You are on page 1of 14

I.

INTRODUCTION RHOMBO CASE

In this project, we will do a history matching and forecasting in Rhombo field. Rhombo
characteristics and properties are given bellow:

1. Geometry
 Top reservoir at 1960 m TVDSS
 Reservoir thickness of 50 m
 Thickness = 37.7 m

Figure 1. Rhombo Reservoir Model

2. Petrophysics
 Net porosity = 20.1 %
 Net permeability = 62.1 mD
3. Fluid Properties
 Oil properties
 Stock tank oil density= 849.7 kg/m3  Oil volume factor = 1.15 vol/vol @Psat
 Gas solution factor = 124.1 m /m3 3  Compressibility = 0.5 x 10-4 bar-1
 Saturation pressure = 220 bara  Viscosity = 1.20 cP @ Psat
 Gas properties
 Stock tank oil density = 0.9 kg/m3
 Gas volume factor = 0.0059 rm3/m3 @220 bara
 Viscosity = 0.026 cP @220 bara
 Water Properties
 Water density = 1000.5 kg/m3  Formation volume factor = 1.01 vol/vol
 Compressibility = 0.44 x 10-4 bar -1 @250 bara
 Viscosity = 0.481 cP
4. Initial State
 Initial pressure =250 bars @2000 m TVDSS
 Water oil contact = 2160 m TVDSS (assumed)
5. Aquifer Activity
Unknown
6. Production data
 Well P3 put in to production for 1462 days (4 years). Its production data history are
shown in the handbook.

II. OBJECTIVES

The objectives of this project are:

1. Find the production mechanism


2. Identify the two most influent parameters by running the sensitivity test.
3. Determine the value of these 4 parameters corresponding to the best match with
numerical and analytical aquifer.
4. Do forecasting for the next 9 years.
5. Optimize the oil production.

III. RESULTS AND DISCUSSION

a.) Production mechanism

When we run the history data of the rhombo field, we can find the mechanism of
production as graph in the result section. This graph can explain how much oil production in
every mechanism that held in the system as shown bellow.

From Figure 2, we can find that at the first, oil expansion is the most important
driving force, but it is decrease by the time. In other side, we can see also that water give an
effect also in the beginning and it is increasing significantly than before. Gas does not exist
at first (because reservoir pressure is higher than bubble point pressure), but it will affect
the production until 4 years later. And rock expansion does not effect too much.

According to that, we know that the driving mechanism from this history data is
water drive. This reservoir has a strong water drive in the system. And we know also the
GOR will be increasing as much as dissolved gas takes place in the driving mechanism. So,
for the simulation, we will concern in the water drive mechanism to match the history data.
water

oil gas

rock

Figure 2. The driving mechanism of Rhombo field.

b.) Identify the two most influent parameters

After knowing the driving mechanism for this reservoir, we will investigate the effect of
changing some parameters such as: aquifer volume, permeability in the lowest layer, anisotrophy
ratio (kv/kh), and maximum relative permeability of water by running a sensitivity test. The
production datas observed for each changing parameter are: cumulative oil production, final liquid
rate, final BHP, water breakthrough, final water cut, and final gas oil ratio. First, we will define the
base case and compare each changing parameter to the base case, then choose the two most
influent parameters giving the biggest difference (according to the base case). The scenario
of each case can be seen in this following table :

Table 1. The scenario for identifying the two most influent parameters
No Case Aquifer Volume (MULTPV) Layer 5 (MULTX) Kv/Kh Krw maximum
0 Run (0) 50 1 0.05 0.3
1 Low aquifer 1 1 0.05 0.3
2 High aquifer 100 1 0.05 0.3
3 Low MULTX 50 0.2 0.05 0.3
4 High MULTX 50 20 0.05 0.3
5 Low Kv/Kh 50 1 0.01 0.3
6 High Kv/Kh 50 1 0.1 0.3
7 Low Krw 50 1 0.05 0.2
8 High Krw 50 1 0.05 0.4
Here is a short explanation about each parameter :

a. Aquifer volume

We tried simulating a water drive mechanism, so we need to make a great amount of


water volume in the aquifer (infinite act of acquifer). We will command Eclipse to increase
the aquifer pore volume by using MULTPV.

By doing MULTPV sensitivity, we can know the behaviour of oil rate production, gas oil
ratio, water cut, and bottom hole pressure. Increasing the value of MULTPV will increase oil
production rate and decrease GOR. It is because the increasing MULTPV value will give
higher energy to support the resevoir pressure (like a pressure maintenance). Oil production
will increase because of water expansion. This water expansion also can maintain the
reservoir pressure to bubble point pressure, so the GOR will decrease.

b. Permeability in the lowest layer

The next step is we want to know effect of changing the permeability in the lowest
layer. By doing this simulation, we can know the behaviour of water drive filling at the
bottom layer (because water can be much more mobile in the bottom layer). We can
command Eclipse by giving MULTX features. We changed the value of MULTX from 0.2 – 2.0.
The results are : cumulative oil production, BHP, and water cut performance have a positive
relationship, and FGOR has a negative relationship. Increasing MULTX give a better
distribution of water below the reservoir. It can mantain the reservoir pressure into bubble
point pressure (WBHP is stabilized) and give a good oil production. A big permeability in the
lowest layer will decrease the gas oil ratio because the reservoir pressure is maintained, but
will increase the water cut because water becomes much more mobile.

c. Anisothrophy Ratio (Kv/Kh)

Anisothrophy means the permeability of fluid to flow to every directions is not the
same. It means also that reservoir have a heterogenetic composition. Anisotropy < 1 means
horizontal permeability is better than vertical, and anisotropy >1 means that the vertical
permeability is better than the horizontal. Anisotropy = 1, means isotropy, vertical and
horizontal permeability are the same.

By running a sensitivity test, we can know that increasing anisotrophy will increase the
cumulative oil production, decrease gas oil ratio, and no significant difference with BHP.

d. Maximum Relative Permeability of water

Water relative permeability will determine amount of water flowing to the reservoir
and pushing the hydrocarbon to the well. Greater value of Krw will give bigger energy to the
system so it can stabilize the reservoir pressure, improve cumulative oil production, and
reduce the gas coming out from the solution. The great amount of Krw can also cause an
earlier water breakthrough. From sensitivity test, we can see the effect of changing water
relative permeability, they are a positive relationship with FOPT and FWCT, and negative
relationship with FGOR and WBHP. Increasing the relative permeability of water will make
the water to flow more rapid in the reservoir, these will make the water encroaching much
more so that the oil produce much more better, FGOR will decrease because the pressure in
the reservoir are preserve better.

Here are the comparison of all case :

Water cut performance (FWCT, %) for


Cumulative oil Production each case can be seen in this table below :
1

0.8
Base case 0.015543
FOPT ( 10E6 sm3)

0.6 MULTX 0.2 0.02107


MULTX 2.0 2.884534
0.4
Krwr 0.2 0.010453
Krwr 0.4 0.020825
0.2
PERMZ 0.01 0.015602
0 PERMZ 1.0 0.016395
Base Case MULTPV1 MULTPV100
MULTPV 1 0.019019
MULTX 0.2 MULTX 2.0 PERMZ 0.1
MULTPV 100 0.015132
PERMZ 0.01 Krw 0.2 Krw 0.4

WBHP Gas Oil Ratio


140 700

600
Gas Oil Ratio (sm3/sm3

135
500
WBHP (barsa)

130 400

300
125
200
120
100

115 0
Base MULTPV 1 MULTPV 100 Base case MULTPV 1 MULTPV 100
MULTX 0.2 MULTX 2.0 PERMZ 0.1 MULTX 0.2 MULTX 2.0 PERMZ 0.1
PERMZ 0.01 Krwr 0.2 Krwr 0.4 PERMZ 0.01 Krwr 0.2 Krwr 0.4

Figure 3. Sensitivity test

From the 3 graphs above, we can conclude that the two most influent parameters are :
aquifer pore volume (MULTPV) and permeability in the lowest layer (MULTX). It gives a
significant difference if we compare it to the base case.
c.) Determine the value of these 4 parameters corresponding to the best match with
numerical and analytical aquifer.

The primary objective of history matching are to test and to improve the reservoir
model. History matching will contribute to an understanding of the current status of the
reservoir, including fluid distribution, fluid movement, and verification or identification of
the current depletion mechanism.

Aquifer needs to be taken into account in history matching because during production,
the change in pressure results in water influx. If the pressure in an aquifer can be
calculated, the resulting volumetric changes can be determined from the pressure/volume
compressibility relationship.

There are two ways to model an aquifer:

 Numerical aquifers
 Analytical aquifers

In numerical aquifer, the aquifer is represented by a set of cells in the simulation which
may then connect to specified faces of the reservoir. In analytical aquifers, the aquifer is
considered as one whole unit and in this case the aquifer uses the Carter-Tracy model. In
this history matching, both models were applied to evaluate the best scenario.

Numerical/Gridding Aquifer

The first step in history matching is to do the pressure match. In pressure match, the
control mode is set to be “RESV”. It means that reservoir fluid volume rate is calculated
corresponding to the observed phase flow rates using the reservoir pressure. That pressure
determines how much volume in the reservoir that can be recovered. The pressure of the
model has to be matched to pressure history because flow directions are not correct if
pressure is not correct.

To match the pressure two most influent parameters from sensitivity test can be
changed to find the best match as explained before. MULTPV parameter is the pore volume
aquifer multiplier. Higher value of MULTPV for aquifer represents bigger aquifer volume.
Increasing the value of MULTPV will increase oil production rate, decreasing GOR, and give a
little effect of reducing BHP. MULTX parameter represents permeability in the bottom of
reservoir. So, by increasing this value we give a stronger water drive to the reservoir. Hence,
high value of MULTX will increase oil production rate, decreasing GOR, and increase the
BHP.

The two parameters is not enough to make a satisfactory result. Water cut remains as
a problem to be matched. We found out that to find the best match in water cut, water
relative permeability parameter needs to be adjusted. High value of Krw will result in earlier
water breakthrough and on the contrary low value of Krw will result in latter water
breakthrough. After best match in pressure is found, the control mode is changed into
“ORAT”. It means that the well will be produced based on the observed oil rate and the
other phases will be produced according to their mobility ratios. The same steps in pressure
match is applied. The match in water cut is still not satisfactory, but if the other parameters
were changed, it would give poorer result to the other match. If we try to go closer to
history, the water cut will be crossed in future time which will give a significant effect on the
forecasting future performance. So, in order to get a good match in FGOR, FOPT, and WBHP,
we put FWCT as the least priority. The parameters that give the most satisfying result in
history matching with numerical aquifer are: MULTPV=100, MULTX= 2, Krw=0.3,
Kv/Kh=0.07. The results can be seen in Figure 4 to Figure 7.

Current model

History

Figure 4. History matching of FGOR

Current model

History

Figure 5. History matching of FOPT


History
Current
model

Figure 6. History matching of FWCT

Current model

History

Figure 7. History matching of WBHP

Analytical Aquifer
In analytical aquifer first thing to do is pressure matching. We will control the scenario
in ‘RESV’ mode. Because we have to determine the pressure in which it will affect to get as
much as oil volume in the reservoir.
How to control the pressure is by changing 4 parameters which are permeability,
porosity, compressibility, and krw until we get the best scenario. We match the current
model with the history by using these graphs: WBHP (for pressure), FGOR (for gas oil ratio),
FOPT (for total oil production), and FWCT (for watercut). After we have the best scenario,
what we do is changing the control mode from ‘RESV’ mode to ‘ORAT’ mode. It is to honor
the production rates. We do the same procedure like ‘RESV’ mode, until we get the best
scenario.
The best scenario for analytical aquifer in ‘ORAT’ mode is : k = 200 mD, Ø = 0.4, Cr = 6
x 10-4 psi-1, Krwr = 0.35. The results can be seen in Figure 8 and Figure 9

History FWCT

History WBHP

Current model
WBHP

Current
model FWCT

Figure 8. History matching of WBHP and FWCT

Current model
FGOR

History and
current model
History FGOR FOPT

Figure 9. History matching of FGOR and FOPT

We found that porosity doesn’t give a significant change in history matching. While
changing in permeability will make the fluid flow easier and the aquifer will give a better
pressure support to the reservoir. Higher value of compressibility means higher aquifer
volume. And the last, Krwr, have already explained in numerical/gridding aquifer.

Then we compare between analytical’s best scenario with numerical’s best scenario. The
parameter is which of the FWCT graph shows the “best” linear line between the current
model and history. We can see from both graphics that the numerical give the best linear
line, in which it is not crossing between the current model and history. This is very
important, because the next step we’re going to do is forecasting the oil production. If we
have a crossing line between current model and history, it will give an incorrect prediction
of future performance that will be far from reality. With this consideration, we choose
numerical/gridding scenario to be forecasting in the next step.

d.) Do forecasting for the next 9 years

After we have the best scenario for history matching, the next step is forecasting the data to
predict oil production from year 2007-2016. The best scenario we used is:
Aquifer MULTPV : 100
Layer 5 MULTX : 2.0
Kv/Kh : 0.07
krw max : 0.3
with ORAT-gridding aquifer.
Here, we will have three modified cases from the original base case. The purpose is to find the most
optimum production guideline.
First case, we defined BHP > 80 bar, WCT < 90%, uptime = 85%, and no production guideline which
means we didn’t set any gas rate for the production.

Figure 10. Forecasting of FLPR with no production guideline

Second case, we used same production constraints, only here we used production guideline and set
the gas rate up to 50000 m3/day. The purpose of giving production guideline is to control the amount
of gas and oil produced.
Figure 11. Forecasting of FLPR with gas rate production guideline

Figure 12. Forecasting of FGOR with gas rate production guideline

We can see from Figure 11 and Figure 12 that when we set the gas rate to 50000 m3/day the
gas oil ratio and the liquid rate both are decreasing. Normally, when the gas oil ratio
decreasing, the liquid rate is also decreasing. This phenomenon is because of the pressure
difference between the well bottom hole pressure and well block pressure 9 (WBP9) as we
can see from the graph below. ΔP between WBHP and WBP9 for gas rate 50000 m3/day is
smaller than ΔP between WBHP and WBP9 for no production guideline.
Figure 13. Forecasting of WBHP with gas rate production guideline

Third case, is still with the same production constraints. Only we modified the production
guideline by set the gas rate vary until we get the most optimum, which means almost the
same or even higher than FLPR than the one without production guideline (case 1).

Figure 14. History matching of FGOR and FOPT

From graph above we can see that gas oil ratio is proportional to liquid production. The
higher gas oil ratio, the higher liquid rate is. At the point we set 200000 m3/day, the liquid rate
give the same number with when it is not set to any rate. So, in the end, we use this gas rate
because if we set a higher gas rate, the liquid rates will not increase anymore because the
maximum amount of gas that can be produced by the reservoir apparently is only up to
200000 m3/day. After we get a satisfying result from forecasting, the next step is to define the
best perforation intervals for production optimization.

e.) Optimize the oil production


After determined the production guideline (BHP< 50 bars, WCT <95%, uptime=95%),
production is optimized by setting the perforated intervals. The perforation is done
on layer 1 to layer 5 with the combination of : 1-1, 1-2, 1-3, 1-4, 1-5, 2-2, 2-3, 2-4, 2-
5, 3-3, 3-4, 3-5, 4-4, 4-5, and 5-5. In determining the best perforation combination,
the parameters to be considered are: FGOR, FWCT, FOPT, and WBHP. The goal of
optimizing the perforation is to find the best combination that gives the highest
production of oil and the lowest production of gas and water cut. But in reality,
higher oil production will give higher gas production and higher water cut as an
effect. The difference between static pressure and flowing pressure gives potential
to the flow of oil. So, higher ΔP will also give higher production of oil.
Of all the perforation combinations, the option is narrowed down to two
combinations which gives relatively high production of oil and reasonable amount of
water cut and GOR.

Parameters Original perforated Proposed perforated intervals


intervals
Layer 1-5 Layer 1-4 Layer 3-5
FOPT (sm3) 1873710.1 1873583.3 1890783
FWCT (%) 86.7 83.43 87.04
FGOR 630.55 513.57 620.15
(sm3/sm3)

The chosen perforated intervals is perforation from layer 1-4. Compared to


perforations from layer 3-5, perforation from layer 1-4 gives a good production
amount of oil produced and a significantly low production of water and gas. It can be
seen from the table and from the graph below, that with small difference amount of
oil produced, the water cut in perfotration layer 1-4 is only 83.43% compared to
87.04%, and also the gas production is much lower.

Oil production GOR

Water cut

Figure 15. Perforation optimization


IV. CONCLUSIONS
1. From the two history matching using analytical aquifer and numerical aquifer, we
can conclude that history matching using gridding aquifer gives a better result
correspond to the water cut
2. History matching using analytical aquifer show a crossing curve with history curve.
Normally, the curve must be parallel to the history. The crossing curve will give a
significant value when we do forecasting. That’s why we don’t choose analytical
aquifer for forecasting.

V. REFERENCES
Mattax, C.C., and Dalton, R.,L.,. 1990. Reservoir Simulation, SPE Monograph Series.
Madaoui, K. Presentation Reservoir Engineering Oil & Gas Development, Total
Professeurs Associes. Bandung: Desember 2008
Shaktikumar, S. and Aguilera M., Reservoir Simulation, Assiciate Professor ENSPM
(IFP School), February 2009
Eclipse 2008 Manual

You might also like