Predicting of Oil Water Contact Level using Material Balance Modeling
of a Multi-tank Reservoir
Muslim Abdurrahman
1
, Bop Duana Afrireksa
2
Hyundon Shin
2
, Adi Novriansyah
1,3
1
Petroleum Engineering Department, Universitas Islam Riau, Pekanbaru, Indonesia
2
Department of Energy Resources Engineering, Inha University, Incheon, South Korea
3
Department of Energy and Mineral Resources Engineering, Sejong University, Seoul, South Korea
Keywords:
Oil Water Contact, Material Balance, Tank Model, Sand Production, Prediction, Reservoir Modeling.
Abstract:
Nowadays, the increase in water production becomes a problem in the oil and gas industry. Besides being a
problem, it also becomes extra energy to produce oil and gas. OWC is one of the keys for water production
determination for each layer. If the perforation at production well is at OWC or below OWC, the production
will be 100% water. In general, the log is used to determine OWC. Besides with log, tank modeling from
the material balance equation is also used to determine OWC. WH field located 15 km from Bangko Field in
Riau. This primary field has high water production with 97% water cut. Before tank modeling starts, each
layer needs to be analyzed based on its reserves, production cumulative and remaining reserves to determine
the productive layer, which can be developed in the future. Prediction can be done when history matching and
calibration process for both historical data and simulated data by software. Prediction ends in August 2021,
which is the end of development contract in WH field. From the results, it can be determined that from C sand,
the OOWC and COWC are at 2922 ft and 2883 ft with the cumulative oil production is 6.78 MMSTB. From
E sand also can be determined the OOWC at 2368 ft and COWC at 2325 ft with the cumulative oil production
is 14.57 MMSTB. From K sand, the OOWC is at 2002 ft and COWC at 1911 ft with the cumulative oil
production is 13.5 MMSTB. L sand the OOWC is at 2243 ft and COWC at 2191 ft with the cumulative oil
production is 29.17 MMSTB. From the analysis, K sand has the most significant OWC movement, which is
91 ft and it is also validated with the current log data. This sand needs more care to maintain water production.
1 INTRODUCTION
Water production is one of the common problems of
the past few years (Hudiman and Permadi, 2016).
Water production is also one of the dilemmas in oil
and gas industries, on the other side water is known
as an energy source in reservoir flow (Daneshy, 2006).
Production well at the beginning of development has
a bigger oil production than water does. As time goes
by, oil production will decrease because of several
things, there are formation damage, pump mechanical
failure, etc. This also caused by the increase in wa-
ter production (increasing of water cut), where water
movement is faster than oil. With this water produc-
tion, it can decrease production efficiency and profit
for the oil and gas company.
The method that has been used to maintain wa-
ter production is by doing workover jobs, one of the
jobs is by closing the zone, which is not productive
or it has 100% water cut which called water shut off
(Noordin, 2009). Water shut off method can be done
by using a mechanical method (packer), cementing
(squeeze), or using chemical mixtures. These meth-
ods can be used in order to maintain water production
so it will increase oil production with low expendi-
tures (Stashin, 1989).
Oil water contact is the key to determine water
production when the production reaches 100% water
cut, OWC must be at or above the perforation. Log-
ging is the common method to determine OWC po-
sition either the original one (OOWC) or even cur-
rent position of OWC (COWC). Besides that, there
are several methods to determine OWC position, there
are RFT, DST, and other good tests. The following
methods including logging data are costly and have
some limitations especially in certain reservoir issue
(Ghahri et al., 2013). Material balance is a low-cost
approach for determining OOWC or even COWC po-
sitions (Nwaokorie and Ukakuku, 2012). By material
balance also we can study the movement of OWC it-
Abdurrahman, M., Afrireksa, B., Shin, H. and Novriansyah, A.
Predicting of Oil Water Contact Level using Material Balance Modeling of a Multi-tank Reservoir.
DOI: 10.5220/0009404603310336
In Proceedings of the Second International Conference on Science, Engineering and Technology (ICoSET 2019), pages 331-336
ISBN: 978-989-758-463-3
Copyright
c
2020 by SCITEPRESS Science and Technology Publications, Lda. All rights reser ved
331
self.
Material balance is one of several methods used
estimating reserves for oil and gas reservoir and thus
allows for making the critical decisions concerning
development plans and strategies regarding the reser-
voir. It is also the simplest way to express the conser-
vation of mass in a reservoir. The material balance is
zero-dimensional, meaning that it is based on a tank
model and does not take into account the geometry of
the reservoir, the drainage areas, the position, and ori-
entation of the wells. The other uses of this concept
are to determine the size of an aquifer, encroachment
angle of the aquifer, estimate the depth of fluid con-
tact, etc (Dake, 1983).
The material balance equation mathematically de-
fines the different producing mechanisms which ef-
fectively relates the reservoir fluid and rock expansion
to the substance of fluid withdrawal. Several methods
have been developed and published applying the ma-
terial balance equation to the various types of reser-
voirs and solving the equation to obtain the initial oil
in place (N) and the ratio of the initial gas to oil (m)
in the reservoir (Havlena and Odeh, 1963). For wa-
ter drive reservoir diagnostic plot, Campbell plot is
used to determine the energy of the aquifer and the
OOIP itself by using F/Eowf vs Np plot (Campbell
and Campbell, 1978).
The general material balance equation for an oil
reservoir is expressed as:
F = NE
t
+W
e
(1)
Where the underground withdrawal F equals to the
production of oil, water, and gas corrected to reservoir
condition:
F = N
p
(B
o
B
g
R
s
) + B
g
(G
p
G
i
)
+ (W
p
W
i
) B
w
(2)
And the original oil in place is N stock tank barrels
and E is the unit per unit expansion of oil (and its dis-
solved gas), connate water, pore volume compaction,
and the gas cap:
E = (B
o
B
oi
) + (R
si
R
s
) B
g
+ m
B
oi
B
g
B
gi
1
+ (1 +m) B
oi
S
W c
C
w
+C
f
1 S
wc
(P
i
P)
(3)
WH field is a primary field, which located in Riau
Province. This field discovered in July 1972 with the
OOIP is 184.457 MMSTB. In February 2017, the av-
erage water cut of this field reached 97%. High water
cut becomes a dilemma in this field.
The purpose of this paper is making the tank
model of each most productive layer from WH field
by using IPM MBAL software and predict the OWC
movement until August 2021, which is the end of the
contract for the WH field development. The predic-
tion is used to determine the sand, which has a sig-
nificant movement of OWC. The log data is needed
to validate the OWC movement for each productive
sand.
2 GEOLOGY AND RESERVOIR
CONDITION
WH is located at Central Sumatera Basin, Indone-
sia, at Bangko Area in Riau Province. This for-
mation consist of Brown Shale Formation at Pe-
matang Valley as the source rock. The lithofacies of
Brown Shale Formation is carbonaceous and algal-
amorphous (Katz and Mertani, 1989). Where algal-
amorphous is oil prone at the upper and middle part
of Brown Shale Formation (Aman, Kamba, and Ran-
gau). Carbonaceous is the gas and light condensate
prone, which located at Kiri, Aman, Kamba, and Ran-
gau. The transition facies between algal-amorphous
and carbonaceous is also located at Aman, Kamba,
and Rangau. Pematang group (fine and medium sand-
stone from Upper Red Formation) and Sihapas Group
come as reservoir rock after the primary migration to
the hinge margin basin caused by the Pematang to-
pography, which is asymmetric. The result is, reser-
voir rocks along steep fault scarp margin and hinge
margin, which formed Telisa, Duri, Bekasap, Bangko,
Pematang, and Petani formation with a total of thick-
ness reached 3300 ft.
Figure 1: WH Field Map
ICoSET 2019 - The Second International Conference on Science, Engineering and Technology
332
WH field reservoir properties from the log data,
core, single well-tracer, and volumetric data are as
follows:
Table 1: WH Field Reservoir Properties
Formation GOR, SCF/STB 26.4
Oil Gravity, API 34.5
Gas Gravity, sp. Gravity 0.8
Water Salinity, ppm 20000
Connate Water Saturation, % 21
Porosity, % 25
3 METHODOLOGY
In this section, the methodology, which applied in
this paper will be discussed in order to build the sand
predictive material balance equation models by using
IPM – MBAL software.
Figure 2: General IPM - MBAL Workflow
3.1 Data Gathering
Proper data acquisition has to be carried out in or-
der to build a good material balance equation model
or MBAL model. Most of these data are acquired
at the early phase of field development. Either us-
ing well tests (RFT, MDT, Swab, PBU, etc) or core
test (RCAL or SCAL) data acquired are, Pressure,
Production data, PVT, Rock properties, OOIP from
the volumetric calculation, and PV fraction vs depth.
Porosity, permeability, and water connate saturated
also are obtained from existing well logs and core
data. Original oil in place (OOIP) obtained by calcu-
lating the rock properties (porosity, water connate sat-
uration, formation volume factor) and net pay thick-
ness and area from well-logs to get the OOIP math-
ematically. Effort should be made in order to under-
stand the uncertainties related to the reservoir param-
eters, which used to calculate OOIP. In cases when the
MBAL initialize volumes are different from the vol-
umetric calculated volumes, basically due to the high
uncertainty of the MBAL data which is used in the
simulation.
3.2 Sand Selection
Sand selection is needed to filter which sand is suit-
able to model and develop in the future. The screen-
ing criteria of this section initial volumetric OOIP,
production cumulative, and remaining reserves. In
this case, when the remaining reserves are too low for
a layer, it will not profit to develop. C, E, K, and L
are the selected sand based on these screening crite-
ria, which are suitable to model and develop.
3.3 Material Balance Model
The understanding of building a material balance
model for each productive layer is needed to make a
sand predictive model in material balance. It requires
basic and fundamental knowledge related to the reser-
voir structure, type, and the aquifer effect to the reser-
voir itself. Several analytical models of the aquifer
were tested in a bid to model the geometry of the
reservoir. Carter Stacy, Van Everdingen, Van Everdin-
gen modified, Hurst-Van Everdingen modified, etc
are the available aquifer models at the software. Af-
ter aquifer model selection (in this case, Hurst-Van
Everdingen modified model was selected), the model
already established to connect the reservoir volume.
The predicted OOIP which generated by the software
can be compared with the volumetric OOIP. In this
case, the generated OOIP is matched to the volumetric
OOIP for all layers (see Fig 2 for initialization model
plots).
3.4 History Matching
With the aquifer model being the key of uncertainty,
encroachment angle, ReD, aquifer permeability, and
inner/outer ratio were regressed upon the reservoir
pressure history matching process and production
data assuming reservoir volume reproduced to stock
tank condition. The regression needs to be done re-
peatedly until the deviation is lower than 5. It needs
to be done in order to validate the model due to the
aquifer model uncertainties.
3.5 Simulation
At this part, reservoir pressure over time is simu-
lated from the production history data. This simulated
reservoir pressure is compared to the measured reser-
voir pressure at the field from the input data to see
the MBAL model could replicate the actual or current
reservoir pressure which is given by the same reser-
voir energy and properties (see Fig 3 to Fig 6). Sim-
Predicting of Oil Water Contact Level using Material Balance Modeling of a Multi-tank Reservoir
333
ulated OWC from the MBAL were calibrated with
logged OWC for modeled sands (Fig 7).
Figure 3: IPM – MBAL Initialization Output
Figure 4: Pressure and Cumulative Production History
Match from K Sand
Figure 5: Pressure and Cumulative Production History
Match from L Sand
Figure 6: Pressure and Cumulative Production History
Match from E Sand
3.6 Calibration
Material balance model calibration is needed to match
the end of history matching point with the prediction
Figure 7: Pressure and Cumulative Production History
Match from C Sand
starting point in order to make prediction more vali-
dated. In this section, pseudo-prediction will be gen-
erated by using the prediction tool. Since the goal is to
predict using tank model, a well prediction model was
not used in this case. For the constraint, history pro-
duction rate and time will be used to generate pseudo-
prediction to calibrate the model. Once both points
matched, prediction can be generated next.
3.7 Prediction
After the model already matched and validated, the
next thing is the prediction of the field performance.
Prediction generated until the end of contract of this
field development (August 2021). The models were
further calibrated by running pseudo-prediction for
existing sands. Results were compared with the out-
come from another method in determining the height
of OWC as shown in Fig 8.
Figure 8: OWC Prediction
4 RESULT
Various results were discussed during the study which
involved saturation reservoir with concurrently oil
production from the oil rim. Well logs will be adopted
to verify results from MBAL models. Table 3 shown
material balance results the OWC from MBAL has
compared well with the log data. For the production
forecast, it predicted using no well prediction which
ICoSET 2019 - The Second International Conference on Science, Engineering and Technology
334
assumpted the sand production rate is decline natu-
rally due to the pressure loss at the reservoir. Pre-
diction rate will be generated by software as long
the reservoir pressure and aquifer is enough to pro-
vide energies. From the result, K sand has significant
movement of OWC, the contact moves from 2002 ft
at 1973 to 1911 ft at 2012. This 91 ft movement
in 48 years from prediction makes this sand needs
more concern due to the water production mainte-
nance. The other sand has a certain movement less
than 55 ft in 48 years.
Table 2: Predicted OOWC vs Log OOWC
Sand
MBAL
OOWC
(ft)
Log
OOWC
(ft)
Error (%)
C 2922 2925 0.103
E 2368 2366 0.085
K 2002 2002 0.000
L 2243 2246 0.134
Table 3: Predicted COWC vs Log COWC
Sand
MBAL
COWC in
2021 (ft)
MBAL
COWC in
2014(ft)
Log COWC
in 2014 (ft)
C 2922 2925 0.103
E 2368 2366 0.085
K 2002 2002 0.000
L 2243 2246 0.134
5 CONCLUSION AND
RECOMMENDATION
Sand predictive Material Balance Models have
been proved to be a quick alternative tool to de-
termine OWC movement as reservoir simulation
in the sand analysis.
Good surveillance acquisition data is needed to
provide input data. The accuracy of each data
needs to be concerned as pre-requisite to make
validate models.
Sand K has the most significant move of OWC
due to water production maintenance. It reached
91 ft in 48 years of prediction. The other sands
have certain movement below 55 ft.
Lift tables are needed and also validated to make
well predictive models.
REFERENCES
Petroleum Experts IPM-MBAL Manual.
Campbell, R. A. and Campbell, J. M. (1978). Mineral prop-
erty economics. Petroleum Property Evaluation, 3.
Dake, L. P. (1983). Fundamentals of reservoir engineering.
Elsevier.
Daneshy, A. A. (2006). Selection and execution criteria for
water-control treatment. In SPE Symposium and Ex-
hibition on Formation Damage Control, Los Angeles.
Ghahri, P., Berthereau, G., Milner, S., Orta, M. E., Sikan-
dar, A. S., et al. (2013). Estimated fluid contact using
material balance technique and volumetric calculation
improves reservoir management plan. In SPE Offshore
Europe Oil and Gas Conference and Exhibition. Soci-
ety of Petroleum Engineers.
Havlena, D. and Odeh, A. S. (1963). The material balance
as an equation of a straight line. Journal of Petroleum
Technology, pages 896–900.
Hudiman, A. and Permadi, B. Y. (2016). Analisa penentuan
laju air produksi yang optimum untuk memperlambat
water coning di lapisan tipis. JTMGB, 10(1):17–22.
Katz, B. J. and Mertani, B. (1989). Central sumatra a
geochemical paradox. In Proc 18th Indon Pet Assoc
Ann Con, volume 1, pages 403–425, Jakarta.
Noordin, F. M., e. a. (2009). Case study: Water shut off
mechanism in small, remote platform-process & chal-
lenge. In SPE European Formation Damage Confer-
ence, pages 27–29, Netherlands.
Nwaokorie, C. and Ukakuku, I. (2012). Well predictive ma-
terial balance evaluation: A quick tool for reservoir
performance analysis. In SPE Nigerian Annual Inter-
national Conference and Exhibition, Abuja.
Stashin, K. (1989). An analytical approach to determining
oil/water contact rise at utikuma field. In 40th Annual
Technical Meeting of The Petroleum Society, Banff.
Petro Society of CIM.
Predicting of Oil Water Contact Level using Material Balance Modeling of a Multi-tank Reservoir
335
APPENDIX
API : American Petroleum Institute
Bo : Current oil volume factor
Boi : Initial oil volume factor
Bg : Current gas volume factor
Bw : Current water volume factor
Cf : Formation compressibility
COWC : Current Oil Water Contact
Cw : Water compressibility
DST : Drill Stem Test
Et : Total expansion of fluid
F : Fahrenheit
FT : Feet
Gi : Cumulative gas injection
Gp : Cumulative gas production
IOIP : Initial Oil in Place
IPM : Integrated Production Modeling
M : Gas oil Ratio
MBAL : Material Balance Modeling
Software
MSTB : Thousand Stock Tank Barrel
MMSTB : Million Stock Tank Barrel
N : Initial Oil in Place
OOIP : Original Oil in Place
OOWC : Original Oil Water Contact
OWC : Oil Water Contact
PBU : Pressure Build-Up Test
ppm : Part per Million
PSIG : Pound Square Inch Gauge
PV : Pore Volume
PVT : Pressure Volume Temperature
RCAL : Routine Core Analysis
RFT : Repeat Formation Test
Rs : Current solution gas oil ratio
Rsi : Initial solution gas oil ratio
SCAL : Special Core Analysis
SCF : Standard Cubic Feet
STB : Stock Tank Barrel
Swc : Connate water saturation
We : Water influx
Wi : Cumulative water injection
Wp : Cumulative water production
ICoSET 2019 - The Second International Conference on Science, Engineering and Technology
336