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

net/publication/337885138

AN UPDATED CONCEPTUAL MODEL OF THE TOMPASO GEOTHERMAL FIELD

USING NUMERICAL SIMULATION

Conference Paper · November 2019

CITATIONS READS

0 113

5 authors, including:

Ade Lesmana Heru Berian Pratama

7 PUBLICATIONS   4 CITATIONS   
Bandung Institute of Technology
46 PUBLICATIONS   49 CITATIONS   
SEE PROFILE
SEE PROFILE

Ali Ashat Nenny Miryani Saptadji

Bandung Institute of Technology Bandung Institute of Technology
45 PUBLICATIONS   36 CITATIONS    35 PUBLICATIONS   35 CITATIONS   

SEE PROFILE SEE PROFILE

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

dieng 2 project View project

Reservoir Modeling View project

All content following this page was uploaded by Ali Ashat on 11 December 2019.

The user has requested enhancement of the downloaded file.

 AN UPDATED CONCEPTUAL MODEL OF THE TOMPASO GEOTHERMAL FIELD
USING NUMERICAL SIMULATION
Ade Lesmana1, Heru Berian Pratama1, Ali Ashat1, 2, Nenny Miryani Saptadji1 and Fatah Gunawan3
1Geothermal Master Program, Institut Teknologi Bandung, Jalan Ganesha 10, Bandung
2Department of Earth Resources Engineering, Faculty of Engineering, Kyushu University, Fukuoka 819-0395, Japan
3PT Pertamina Geothermal Energy, Head Office, Skyline Building 14th Floor, MH Thamrin St. no.9, Central Jakarta
adelesmana21@gmail.com

Keywords: Tompaso, TOUGH2, natural state, conceptual The application of numerical simulation to update conceptual
model models has, for example, been conducted previously by Ashat
et al., (2019), Supijo et al., (2018) and Kurniawan et al.,
ABSTRACT (2018). The aim of this study was to develop a numerical
Recent geoscience studies and well data suggest that the natural-state model which can represent the natural reservoir
Tompaso geothermal field may have two separate reservoirs. condition of the Tompaso geothermal field. Results of the
The first reservoir is located around the LHD-27 well cluster model could then be used to generate an alternative
in the Pinabetengan area and it has supplied 2×20 MWe at conceptual model. The natural state model and the conceptual
units 5 and 6 of the Lahendong power plant since the year model developed in this study have two separate reservoirs in
2016. The second prospect is located around the Tempang the Tompaso field.
area, which is associated with the Soputan fault. A new
conceptual assessment of the Tompaso area is presented
based on a numerical reservoir model which was constructed
using the TOUHG2 simulator and the EOS1 equation of state.
The reservoir model includes two high-temperature upflows
at the base of the model, including one below the Masam
Crater and one below the Toraget area. The locations of these
upflows were chosen based on the apprearance of alteration
zones. A natural state model was developed based on
downhole pressure and temperature data from ten wells.
Based on this natural state model, an alternative conceptual
model was proposed where the second geothermal prospect
area was added. In the first prospect area, the updated
conceptual model has the geothermal fluid coming from the Figure 1: Location of the Tompaso geothermal field in
upflow zone located below the Masam Crater and then North Sulawesi, Indonesia (Google maps, 2019).
moving north-northeast towards the Kawangkoan area. In the
second prospect, the updated conceptual model has 2. RESERVOIR CHARACTERISTICS
geothermal fluid flowing from an upflow zone under the The permeability structure of the Tompaso field was studied
Toraget area and expected moving to the northeast towards by Sardiyanto et al., (2015). The North Sulawesi province is
the Passo area. located at the convergence of three tectonic plates, the
Eurasian, Pacific and Australian plate. Due to those
1. INTRODUCTION conditions, there was double subduction of the Maluku sea
The Tompaso geothermal field is situated within the Tompaso which caused the formation of the Minahasa-Sangihe and
district, Minahasa Regency, North Sulawesi, or around 60 km Halmahera volcanic arcs. The activity of the Minahasa-
from Manado City (Handoko, 2010), as shown in Figure 1. Sangihe volcanic arc resulted in the formation of two major
The field is managed by PT Pertamina Geothermal Energy. strike-slip faults named Menado-Kema and Amorang-
Exploration of the Tompaso field began in 1982 to meet Malompar that are NW-SE trending. The Tompaso
electricity demand for North Sulawesi. Three exploration geothermal field is located between those faults (Figure 2).
wells and five development wells were drilled from 2008 to
2009 to confirm the geology and the size of the geothermal
resource. Early exploration suggested a liquid dominated
resource with a temperature range of 250°C to 300°C
(Gunawan et al., 2015). A resource assessment was
performed by Handoko (2010) using a Monte Carlo
probabilistic approach to estimate the amount of stored heat
in the system. The assessment suggested that the electricity
potential for 30 years of operation could be 39 MWe (50%
probability). Since the year 2016, the field has provided 2×20
MWe through units 5 and 6 of the Lahendong geothermal
power plant (Directorate Geothermal and Center of Mineral,
Coal and Geothermal Resources, 2017), with the main
production zone located around the LHD-27 well cluster
(Handoko, 2010). Figure 2: Geological structure model of Tompaso field
(Sardiyanto et al., 2015).

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand
Figure 3: Geological map of the Tompaso area (modified from Prasetyo et al., 2015; well R1 and R2 azimuth from Gunawan,
2016).
A structural map of the Tompaso geothermal field is • Manifestations that appear on the edge of the Tondano
presented in Figure 3. The ENE-SWS trending Soputan Fault lake and Remboken village are dominated by
is a major strike-slip fault in this area. According to previous bicarbonate hot springs.
study by Sardiyanto et al. (2015), the fault controls fluid
movement from the upflow zone in the Riendengan-Sempu Two prospects with resistivity values of 5–10 ohm-m were
area to the Tempang-Toraget area. At the center of the field, inferred based on a magnetotelluric (MT) resistivity study
the Tompaso Fault is oriented in a NNE-SSW direction, (Handoko, 2010). The main prospect is located under the
which is a dominant fault orientation for other minor faults in Masam Crater (Mount Rindengan) and the second prospect is
that area. It appears that those faults provide the main located in the Tompaso-Tempang area. The top of the
permeability in the Tompaso field. reservoir in the Masam Crater reservoir is around sea level,
and in the Tompaso - Tempang area the top of reservoir is
Based on geological evidence, it appears that the magmatic around 500 metres below sea level. The MT study suggested
heat source in the area moved over time from the north-east a conductive layer in both prospects with a value between 3–
to the south-west, from Mount Lengkoan to Mount Senpu, 5 ohm-m, which indicates argillic altered rocks (clay) about
Mount Rindengan-Maniporok-Aesoput and finally Mount 750–1000 m thick.
Soputan. The heat source that currently exists is relatively
shallow and derived from Mount Soputan (Handoko, 2010). Selected well locations and well tracks in the Tompaso field
are also shown in Figure 3. Eight wells which were drilled
According to Handoko (2010), the manifestation distribution from the year 2008 to 2009 are distributed in three clusters.
in the Tompaso field can be divided into four different They are cluster LHD-26 (wells LHD-26 and LHD-30),
regions: cluster LHD-27 (wells LHD-27, LHD-31, LHD-33, and
LHD-24) and cluster LHD-32 (wells LHD-32 and LHD-35)
• Manifestations in the Tempang – Tompaso – Mount (Handoko, 2010). Figures 10 show interpreted natural-state
Umeh area are dominated by chloride hot springs, mud temperature observations in the wells of those clusters. Well
pools, acidic hot springs, and steaming ground. LHD-26 was drilled down to -900 masl and showed a
conductive temperature profile. The maximum temperature
• Manifestations in the Masam Crater – Mount was 250°C. Well LHD-30 had a convective temperature
Riendengan area are dominated by fumaroles, solfatara, profile at an elevation of -750 to -1000 masl, witht a
and hot springs. temperature around 240°C. Typical convective profiles were
found at three wells of cluster LHD-27. Their temperatures
• Manifestations in the Kawangkoan – Mount Emung reached 250°C at elevations of -100 to -500 masl. The other
area are dominated by bicarbonate and sulfate hot well (LHD-33) had a convective profile at elevations between
springs -950 to -1100 masl. Then at -900 masl the temperature
suddenly dropped from 200°C to 125°C. Wells LHD-32 and
LHD-35 had comparatively low temperatures.

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand
Figure 4: The Subsurface condition of Tompaso field (Prasetyo and Tri Handoko, (2009) in Handoko (2010)).
Two additional wells, R1 and R2, were drilled in the Tempang Crater, as indicated by the large number of high-temperature
area in 2014 (see Figures 3 and 11). The maximum vertical manifestations at the crater.
depth of these wells is about 2600 m. A massive convective
zone was found in well R1 at an elevation of -100 to -1940 3. NUMERICAL MODEL
masl (Gunawan, 2016). The temperature of the convective 3.1 The Model Grid
zone was about 280°C. In well R2 a convective profile was
The numerical model of the Tompaso geothermal field was
found at elevations between -1500 masl and -1825 masl with
built using the TOUGH2 simulator with the EOS1 equation
its temperature reaching 330°C. Considering the high-
of state module. The model covers a 39 km2 area and the grid
temperature profile of both wells, it is possible that there is
was rotated to align it with the dominant fault trend (Figure
another heat source or high-temperature inflow around this
6). The top of the model was constructed to follow the ground
area. In addition, there is a difference in pressure gradient of
topography. The model has 17 horizontal layers extending
well R1 from other well in cluster LHD-26 dan LHD-27,
from 1125 to -2400 masl. The 13 lowest layers have the same
which also indicates that there are two separate reservoir in
thickness of 200 m. One additional 50 m thick layer was
Tompaso field (see Figure 5).
added for the atmospheric boundary conditions. All model
grid blocks are rectangular and their horizontal dimensions
Well's Pressure Gradient vary from the largest ones being 300×300 m to the smallest
Pressure (Bar) ones being 100×100 m in size around wells. The total number
0 50 100 150 200 250 of grid blocks was 38,502. Figure 7 shows the 3D model and
1000
grid system of the Tompaso field.
500 LHD-26
Elevation (masl)

LHD-27
0
LHD-34
-500 R1
-1000
-1500
-2000

Figure 5: The difference in pressure gradient of well R1

from other wells in cluster LHD-26 and LHD-27.
Prior to this study there was no complete conceptual model
available for the Tompaso geothermal field. A verticals cross-
section showing isothermal temperature profiles is given in
Figure 4 for selected wells in the western part of the Tompaso
field. The conceptual model has geothermal fluid moving
from below the Masam Crater to a shallow depth below
cluster LHD-27. The reservoir temperature is 240°C while the
top of the reservoir is around -100 masl below cluster LHD- Figure 6: Top view of the Tompaso field model.
27. It appears that the upflow zone is located under Masam

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand
 Table 1 shows the calibrated permeabilities for each assigned
material or rock type. K xy indicates the rock permeability in
lateral direction, while K z shows the permeability in vertical
direction, in millidarcy (mD). The permeability of reservoir
material ranges between 3 to 100 mD. The rock
characteristics curve, relative permeability, uses Corey’s
curve and has been assigned to all rock material.

Other material properties were uniform across the model

(e.g., rock density and specific heat were set uniformly as
2600 kg/m3 and 1000 J/(kg·K). The material distribution of
the Tompaso 3D model in proposed conceptual model slicing
is shown in Figure 9.

Tabel 1: Material properties.

Material K xy (mD) K z (mD)
BOND 0.0001 0.0001
Figure 7: The 3D model and grid system of the Tompaso BASE 0.01 0.01
field. ATM 100 100
CAPRK 0.001 0.001
3.2 Initial and Boundary Conditions
HEAT 100 100
The top layer was set constant at 1 bar and 25°C to represent RES1 6 3
atmospheric conditions. The side boundaries are assumed to RES2 10 5
be no-flow condition. Thus material with permeabilities less RES3 80 40
than 0.001 mD were assigned for the side boundary. RES4 60 30
RES5 100 50
At the bottom boundary, two high-temperature inflow zones
were assigned, while other areas of the bottom boundary were RES6 100 100
treated as impermeable basement formations. The first inflow IMPR1 0.001 0.001
was located below the alteration zone of Masam Crater, where IMPR2 0.0001 0.0001
13 blocks had a temperature of 360°C and a pressure of 240
bar as a fixed state. The second inflow zone was located
below the alteration zone of the Toraget area, where 20 blocks
were used with a constant temperature of 350°C and a
pressure of 250 bar. A volume factor of 1×1030 also was
assigned to both high-temperature inflow area. Several NNE-
SSW trending faults were represented in the model, including
the Soputan Fault. These faults were referenced when
assigning rock types in the model. Figure 8 shows the location
of high-temperature inflow (red colour) and assigned faults
(black colour) in the 3D model of the Tompaso field.
Definition of other color refer to Tabel 1.

Figure 9: Material or rock-type distribution.

4. MODEL VALIDATION
Natural-state simulations were carried out by running the
model up to a large simulation time to reach a steady-state
condition. The final calibrated natural-state model was found
by adjusting model permeabilities until the model gave a
reasonable match to downhole pressure and temperature
observations.

4.1 Pressure and Temperature Matching

The simualted pressures and temperatures given by the
Figure 8: Bottom boundary and assigned faults. calibrated model were validated using observations from
3.3 Material Properties eight wells around the Pinabetengan area (the LHD wells) and
two wells in the Tempang area (wells R1 and R2). However,
Rock types or rock formations were assigned to areas in the
model to be representative of rock properties, such as only seven wells were thought to penetrate the reservoir zone
according to pressure and temperature downhole logging
porosity, permeability, density, and specific heat. The most
important model parameters in natural state models include data. Wells LHD-32 and LHD-35 did not reach the reservoir
zone while the logging data of well LHD-33 appeared to be
rock permeabilities and the bottom boundary conditions
(O’Sullivan et al., 2001 and Ashat et al., 2019). The unreliable since it’s temperature decrease lower than normal
gradient temperature. Thus those three wells were not
permeability will, in part, control the temperature and
pressure distribution by controlling the flow of heat and mass considered in the validation step. Figures 10 and 11 compare
observed pressures and temperatures with values given by the
in the model.
calibrated model.

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand
 Figure 11: Pressure and temperature matching results for
cluster R1 and R2.
Figure 12 and 13 shows overall pressure and temperature
matching from 7 wells, which were evaluated at liner section
only. Data gathered on 45° line indicated a proper matching
was achieved. Regarding the significant temperature drop at
elevation -1400 masl of R2 well, there is no information to
explain this condition. However, since the distance between
well R1 and R2 less that 300 m, it was assumed that both wells
penetrate the same formation. Then the model’s temperature
profile of both well’s become similar.According to
Seyedrahimi-Niaraq et al. (2019), the result of pressure and
temperature matching can be evaluated using a root-mean-
square-error (RMSE) measure by using the following
formula:

1
𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 = �N ∑N
i = 1(X Data,i − X Model,i )
2 (1)

Here N is the number of observations, XData,i is the its

observation and XModel,i is its corresponding simulated value.
Applying the RMSE method separately to each validation
well, the RMSE for temperature matching ranged from 4.1 to
20.2°C for temperature observations and from 1.2 to 18 bar
for pressure observations (Figure 14).

Figure 12: Comparison of well’s temperature data with

simulation result for natural state model.
Figure 10: Pressure and temperature matching results for
cluster LHD-26, LHD-27 and LHD-32.

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand
 On the other hand, the heat and fluid which comes from
beneath the Toraget area is shown in the model to move
relatively straight up due to the Soputan Fault. At the surface
of this area, there is an abundance of thermal manifestations.
The appearance of bicarbonate warm springs on the edge of
the Tondano lake in the Passo area suggests an outflow zone
for this second prospect (Figure 3). There was no indication
of a steam zone in the model which is consistent with
downhole logging data. Based on this, the reservoir of the
Tompaso field can be categorized as a single-phase liquid
system.

Figure 13: Comparison of well’s pressure data with

simulation result for natural state model.

Figure 16: Fluid flow direction in the model.

5. AN UPDATED CONCEPTUAL MODEL
We used the calibrated 3D numerical model to create an
updated conceptual model of the Tompaso geothermal field
(Figure 17). For the location of the A-B-C model slice used
in Figure 16 refer to Figure 6.

Based on the interpretation of geophysical studies and

Figure 14: RMSE of the model at natural state conditions. downhole well data, there are appear to be two separate
reservoirs in the Tompaso area. The first prospect is
4.2 Heat and Mass Flow Directions associated with the Masam Crater. The heat is estimated to
Figure 15 shows 3D isothermal profiles and heat flow come from beneath the alteration zone of the Masam Crater.
directions for the calibrated model, while the mass flow The top of the reservoir in this area is estimated to be at an
directions are shown in Figure 16. The heat and fluid comes elevation of -700 to -800 masl. The reservoir extends to the
from beneath the Masam Crater and the Toraget alteration north-east towards the Pinabetengan area where the LHD-27
zone. The appearance of fumarole indicates that both areas well cluster is located. In that area, the top of the reservoir is
include upflow zones of the Tompaso field (see Figure 3). found at a shallower depth at an elevation of about -100 masl.
From the Masam Crater, the heat and fluid moves to the east The temperature of the reservoir is around 250 to 270°C. The
and up to a shallower depth where the Tompaso fault is caprock thickness varies from 750 to 1000 m. Between the
located. The high-temperature region ends around cluster reservoir and caprock, there is a 100 m thick transition zone .
LHD-27 and forms a “tongue” shape. This result is consistent The upflow zone is located below the Masam Crater while the
with the temperature profiles at wells LHD-27, LHD-31, and outflow zone is estimated to be at the Kawangkoan area
LHD-34. The appearance of bicarbonate warm springs in the towards the north side.
Kawangkoan area to the north is estimated to be caused by an
outflow area of fluid flowing from underneath the Masam The second prospect is located in the Tempang area which is
Crater (Figure 3). associated with the Soputan Fault. The heat comes from
beneath the alteration zone of the Toraget area. In the model,
the top of the reservoir is set at an elevation of -200 masl and
the temperature of the reservoir ranges between 290 and
320°C. The caprock thickness is estimated to be around 1000
m. The upflow zone is located in the Toraget area while the
outflow zone is estimated to be at the edge of the Tondano
lake in the Passo area.

The typical temperature profile of well LHD-31 and LHD-34

indicates that both these wells are located at the edge of the
reservoir. It appears that the fault which is located between
cluster LHD-27 and well R1 in the Tompaso area seperates
the two prospects from each other. Because the well’s
temperature suddenly decrease when the trajectory intersects
the fault. According to Gunawan (2016), the recharge area is
Figure 15: Temperature isotherms and heat flow vectors estimated to be around Tompaso prospect area and possibly
for the calibrated model. from the Tondano Lake.

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand
Figure 17: Proposed conceptual model of Tompaso field based on numerical modelling.
6. CONCLUSIONS REFERENCES
• A numerical model describing the Tompaso geothermal Ashat, A., Pratama, H. B. and Itoi, R.: Updating Conceptual
field has been successfully developed and resulted in a Model of Ciwidey-Patuha Geothermal Using Dynamic
good match to natural-state conditions (based on Numerical Model, IOP Conf. Ser. Earth Enviromental
available data and given assumptions). Sci., (April), doi:10.1088/1755-1315/254/1/012010,
(2019).
• For each validation well, the root-mean-square error for
the observation matches ranges from 4.1 to 20.2°C for Directorate Geothermal and Center of Mineral Coal and
temperature matches and from 1.2 to 18 Bar for Geothermal Resources: Potensi Panas Bumi Indonesia,
pressure matches. Directorate General of New Energy, Renewable Energy
and Energy Conservation, Jakarta., (2017).
• The model is able to depict the appearance of two Gunawan, F.: Rencana Pengembangan Panas Bumi Lapangan
separate reservoirs at the Tompaso field. In the model, Tompaso Sulawesi Utara dengan Simulasi Reservoir,
the first reservoir is shown to be located around the Institut Teknologi Bandung., (2016).
Masam Crater and the second reservoir is shown to be
located in the Tempang-Toraget area. Gunawan, F., Hastriansyah, G., Prabowo, T. and Zuhro, A.
A.: Remedial Work in a Geothermal Well Case Study:
• An updated conceptual model has been generated based Sendangan-3, Tompaso Project, in Proceeding World
on the numerical model. Geothermal Congress 2015, pp. 1–6, Melbourne,
Australia., (2015).
• The Tompaso geothermal field can be categorized as a
Handoko, B. T.: Resource Assessment of Tompaso
single-phase liquid reservoir and high-temperature.
Geothermal Field , Indonesia, United Nation Univ.
7. RECOMMENDATIONS Geotherm. Train. Program. Reports, (30), 647–674,
(2010).
Further validation with production data (history matching)
will increase the model’s accuracy. Considering the high- Kurniawan, I., Sumartha, A. G. A., Wiradinata, R.,
temperature zone and good permeability in the Tempang- Nandaliarasyad, N., Sutopo, Pratama, H. B. and Prabata,
Toraget are, there is a possibility of another productive area T. W.: Updating Conceptual Model of Cisolok-
in the south of Toraget controlled by the Soputan Fault. Cisukarame Geothermal Field, West Java, Indonesia, in
Proceeding The 6th Indonesia International Geothermal
ACKNOWLEDGMENTS Convention & Exhibition (IIGCE) 2018., (2018).
The author would like to thank his family for their sacrifice O’Sullivan, M. J., Pruess, K. and Lippmann, M. J.: State of
and understanding so that this paper could be finished. The the Art of Geothermal Reservoir Simulation,
author also would like to thank Ridwan, Claudio, Zhafira and Geothermics, 30, 395–429, (2001).
Firdaus from Geothermal Master Program ITB for direct
assistance that helped to finish this paper.

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand
 Prasetyo, I. M., Koestono, H. and Thamrin, M. H.: Clay Sabalan Geothermal System, Geothermics, 77(August
Alteration Study from Wells of Tompaso Geothermal 2018), 42–61, doi:10.1016/j.geothermics.2018.08.009,
Field, North Sulawesi, Indonesia, in Proceeding World (2019).
Geothermal Congress 2015, Melbourne, Australia.,
2015. Supijo, M. C., Wahjono, A. D., Lesmana, A., Harahap, A. H.,
Sutopo, Pratama, H. B. and Prabata, T. W.: Updating
Sardiyanto, Nurseto, S. T., Prasetyo, I. M., Thamrin, M. H. Conceptual Model Using Numerical Modelling for
and Kamah, M. Y.: Permeability Control on Tompaso Geothermal Green Field Prospect Area in Atadei , East
Geothermal Field and Its Relationship to Regional Nusa Tenggara , Indonesia, in Proceeding The 6th
Tectonic Setting, in Proceeding World Geothermal Indonesia International Geothermal Convention &
Congress 2015, Melbourne, Australia., (2015). Exhibition (IIGCE) 2018., (2018).
Seyedrahimi-Niaraq, M., Ardejani, F. D., Noorollahi, Y.,
Porkhial, S., Itoi, R. and Nasrabadi, S. J.: A Three-
dimensional Numerical Model to Simulate Iranian NW

Proceedings 41st New Zealand Geothermal Workshop

25-27 November 2019
Auckland, New Zealand

View publication stats