Geothermics 106 (2022) 102571

Contents lists available at ScienceDirect

Geothermics
journal homepage: www.elsevier.com/locate/geothermics

Investigation of the main reservoir of the Tulehu geothermal system

(Indonesia) using 3-D inversion of MT data
Yunus Daud a, b, *, Wambra Aswo Nuqramadha c, Fitrianita a, Fikri Fahmi c, Suwijanto Tarmidi c,
Magdalena Ari Tifani c, Tony Widiatmoro d, Hendra Yu Tonsa d, Yudistian Yunis e
a
Geothermal Exploration Magister Program, Graduate Program of Physical Sciences, Department of Physics, Faculty of Mathematics and Natural Sciences, Universitas
Indonesia, Kampus Depok 16424, Indonesia
b
Geothermal Research Center (GRC), Institute for Sustainable Earth and Resources (I-SER), Faculty of Mathematics and Natural Sciences, Universitas Indonesia,
Kampus Depok 16424, Indonesia
c
PT NewQuest Geotechnology, Pesona Khayangan Estate DC12, Depok 16411, Indonesia
d
PT PLN (Persero), Jalan Trunojoyo Blok M – I No 135 Kebayoran Baru, Jakarta 12160, Indonesia
e
PT PLN Gas & Geothermal, Gedung KMO Lantai 8, Jalan Kyai Maja No. 1, Kebayoran Baru, Jakarta Selatan 12120, Indonesia

A R T I C L E I N F O A B S T R A C T

Keywords: Tulehu geothermal prospect area is located in Maluku Province, Indonesia. The existence of the geothermal
Magnetotellurics system in this area is recognized by the appearance of hot springs on the western side of Mt Eriwakang and the
Geothermal northern side of Mt. Huwe. Most of these hot springs occur along inferred geological structures. The geological
Main reservoir
conditions of this area are dominated by Quaternary volcanic rocks, limestone, alluvium, and metamorphic
Tulehu
Indonesia
rocks. Several authors have conducted several studies to propose various geothermal conceptual models, from
3-D inversion structurally controlled to volcanic-hosted geothermal systems. However, the geoscientific data supporting the
hypotheses are not comprehensive yet. There are still gaps in the data coverage as well as in the interpretation of
geological, geochemical, and geophysical data, so a comprehensive study should be done before deciding on
further stages. Therefore, this paper proposes a new conceptual model of the Tulehu geothermal area by reducing
the existing data gap. Interpretation of remote sensing data, geological field observations, age dating analysis,
reanalysis of geochemical data, and additional MT data covering Mt. Eriwakang are conducted to reconstruct the
conceptual model. Graben-like structures are identified based on the geological data where the product of Mt.
Eriwakang is in the middle. Based on the age dating data, the product of Mt. Eriwakang erupted in the Qua­
ternary time (413±6 kA), so it is still quite potential as a heat source. This indication is also supported by the
trend of Cl, HCO3, and SiO2, which describe the possible flow from Mt. Eriwakang to the surrounding hot springs.
Based on the Na-K-Mg and Silica geothermometers, the reservoir temperature is estimated at 210-240 ◦ C. A
three-dimensional (3-D) MT inversion was applied to all the existing and new MT data to obtain a 3-D subsurface
resistivity model. The oceanic bathymetry data was also incorporated to mitigate the possible coastal effect from
the surrounding seawater to the MT data. The new 3-D MT inversion results show a good agreement with the
geological and geochemical indications. The updome-shaped pattern of the subsurface resistivity distribution
beneath Mt. Eriwakang supports the geological and geochemical data analysis results concerning the indications
of the prospects area around Mt. Eriwakang. These interpretations can strengthen the recommendation to focus
on the area around Mt. Eriwakang for further development.

1. Introduction eastern part of Indonesia. Exploration in the Tulehu geothermal area is

quite challenging. The existence of geothermal resource potential can
Tulehu geothermal area is located in Maluku Province, Ambon Island only be recognized by several hot springs discharged at the inner part
(Fig. 1). It is one of the earliest developed geothermal fields in the and the northeastern shore of the Tulehu Peninsula, as well as the fossil

* Corresponding author at: Geothermal Exploration Magister Program, Graduate Program of Physical Sciences, Department of Physics, Faculty of Mathematics and
Natural Sciences, Universitas Indonesia, Kampus Depok 16424, Indonesia.
E-mail address: ydaud@sci.ui.ac.id (Y. Daud).

https://doi.org/10.1016/j.geothermics.2022.102571
Received 18 October 2021; Received in revised form 29 August 2022; Accepted 12 September 2022
Available online 22 September 2022
0375-6505/© 2022 Elsevier Ltd. All rights reserved.
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 1. Tulehu geothermal area, Ambon Island (Indonesia).

Fig. 2. Estimated faults based on geological and geophysical studies conducted by JICA (JICA, 2011). Permeable zones (orange color) are interpreted as
geothermal reservoirs.

sinter cone, which is situated at the North Suli (Fig. 4) (Marini and Since there is no geophysical data, the specific interest areas cannot be
Susangkyono, 1999). There are no impressive manifestations, such as delineated. Meanwhile, JICA (2011) proposed a prospective area on the
high-temperature fumarole or steaming ground. Moreover, the occur­ northwestern side of Mt. Eriwakang. The geothermal reservoir is ex­
rence of volcanic products, such as the Mt. Salahutu complex, Mt. Eri­ pected to extend along the Banda-Hatuasa, Banda, and R2 faults (Fig. 2),
wakang, and Mt. Huwe, are not active volcanoes. Mt. Eriwakang and Mt. although the location of its up-flow center is uncertain (JICA, 2011). A
Huwe have also been covered by coral reefs that seem to be a part of a similar hypothesis with JICA (2011) was also indicated by Nasution
submarine volcano in Early Quaternary time, which have no informa­ et al. (2015). The indicated faults are previously considered a productive
tion on their last activities (Nasution et al., 2015). Therefore, an zone. The faults are expected to be a conduit of a heat transfer medium
exploration program should be conducted thoroughly concerning any from the heat source to the surface and a migration channel system to
possibilities to determine the main reservoir area. the superficial level, where convective heat may occur below the
Several previous studies have proposed the conceptual models of the Banda-Hatuasa-Sila complex. Meanwhile, the heat sources are probably
Tulehu geothermal system (e.g., Marini and Susangkyono, 1999; JICA, intrusive rocks below the structural complex and Mt. Eriwakang old
2011; Nasution et al., 2015). Based on fluid geochemistry analysis, volcano (Nasution et al., 2015). Unfortunately, since the MT stations are
Marini and Susangkyono (1999) indicated potential sites for deep dril­ only distributed over the structural complex toward Salahutu volcanic
ling within two areas: close to Pulau Batulompa and Sila-Telaga Biru. complex (JICA, 2011; Nasution et al., 2015), the hypothesis regarding

2
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 3. Tectonic map of eastern Indonesia (Abdullah et al., 2000). The red mark indicates the location of the Tulehu geothermal area.

Fig. 4. Geological map of Tulehu geothermal area based on remote sensing and geological observations.

3
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 5. Distribution of thermal manifestation over Tulehu geothermal area. Numbers are surface temperature (red) and pH (blue).

Mt. Eriwakang has not been answered yet. areas, the volcanic products are also covered by coral limestone. Based
This paper proposes a new opinion to the conceptual model of the on Nasution et al. (2015), the areas covered by limestone are Quaternary
Tulehu geothermal system by completing the survey data with remote eruption centers, such as Mt. Eriwakang and Mt. Huwe.
sensing data interpretation and geological observation (including rock
dating), reanalysis of the geochemical data, and additional MT data. The 2.2. Local geology
remote sensing and geological data observations are needed to recon­
struct the geological “play” concept from a regional to a local point of Before geological field observation was conducted, a tentative
view. Reanalysis of fluid chemistry is focused on investigating hydro­ geological map had been produced based on remote sensing data
logical systems (possible upflow and outflow) with still involving interpretation to guide the observation target as well as to deliver an
reservoir temperature estimation and possible seawater mixing. Mean­ efficient field survey. In addition, the interpretation is also necessary to
while, the additional MT data is concentrated to cover the data distri­ investigate the geological “play” concept from a regional point of view
bution gap in Mt. Eriwakang and its surroundings. in the Tulehu geothermal area. The images used in this analysis are
DEMNas (Indonesian National Digital Elevation Model) and Landsat.
2. Geological setting After the field observation, the tentative geological map was updated to
the “final” geological map (Fig. 4).
2.1. Regional geology Four eruption centers were recognized: Mt. Salahutu, Mt. Huwe, Mt.
Eriwakang, and Mt. Tengah-Tengah (Fig. 4). Mt. Salahutu, as a part of
Tulehu geothermal field is tectonically located in Banda Arc, in the Mt. Ambon situated on the northwestern side of Mt. Eriwakang, was
south of Ceram trough, which is affected by the collision of Eurasian, formed in Late Pliocene time (Tjokrosapoetro et al., 1993). Meanwhile,
Indo-Australian, and Pacific plates (Fig. 3). The interaction between the the other three volcanic products probably occurred in Quaternary time.
three plates since the Late Miocene, having affected tectonic activities Thermoluminescence (TL) dating analysis was performed in the dacitic
such as thrusting, uplift, and erosion, is now reflected in geological rock samples to know the absolute age of the other three Quaternary
structures present in the region. Quaternary volcanic complexes from products (Fig. 4). The results show that the Huwe lava flow has an age of
the Ambon Island, Haruku island, to the eastern islands (Fig. 1), are 806±2 ka, while the Eriwakang pyroclastic flow and Tengah-Tengah
caused by the northern subduction of the Indo-Australian plate (western lava flow show 413±6 ka and 237±9 ka, respectively.
Irian Jaya) beneath the Ceram trough (Honthaas et al., 1999; Nasution Lineaments and faults are identified dominantly along NW-SE and
et al., 2015). NE-SW directions. Banda fault and Eriwakang fault (Fig. 4), which are
Tjokrosapoetro et al. (1993) inferred that Salahutu volcanic product located at western and eastern flanks of Mt. Eriwakang respectively, are
is the oldest rock in Tulehu geothermal area. These products are inferred to form a graben structure. Banda Fault is a normal fault with
distributed widely over Ambon Island (Fig. 1). It is composed of lavas, south-eastward dipping (Tjokrosapoetro et al., 1993). Eriwakang fault
igneous rocks (andesite and dacite), and pyroclastic rocks (breccia and and Huwe fault are normal faults with north-westward dipping (Nasu­
tuff). Some of Salahutu’s volcanic products are covered by coral lime­ tion et al., 2015). Furthermore, the Tengah-Tengah fault is also a normal
stone, dominantly found around the beach. Between the Tulehu and Suli fault with north-westward dipping marked by an escarpment between

4
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 6. (A) Mg vs Cl and (B) Na vs Cl diagram. The geothermal mixing line is inferred by the hot springs mainly located at the center of the island, while the seawater
mixing line is reconstructed by the hot springs at the coast.

lower and higher topography. contamination by seawater is something that needs to be identified
Graben structure, in principle, is an open structure that becomes a before using these data for further interpretation regarding the
path for geothermal fluid to emerge on the surface (Wohletz and geothermal system. The combination of three chemical components (Na,
Heiken, 1992). The graben structure in this area is also supported by the Cl, and Mg) explains the possibility of mixing or contaminating seawater
occurrence of hot springs and other thermal manifestations along the in two diagrams, namely Na vs Cl and Mg vs Cl diagrams (Fig. 6).
inferred faults. Generally, there is a weak zone in the center of the The two diagrams (Fig. 6) show that the hot springs are divided into
graben structure consisting of many fractures as a pathway for the two trend lines: the geothermal mixing line and the seawater mixing
subsurface fluids, including magma intrusion, to reach the surface. Due line. Samples originating from the Tulehu geothermal area are located in
to the permeable area, many geothermal systems are usually formed in the geothermal mixing line area, clearly different from the seawater
this zone. Therefore, the existence of Mt. Eriwakang suggests that there mixing line. The concentration of Cl from the hot springs in the
is a heat source for a geothermal reservoir, which can be an exploration geothermal mixing line is below 6000 ppm, different from the two
target in the Tulehu area. samples forming the seawater mixing line, indicating Cl concentrations
of more than 14,000 ppm (Fig. 6). This result indicates that the fluid
3. Geochemical evidence appearing on the surface of the Tulehu geothermal area does not expe­
rience a mixing reaction with seawater. In addition, since the hot springs
The thermal manifestations mostly discovered in Tulehu geothermal are mostly located in the mature water region on the Cl-SO4-HCO3
area are hot springs (Fig. 5). In addition, an inactive fumarolic field with ternary diagram (Marini and Susangkyono, 1999), liquid geo­
a fossil sinter cone is also mentioned by Marini and Susangkyono (1999) thermometer calculation is also applicable.
and Nasution et al. (2015) at the Banda village (Figs. 4 and 5). Marini In the Tulehu geothermal conceptual model, two key drivers would
and Susangkyono (1999) and Japan International Cooperation Agency be confirmed through geochemical data: reservoir temperature and
JICA 2011 provide a chemical composition table of the available hot hydrology (upflow-outflow). For reservoir temperature estimation,
springs, which is then summarized in this paper to deliver geo­ enthalpy (H) vs chloride (Cl) mixing graph and Na-K-Mg trilinear graphs
thermometer calculation and geochemical interpretation. The Tulehu were used (Fig. 7). The Enthalpy (H) vs Cl mixing graph gave the
geothermal area is located near the coast; even some manifestations of reservoir temperature around 210 ◦ C, while the Na-K-Mg graph gave
hot springs appear on the shore. Therefore, the process of mixing or 220-240 ◦ C. In addition, a similar indication is presented by Marini and

Fig. 7. (Left) Graph of enthalpy (H) and chloride (Cl) mixing and (right) triangle graph of Na-K-Mg to analyze reservoir temperature.

5
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 8. The inferred contour of chloride (Cl) content (in ppm) from the hot springs in the Tulehu geothermal area.

Fig. 9. The inferred contour of bicarbonate (HCO3) content (in ppm) from the hot springs in the Tulehu geothermal area.

6
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 10. The inferred contour of silica (SiO2) content (in ppm) from the hot springs in the Tulehu geothermal area.

Susangkyono (1999), which has a range of 230–245 ◦ C. 4.1. Overview of MT data

Although there are no common manifestations that appear in the
upflow zone (such as fumaroles, steaming ground, and steam-heated There are two sets of existing MT data in the Tulehu geothermal area
acid-sulfate water), the content of chloride (Cl), bicarbonate (HCO3), which were obtained in 2009 (PT. PLN Geothermal and PT. Geo ACE,
and silica (SiO2) can provide an overview of the main reservoir location 2010) and 2011 (JICA, 2011). In 2009, 45 MT stations were acquired
of Tulehu geothermal system. Based on the chloride (Cl) content, the Sila and mostly distributed around the southern vicinity of the Mt. Salahutu
hot spring, which has the highest Cl content, is probably situated closer area. Several MT stations are aligned in a profile crossing Mt. Eriwakang
to the main reservoir (Fig. 8). The Cl content then gradually decreases and Mt. Huwe (Fig. 11). In 2011, another MT measurement was carried
towards the hot springs located on the west, north, and east sides of Sila. out for 34 stations and arranged to focus on the occurrence of surface
The Cl contour, which opens to the south, indicates the presence of the manifestations along the Banda Fault based on the previous hypothetic
main reservoir (Fig. 8). The same indication is also shown by the HCO3 model (Fig. 11). Measuring stations were placed in a grid configuration
contours (Fig. 9) and SiO2 content (Fig. 10), although the trend is the with a station separation from approximately 700 m to 800 m. The MT
opposite for HCO3 (from low to high content). Therefore, geochemical impedances were estimated for more than 50 frequencies in the range
data provide the same indication as geology, that is, the possibility of the from 320 to 0.001 Hz. The result of the MT investigation was reported in
main reservoir in the Tulehu geothermal area is leading to Mt. JICA (2011) and Nasution et al. (2015). However, in this study the MT
Eriwakang. data has been reprocessed from the time series format in order to get
better data quality. Overall, all of the reprocessing results show good to
4. Magnetotelluric investigations excellent quality even without remote reference data available to the
authors.
Magnetotelluric (MT) is a frequently used geophysical method to As described above, the analysis of the geological and geochemical
delineate geothermal reservoirs (e.g., Árnason et al., 2010, Train­ data has suggested that the possible heat source of the geothermal sys­
or-Guitton et al., 2014, Gao et al., 2018, Daud et al., 2019). Resistivity tem in the Tulehu area is probably correlated with the volcanic activity
variation derived from MT inversion result provides subsurface infor­ of Mt. Eriwakang. However, the coverage of existing MT stations in 2009
mation regarding conductive clay cap distribution, upflow-outflow and 2011 was limited for the eastern part of Mt. Eriwakang to confirm
location, and reservoir top and boundary indication. This investigation this hypothesis. We therefore carried out an additional MT survey in
was carried out to delineate the location of the most promising reservoir, 2020 under the 2020 International Indexed Publications (IIP) Research
which, based on geological and geochemical analysis, is suspected to be Grant from Universitas Indonesia. The new 20 MT stations were
under Mt. Eriwakang. Because the existing MT data obtained in 2009 distributed around Mt. Eriwakang to fill the gap in the existing data
and 2011 did not cover the eastern side of Mt. Eriwakang sufficiently, we (Fig. 11). The data acquisition was conducted using a single MT
have conducted additional MT measurements in 2020 to strengthen the equipment from Phoenix Geophysics. Therefore, the MT data was also
hypothesis. processed without the remote reference in a 320-0.001 Hz frequency
range.

7
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 11. MT station distribution in Tulehu geothermal area. The blue, orange, and yellow dots represent MT stations acquired in 2009, 2011, and 2020.

Fig. 12. Apparent resistivity and phase curves at station MT-30 before and after cross-power selection.

4.2. Data processing and quality techniques proposed by Jones and Jödicke (1984) and Rousseeuw
(1984) to down-weight the data influenced by noises. A cross-power
Processing and reprocessing of the time-series data were conducted selection was performed to optimize the impedance data’s quality,
for all of 2009, 2011, and 2020 data using the software of Phoenix smoothness, and realistic level for the last step. Fig. 12 shows an
Geophysics to calculate the MT impedance. It utilizes robust processing example of MT data from the station MT-30 before and after cross-power

8
Y. Daud et al. Geothermics 106 (2022) 102571

Table 1 selection.
Parameter settings for the 3-D MT inversion. Data quality was classified based on the error bar and smoothness of
3-D MT inversion parameter the curves as used by Daud et al. (2019). In general, the MT data in the
Number of Sites 70 Tulehu geothermal area have good to excellent data quality. Finally, 70
Block Size (m) 200 × 200 × 10 MT stations, out of the total 99 stations, were selected for further 3-D
Number of Padding Blocks (x,y) 10 × 10
inversion by considering the data quality and distribution.
Padding Factor (x,y,z) 1.5 × 1.5 × 1.5
Number of Blocks (x,y,z) 64 × 70 × 22 Prior to MT inversion, a final data correction related to the static shift
Initial Resistivity Model Homogeneous 100 Ωm effect should be thoroughly investigated (Árnason, 2015). The MT static
Ocean Model / Resistivity Value Yes / 0.1 Ωm shift is a galvanic distortion effect that locally shifts apparent resistivity
Mesh Rotation 0 sounding curves by a scaling factor independent of the frequency,
Data Rotation 0
keeping the phases unchanged (e.g., Tournerie et al., 2006). In this
Impedance Full Tensor
Number of Periods 16 study, the geostatistical averaging method was used to correct the static
Frequency Range 320-0.01 Hz shifts (Berdichevsky et al., 1980; Beamish and Travassos, 1992). The
Error Floor (ZxxZyy / ZxyZyx) 5% / 5% method first computes an average apparent resistivity (at the same
frequency) of several MT stations surrounding the static shift disturbed
MT station. The referenced curve resulting from the averaging process is

Fig. 13. Comparison of observed and calculated data of Zxx, Zxy, Zyx, Zyy components at several MT stations.

9
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 14. RMS misfit for each MT station by the 3-D MT inversion.

used to correct the disturbed MT station. All impedance components low-frequency data. These results further confirm the existence of the
value (Zxx, Zxy, Zyx, and Zyy) were then recalculated for all disturbed main resistive up dome anomaly in the inversion model.
MT stations after the correction process. The whole processing steps
should be conducted appropriately to reduce uncertainty in the 5. Results and discussions
modeling result (Daud et al., 2017a).
The 3-D MT inversion results show a good agreement with the
4.3. Three-dimensional inversion geological and geochemical indications. Additional MT data that fill the
gap over Mt. Eriwakang gives a realistic subsurface resistivity model
3-D MT inversion was performed using WSINV3DMT code that uti­ beneath the volcanic product. Fig. 16 shows the resistivity distribution
lizes a data-space variant of Occam’s approach (Siripunvaraporn et al., maps at 0 m (sea level), − 500 m, − 1000 m, − 1500 m, − 2000 m, and
2005). The full components of the impedance tensor (Zxx, Zxy, Zyx, Zyy) − 2500 m elevations. Indication of the conductive layer can be well-
were used for the inversion. The ocean bathymetry data was included in recognized from the superficial level to about 2000 m.b.s.l. Its pres­
the modeling to mitigate the coast effect due to the presence of seawater ence gradually diminishes at greater depths. The distribution dominates
near the survey area. The more detailed 3-D inversion parameters used the survey area, except in the northwestern part, where the oldest
in this study are shown in Table 1. The final model from the 3-D geological products are located. At the superficial level, the conductive
inversion result has an RMS misfit of 8.71. Fig. 13 shows several sam­ layer is also correlated with the existence of surface thermal
ples of observed and calculated MT curves, while Fig. 14 shows an RMS manifestations.
map resulting from the 3-D resistivity model. In general, the RMS value Two cross-sections are provided for giving a vertical view of re­
for each station in the interest area, around Mt. Eriwakang, varies from 6 sistivity distribution. Line 1 in Fig. 17 is oriented to NW-SE, crossing the
to 10. Moreover, a higher RMS value (>10) arises on several MT stations main NE-SW structure (Banda Fault) as well as Mt. Eriwakang. While
in the margin of MT station distribution, especially on the north side Line 2 is perpendicular to Line 1, connecting Sila hot spring (SL-1) and
which has a minimum distribution of MT stations. Mt. Eriwakang, as shown in Fig. 18. A high-low-high resistivity pattern
A sensitivity test was performed to verify the reliability (accuracy) of as a common pattern found in a geothermal area can be observed in both
the main resistive up dome anomaly in the inversion model. Resistivity vertical sections. There is a high resistivity layer in the range of about
values >100 Ωm in the inversion model were replaced by 20 Ωm below 100 – 200 Ωm at the deeper part, overlaid by slightly lower resistivity of
− 500 m elevation. 3-D forward modeling was then performed on the about 20-100 Ωm, where the geothermal reservoir is usually associated.
modified model and compared its response with both the observed and In the middle part, a conductive layer (< 10 Ωm) probably correlated
calculated data from the inversion result. Fig. 15 illustrates the sensi­ with the geothermal system’s clay cap is situated in this area. The
tivity test model and its results along Line 1. The sensitivity test result thickness of the conductive layer varies from 200 to 2000 m. While the
shows that the calculated data are sensitive to the deeper part of the overburden layer is mostly dominated by a high resistivity anomaly (>
modeled structure as indicated by a change of MT response, especially in 100 Ωm), excluding several areas that may be affected by the occurrence

10
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 15. (Left) Illustration of 3-D models for the sensitivity test and Line 1 location map. (Right) MT responses at stations along Line 1: dots are MT responses from
the observed data, solid lines are from the inversion model, and dashed lines are from the sensitivity test model.

of surface manifestations and altered rocks. 2001). Strongly developed block faults can localize extrusive features
Three updome-shaped patterns with different compositions of the such as domes. The topographically high recharge areas and the pro­
middle conductive layer and deep resistive body are observed in the found permeable graben influence the graben’s hydrology surrounding
study area (Figs. 17 and 18). The updome-shaped pattern is commonly horst blocks. Such a situation promotes not only deep circulation of
found in the geothermal system as an indication of a promising zone. meteoric water but also the development of extensive hydrothermal
However, understanding this pattern as well as other supporting data convection along the margins of the intrusive body below the dome
should be considered before deciding a drilling target to this feature. The (Wohletz and Heiken, 1992). Accordingly, Mt. Eriwakang could be one
first pattern (D1) can be seen in the Salahutu region, in the northwestern of the most prospective areas. From the geochemical side, the contour of
part of Line 1 (Fig. 17). The deep resistive body shows the highest re­ Cl, HCO3, and SiO2 are analyzed to identify possible main reservoir.
sistivity value (more than 250 Ωm), while the conductive layer is the Since there is no upflow-type manifestation, such as fumarole and
thinnest. This pattern is commonly observed over an older geological steaming ground, those chemical contents could be used to reconstruct
segment in a geothermal field. The second pattern (D2) can be discov­ the existence of the main reservoir. Although Sila hot springs show
ered beneath KBJ-2 hot spring and fossil sinter cone (Fig. 17). The value characteristics of outflow hot springs, their location is probably closest
of the deep resistive body (in the range of 150 – 200 Ωm) is lower than to the main reservoir. The chloride (Cl) and silica (SiO2) contents
that of D1. The conductive layer, however, is thicker. The third pattern decrease, while the bicarbonate (HCO3) content increases from Sila to
(D3) can be observed beneath the Mt. Eriwakang region (Fig. 18). Its the surrounding hot springs and opens toward Mt. Eriwakang. These
shape looks like D2 but with a lower value of the deep resistive body phenomena support the possible existence of the main reservoir around
(less than 150 Ωm). Different resistivity value of the deepest layer Mt. Eriwakang.
mainly represents a different degree of the subsurface temperature. The The 3-D MT inversion result then confirms the indication from the
deepest resistivity layer of the current high-temperature geothermal geological and geochemical data. Two updome-shaped patterns, D2 and
system has mostly less than 100 Ωm (Daud et al., 2017b). D3, exist in this zone, where D2 is located close to Banda Fault. The
The new interpretation of geological and geochemical evidence as updome structure, which is similar to D2, was obtained by the 3D MT
well as the 3-D MT inversion result give an insight regarding geothermal inversion in JICA (2011) and Nasution et al. (2015), with a more shifted
resource occurrence over the Tulehu geothermal area. The existence of position to the north. However, based on the resistivity value of the
the main reservoir around Mt. Eriwakang has become more probable deepest layer, this zone is probably not the highest temperature in the
than before. On the geological side, remote sensing data interpretation study area. The existence of fossil sinter cones also indicates that there
indicates that a pair of NE-SW structures at the western and eastern sides was possibly a geothermal system in the recent past (Marini and Sus­
of Mt. Eriwakang forms a graben-like structure. Full grabens offer a angkyono, 1999). Meanwhile, the third updome-shaped pattern, D3, is
perfect setting for developing exploitable geothermal systems (Bogie, located beneath Mt. Eriwakang (Fig. 18). As the only young volcanic

11
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 16. Resistivity distribution maps derived from 3-D MT inversion in Tulehu geothermal area.

12
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 17. Resistivity Section in Line 1 derived from 3-D Inversion. D1 and D2 indicate updome-shaped resistivity patterns.

Fig. 18. Resistivity Section along Line 2 derived from 3-D Inversion. D3 indicates an updome-shaped resistivity pattern.

dome inside the graben, the heat flow beneath Mt. Eriwakang is possibly Huwe complex). The reservoir temperature of the most recent
high. It is also supported by slightly lower resistivity of the deepest layer geothermal system is about 210–240 ◦ C, estimated from Na-K, K-Mg,
than those of D2 and D1. The evolution of heat could be inferred from and Silica geothermometers. The base of conductor (BOC), or mostly
the resistivity map at the deeper level (more than 2000 m.b.s.l) (Fig. 16). identified as the top of reservoir, is situated about 600-1000 m.b.s.l,
It is started from Salahutu volcanic complex (D1), then continued to the varies from center to margin of reservoir. Based on this study, it can be
fossil sinter cone (D2) and Mt. Eriwakang (D3). The high resistivity trend recommended that further development of this geothermal prospect
continues to Mt. Tengah-Tengah by following the NE-SW lineament. should be focused on in this area.
This possible evolution is supported by age dating results showing that
Mt. Tengah-Tengah (237±9 kA) is younger than Mt. Eriwakang (413±6 6. Conclusions
kA). However, additional MT surveys should also be conducted over the
Mt. Tengah-Tengah and Mt. Huwe complex to confirm this indication. Remote sensing data interpretation, age dating analysis, and
Such survey could also confirm the possible geothermal prospect geochemical data interpretation have been carried out to identify the
beneath this complex and its relation to the occurrence of hot springs in geological “play” concept from the regional scale to the local one
the Tulehu-Batulompa complex in the northern part. In addition, further regarding reservoir conditions. Furthermore, additional MT data were
study should also include gravity surveys to confirm the existence of a collected during the 2020 IIP Research Program to fill the data gap in the
graben structure. suspected promising zone beneath Mt. Eriwakang. A graben-like struc­
A conceptual model is then produced to describe the possible ture is identified in the center of the Tulehu geothermal area. This
geothermal system in Tulehu geothermal system. Two possible upflow structural setting encourages the potential recent geothermal reservoir
zones located below the fossil sinter cone and Mt. Eriwakang (Fig. 19), beneath Mt. Eriwakang. Age dating results also indicate a mature
respectively. The outflow of both systems is probably toward the series geothermal system in Mt. Eriwakang (413±6 kA). Possible upflow in Mt.
of hot springs along the Banda Fault and its surroundings. Since the Eriwakang is also strengthened by Cl, HCO3, and SiO2 contour from
location of both upflows is so close to each other, it is not easy to define several hot springs in its surrounding. The temperature of the most
the boundary as well as the different heat sources between them from recent geothermal system is estimated at 210–240 ◦ C, derived from Na-
the MT imaging. However, based on geological lineament, they may be K, K-Mg, and Silica geothermometers. The 3-D inversion of MT data
separated by the NE-SW structure situated between them. The ground­ confirms the geological and geochemical indications. An updome-
water recharge of the system most probably originated from their horst shaped pattern can be observed below Mt. Eriwakang, with a BOC of
structures at the western (Salahutu Complex) and eastern part (Mt. about 600-1000 m.b.s.l. The other updome-shaped resistivity patterns

13
Y. Daud et al. Geothermics 106 (2022) 102571

Fig. 19. Conceptual model of Tulehu geothermal system. The cross-section is a combination of Line 1 and Line 2 to illustrate the second (D2) and third (D3) re­
sistivity patterns (see inserted map).

can also be recognized at the Salahutu complex and beneath the fossil Data availability
sinter cone. However, the subsurface temperature could be less than the
most recent reservoir. The authors do not have permission to share data.
Since the limitation of the 2020 IIP Research Program, the recom­
mendation of an additional MT survey cannot be entirely done. Addi­
tional MT data should also be completed on the eastern side, toward Mt. Acknowledgments
Huwe and Mt. Tengah-Tengah, to confirm either the possible reservoir
extension from Eriwakang or a separated system. A gravity survey The work was funded by the 2020 International Indexed Publications
should also be performed to ensure the existence of a graben structure (IIP) Research Grant from Universitas Indonesia. We acknowledge the
indicated by geological interpretation. Some development constraints, support from NewQuest Geotechnology for using their MT data pro­
such as scaling possibilities, should also be confirmed through further cessing and 3D inversion facilities. Support from PLN for permission to
study. use the previous MT data is highly appreciated. Thanks to Ms. Diajeng
Liati and Ms. Nurdianasari Latuconsina for their assistance during MT
CRediT authorship contribution statement field studies. Thank also to Ms. Fadhila Friany Putri for her assistance in
formatting the paper. We also thank anonymous reviewers for their
Yunus Daud: Conceptualization, Methodology, Validation, Super­ careful reading of the manuscript and their constructive remarks, which
vision, Writing – review & editing. Wambra Aswo Nuqramadha: improved and clarified this paper.
Formal analysis, Writing – original draft. : Investigation. Fikri Fahmi:
Software, Visualization. Suwijanto Tarmidi: Formal analysis. Magda­ References
lena Ari Tifani: Investigation. Tony Widiatmoro: Resources. Hendra
Yu Tonsa: Resources. Yudistian Yunis: Resources. Abdullah, C.I., Rampnoux, J.-.P., Bellon, H., Maury, R.C., Soeria-Atmadja, R., 2000. The
evolution of Sumba Island (Indonesia) revisited in the light of new data on the
geochronology and geochemistry of the magmatic rocks. J. Asian Earth Sci. 18,
Declaration of Competing Interest 533–546. https://doi.org/10.1016/S1367-9120(99)00082-6.
Árnason, K., Eysteinsson, H., Hersir, G.P., 2010. Joint 1D inversion of TEM and MT data
and 3D inversion of MT data in the Hengill area, SW Iceland. Geothermics 39, 13–34.
The authors declare that they have no known competing financial https://doi.org/10.1016/j.geothermics.2010.01.002.
interests or personal relationships that could have appeared to influence Árnason, K., 2015. The static shift problem in MT soundings. In: Proceedings World
the work reported in this paper. Geothermal Congress 2015. Melbourne, Australia. Paper 13031.
Beamish, D., Travassos, J.M., 1992. A study of static shift removal from magnetotelluric
data. J. Appl. Geophys. 29, 157–178. https://doi.org/10.1016/0926-9851(92)
90006-7.

14
Y. Daud et al. Geothermics 106 (2022) 102571

Berdichevsky, M.N., Vanyan, L.L., Kuznetsov, V.A., Levadny, V.T., Mandelbaum, M.M., Japan International Cooperation Agency (JICA), 2011. JICA Preparatory Survey for
Nechaeva, G.P., Okulessky, B.A., Shilovsky, P.P., Shpak, I.P., 1980. Geoelectrical Tulehu Geothermal Power Plant. Final Report (Summary Report) to PT PLN
model of the Baikal region. Phys. Earth Planet. Interface 22, 1–11. https://doi.org/ (Persero). JICA Annual Report.
10.1016/0031-9201(80)90095-3. Marini, L., Susangkyono, A., 1999. Fluid Geochemistry of Ambon Island (Indonesia).
Bogie, I., 2001. Permeability in geothermal system associated with continental rifts. In: Geothermics 28, 189–204. https://doi.org/10.1016/S0375-6505(99)00003-6.
Proceedings 23rd NZ Geothermal Workshop, pp. 45–50. Nasution, A., Aviff, M., Nugroho, S., Yunis, Y., Honda, M., 2015. The preliminary
Daud, Y., Nuqramadha, W.A., Fahmi, F., Pratama, S.A., Rahman, K.R., Subroto, W., conceptual model of tolehu geothermal resource, based on geology, water
2017a. Discovering “Hidden” geothermal reservoir in Blawan-Ijen geothermal area geochemistry, MT, and drilling. In: Proceedings World Geothermal Congress 2015.
(Indonesia) using 3-D inversion of MT data. In: Proceedings of Geo Electromagnetic Melbourne, Australia. Paper 11011.
Workshop. Februari, 2017. Bandung, Indonesia, pp. 20–24. PT. PLN Geothermal and PT. Geo ACE, 2010. Additional geoscientific survey of tulehu
Daud, Y., Darma, S., Nuqramadha, W.A., Pratama, S.A., Fahmi, F., 2017b. Applying geothermal prospect. Unpublished Internal Report.
innovations in MT technology for reducing geothermal exploration risks. In: Rousseeuw, P.J., 1984. Least median of squares regression, 79, 871–880. https://doi.
Proceedings of 39th New Zealand Geothermal Workshop. 22-24 November 2017. org/10.1080/01621459.1984.10477105.
Rotorua, New Zealand. Siripunvaraporn, W., Egbert, G., Lenbury, Y., Uyeshima, M., 2005. Three-dimensional
Daud, Y., Nuqramadha, W.A., Fahmi, F., Sesesega, R.S., Fitrianita, Pratama, S.A., magnetotelluric inversion: data-space method. Phys. Earth Planet. Inter. 150, 3–14.
Munandar, A., 2019. Resistivity characterization of the arjuno-welirang volcanic https://doi.org/10.1016/j.pepi.2004.08.023.
geothermal system (Indonesia) through 3-D magnetotelluric inverse modeling. Tjokrosapoetro, S., Rusmana, E., Achdan, A., 1993. Geological Map of the Ambon,
J. Asian Earth Sci. 174, 352–363. https://doi.org/10.1016/j.jseaes.2019.01.033. Maluku Sheet. Geological sheet 2612–2613. Geological Research and Development
Gao, J., Zhang, H., Zhang, S., Chen, X., Cheng, Z., Jia, X., Li, S., Fu, L., Gao, L., Xin, H., Center, Bandung, Indonesia.
2018. Three-dimensional magnetotelluric imaging of the geothermal system beneath Trainor-Guitton, W.J., Hoversten, G.M., Ramirez, A., Roberts, J., Juliusson, E., Key, K.,
the Gonghe Basin, Northeast Tibetan Plateau. Geothermics 76, 15–25. https://doi. Mellors, R., 2014. The value of spatial information of for determining well
org/10.1016/j.geothermics.2018.06.009. placement: a geothermal example. Geophysics 79, W27–W41. https://doi.org/
Honthaas, C., Maury, R.C., Priadi, B., Bellon, H., Cotten, J., 1999. The Plio–Quaternary 10.1190/GEO2013-0337.1.
Ambon Arc, eastern Indonesia. Tectonophysics 301, 261–281. https://doi.org/ Tournerie, B., Chouteau, M., Marcotte, D., 2006. Magnetotelluric static shift: estimation
10.1016/S0040-1951(98)00227-3. and removal using the cokriging method. Geophysics 72, F25–F34. https://doi.org/
Jones, A.G., Jödicke, H., 1984. Magnetotelluric transfer function estimation 10.1190/1.2400625.
improvement by a coherence-based rejection technique. SEG Technical Program Wohletz, K., Heiken, G., 1992. Volcanology and Geothermal Energy. University of
Expanded Abstracts 1984, 51–55. https://doi.org/10.1190/1.1894081. California Press, Berkeley.

15