me STOIIP is 350MMSTB. And from table 2B,
Encro 360 352.79 358.56 279.69 the -van Everdingen-Modified aquifer model
achm 1 9 8 gave a STOIIP of 364.106MMSTB;
ent Fetkovich steady-state model gave
angle 450.371MMSTB, Fetkovich semi-steady an
(degr estimated STOIIP of 444.561MMSTB, and
ees) the Carter-Tracy aquifer model a STOIIP of
Aquif 35md 40.42 35.993 48.873 373.835MMSTB.
The slight difference in the estimated 1971). But this is only true for the Fetkovich
STOIIP of the Hurst-van Everdingen- steady state aquifer model and not always
Modified and the Carter-Tracy aquifers as the case for the Fetkovich semi-steady state.
observed from table 2A and 2B agrees with This trend is shown in the calculated aquifer
Carter R.D and Tracy G.W (1960) on their volumes obtained from the modeled and
work where they highlighted that “the history matched parameters in tables 2A and
principal difference between the Carter- 2B for reservoirs E-1000X and G-9200Y.
Tracy method and that of Hurst is that, over The Fetkovich steady state model gave a
finite intervals of time, constant oil- calculated aquifer volume of 37985MMft3
production rates are assumed by Hurst which was lesser than the Hurst-van
whereas constant water influx rates are Everdingen-Modified model’s
assumed in the presently described method. 39722.2MMft3 and the Carter-Tracy’s
Hence, the slight difference in estimated 39194.7MMft3 for reservoir E-1000X
volumes of Oil in Place and calculated confirming Fetkovich’s statement to be true.
aquifer volume for the history matching”. Whereas the Fetkovich semi-steady state
Joao B. M and Osvair V. T (2007), alerted model’s larger calculated aquifer volume of
on the inadequacy of the use of the 41486.8MMft3 proved the statement not to
Fetkovich aquifer model when the ratio be true and all encompassing. For reservoir
between the radius of the aquifer and the G-9200Y; the Fetkovich steady state model
radius of the reservoir (outer/inner radius gave a calculated aquifer volume of
ratio) is below 12. This confirms the large 72138.7MMft3, whereas the Hurst-van
STOIIP estimates the Fetkovich steady and Everdingen-Modified and Carter-Tracy
semi-steady aquifers models give and the models gave calculated aquifer volumes of
large disparity with the geologic estimated 209873MMft3 and 212429.2MMft3
STOIIP as shown in the tables 2A and 2B respectively. Similarly, the Fetkovich semi-
for reservoirs E-1000X and G-9200Y. steady aquifer model gave a large calculated
However, because Fetkovich neglects the aquifer volume of 254060MMft3.
early transient time period in these Van Everdingen and Hurst (1949) in their
calculations, the calculated water influx will work stated that their model is basically a
always be less than the values predicted by solution of the radial diffusivity equation;
the previous two models (Fetkovich M.J.: hence, it yields an accurate estimate of water
encroachment for practically all flow system of chart to model various reservoirs.
regimes, provided that the flow geometry is Also, the boundary conditions chosen are in
actually radial. This therefore, confirms and sync with the space-time relationship that
serves as reason for the model being better characterizes the diffusivity equation. The
at water influx calculation and consequently conditions are - constant terminal rate and
reserve estimation since it agrees with constant terminal pressure. For the constant-
Ahmed .T; (2000) on his opinion on terminal-rate boundary condition, the rate of
“reservoirs with active water drive, water water influx is assumed constant for a given
influx into the reservoir contributes the period and the pressure drop at the reservoir-
highest primary drive mechanism to aquifer boundary is calculated. For the
production”. constant-terminal-pressure boundary
condition, a boundary pressure drop is
The Hurst-van Everdingen-Modified and
its Variations assumed constant over some finite time
period, and the water influx rate is

This model and the variations to it is the determined.

most suitable correlation that models fluid

flow in porous medium. This is because it is
a Laplace transform to the diffusivity CONCLUSION

equation. The Diffusivity equation is a 1. The STOIIP for reservoir E-1000X

continuous equation in both time and space; using the Hurst-van Everdingen-
and discretizing either time or space gives Modified model is 195.218MMSTB
time slices or space slices of the reservoir. and 364.106MMSTB for reservoir
Because of this, flow can be modeled as G-9200Y.
long as either initial or boundary conditions 2. The STOIIP for reservoir E-1000X
are given to suit the second order partial and G-9200Y using the Fetkovich
differential equation (PDE). Similar to steady state model is
reduced pressure and temperature values 239.878MMSTB and
used in z-factor determination (for gases), 450.371MMSTB respectively.
dimensionless form of diffusivity equation 3. The STOIIP for reservoir E-1000X
simplifies the flow equation considerably and G-9200Y using the Fetkovich
and makes it possible for the use of a single semi-steady state aquifer model is
 239.948MMSTB and Fetkovich’s statement that
444.561MMSTB respectively. “because Fetkovich neglects
4. Using the Carter-Tracy aquifer the early transient time period
model, the STOIIP for reservoir E- in these calculations, the
1000X and G-9200Y is calculated water influx will
193.81MMSTB and always be less than the values
373.835MMSTB respectively. predicted by the previous two
5. The following constraints were models (Hurst-van
observed in the use of the models; Everdingen-Modified and
I. Joao B. M and Osvair V. T Carter-Tracy). This greatly
(2007), alerted on the affects its competence in
inadequacy of the use of the making accurate estimates of
Fetkovich aquifer model OIIP as the presence of an
when the ratio between the aquifer greatly tells of the
radius of the aquifer and the quantity of oil initially in
radius of the reservoir place in the reservoir.
(outer/inner radius ratio) is 6. The Hurst-van Everdingen-Modified
below 12. This was proved to model should be used.
be true in this study from the 7. I also recommend that the model
excessively large STOIIP should be compared with the Carter-
estimates made by the models Tracy model.
for the outer/inner radius of
10 and 8 for reservoirs E- ACKNOWLEDGEMENTS
1000X and G-9200Y The authors thank Niger Delta University
respectively. and the department of Petroleum
II. Also, Fetkovich semi-steady Engineering for allowing us carryout this
model is not adequate for research.
predicting water influx
volumes as it seemingly gives
larger calculated aquifer
volumes which contradicts
REFERENCES Edition, Prentice – Hall Inc.
Ahmed .T, (2000) “Reservoir engineering Eaglewood Cliffs, NJ 56-65.
handbook”, Gulf Professional
Dake .L (1994) “The Practice of Reservoir
Publishing, Houston (TX)
Engineering”, Elsevier publisher,
Baker .R (2003) “PVT Error Analysis for Amsterdam, 94-140
Material Balance Calculation”,
Eugene .E, (2004) “Use of Material Balance
Paper presented at the Canadian
to Enhance 3D Reservoir
International Petroleum Conference,
Simulation”, paper SPE 90362
Calgary, Alberta.
presented at the SPE Annual
Buckley J.M. and Leverett M.C., (1942) Technical Conference and Exhibition
“Mechanism of fluid displacement in held in Houston, Texas, USA.
sands”, American Institute of
Hilg-Economides, C.A. and Ramey, H.J., Jr.
Mining, Metallurgical and Petroleum
(February 1981) “Transient Rate
Engineers Transactions, 146, 107-
Decline Analysis for Wells Produced
116
at Constant Pressure,” SPEJ, 98-
Carlos A. and Jose R (2007) “Pressure & 104.
PVT Uncertainty in Material
Falder, D.D., (1997) “Advanced Decline
Balance Calculations”, paper
Curve Analysis in Vapor-Dominated
107907 presented at the SPE
Geothermal Reservoirs”, SPE
Conference in Buenos, Argentina.
38763, presented at the 1997 SPE
Carter, R.D. and Tracy, G.W. 1960. “An Annual Technical Conference and
Improved Method for Calculating Exhibition, San Antonio, TX, USA
Water Influx”, Trans., AIME 219:
Ferruh Demirmen “Reserves Estimation:
415
The Challenge for the Industry” JPT
Craft, .B, Hawkins .M, revised by Terry .R. (May 2007)
(1991) “Applied Petroleum
Fetkovich, M.J., Vienot, M.E., Bradley,
Reservoir Engineering”, Second
M.D., and Kiesow, U.G. (December
1987) “Decline-Curve Analysis
 Using Type Curves−Case Histories”, Mattar L. and D. Anderson (2005)
SPE Formation Evaluation 637-656 “Dynamic Material Balance (Oil or
Gas in place without shut-ins)”,
Fetkovich, M.J. 1971. “A Simplified
paper SPE presented at the 6th
Approach to Water Influx
Canadian International Petroleum
Calculations—Finite Aquifer
Conference, Calgary, Alberta,
Systems”, J Pet Technol 23 (7): 814–
Canada.
28. SPE-2603-
PA http://dx.doi.org/10.2118/2603- Mattar L. and McNeil (1998) “The Flowing
PA Gas Material Balance” journal of
JCPT, vol 37, page 2.
Fair, Jr: (1994) “A Statistical Approach to
Material Balance Methods” paper M Abu El Ela (2007) “Reservoir
SPE 28629 presented at the SPE 69th Characterization from Material
Annual Technical Conference and Balance Results Analysis”, paper
Exhibition held in New Orleans, SPE 108648 presented at the SPE
L.A, USA. International Oil Conference and
Exhibition held in Veracruz, Mexico.
Havlena .D and Odeh .A. (1964) “The
Material Balance as an Equation of M.Abu El Ela (2007) “Data Analysis
a Straight Line”, paper 559 Methodology for Reservoir
presented at the University of Management”, paper SPE 106899
Oklahoma SPE Production Research presented at the SPE EAGE Annual
Symposium, Norway. Conference and Exhibition held in
London, United Kingdom.
Justina and Onyekonwu M. (2005)
“Determination of Reservoir Extent Ela, M. A. E., (2007) “Reservoir
Using Material Balance”, paper SPE characterization from material
98811 presented at the 29th Annual balance results analysis”, in Proc.
SPE International Technical International Oil Conference and
Conference and Exhibition held in Exhibition, Veracruz Mexico: SPE
Abuja, Nigeria. 108648.
MBALTM Manual, (2007) Reservoir Walsh .M.P (1999) “Effects of Pressure
Engineering Tool Kit, User Guide, Uncertainty on Material Balance
Version 10.5. Plots”, paper SPE 56691 presented
at the SPE Annual Technical
Mc Ewen, C (1961) “Material Balance
Conference and Exhibition held in
Calculations with Water Influx in the
Houston, Texas.
Presence of Uncertainty in
Pressures”, paper SPE 225 Walsh, M.P., Ansah, J. and Raghavan, R.
presented at the 32nd Annual (1994) “The New, Generalized
California Regional Meeting of SPE, Material Balance as an Equation of
Bakersfield, California. a Straight Line”, Part 2 -
Applications to Saturated and Non-
Pletcher, J.L (2002) “Improvements to
Volumetric Reservoirs. Presented at
Reservoir Material Balance
the Permian Basin Oil and Gas
Methods”, paper SPE 75354,
Recovery Conference, Midland,
presented at the SPE Annual
Texas, 16-18 March 1994 SPE-
Technical Conference and Exhibition
27728-MS
held in Dallas.
http://dx.doi.org/10.2118/27728-MS

Turhan, Y and Ahmadreza, K (2007) “An

Analytical Bottom Water Drive
Aquifer Model for Material Balance
Analysis”, paper SPE 103283
presented at the SPE Annual
Technical Conference and
Exhibition, San Antonio, Texas

Van Everdingen, A.F. and Hurst, W., (1949)

“The Application of the Laplace
Transformation to Flow Problems in
Reservoirs”, Trans., AIME 186, 305