Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

Geothermics 82 (2019) 97–120

Contents lists available at ScienceDirect

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

Geologic controls on the Muara Laboh geothermal system, Sumatra, T


Indonesia

Jim Stimaca, , Agus M. Sihotanga, Wildan Mussofana, Marino Baroeka, Clay Jonesb,
Joseph N. Mooreb, Axel K. Schmittc
a
Supreme Energy, Menara Sentraya 23rd fl., Jl. Iskandar Syah Raya, Jakarta, 12160, Indonesia
b
Energy and Geoscience Institute, 323 Wakara Way, Salt Lake City, UT, 84108, United States
c
Institute for Earth Sciences, Heidelberg University, Im Neuenheimer Feld 234-236, 69120, Heidelberg, Germany

A R T I C LE I N FO A B S T R A C T

Keywords: An integrated review of the stratigraphy and alteration of the Muara Laboh geothermal system suggests that
Gamma-ray (GR) log sector collapse and subsequent erosion of the Patah Sembilan volcano may be responsible for significant changes
Stratigraphy in permeability and fluid circulation during its evolution. This sector collapse is relatively old, probably oc-
Zircon U-Pb dating curring more than 41,000 years ago. It is likely that unroofing of part of the clay cap of the system due to edifice
Alteration
collapse and subsequent erosion, resulted in extensive boiling, fluid loss, and precipitation of quartz, calcite, and
Paragenesis
other minerals that reduced permeability in the upper part of the SW reservoir. Vein paragenesis and fluid
Fluid inclusions
System evolution inclusion data show that the geothermal system was hotter and more saline in the past. Cooling is most pro-
nounced in the shallow NE sector. Conversely the SW reservoir is currently close to fluid inclusions temperatures
but has significantly lower salinity, indicating an increased fraction of meteoric recharge.
The reservoir section includes a wide variety of volcanic and intrusive rocks ranging in age from < 0.5 to 96
Ma. The NE reservoir is hosted mainly in the lower part of the Patah Sembilan Andesite and underlying Upper
Silicic Formation, whereas the SW reservoir is hosted by an underlying intrusive complex. At least two episodes
of intrusion, in the Cretaceous and Miocene were dated. These intrusions are probably hosted by Miocene and
Early Tertiary to Mesozoic age volcanics of andesitic to basaltic composition. Late-stage dikes, inferred from
cuttings and image logs, may provide an indication of more recent intrusion.

1. Introduction and fluid inclusion measurements on vein samples. We combine the


results of surface mapping with this subsurface information to constrain
1.1. Background the evolution of the Muara Laboh geothermal system. A companion
paper describes other geoscientific aspects of the system, such as geo-
This paper provides a detailed first look at some important geologic physical and geochemical constraints on the conceptual model of the
constraints on the Muara Laboh geothermal reservoir, located in system (Stimac et al., 2019).
Sumatra, Indonesia. The system lies along the Great Sumatra Fault
(GSF), an important structural control on geothermal activity on the 1.2. Regional geologic context
island. Quaternary to Pliocene volcanic vents are concentrated within a
few km of the Siulak segment of the GSF, highlighting the close asso- The island of Sumatra is composed of a craton-like core of Paleozoic
ciation of tectonism and magmatism in this area (Mussofan et al., rocks (Sundaland Craton), and later accreted terranes of Mesozoic age,
2018). that were assembled as a consequence of prolong subduction (Barber
Some aspects of the geologic context for the system, such as regional et al., 2005a). The ancestral Tertiary and modern volcanic arcs run the
stratigraphy, the dominant structural orientations and mineral para- length of the island, generally within 30 km of the Great Sumatra Fault
genesis of open fractures, have been addressed in recent conference (GSF) system, a highly segmented right-lateral strike-slip system
papers (Baroek et al., 2018; Mussofan et al., 2018). We expand on these (Barber et al., 2005b; Gasparon, 2005; Sieh and Natawidjaja, 2000). A
topics here by the addition of new age constraints on the stratigraphy, series of thick Tertiary sedimentary basins formed E of this fault zone in


Corresponding author.
E-mail address: sgcgeo00@gmail.com (J. Stimac).

https://doi.org/10.1016/j.geothermics.2019.06.002
Received 5 April 2019; Received in revised form 31 May 2019; Accepted 1 June 2019
0375-6505/ © 2019 Elsevier Ltd. All rights reserved.
J. Stimac, et al. Geothermics 82 (2019) 97–120

hornfels, meta-greywacke and marble (Pbl) constitutes the oldest rocks


known in the Muara Laboh area. Locally the Barisan Formation overlies
the Siguntur Formation (Ps), consisting of quartzite that was intruded
by Cretaceous Granite (Kgr). This basement sequence outcrops 8–10 km
N and E of the Muara Laboh field and could possibly represent the
western margin of the Sundaland Craton. Isolated outcrops interpreted
as Kgr may represent erosional promontories partially buried by
younger Tertiary units since their host rocks are not represented. Later
we present well data indicating that volcanic-intrusive sequences of
Early Tertiary (Paleocene to Eocene) and Mid- to Late-Mesozoic age
host Miocene and Cretaceous age plutons, respectively.
The Painan Formation (Tomp), consisting of mixed volcanic and
sedimentary rocks, unconformably overlies the Pre-Tertiary basement.
This volcanic sequence consists mainly of andesitic to dacitic lava,
breccia, tuff, and volcaniclastic deposits. Sedimentary rocks include
arkose, bituminous shale, shaley coal, and tuffaceous sandstone (Rosidi
et al., 1996). The mid-Tertiary Painan Formation outcrops in the NW,
forming the W side of fault scarps bounding the Muara Laboh basin. K-
Ar dates on volcanic rocks of this formation yield ages ranging from 23
to 14 Ma (Late Oligocene to Middle Miocene), which is also supported
by Dicotylendenblad biomarker fossils (Bellon et al., 2004; Rosidi et al.,
1996). In the Middle Miocene, granitic (Tgr) and granodioritic (Tgdr)
rocks intruded the Barisan and Painan Formations. These intrusive
Fig. 1. Regional tectonic setting of Sumatra (modified from Darman and Sidi, bodies outcrop to the W and S of Muara Laboh and extend to the S of
2000). The tectonic elements from left to right are the Indo-Australian plate,
Mt. Kerinci following the main trend of the GSF. Their exposure implies
Sunda Trench, toe of the Accretionary Wedge, Fore Arc Ridge, and Mentawai
extensive uplift and erosion of these formations occurred during Middle
Fault System (MFS). These elements are collectively referred to as Sumatran
sliver plate. The Barisan Mountains, active volcanoes (red triangles), Sumatran Miocene to Recent time. Like the Kgr outcrops, these Tertiary granitoids
Fault System and major Tertiary to modern sedimentary basins lie further east. must represent erosional highs buried by Quaternary deposits (Fig. 2).
Outcrops interpreted as Plio-Pleistocene to Holocene volcanics,
ranging from intermediate (Qyu) to silicic composition (Qou and Qol),
the backarc environment (De Smet and Barber, 2005).
are widespread west of the GSF and largely undated (Rosidi et al.,
With a main direction of N30 °W and the total length of approxi-
1996). The younger volcanic centers whose activity may have spanned
mately 1900 km the GSF accommodates a significant amount of the
from the Pleistocene to Holocene can be recognized by relatively well-
oblique plate convergence by partitioning it into right-lateral strike-slip
preserved volcanic landforms comprising stratocones, domes, and lava
motion (Darman and Sidi, 2000; Sieh and Natawidjaja, 2000). Fig. 1
flow sequences. As outlined by Mussofan et al. (2018), young volcanic
also shows the Sumatran “forearc sliver plate”, a 300-km-wide strip of
vent complexes are concentrated all along the Siulak Fault, including
lithosphere between the GSF and the Sumatran forearc deformation
from NW to SE, Mt. Bangko, Anak Patah Sembilan, Patah Sembilan,
front (Sieh and Natawidjaja, 2000). Oblique convergence is also ac-
Kapur, and Kerinci (Fig. 2). Patterns of flow are generally from these
commodated by the offshore Mentawai Fault System through right
high-elevation vents in the S and SW toward the N and NE into the
lateral displacement.
Muara Laboh basin. Field mapping indicates that recent eruptive pro-
Large faults grow by connection of shorter, well established seg-
ducts consist mainly of andesitic lava, breccia, tuff, and related volca-
ments through time. Mapping of the GSF using topography and aerial
niclastic deposits. The more distal deposits derived from these sources
photographs indicates that it is highly segmented, and that other
consist dominantly of volcaniclastic equivalents of these rocks.
structures must also be accommodating right-lateral offset (Sieh and
Patah Sembilan (PS) is the nearest stratocone to the Muara Laboh
Natawidjaja, 2000). According to these authors the GSF is composed of
geothermal system. It has a prominent double crater that suggests
19 segments that have distinct patterns of seismicity. As many as 13
multiple volcanic vents and sector collapses through time (Mussofan
pull-apart basins have been defined in stepovers of the GSF master fault
et al., 2018). The superheated PS fumarole lies in the southeasterly PS
segments (Muraoka et al., 2010).
crater (Fig. 2). Its gas chemistry, including relatively low helium iso-
About 46% of active volcanism on Sumatra occurs within 10 km of
tope ratio (R/Ra = 1.95), implies that the Muara Laboh geothermal
the GSF (Acocella et al., 2018). The close relationship between the GSF
system may extend into the PS crater area (Stimac et al., 2019). As also
and modern volcanoes of the Sumatran arc (Fig. 1) has promoted for-
described in that paper, the craters are inside a protected forest en-
mation of numerous high-temperature geothermal systems on the is-
vironment, limiting future exploration of this resource area.
land (Hochstein and Sudarman, 1993; Muraoka et al., 2010). The
Muara Laboh basin is a prime example, formed in a right-stepover be-
1.3.1. Most recent volcanic deposits
tween the Suliti and Siulak GSF segments (Sieh and Natawidjaja, 2000).
Review of LiDAR imagery and limited field mapping of the project
Many other Sumatran geothermal systems are located along or near the
site provide some new information about the youngest volcanic and
GSF, either in pull-apart basins (Sarmiento et al., 2019 or along ex-
sedimentary deposits in the area. A secondary summit, now partially
tensional bends of the main or subsidiary fault segments (e.g., Hickman
truncated by the northwesterly PS crater, was inferred as a satellite vent
et al., 2004; Drestanta et al., 2018).
complex and named Anak Patah Sembilan (Qaps in Figs. 2 and 3).
Moderately well-preserved lava flow morphologies extend from this
1.3. Surface stratigraphy high relief satellite vent area to the N. The Qaps andesitic lavas are
relatively unaltered but lack datable minerals or carbon. They appear to
The regional stratigraphy of the Muara Laboh area includes for- be slightly younger than the main PS edifice based on their less ad-
mations deposited from Paleozoic to Recent time (Mussofan et al., vanced state of erosion and alteration.
2018; Rosidi et al., 1996; Fig. 2). A more detailed geologic map of the project area was made based on
The Paleozoic Barisan Formation (Pb) consisting of slate, phyllite, review of LiDAR imagery augmented by local field mapping. The most

98
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 2. Geologic map and simplified stratigraphy of the Muara Laboh region (modified from Rosidi et al., 1996; Mussofan et al., 2018).

99
J. Stimac, et al. Geothermics 82 (2019) 97–120

southeasterly subcraters separated by a large mass of lava flows (Qps1)


that appear to have erupted after the craters formed. Two prominent
lobe-shaped sedimentary deposits have been identified downstream
that were clearly sourced from the PS crater area (Qps2a and 2b in
Fig. 3). Based on their overall shape, extent, and internal characteristics
in the immediate project site, they are interpreted as debris flow se-
quences.
The Muara Laboh development area (well pads, roads and power
plant) is near the western margin of the debris flow deposits where they
lap onto older and more structurally deformated Anak PS lavas. The
debris flow deposits are unbedded, matrix supported, and very poorly
sorted. Near the Pad A and H areas they contain boulders up to 3 m in
long dimension, mostly of andesitic composition, in a matrix of smaller
rock fragments, crystals, ash and clay (Colvin and Mussofan, 2017;
Stimac and Mussofan, 2017). The maximum thickness of the debris flow
sequence is unknown because its base is not exposed, but locally it
reaches about 30 m. The continuity of this deposit across the area of the
Muara Laboh drilling pads and power plant site suggests that it may
represent either a series of significant debris flows formed by gradual
downcutting of the crater area, or alternatively they may represent the
marginal parts of one or more debris avalanches resulting from a sector
collapse of the PS crater to the NE.
The lack of clearly identified megabreccia blocks or areas of hum-
mocky topography characteristic of debris avalanches make it more
likely these deposits formed after the main sector collapse, but may also
be explained by the mapped area being at the edge of a debris ava-
lanche deposit; large blocks tend to be centralized along the avalanche
path (Belousov et al., 2007). Based on the LiDAR image, large blocks
and hummocks may be present at more distal locations, but these have
not been verified by mapping. It is most likely that these debris flow
deposits may have formed in multiple events related to the eruption of
post-collapse lavas and as the PS craters were progressively excavated
over geologic time. In this scenario the observed deposits post-date the
main collapse event and overlie and obscur the main debris avalanche
deposits.
The debris flow sequence interpreted in Fig. 3 is locally overlain by
a series of fallout and ash-flow tuffs in the development area (Mussofan
et al., 2018). These tuffs lie essentially at the surface, but are locally
overlain by additional debris flows. Their outcrop is too poor and ir-
regular to map, but in the A Pad area they includes a sequence of fallout
tuffs that is approximately 4 m thick and composed of three distinct
subunits where it overlies the debris flow deposits (Colvin and
Fig. 3. Detailed geologic map based on LiDAR image of development area. Map Mussofan, 2017). This same tuff sequence is 1–3 m thick in excavations
shows interpretation of the crater (black hatured line), area of Anak Patah near E Pad, were it was carbon-dated (Mussofan et al., 2018). The
Sembilan (Qaps) vents and lavas, likely post-crater lavas (Qps1), and post-crater fallout tuff sequence was interbedded with up to one meter of com-
debris flow deposits (Qps 2a and 2b), and fumaroles. Yellow circles mark the pressed woody material that was encapsulated by debris flow deposits.
locations of 14C date samples. Pad locations marked by blue squares. A non-welded to weakly sintered andesitic ash-flow tuff underlies the
fallout tuff near F Pad but was not observed at other locations, possibly
significant feature in this map area is a large double crater truncating because it represented a valley fill deposit. It contains charcoal that was
the summit area of PS volcano (Fig. 3). The size and depth of the double dated from that location (Fig. 3).
crater make it very likely that it formed by one or more sector collapse
events, although clearly related debris avalanche deposits have not
been confirmed. The crater is divided into northwesterly and

Table 1
14
Carbon dates on surface samples.
Sample ID Material Carbon Mass Carbon Yield d13C value (per mil) F (d13C) dF (d13C) 14C age (BP) d14C age

8-12-17-C1 Tan andesite ash-flow tuff charcoal 2.12 57.7 −23.68 0.0061 0.001 40900 1300
8-12-17-C2 charcoal 2.63 54.9 −23.414 < 0.0023 NA > 48800 NA

8-14-17-1C White dacite fallout tuff wood 2.00 55.6 −25.473 < 0.0021 NA > 49500 NA
8-14-17-2C-1 wood 2.00 41.7 −24.677 0.0155 0.001 33490 520
8-14-17-2C-2 wood 1.42 38.9 −24.227 0.0133 0.001 34680 600

Notes: The fraction of modern carbon and radiocarbon age were calculated as weighted averages of combined machine runs to reduce overall error. A small sample
correction was applied to samples with a carbon mass less than 0.50 mg.

100
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 4. Well locations and data collected (revised from Baroek et al., 2018). Abbreviations: Extended range micro-imager (XRMI) log; GR, gamma-ray log; BHPT,
borehole pressure-temperature log.

1.4. Carbon dating gamma-ray (GR) logs, and borehole resistivity images (Stimac, 2019).
U-Pb zircon ages were determined for selected silicic volcanic and in-
Carbon dates were obtained on the tuffs described above as sum- trusive rocks to constrain the timing of igneous activity, and their re-
marized in Table 1 and Mussofan et al. (2018). The ages were de- lationships to regional tectonic events (Table 2). The detailed analytical
termined by the University of Arizona AMS Laboratory. Charcoal and methods used, and data gathered are summarized in Appendix 1.
wood from the tuffs yielded ages of ∼34 to 41 ka, respectively, for the
two tuffs (Fig. 3). Since these tuffs directly overlie debris flows that may
2.1.2. Reservoir rock types and formations
be related to or post-date the sector collapse of the PS volcano, these
Wells drilled on the A Pad lost cuttings returns at about 600 m asl,
dates provide a minimum age of the inferred collapse event. For re-
so cores and limited logging coverage provide the main data on re-
ference, this would make this sector collapse relatively old compared to
servoir rocks (Fig. 4). Wells in SW Muara Laboh drilled from H and F
documented collapse events at Gunung Wayang (> 7450 years) or
Pads had full or partial returns of cuttings until the deep reservoir was
collapse of Galunggung volcano near Karaha-Bodas at 4200 years ago
reached at between 200 and 800 m bsl. At this point circulation was lost
(Bogie et al., 2008; Moore et al., 2002), and it is likely somewhat older
and cores were taken to characterize the deepest levels of the reservoir.
than this minimum. Further work may provide new insights into the age
Borehole image and GR logs were also run and provide a more con-
of PS sector collapse, post-collapse lavas, and debris flow and tuff de-
tinuous record of likely formation rock composition and deposit type.
posits described here.
Well H3 provides a good example of the data obtained and integrated to
interpret subsurface formations. The well location is shown in Fig. 4,
2. Reservoir rock types and ages and the simplified composite log in Fig. 5.
Using the approaches described above, the rock sequence was di-
2.1. Well stratigraphy vided into five major formations, with a layer of young alluvial deposits
present as the sixth formation in some northerly wells (Figs. 6 and 7).
The stratigraphy of the Muara Laboh reservoir presented here in- The deepest of these formations were cut by a variety of intrusions.
tegrates the surface geologic mapping and 14C dates described above From bottom to top, the sequence consists of the Lower Andesite/Dacite
with information from 15 deep wells. Wellhead elevations ranged from and Lower Basalt; the Lower Rhyolite/Dacite; the Middle Andesite/
about 1100 to 1600 m above sea level (asl), with data extending to Dacite; the Upper Rhyolite/Dacite, the Upper Dacite/Andesite, and the
about 1300 m below sea level (bs) in the deepest well penetration Patah Sembilan Andesite (Figs. 5 and 6). As described later these in-
(Fig. 4). formal units were assigned likely ages and correlated with regional
formations based on zircon dating and inferred depositional and in-
2.1.1. Well sampling approach trusive relationships.
Borehole logs and data from cuttings and cores were integrated to Based on the distribution and thickness of rock types, it is likely that
constrain rock types and their structural properties and to infer the the Muara Laboh Basin formed in concert with Plio-Pleistocene vol-
orientations of permeable fractures determined by pressure-tempera- canism comprising the PS Andesite and Undifferentiated Silicic
ture-spinner (PTS) and pressure-temperature (PT) logs (Fig. 4). These Formation (USF). The USF is thickest in the SW (F and H wells) near the
data were collected on most wells where conditions allowed (Baroek Siulak segment of the GSF and thins to the N and E (Pads D and E)
et al., 2018). A total of 254 thin sections were examined to determine towards the Suliti Fault segment (Fig. 6). This fits with previous de-
primary rock types and hydrothermal alteration assemblages, and scriptions of the Muara Laboh basin as being asymmetrical with the
specific intervals were selected for geochronology and fluid inclusion thickest accumulations of silicic material near the Siulak Fault, and
studies. Primary deposit types and rock compositions were interpreted more andesitic and alluvial accumulations in the wider eastern basin
from wellsite descriptions, detailed petrography on selected samples, near the Suliti Fault, based on gravity and initial well results (Mussofan

101
J. Stimac, et al.

Table 2
Descriptions of zircon date samples.
Sample Sample Type Vertical Depth (m) Elevation (m asl) Age (Ma) Rock Description

ML-H2OH (999 m) Cuttings 978 591 0.174 ± 0.020 Partially welded rhyolite ash-flow tuff with sparse plagioclase and sericitized biotite phenocrysts, and fine-grained vapor phase devitrification
and spherulites. Transitional to phyllic alteration with calcite, quartz, sericite and chlorite replacement. Sample interval also contains 5%
andesitic lava lithic fragments.
ML-H1OH (519 m) Cuttings 513 1056 0.534 ± 0.018 Hornblende dacite lava flow breccia. Some fresh hornblende phenocrysts, mostly altered with FeTi oxide rims. Transitional alteration with
chlorite, quartz, and mixed-layer clay replacement.
ML-H2OH (1050 m) Cuttings 1019 550 0.559 ± 0.048 Partially welded rhyolite ash-flow tuff with spherulitic devitrification. Transitional to phyllic alteration with calcite, quartz, sericite and
chlorite replacement.
ML-A3 (955 m) Core 858 573 2.93 ± 0.05 Devitrified to originally glassy silicic non-welded ash-flow tuff with 5% large (> 5 cm) intrusive fragments. Tuff contains phenocrysts of
plagioclase, quartz, biotite (chloritized), possibly pyroxene, now altered to chlorite. Lithic fragments and crystalline debris derived from
intrusive rocks. Strong phyllic alteration with open veining. Intrusive lithics are coarse-grained biotite granite with plagioclase, k-feldspar,
quartz, minor amphibole, pyroxene, FeTi oxides, zircon, and titanite.
ML-H4 (3101 m) Core 2824 −1255 20.79 ± 0.3 Medium grained diorite with hornblende, biotite and pyroxene in addition to plagioclase and quartz, accessory titanite, apatite and zircon.
Locally sheared with minor open space and alteration selvages around fractures. The darker part of the rock has a higher percentage of biotite
and amphibole. This rock is much less altered and deformed than the granodiorite.
ML-H4 (2584 m) Core 2368 −799 96.1 ± 1.4 Granodiorite with plagioclase, quartz, k-feldspar, and biotite as major minerals and apatite, zircon, and FeTi oxides as accessory minerals.
Local andesitic enclaves. Enclaves could be comagmatic with granodiorite, or older assimilated country rock. strongly sheared and altered
with pervasive sericitization of feldspar.
Notes:

102
Well/Location Sample Depth Unit
ML-H1OH 519 Hb dacite lava
ML-H2OH 999 Silicic Tuff sequence near top
ML-H2OH 1050 Silicic Tuff sequence near bottom, rhyolite
ML-A3 core 1 955 Lithic NW Dacite ash-flow tuff
ML-H4 core1 2581 to 2590 Granodiorite
ML-H4 core2 3100 to 3103 Diorite (deepest, may not be oldest)
H1 519 is at a shallower level than H2 999, 1050 by roughly 500 (H1 & H2 wells drilled near each other).
H1 519 looks like a Hb dacite lava while H2 999, 1050 are ash flow tuffs.
I expected H1 519 to be younger, but it is older than H2 999 and slightly younger than H2 1050. I see two possibilities, but I’m sure there are
more:
1. Since the wells are on the same pad but deviated there could be a fault offset bringing H1 519 up relative to H2
2. The age on H2 999 could spurious (somehow contaminated by young zircon), then H1 519 would be slightly older than H2 1050
I will look back at the samples, and the gamma-ray logs of the two wells and think more about this.
A3 955 core (also an ash flow tuff like H2 999, 1050) was expected to have about the same age as H2 999 & 1050 based on log correlations,
but is much older. I see a couple possibilities here:
1. A completely different tuff and/or fault explanation similar to above
2. A3 zircons are inherited older zircons from plutonic rock xenoliths in the tuff. Recall that I mentioned they are very abundant in the rock.
Based on the U-Pb tab of your spreadsheet all the zircons were about the same age except one much older, so this would entail that the tuff
magma had no zircon of its own and all were inherited. Seems a bit unlikely?

Notes: negative values indicate elevations bsl.


Geothermics 82 (2019) 97–120
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 5. Simplified composite log of well H3 with casing configuration, major rock types, alteration assemblages, Methylene Blue Index (MEB), permeable zones and
respective injectivity index (II) determined from PT-spinner log interpretation, orientations of fractures within a maximum ± 20 m interval around the interpreted
permeable zone, and interpreted formations. Locations of thin sections indicated by diamonds. Formations and unit names are informal to the Muara Laboh reservoir
area. See Fig. 6 for likely regional correlation of these informal units.

et al., 2018). A similar asymmetric basin related to strike-slip faulting multiple small vents complexes rather than large single caldera erup-
was described by Busby and Bassett (2007). They presented a model for tions. This model fits well for the Muara Laboh basin, which is domi-
coupled volcanism and sedimentation along the releasing bends of nated by volcanic vents and thick proximal volcanic accumulations
strike-slip faults. In their model basin fill thins and volcanism decreases along the Siulak Fault (SW margin of basin), and sedimentary and distal
markedly away from the master strike-slip fault (“deep” end), where volcanic deposits along the Suliti Fault (NE margin of basin).
subsidence is greatest, toward the basin-bounding normal faults
(“shallow” end). Talus cone alluvial fan deposits are largely restricted
2.2. U-Pb zircon ages
to the master fault-proximal (deep) end of the basin. Volcanic centers
are localized along the master strike-slip fault and its splays. Volcanism
2.2.1. Sampling and uncertainty in zircon sources
along the master fault keeps it overfilled, pushing sedimentation into
Six samples of silicic volcanic and intrusive rocks were selected for
the shallow basin. Extension allows frequent eruption that favors
SIMS U-Pb zircon dating (Table 2). The primary objectives of the dating

103
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 6. Summary of subsurface stratigraphic formations based on regional correlation (left) and informal units (right). Detailed stratigraphy of selected wells from
NNW to SSE with subunits determined mainly from GR logs and detailed petrography. Abbreviations: bas, basalt; and, andesite; dac, dacite; rhy, rhyolite; INT-GD,
granodiorite; INT-D, and diorite.

were to determine the ages of 1) intrusive rocks encountered deep in likely derived from older rocks (inherited xenocrysts). In some cases
SW wells, and 2) the youngest packages of silicic rocks that underlie inheritance might be from lithic debris of slightly older, and/or possibly
Quaternary andesitic volcanics from the PS volcano and satellite vents. coeval intrusive rocks (e.g., A3 958), but in other cases, it was clearly
Despite extensive alteration of the primary minerals in the samples, derived much older rocks (H2 999, H2 1050).
zircon crystals appear unaffected (either in the form of dissolution or by
overgrowths visible under cathodoluminescence). See Appendix 1 for
2.2.2. Estimated sample ages
further details.
No attempt was made to date the PS Formation since these andesitic
Two samples of intrusive rocks were dated from well H4 and four
rocks generally do not contain zircon or potassium-rich minerals sui-
samples of silicic lava and tuff were dated from wells A3, H1, and H2.
table for geochronology. Instead emphasis was placed on dating the
The intrusive samples are from a deep plutonic complex intersected in
underlying silicic tuff sequence. The silicic sequence, dominated by ash-
SW wells. The volcanic samples represent the upper part of a thick
flow tuffs, typically has GR counts from ∼70 to 150 API versus ∼40–65
sequence of silicic tuffs, lavas, and volcaniclastic rocks that underlies
API for andesite. Four rock samples were dated from wells A3, H1, and
the PS Andesite Formation, and is informally named the
H2 from high GR intervals of the USF (Fig. 8 and Tables 1, 3 and A1).
Undifferentiated Silicic Formation or USF (Fig. 6).
Based on the interpreted ages of these samples the USF appears to have
Milicich et al. (2013) summarized challenges of dating zircons from
been deposited in the interval from ∼3 to 0.5 Ma. Underlying the USF
well samples. Ideally zircons represent magmatic material such as a
are several older packages of mafic to silicic volcanics that are cut by a
lava, intrusion, or pumice from a tuff. However, such unambiguous
variety of intrusive rocks. The dominantly andesitic volcanic sequence
samples can rarely be obtained from wells. Bulk pyroclastic material
underlying the USF is correlated with the Painan Formation (23.7 to
contains fragments of accidental lithics and their disaggregation pro-
14.3 Ma) based on its composition, stratigraphic position, and con-
ducts, and even lava and intrusive rocks may contain xenoliths of for-
tained fossiliferous limestone fragments. Underlying the Painan For-
eign material or remnants of crustal assimilation (xenoliths, xeno-
mation are additional sequences of volcanic and volcaniclastic rocks cut
crysts). Core samples at least represent a known depth and rock type,
by Early Miocene (20.8 Ma) and Late Cretaceous (96.1 Ma) intrusive
and a lack of downhole contamination can be assumed. On the other
rocks. The ages of host rocks for these intrusions were not directly
hand well cuttings have the possibility of downhole contamination from
constrained by zircon dating, but as discussed in more detail below, the
the open-hole section potentially introducing younger zircons from
lack of any Paleozoic zircon ages suggests that they may be of Early
above the sampled interval. Sampling near the previous well casing
Tertiary and Mesozoic age. Each dated sample is discussed in more
point can potentially minimize this issue by reducing the open hole
detail below.
section. We used SEM cathodoluminescence (CL) and petrography to
assess the uncertainty introduced by less than ideal well samples.
Descriptions of samples selected for zircon dating and their posi- 2.2.2.1. ML-H1 519. Based on inferred stratigraphic position, sample
tions within the well stratigraphic sequence are given in Table 2 and H1 519 was expected to be the youngest unit, being several hundred
Fig. 7, respectively. Core samples were selected in preference to well meters shallower than the other three USF samples (Fig. 7). Even
cuttings, but too few cores of silicic units were available to uniformly though H1 519 was a cuttings sample, it has a coherent zircon
follow this approach. All samples were selected based on their inter- population with an age of 0.534 ± 0.018 (MSWD = 0.7, n = 36).
preted stratigraphic position and petrographic evidence of containing This lava flow represents one of the uppermost units of the USF and
zircon. Cuttings samples were selected from intervals where a single provides a minimum age for this episode of volcanic activity. This unit
rock type made up 90% or more of the sample. As will be discussed, can be correlated to approximately the same stratigraphic horizon in all
silicic ash-flow tuffs of the area commonly contain zircon that was most wells drilled from H and F pads (and possibly from A3) and appears to
overlie the other dated tuff samples. One zircon was significantly older

104
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 7. Locations of zircon date samples relative to main stratigraphic formation and combining those below the USF for simplicity.

than the rest, with an age of 1.23 Ma. This is similar in age to one of the is unique within the broader population (i.e., not representative of a
zircons from the A3 sample, probably representing intrusive material larger population grouping).
formed during earlier eruptions of the USF.

2.2.2.4. ML-A3 955. A3 955 yielded a homogeneous population of 19


2.2.2.2. ML-H2 999. Two tuff samples from ML-H2 were from the zircons with ages from 1.65 to 3.13 Ma, and one zircon dated at 20.7
lower part of the USF about 50 m apart in depth (Fig. 7). The upper Ma (not considered in the age). The population yielded a mean age of
sample, H2 999 yielded only 9 zircons, 8 of which were derived from 2.93 ± 0.05 Ma (MSWD = 1.6, n = 19). The determined age is older
much older rocks. Seven fell in the range from 48.3 to 59.4 Ma, and one than the anticipated age of the unit based on GR log and petrographic
was 96.1 Ma. The remaining single zircon was dated twice, giving ages correlation with other wells. The single zircon dated at 20.7 Ma falls
of 0.167 and 0.183 Ma (mean age of 0.174 Ma). This age is the youngest within the age range of zircons dated from the H4 diorite and was likely
determined from subsurface samples but is inconsistent with other data. inherited from volcanics or intrusions of this age.
It is likely to represent contaminating material from shallower in the This sample was taken from core, but the determined age is much
well and is not considered further. older than stratigraphically equivalent silicic tuffs from well H2. After
detailed petrographic review, it was determined that this ash-flow tuff
2.2.2.3. ML-H2 1050. The second sample from H2, H2 1050, yielded a contains abundant intrusive lithic fragments. These intrusive lithics are
somewhat similar zircon population as H2 999, with 17 of the 20 considered to be the most likely source of all the zircons. The intrusive
zircons being derived from much older rocks, and three zircons being source could be either older or broadly coeval with the tuff’s magmatic
much younger and yielding an age of 0.559 ± 0.048 (MSWD = 1.3, source, but it is considered most probable that the dated zircons were
n = 3) using these three zircons. This age is consistent with H1 519 in derived from intrusive precursors of the host eruption, and not to re-
terms of the established stratigraphy. Of the older zircons, 15 have the present the eruption age of the host tuff. If the population represents the
same age as the granodiorite intrusion dated in H4 (91–100 Ma). Two source magma of the tuff, then the zircon dated at 1.65 Ma may re-
other older zircons were dated at 6.81 and 34.8 Ma. Each of these ages present the eruption age of the tuff or nearly coeval but older intrusive

105
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 8. Tera-Wasserburg concordia diagrams showing


SIMS U-Pb zircon data for Muara Laboh samples.
Zircon results are plotted uncorrected for common Pb
due to their young age and the resulting uncertainty of
reliably determining radiogenic Pb for individual
analyses. Instead, a linear regression was applied with
a fixed-intercept (207Pb/206Pb = 0.836) corresponding
to the Stacey and Kramers (1975) model Pb-isotopic
composition at 0 ka, which is representative for
common Pb in recent igneous rocks. Resulting con-
cordia intercept ages are indicated. Open symbols are
ages excluded from regression, because they may re-
present xenocrysts, contamination from material de-
rived higher up in the well, or be affected by minor Pb-
loss (in sample H4 2584).

debris, or a distinct lithic clast age. The age of 2.93 Ma is adopted as a magmatic episode began about 100 Ma and continued to about 94 Ma,
maximum age on the USF magmatic episode pending additional data culminating between ca. 88 and 84 Ma. The mean age is 96.1 ± 1.1 Ma
that will be needed to close this knowledge gap. As described later, (MSWD = 1.4, n = 17) excluding three slightly younger analyses which
comparison with the zircon populations from tuffs samples H2 999 and are likely from crystals affected by minor Pb-loss (see Appendix 1). As
H2 1050 indicate that the A3 955 tuff is unlikely to be derived from the noted above, this population of zircons is also represented in H1 and H2
same eruption since the populations have no overlap in age. tuff samples, probably from xenolithic debris since the underlying
granodiorite had not yet been drilled at the time of their sampling. As
mentioned above, the tuff from A3 (A3 955) did not contain any
2.2.2.5. ML-H4 2584. Sample H4 2584 has a homogeneous zircon Cretaceous-age zircon and therefore intrusions of this age may not be
population that corresponds to a Cretaceous plutonic episode found in present beneath that well. This also implies that the tuff sampled in A3
other parts of Sumatra (Cobbing, 2005). It appears that this protracted

Table 3
U-Pb zircon ages for well samples.
Sample Age (Ma) ± (Ma) Statistics

ML-H2OH (999) Rhyolite ash-flow tuff 0.174 0.020 (MSWD = 0.7, n = 2a)
ML-H2OH (1050) Rhyolite ash-flow tuff 0.559 0.048 (MSWD = 1.3, n = 3)
ML-H1OH (519) Hb-dacite lava 0.534 0.018 (MSWD = 0.7, n = 36)
ML-A3 core 1 (955) Rhyodacite lithic ash-flow tuff 2.93 0.05 (MSWD = 1.6, n = 19)
ML-H4 core 1 (3101) Diorite 20.8 0.3 (MSWD = 0.7, n = 18)
ML-H4 core 1 (2584) Granodiorite 96.1 1.4 (MSWD = 2.7, n = 18)

Notes: see Table A3 for all age determinations.


a
Mean age of two analyses of a single zircon. See Fig. 8 for graphical display of dates used in these interpretations.

106
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 9. Schematic stratigraphic section. The section shows the stratigraphy within the Muara Laboh Basin constrained by dating on the left (between Siulak and Suliti
Fault segments) and contrasts it with the likely section to the E of the basin on the right.

is from a different eruptive source and event than the tuffs sampled in 10) 0.167 & 0.183 [Pleistocene; two analyses of a single crystal, mean =
(n=1) 0.174]
H1 and H2. This indicates multiple silicic ash-flow tuff eruptions from
different sources and vents comprise the USF volcanic episode.

2.2.2.6. ML-H4 3101. Sample H4 3101 yielded a homogeneous These groupings suggest that most igneous activity occurred in the
population of 19 zircons with a mean age of 20.79 ± 0.3 Ma Early-Late Cretaceous (85–100 Ma), Paleocene-Eocene (48–63 Ma),
(MSWD = 0.7, n = 18) with one outlier of 63.1 Ma. Sample A3 955 Miocene (18–22 Ma), Pliocene (6.8, 1.2–3.1 Ma) and Pleistocene (0.46-
also had one zircon in this age range (20.7 Ma). 0.65 Ma). With the exception of the Paleocene and Eocene interval,
Fig. 9 depicts possible stratigraphic relationships among the dated these episodes largely fall within previously recognized plutonic-vol-
samples that are consistent with constraints from GR logs and petro- canic events in Sumatra (Bellon et al., 2004; Crow, 2005; Cobbing,
graphy. The age of some formations has been inferred from the range of 2005; Rosidi et al., 1996).
zircon ages, and the most likely host rocks for plutonic sequences. These It is notable that there were no zircons of Paleozoic age dated even
interpretations should be viewed as preliminary since only a few sam- though Paleozoic rocks are mapped on the NE side of the GSF about
ples were dated. It would not be surprising if further dating would 10 km to the N. This is consistent with our geologic map (Fig. 2) and
necessitate revision of this initial effort. other studies suggesting that basement rocks are largely limited to the
east side of the GSF (Rosidi et al., 1996). It is also consistent with a
2.3. Zircon age distribution and possible stratigraphic correlations marked decrease in Bouguer gravity along the eastern edge of the
Muara Laboh basin (see Fig. 6 in Mussofan et al., 2018). No samples
Since no previous geochronological studies exist for the Muara were available from exploration well C1, which is located in the far NE
Laboh area, we take the approach of considering the entire dated zircon of the drilled area and near the projected trace of the Suliti Fault seg-
population as a guide to reservoir rock ages and regional affiliations, ment (see Fig. 4). This well may have penetrated Paleozoic rocks based
when viewed in the context of other local and regional geological stu- on descriptions of foliated granitoid and metamorphic rocks from about
dies (Bellon, 2004; Cobbing, 2005; Rosidi et al., 1996; Figs. 2 and 3). 1002 m to its total depth (ITB, 2016), however, these materials may
This approach is, by its nature, rather speculative and is thereby subject also be resedimented into younger alluvial deposits from the nearby
to verification by further study. Paleozoic outcrops.
The overall age distribution of 123 dated zircons ranges from 0.167 Petrography of well cuttings and core suggests that the Late
Ma to 99.9 Ma, highlighting that the Sumatran crust underlying the Cretaceous granitoids intruded a basaltic to andesitic composition
Muara Laboh Basin is at least 100 Ma, and was constructed from mul- volcanic complex with lesser silicic intervals (Fig. 6). Unless this se-
tiple igneous events (Appendix 1 and Figs. 9 and 10). Ages group into quence overlaps in age with the intruding plutons, it appears that no
the following intervals from oldest to youngest with significant gaps: zircon was derived from these formations, possibly because of their
dominant mafic to intermediate composition. Based on regional corre-
lation this sequence may represent either the Jurassic to Early Cretac-
1) 84.5 to 99.9 [Late Cretaceous] eous Woyla Formation, or Late Cretaceous volcanism related to the
to(n=36) contained intrusions (Barber et al., 2005a,b; Cobbing, 2005).
2) 58.4 to 63.1(n=5) [Paleocene]
Unconformably overlying the intruded Mesozoic volcanic-intrusive
3) 48.3 to 50.7(n=3) [Eocene]
4) 34.8 (n=1) [Latest Eocene] complex, is a sequence of more silicic tuffs, lavas, and related volca-
5) 18.4 to 22.0 (n=20) [Miocene] niclastic rocks that are grouped as the Lower Rhyolite-Dacite Formation
6) 6.81 (n=1) [Pliocene] (LRD). This formation also includes some andesitic to basaltic intervals
7) 2.77 to 3.13(n=18) [Pliocene]
or is cut by dikes of that composition. These rocks are commonly oxi-
8) 1.23 to 1.65(n=2) [Pleistocene]
9) 0.463 to [Pleistocene] dized with abundant groundmass hematite and chlorite. Volcanic tex-
0.652(n=38) tures are generally not well preserved but are still recognizable in some
samples. No ages were determined on this sequence, but it is inferred to

107
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 10. Histogram of all dated zircons. The number of zircons dated in a given age range in parentheses.

be Paleocene or Eocene based on its position between Cretaceous and descriptions. Veins showing open-space textures in core samples were
Miocene intervals and inherited zircon xenocrysts of this age observed selected for paragenetic and fluid inclusion studies because they are
in younger tuff samples. Based on regional studies, it is possible that likely to reflect the recent evolution of temperature and fluid chemistry.
these rocks are correlative with the Paleocene Kikim Formation or Mineral abbreviations used in the following figures and text are given in
unrecognized Eocene to Oligocene formations (Crow, 2005). Table 4 and the paragenesis of samples is summarized in Table 5.
Overlying the LRD is a sequence dominantly of andesitic composi-
tion that is correlated with the Painan Formation exposed at surface in
3.1. Distribution of alteration assemblages
the area. The Painan Formation has been dated regional at ∼24 to 14
Ma (Bellon et al., 2004; de Smet and Barber, 2005; Rosidi et al., 1996).
In general, the alteration patterns observed in the Muara Laboh
In the Muara Laboh wells it consists primarily of coarse-grained lavas or
geothermal system are similar to those documented from other liquid-
dikes with lesser tuffaceous and volcaniclastic interbeds. One of the
dominated systems with temperatures of 230 °C or greater (Browne,
dated intrusive samples lies within the age range of the Painan For-
1978a, b; Hedenquist et al., 1992; Milicich et al., 2018; Moore et al.,
mation and may be a dioritic equivalent of this extrusive sequence, but
2001; Reyes, 1990; Stimac et al., 2015). The shallow cap of the system
the intrusion cuts older rocks of the Cretaceous volcanic/intrusive
consists of smectite-rich to transitional (mixed-layer clays and chlorite)
complex at the location sampled.
argillic alteration. The permeable reservoir alteration consists of over-
lapping propylitic (Ep-Chl-Qtz-Py) and phyllic (Ser-Qtz-Py) alteration
zones, however, as described below, Ep is largely absent from the
3. Hydrothermal alteration and fluid inclusions
shallow steam cap. Silicic formations typically contain higher propor-
tions of Ser, Qtz, and Ad, whereas intermediate and mafic composition
Alteration mineralogy was determined from wellsite descriptions,
rocks are dominated by calc-silicate minerals such as Ep and Chl. Hy-
methylene blue (MEB) analysis, detailed petrography and XRD analysis
drothermal Amph and rare garnet are present in the vicinity of intru-
of selected samples. MEB analysis was particularly useful in estimating
sions and where temperatures are currently in excess of about 280 °C.
the top of the reservoir during drilling in the NE reservoir. In general
The general shape of the clay cap and underlying propylitic zone in
MEB index < 10 showed good agreement with the base of the smectite-
relation to current subsurface temperature has been described else-
rich clay cap as determined from resistivity (see Figs. 5 and 12 in Stimac
where (Dyaksa et al., 2016; Baroek et al., 2018; Stimac et al., 2019).
et al., 2019). Bulk alteration and vein mineral assemblages and para-
These workers showed that the occurrence of vein Ep (defining the
genesis were mostly interpreted from thin sections and wellsite cuttings
propylitic zone) is a poor predictor of the top of the permeable reservoir
at Muara Laboh, both in the NE and SW sectors (Fig. 11), but for dif-
Table 4
Alteration Mineral Abbreviations. ferent reasons. In the NE, the top of the reservoir is better predicted by
the occurrence of vein Wai or Qtz-Preh that seem to have formed re-
Mineral Abbreviation cently at temperatures of < 240 °C. Epidote is found at deeper levels in
Adularia Ad this area even though current temperatures are no higher than at
Amphibole Amph shallow levels. This Ep is typically overgrown by later Wai. Calcite is
Anhydrite Anhy also uncommon at shallow levels in the NE but is present along with
Calcite Cc Anhy locally at deeper levels.
Chlorite Chl
Epidote Ep
In the SW, there is a significant interval of the upper propylitic zone
Prehnite Preh that is currently less permeable than would be expected. This appears to
Pyrite Py be related mainly to infilling of early-formed Ep ± Ad veins with later
Quartz Qtz Qtz ± Preh ± Wai or Cc ± Qtz, sealing once permeable fractures.
Sericite/Illite Ser
Calcite is particularly abundant in the sealed interval (Fig. 11). Calcite
Titanite Tit
Wairakite Wai infilling is dominant in the uppermost propylitic zone with Qtz ± Preh
becoming more abundant with depth (Mussofan et al., 2018). Late-stage
Note: albite is commonly present but not described. veins filled exclusively by Cc are also common in the SW. These

108
J. Stimac, et al. Geothermics 82 (2019) 97–120

Table 5
Characteristics of samples analyzed for fluid inclusions.
Sample and Measured Depth Elevation (m asl) A (mm) KA (mm) Vein Paragenesis Host Rock
(m MD)

ML-A2 710.13 733-737 1.0 1.5 Cc-Anhy > Wai Andesite lava breccia
ML-A2 714.32 2.0 5.0 Ad?-Tit-Qtz > Wai
ML-A3 959.76 573 2.0 4.0 Ad-Tit > Qtz-Ser-Py Rhyolitic lithic ash-flow
tuff
ML-A1 1552.20 125 5.0 Qtz-Ad- > Ep > Wai-Cc, Anhy-Ep-Wai-Cc Andesite lava breccia
ML-A1 1552.50 4.0 Qtz-Ad > Ep > Wai, Ad-Anhy > Ep > Wai-Cc
ML-H4 2588.78 −799 8.0 Sheared Ep, etc. Granodiorite
ML-H4 2589.14 −800 35.0 Ep-Amph > Cc > Dissolution, Ep-Preh, Ep-Amph > Ep-Qtz,
Shearing > Qtz-Ep-Amph ± Cc
ML-H4 3100.78 −1254 10.0 Amph, Amph-Ad, Amph-Ad-Ep, Tit-Qtz Diorite
Adularia
Amphibole
Anhydrite
Calcite
Epidote
Plagioclase
Prehnite
Pyrite
Quartz
Sericite/Illite
Titanite
Wairakite

Notes: A, open fracture aperture; KA, kinematic fracture aperture (includes fillings); see Table 4 for mineral abbreviations. -, coprecipitation; > , sequential pre-
cipitation.

paragenetic relationships are consistent with extensive boiling, ac- bearing fluid inclusions in the granodiorite from ML-H4 core 2 dis-
companied and followed by ingress of cooler bicarbonate-rich waters cussed above.
(Baroek et al., 2018).
3.2.1. Shallow NE reservoir
3.2. Vein paragenesis and fluid inclusions Veins from the shallowest cored level in the NE reservoir are re-
presented by well A2 at about 710 to 718 m (Fig. 14), where silicic
In this section we describe vein paragenesis and fluid inclusion data andesite lava and breccia are cut by a series of veins and tectonic and
from selected core samples in more detail (Table 5). Since the NE and hydrothermal breccias. Open and partially open veins consist of early
SW reservoir sectors represent different depths and temperatures, they Tit, Ad and Qtz, and later Wai. In some veins an episode of Anhy ± Cal
are described separately, followed by an overview of the system and its precipitation similar to that observed in A1 at 1552 m is enclosed in
inferred recent history. later Wai. Sample A2 710.13 m contained open veins with mono-
Fluid inclusions were analyzed from selected veins with open space mineralic and euhedral Wai. Since Wai is the last mineral to form in the
where paragenesis was already determined (Jones and Moore, 2018). polymineralic veins, we infer that the euhedral Wai was formed re-
Samples were selected from cores of both the shallow NE (A1, A2, A3) cently at temperatures closer to the current condition (Table 5 and
and deep SW reservoir (H4) sectors to determine their homogenization Fig. 14).
temperatures and estimate their salinities. Veins in samples from A2 All fluid inclusions were two phase liquid-vapor at room tempera-
and A3 were mostly open and probably currently active in hydro- tures and secondary in origin. The majority of the Th values fall be-
thermal circulation, whereas veins in the samples from A1 and H4 were tween ∼245 and 260 °C and the Tm ice values between about -0.7 to
mostly sealed and may not have been involved in large-scale fluid cir- -0.2 °C (1.2 to 0.4 wt% NaCl eq) (Table A2, Fig. 12). Anhy and Cc have
culation. retrograde solubility and are expected to precipitate on heating of
More than 231 fluid inclusions were measured in Anhy, Cc, Wai, Qtz downward percolating fluids and/or during mixing, suggesting that this
and Ep (Appendix 1). No temperature homogenization (Th) data were may have been a zone or marginal fluid recharge in the past.
obtained from Wai because these inclusions leaked at elevated tem- Core samples from A3 from 951 to 959 m consisted of a silicic ash-
perature on heating. Homogenization (Th) and ice melting temperatures flow tuff with high GR counts containing quartz, biotite, and plagio-
(Tm ice) are shown in Table A2 grouped by fluid inclusion population. clase phenocrysts in a groundmass with spherulitic devitrification tex-
Two phase, liquid-dominated fluid inclusions containing liquid and tures. Veins with significant open space are common (Fig. 15). Vein
vapor at room temperature were the most common type observed minerals comprise a typical phyllic assemblage with early Ad and Qtz,
(Fig. 12). Vapor-rich inclusions were observed in samples A3 959.76 m along with Ser, Py ± Chl and Tit. Adularia is commonly formed in
and H4 3100.78 m, and inclusions containing two or more daughter boiling environments due in part to an increase in pH that accompanies
salts (halite plus others) where observed in H4 3100.78 m. The majority CO2 loss (Browne, 1970). Some veins show complex multi-stage
of the fluid inclusions measured were secondary in origin. Primary fluid shearing and brecciation with selvages of enhanced open space.
inclusion populations were identified in Qtz growth zones in A Pad Sample A3 959.76 m contains a vein that was filled by early Ad and
wells. 3D populations in A1 and H4 core 1 are potentially primary in later Qtz. Fluid inclusions in this sample were primarily two phase li-
origin and are denoted as such in Table A2. No temperature corrections quid-rich inclusions and less abundant vapor-rich inclusions along grain
were applied to measured homogenization temperatures. boundaries and as secondary inclusions in healed fractures. Primary
Fluid inclusion characteristics will be discussed by cored interval as fluid inclusions in growth zones parallel to the edge of the crystal and
shown in Table 5. The majority of fluid inclusion Th’s match or exceed secondary population in healed fractures were measured in quartz
measured temperatures by ∼10-20 °C, suggesting only modest cooling crystals. The majority of the Th values fall between ∼245 and 260 °C
since fluid trapping (Fig. 13). The exception being the daughter salt- and the Tm ice values between about -1.1 to -0.5 °C (1.9 to 0.9 wt % NaCl

109
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 11. NNE-SSW cross sections of Ep and Cc abundance based on detailed petrography. Scale is 1=≤1-2%; 2 = 2–5%; 3 = 5–15%. Note data are limited by
circulation losses in the shallow NE wells (A pad) which typically occurred by about 600 m asl. Core in A1 and A3 confirm minor Cc and Ep at deeper levels. Yellow
shaded area denotes ≤10 OΩm resistivity in both figures.

eq) (Table A2). The temperature and salinity of the fluids appear to by Wai. The core in A1 was taken near bottom where the well tem-
have decreased with time from those trapped in the primary inclusions, perature shows a slight reversal as it traverses to the SE.
to the secondary inclusions, to measured modern values (Fig. 16). Fluid inclusion populations were measured in Anhy and Wai, the
Vein paragenesis in the deeper part of the NE reservoir is illustrated majority of which are secondary in origin. All fluid inclusions are two
in samples from A1 at 1552 m depth. Tectonic and hydrothermal veins phase liquid-rich at room temperature. Most of the Th values fall be-
and breccia fillings consist of early Ad-Ep ± Qtz followed by Anhy and tween ∼225 and 245 °C and Tm ice values of about -1.0 to -0.2 °C (1.7 to
slightly later Cc, bladed Ep, and finally encapsulated in later Wai, which 0.4 wt% NaCl eq), with one secondary population at higher Th and
completely filled late-stage veins and breccias (Fig. 17 and Table 5). Ad- lower Tm ice (Table A2). The paragenetic relationships and fluid inclu-
Ep-Qtz was precipitated first on vein walls, whereas Anhy and Cc en- sion data are consistent with initial vein formation at 250–260 °C, fol-
close these minerals, suggesting an episode of downflow of sulfate- and lowed by ingress of steam heated waters at temperatures of ∼230 °C,
bicarbonate-rich waters, possibly from the nearby IM fumarole area similar to the modern temperature of 232 °C at this depth.
(see Fig. 4). Anhydrite is locally overgrown by a second generation of In summary, the NE reservoir near Idung Mancung (IM) fumarole
acicular epidote suggesting that temperatures remained high or re- hosts a ∼240 °C steam cap and underlying liquid that represents the
covered at this time. Remaining open space in the vein was later infilled shallowest part of the system (Dyaksa et al., 2016; Situmorang et al.,

110
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 12. Examples of fluid inclusions from Muara Laboh. A) Large liquid-rich fluid inclusion. B) Primary fluid inclusions in a growth zone parallel to the crystal face.
C and D) Secondary fluid inclusions. E) Liquid-rich fluid inclusion with a halite daughter salt and vapor-rich inclusions. F) Vapor-rich fluid inclusions. Sample name,
encapsulating mineral, and magnification are shown at the bottom of each photomicrograph. Field of view is ∼130 μm in all images.

2016). It hosts Wai, Qtz, and Preh (Preh only observed in cuttings) veins 3.2.2. Deep SW reservoir
with open space texture (Mussofan et al., 2018). The shallowest ande- Rocks hosting the deep SW reservoir are mainly older volcanic se-
sitic samples are dominated by open veins with Wai, intermediate depth quences of dacitic to basaltic composition cut by a variety of intrusions.
rhyodacite by Ad-Qtz-Py-Ser, and deeper andesitic samples by early Ad- As described above, the top of the propylitic alteration zone, as defined
Ep→Anhy ± Cc→Ep→Wai. Based on paragenesis and fluid inclusions by the presence of megascopic vein epidote, does not conform to the top
the NE reservoir rocks were hotter in the past (up to ∼260 °C) and of the reservoir due to sealing of fractures. Relict propylitic alteration
probably near the source of boiling upflow. Geothermometry from A and local potassic alteration (not in equilibrium with measured tem-
Pad wells suggests that the upflow local to this area is currently at a peratures) are present in some edge wells drilled for injection where
maximum of about 250 °C (Stimac et al., 2019). The consistent occur- temperatures are currently 180 °C or less (e.g., E1 and E2) and at about
rence of Wai last in the paragenetic sequence suggests that open veins 220–230 °C within the reservoir area. More importantly the upper part
with only Wai such as at A2 711.81 m formed most recently and of the propylitic zone has been partially sealed by late-stage Cc, Qtz,
probably represent current conditions. and Qtz ± Preh ± Cc assemblages (Fig. 18). This suggests that hy-
drothermal circulation has had a long and complex history in this area.

111
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 13. Equilibrated well temperature surveys plotted with fluid inclusion Th. Also plotted is a boiling point to depth curve for pure H2O with the water table at the
surface (solid black line), and a conductive temperatures profile (dashed black line) that is more representative of the cap rocks for wells drilled from the A pad than
the PT surveys, which are affected by upward movement of steam in the wellbore.

In deeper portions of the system (generally below 200 m bsl) high indicate they were emplaced recently, after the bulk of hydrothermal
permeability is associated with open Ep ± Ad ± Amph veins produ- alteration took place. These intrusions are interpreted where abrupt
cing fluid at ≥270 °C. Here, permeability is only locally reduced by changes in GR counts and high-angle contacts are observed on
late-stage infilling by Cal, Qtz, and Preh and shearing (Figs. 19 and 20). resistivity images in the SW reservoir permeable intervals (see Fig. 5
below about 2300 m and Baroek et al., 2018). One such relatively
3.2.2.1. Relationship between intrusion and permeability. Large intrusive unaltered microdiorite intrusion was described from cuttings near the
bodies were intersected in wells H1, H2, H2RD, H4 and F1. Smaller upper permeable zone of H3 (Fig. 18). The weak alteration of this rock
intrusions were encountered in all the deep wells directed to the SW is in strong contrast to other cuttings with pervasive propylitic
reservoir. Fine-grained, small intrusions (mostly dikes) are interpreted alteration, suggesting that it was emplaced late in the alteration
from primary rock textures, and abrupt changes in gamma ray logs that history of the rock.
correlate with sharp contacts and a change from volcanic (variable Samples H4 2588.78 m and 2589.14 m (Fig. 19) contain veins that
resistivity, layered, fragmental) to intrusive rock characteristics (non- were filled by early Act and Ep, followed by Ep and Qtz with late Qtz,
layered, high resistivity, abundant fractures) determined from image Ep, Act and Cc. There is also evidence for recent dissolution of Ad and
logs. The dip of the contact indicates whether the intrusion is a dike Cc (e.g., Fig. 19) in the vicinity of permeable zones. Two phase liquid-
(high angle) or sill (low angle). Permeable zones tend to be common rich fluid inclusions were observed in these samples. Inclusion popu-
near the margins of both large and small intrusions (Baroek et al., lations were measured in Qtz, Cc and Ep. All inclusions in Qtz were
2018). Entries do not appear to be common within the larger mass of secondary. Three-dimensional arrays of inclusions in calcite may be
96.1 Ma coarse-grained granodiorite, which has alteration and primary. Large isolated inclusions in epidote may also be primary. The
deformation textures suggesting it was emplaced and propylitically majority of the Th values fall between ∼305 and 340 °C and the Tm ice
altered early in the history of the area. For example, multiple episodes values between about -2.5 to -0.2 °C (4.2 to 0.4 wt% NaCl equivalent
of Ep ± Preh deposition can be distinguished in some intrusive (eq)) (Table A2).
samples based on Ep color, texture, and degree of deformation. A Veins containing Amph, Qtz, feldspar and opaques cut sample ML-
finer-grained dioritic intrusion dated at 20.9 Ma is considerably less H4 3100.78 m (Fig. 20). Three types of fluid inclusions were observed
altered and deformed. Secondary amphibole and lesser garnet are in this sample: 1) vapor-rich; 2) liquid-rich-one or more daughter
common vein minerals in these intrusions. A few undated fine- crystals; and 3) two phase liquid-rich. Cubic halite crystals and an
grained intrusions contain fresh plagioclase and pyroxene that opaque phase, most likely a sulfide, were commonly observed;

112
J. Stimac, et al. Geothermics 82 (2019) 97–120

the geothermal system is correct, then relatively unaltered dikes de-


scribed in the SW (see Fig. 18) may also be < 41,000 years old since
they apparently post-date most of the hydrothermal alteration asso-
ciated with reservoir formation.
Debris avalanches associated with sector collapse can be generated
by large earthquakes, volcanic eruptions, or degradation of edifice
strength due to progressive conversion of rocks to hydrothermal clay
(Belousov et al., 2007; Ui et al., 2000; van Wyk de Vries and Davies,
2015). Loss in cohesion can be exacerbated by episodic high pore
pressures generated by magmatic-hydrothermal circulation. The
proximity of the Siulak Fault to the PS volcano, its intersections with
NE-trending faults, and extensive clay alteration related to the Muara
Laboh geothermal system, are all factors that likely contributed to
failure of the PS stratocone. The lack of a prominent resistivity low in
the crater is also consistent with removal of the shallow clay cap of the
geothermal system by debris avalanche (see Fig. 10 in Stimac et al.,
2019). However, we cannot eliminate the possibility that the crater
formed by a series of smaller erosional events over a longer time in-
Fig. 14. Summary plot of fluid inclusion populations for which both Th terval. In this scenario, the debris flow deposits dominating the surface
(homogenization temperature) and Tm ice (freezing point depression [FPD]) geology of the area would have been created during high-energy
measurements were obtained from core sample A3 959.76 m. Symbols are storms, or small collapse events over geologic time.
placed at average values of Th and Tm ice, with error bars showing the measured The tuffs that overlie the bulk of the debris flow deposits may
ranges. The shape of the symbol represents the mineral that contains the fluid provide evidence that a late eruptive episode of the volcano led to
inclusions (see legend). An X is used to denote modern measured temperatures sector collapse. Destabilization of a volcanic edifice by emplacement of
and an approximation of salinities based on (total dissolved solids [TDS] in well new magma, resulting in sector collapse and subsequent ash-flow tuff
samples). The relationship between primary and secondary fluid inclusions and
eruption, was vividly demonstrated by events at Mount St. Helens in
modern fluid chemistry suggests a decrease in both apparent salinity and
1980 and Bezymianny in 1956 (Belousov et al., 2007). Tuffs described
temperature with time.
above have some features of the directed blast deposits at these vol-
canoes, including a thick carbon layer that could have resulted from
however, the number and size of the daughter crystals is variable. Fluid tree “blowdown”. Although no detailed volcanological studies of these
inclusions containing an opaque daughter mineral do not homogenize deposits confirm they have the characteristics of a directed blast, the
at temperatures of up to 490 °C; however, halite dissolution tempera- general sequence of debris flow deposits overlain by fallout and ash-
tures ranged from 290 to > 490 °C, indicating salinities of ˜37 to > flow tuffs with locally associated woody layers is consistent with a
58 wt% NaCl eq. The observed fluid inclusion assemblage, consisting of linked sector collapse and volcanic eruption origin. Alternatively the
vapor-rich fluid inclusions and high temperature and high salinity fluid tuffs may be related to post-collapse volcanism of PS volcano.
inclusions with multiple daughter salts, is similar to those reported from
porphyry Cu/Mo/Sn deposits (e.g. Roedder, 1984). This assemblage 4.2. Stratigraphy and zircon dating
likely represents boiling of a magmatic fluid at temperatures and
pressures unrelated to the modern hydrothermal system. However, The subsurface stratigraphy at Muara Laboh has been established
secondary liquid-rich fluid inclusions from this sample that postdate the from well data and age-calibrated by U-Pb zircon dates. Zircon provides
early high temperature event were also measured. Secondary liquid- reliable age information from hydrothermally altered rocks due to it
vapor fluid inclusions give a fairly narrow Th range, averaging ∼305 °C. resistance to hydrothermal alteration, replacement, or recrystallization
Tm ice values range from -1.6 to -0.4 °C (2.7 to 0.7 wt% NaCl eq) (Table (Wilson et al., 2008; Milicich et al., 2013). Dating individual zircons
A2). provides a spectrum of crystallization ages for primary magmatic ma-
terial in these rocks. The peak of zircon crystallization provides a
4. Discussion maximum age limit on the eruption(s) concerned or provides an ap-
proximation to the emplacement age for intrusive rocks, whereas the
4.1. Sector collapse and evolution of the hydrothermal system youngest zircons provide a close approximation to eruption age (e.g.
Milicich et al., 2013; Schmitt et al., 2003). Combining age data with
Some aspects of the Muara Laboh surface geology, such as the origin other geological studies supports further understanding of the evolution
of the PS craters, whether the project site was covered by the resulting of geothermal systems (Moore et al., 2000; Schmitt et al., 2003;
debris avalanche, and the timing of collapse relative to evolution of the Milicich et al., 2013).
geothermal system were not fully resolved by this study. We have In the case of Muara Laboh, relatively little was known about the
presented evidence that the northwesterly PS crater was formed by a subsurface geology prior to this study. Paleozoic formations outcrop on
collapse prior to 41,000 years B.P., but most likely somewhat older than the west side of the Suliti segment of the GSF and appear to extend
this, since a thick sequence of post-collapse lavas appear to partially fill toward Mt. Kerinci as erosional promontories on the east side of the
and spill out of the crater (Fig. 3). Based on the distribution of clay as Muara Laboh basin (Fig. 2). It is not fully known if the edge of the
indicated by resistivity, the collapse events and subsequent erosion Paleozoic craton is coincident with the current GSF as seems consistent
have partially removed the clay cap of the geothermal system (see with gravity data (Fig. 6 in Mussofan et al., 2018), and the stratigraphic
Stimac et al., 2019). This would have resulted in extensive boiling and a section presented here, or whether the Muara Laboh basin is an area of
drop in fluid level of the geothermal system, and we infer that sealing of crustal extension filled at mid-crustal levels by intrusion, and at shallow
the shallower part of the SW reservoir described above is related to levels by volcanism and sedimentation. But the lack of zircon xeno-
these events. These events may have also contributed to formation of crysts of Paleozoic age is consistent with Paleozoic rocks being absent
the modern steam cap in the NE, attendant with cooling and limited from the crustal section underlying the basin. We suggest it is most
ingress of steam heated waters. If this interpretation regarding the re- likely that the Muara Laboh pull-apart basin is floored by heavily in-
lationship between sector collapse of PS volcano and the evolution of truded Mesozoic age crust. It is possible Paleozoic rocks extend to the

113
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 15. A2 vein occurrence and paragenesis. (Top) samples are ∼4 x 6 cm and shown in plane light. Paragenesis of veins is early Anhy-Cc (710.13 m) followed by
Wai (all samples). Middle: Multi-stage veins at 710.13 m. Early Anhy-Cc, infilled by later Wai. Bottom: Open single-stage veins at 711.81 m partially filled by
euhedral Wai. Photomicrographs are shown in both plane (left) and crossed polarized light (right).

west side of the GSF, but it is considered more likely that the edge of the Formation may be present at depth, but not exposed at surface. This
Paleozoic block is marked by the GSF in this area. This would be con- formation is present widely in the Barisan Mountains on both sides of
sistent with other studies suggesting that basement rocks are largely the GSF (Barber and Crow, 2005). According to these authors it consists
limited to the east side of the GSF (Barber et al., 2005b), and that this of oceanic crustal, arc, and marine carbonate assemblages. The oceanic
profound structural boundary played a roll in localizing strike-slip crustal assemblage is comprised of serpentinite, gabbro, basalt lavas
faulting. and pillows, and hyaloclastite with associated volcaniclastics and se-
We infer from well data that the Jurassic to Mid-Cretaceous Woyla diments such as chert, shale and sandstone. The arc assemblage consists

114
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 16. Vein occurrence and paragenesis in A3. Top: multi-stage veins, breccias and selvages.Veins show a relatively simple paragenesis with early Ad and Tit,
followed by Qtz, Ser, and Py. Enhanced open space is found in selvages adjacent to some veins. All samples are∼4 x 6 cm and shown in plane light. Middle: Open
veins at 959.76 m contain early Ad overgrown by Qtz. Matrix alteration includes Py, Ad, Qtz, Chl, and Ser. Bottom: Open vein at 959.76 m with initial precipitation of
Ad and Tit, followed by Qtz. Photomicrographs are shown in plane (left) and cross polarized light (right).

of porphyritic basaltic to andesitic lavas and dikes with associated 4.3. Controls on permeability and system evolution
breccias, tuffs and volcaniclastics. Based on sparse GR log and petro-
graphic data indicating an andesitic to basaltic volcanic sequence Detailed observations from rock samples and wireline logs coupled
hosting Late Cretaceous plutons, the Woyla arc assemblage may be with well tests confirm that fractures provide the main control on re-
present in deep wells. servoir permeability at Muara Laboh (Baroek et al., 2018). Moreover,
those fractures that support flow have orientations related to the pre-
vailing patterns of in situ stresses and faulting expected in a pull-apart

115
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 17. Vein occurrence and paragenesis in A1 at 1552 m. (Top) full thin section photos of complexly fractured zone. Samples are ∼4 x 6 cm and shown in plane
light. Andesite lava with interconnected fractures, irregular branching veinlets and puzzle breccia. Most open space in veins and breccia is infilled by Anhy and later
Wai, but some open space is present in the adjacent andesite (represented by blue dyed epoxy). Bottom: Transect of sealed vein with host rock at left. Paragenesis of
vein is Ad-Ep→Anhy/local shearing→Cc (not seen in this view)→bladed Ep→Wai. Arrow indicates general direction of mineral growth. Photomicrograph in plane
(left) and cross polarized light (right).

basin environment (Mussofan et al., 2018; Stimac et al., 2019). Cuttings in the SW reservoir some fractured regions are occluded by late Cc, Wai,
and core data also constrain a complex history of fracturing and mineral and Qtz ± Anhy. In the NE reservoir, Wai and Qtz-Ser ± Ad ± Tit
infilling. Ad and Ep ± Amph were deposited on fracture initiation and are found at shallow levels and Ad and Ep are found deeper, and pos-
are found in open-space veins in the system upflows. At shallower levels sibly only as relict phases. Paragenesis indicates ingress of steam-heated

116
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 18. SW reservoir alteration. Top left: H2 1548 m epidote encapsulated in calcite (crossed polars). Top right: H2 2400 m epidote encapsulated in Qtz (crossed
polars). Bottom: H3 2319 m relatively fresh microdiorite (plagioclase-pyroxene-FeTi oxides, minor chlorite/amphibole) that may post-date most tectonic deformation
and propylitic alteration (plane polarized light at left, and crossed polars at right).

water has locally reduced permeability. In particular, a significant part much greater depth than a gas free fluid (e.g. Sasada et al., 1986).
of the upper propylitic zone in the SW was impacted by Cc and Qtz Only ∼3.4 wt% CO2 can be dissolved in a fluid before hydrates
infilling of open veins, sealing formerly permeable areas. Later calcite appear on freezing (Hedenquist and Henley, 1985). No such hydrates
veins appear to indicate descent of bicarbonate-rich water, possibly were observed during fluid inclusion freezing runs and therefore these
related to a drop in reservoir pressure and fluid level. Sector collapse of fluids must contain less than 3.4 wt% CO2, which is equivalent to a
the nearby PS volcano, or regional fault rupture provide possible me- freezing point depression of ∼1.5 °C or less. The observed spread in Tm
chanisms for this evolution. Sector collapse has been considered as a ice in samples A2, A3 and H4 core 2 is less than 1.5 °C and could
cause for boildown of other geothermal systems including Wayang therefore be due to CO2 loss alone. The observed Tmice range in A1 and
Windu (Bogie et al., 2008) and Karaha-Bodas (Moore et al., 2002) and is H4 core 1 is greater than 1.5 °C, and therefore, loss of CO2 alone cannot
advocated in this study. account for the observed range in Tm ice. One possible explanation could
Measured temperature and total dissolved solids (TDS) salinity data be that CO2 loss during boiling was followed by mixing with a dilute
were compared to fluid inclusion results and are shown in Figs. 13 and fluid. Vein anhydrite and/or calcite are present in all core intervals
21. The TDS data have been converted to Tm ice by assuming it is a NaCl except H4 core 2 and may record the infiltration of dilute recharge
fluid. The calculated salinities of the modern fluid are consistently waters as these minerals display retrograde solubilities (they precipitate
lower than those suggested by fluid inclusion freezing point depression on heating) and convex solubility curves (they precipitate on mixing).
data suggesting a decrease in salinity with time (Table A2).
The drop in apparent salinity and temperature likely results from
5. Conclusions
boiling of a CO2-bearing fluid, mixing with a cooler dilute fluid or a
combination of these two processes. The boiling of a CO2-bearing fluid
Muara Laboh is situated in a pull-apart basin formed in a stepover
(or effervescence), will result in a rapid and almost complete loss of CO2
between segments of the Great Sumatra Fault. Dilation accommodated
over a small temperature drop without significant steam loss (e.g.
prolonged intrusion at depth and overlying brittle extension and shear
Hedenquist and Henley, 1985; Wilkinson, 2001). CO2 dissolved in H2O
within the pull-apart basin. Quaternary magmatism occurred primarily
will impact Tm ice much like NaCl, with higher concentrations of dis-
along the southern basin margin segment. Ages of surficial tuffs were
solved CO2 resulting in larger freezing point depressions. Modern
obtained using the 14C method and U-Pb ages were obtained on in-
measured CO2 concentrations at Muara Laboh are generally < 1 wt%
dividual zircons from well samples. Tuff outcrops associated with debris
(equivalent to ∼-0.4 °C freezing point depression) and cannot account
flows resulting from sector collapse of the Patah Sembilan volcano
for the observed decrease in salinity. There is evidence of boiling in
range in age from 33 to 41 ka, providing a minimum age of the event.
many of the core samples from Muara Laboh, such as vapor-rich in-
U-Pb dates on zircons from silicic volcanic formations and intrusions
clusions in A3, and vein adularia in samples A1, A2, A3 and H4 core 1.
comprising the Muara Laboh reservoir range from about 0.5 to 96 Ma. A
An important consequence of dissolved CO2 in geothermal fluids is its
dominantly silicic sequence underlying the Patah Sembilan Andesite
impact on the boiling point to depth. A CO2bearing fluid will boil at
Formation was deposited from about 2.9 to 0.5 Ma. Underlying

117
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 19. Vein occurrence and paragenesis in H4 2589.14 m. Top: Part of a sheared vein with vuggy open space towards its center (dashed yellow lines highlight shear
planes). The width of the shear zone is about 35 mm while the area with some unfilled space is about 4–6 mm wide (blue epoxy)(plane polarized light). Middle: Ep
with later infilling of Cc, which may currently be undergoing dissolution at this location (resorbed texture) (plane polarized light at left, cross polars at right). Bottom:
Ep-Qtz intergrowth in open space within a sheared epidote vein (plane polarized light at left, cross polars at right).

volcanic-intrusive packages appear to represent magmatic episodes of fluids suggesting an increasing proportion of meteoric water input.
Miocene (22-18 Ma), Paleocene-Eocene (48–63 Ma), and likely Jurassic Trends in Th and Tmiceare consistent with boiling of a CO2 bearing fluid,
to Early Cretaceous age. Early Miocene diorite (20.8 Ma) and Late mixing with a dilute fluid or a combination thereof. In the deeper SW
Cretaceous granodiorite (96 Ma) intrusions cut underlying andesitic to reservoir boildown and ingress of steam-heated water led to calcite and
basaltic rocks of likely Late Mesozoic age. Weakly altered microdiorite quartz precipitation followed by calcite veining that reduced perme-
dike intrusions are locally associated with permeable zones near the SW ability in the upper propylitic zone. This evolution may be related to
reservoir upflow. These dikes were not dated but their emplacement breach of the system’s clay cap by sector collapse of the Patah Sembilan
appears to post-date the main episode of hydrothermal alteration. volcano, or alternatively, by its more gradual excavation by erosion.
Hydrothermal vein occurrences, paragenesis, and contained fluid
inclusions indicate that the Muara Laboh geothermal system has a long
and complex history. Adularia and epidote were deposited upon frac- Acknowledgments
ture initiation and are found in open veins in the system upflows,
whereas wairakite and quartz ± prehnite dominate in the shallow Hearty thanks to Supreme Energy and their partners, Sumitomo and
steam cap region. Mineralogy and fluid inclusion homogenization Engie, for permission publish this work. Special thanks to Novi
temperatures indicate fracture initiation at temperatures of 300–340 °C Ganefianto for his encouragement to share basic scientific findings that
in the deep SW reservoir to 260 °C in the shallow NE reservoir. The deep might stimulate additional research into the geology of Sumatra. Irene
SW reservoir has remained at high temperature (270–310 °C), whereas Wallis and Anna Colvin helped unravel the well-by-well stratigraphic
the NE reservoir has experienced more cooling (currently 232–240 °C). patterns. We also appreciate Anne Sturm’s assistance with the zircon
Inflow of steam-heated water (222 °C), recorded by Th of inclusions in separation. Finally the original manuscript benefitted from insightful
anhydrite ± calcite, is observed in some NE wells, with subsequent reviews from Erik Layman and an anonymous reviewer.
sealing and reheating. Tmicevalues indicate apparent salinities in both
the NE and SW reservoir areas were higher than the modern reservoir

118
J. Stimac, et al. Geothermics 82 (2019) 97–120

Fig. 20. Sheared veins in diorite in H4 3101.67 m containing late-stage growth of epidote and amphibole in open space (plane polarized light at top, crossed nicols at
bottom).

References

Acocella, V., Bellier, O., Sandri, L., Sebrier, M., Pramumijoyo, S., 2018. Weak tectono-
magmatic relationships along an obliquely convergent plate boundary: Sumatra,
Indonesia. Front. Earth Sci. 6, 3. https://doi.org/10.3389/feart.2018.00003.
Barber, A.J., Crow, M.J., 2005. Chapter 4: pre-tertiary stratigraphy. Sumatra: Geology,
Resources and Tectonic Evolution. pp. 24–53 Geol. Soc. Memoir No. 31.
Barber, A.J., Crow, M.J., de Smet, M.E.M., 2005a. Chapter 14: tectonic evolution.
Sumatra: Geology, Resources and Tectonic Evolution. pp. 234–258 Geol. Soc. Memoir
No. 31.
Barber, A.J., Crow, M.J., Milsom, J.S., 2005b. Chapter 1 – introduction and previous
research. Sumatra: Geology, Resources and Tectonic Evolution. The Geological
Society, pp. 1–15 Geol. Soc. Memoir No. 31.
Baroek, M., Stimac, J., Sihotang, A.M., Putra, A.P., Martikno, R., 2018. Formation and
fracture characterization of the Muara Laboh geothermal system, Sumatera,
Indonesia. Proceedings Geothermal Resource Council. pp. 42.
Bellon, H., Maury, R.C., Sutanto, Soeria-Atmadia, R., Cotton, J., Polve, M., 2004. 65 m.y.-
long magmatic activity in Sumatra (Indonesia) from Palaeocene to Recent. Bull. Soc.
Geol. France 175, 61–72.
Belousov, A., Voight, B., Belousova, M., 2007. Directed blasts and blast-generated pyr-
oclastic density currents: a comparison of the Bezymianny 1956, Mount St Helens
1980, and Soufriére Hills, Montserrat 1997 eruptions and deposits. Bull. Volcanol. 69,
701–740.
Bogie, I., Kusumah, Y.I., Wisnandary, M.C., 2008. Overview of the Wayang Windu geo-
thermal field, West Java, Indonesia. Geothermics 37, 347–365.
Fig. 21. Summary plot of fluid inclusion populations for which both Th Browne, P.R.L., 1970. Hydrothermal alteration as an aid in investigating geothermal
(homogenization temperature) and Tm ice (freezing point depression) mea- fields. Geothermics 2, 564–570.
Browne, P.R.L., 1978a. Hydrothermal Alteration as an Aid in Investigating Geothermal
surements were obtained from the Muara Laboh core samples. Symbols are
Fields. Geothermics (U.N. Symposium on the Development and Utilization of
placed at average values of Th and Tm ice, with error bars showing the measured Geothermal Resources Vol. 2, 564–570 Part 1.
ranges. The symbols are color coded to the well/core sample (see legend at Browne, P.R.L., 1978b. Hydrothermal alteration in active geothermal fields. Annu. Rev.
bottom), and the shape of the symbol represents the mineral that contains the Earth Planet. Sci. 6, 229–248.
fluid inclusions (see legend at upper left). A color-coded X is used to denote Busby, C.J., Bassett, K.N., 2007. Volcanic facies architecture of an intra-arc strike-slip
basin, Santa Rita mountains, Southern Arizona. Bull. Volcanol. 70, 85–103.
modern measured temperatures and an approximation of salinities based on
Cobbing, 2005. Sumatra: Geology, Resources and Tectonic Evolution. Plutonic Episodes
TDS data from well samples. of Sumatra, Geol. Soc. Memoir No. 31. pp. 554–562.
Colvin, A., Mussofan, W., 2017. Follow-up Surface Geology Near Well Pads and Civil
Works. Unpublished Supreme Report, 9p,. .
Appendix A. Supplementary data Crow, M.J., 2005. Chapter 8: tertiary volcanicity. Sumatra: Geology, Resources and
Tectonic Evolution. pp. 98–119 Geol. Soc. Memoir No. 31.
Supplementary material related to this article can be found, in the Darman, H., Sidi, F.H., 2000. An outline of the geology of Indonesia. Indones. Assoc. Geol.
Jakarta 192 pp.
online version, at doi:https://doi.org/10.1016/j.geothermics.2019.06. De Smet, M.E.M., Barber, A.J., 2005. Chapter 7 - tertiary stratigraphy. Sumatra: Geology,
002. Resources and Tectonic Evolution. pp. 86–97 Geol. Soc. Memoir No. 31.
Drestanta, Y.S., Soeda, Y., Drakos, P., Astra, D., Lima Lobato, E.M., 2018. Building a 3D
earth model of silangkitang geothermal Field, North Sumatra, Indonesia.
Proceedings, 6Th International Geothermal Convention and Exhibition.

119
J. Stimac, et al. Geothermics 82 (2019) 97–120

Dyaksa, D.A., Ramadhan, I., Ganefianto, N., 2016. Magnetotelluric reliability for ex- on Geothermal Reservoir Engineering.
ploration drilling stage: study cases in Muara Laboh and rantau dedap geothermal Reyes, A.G., Petrology of Philippine geothermal systems and the application of their al-
project, Sumatera, Indonesia. Proceedings, 41st Workshop on Geothermal Reservoir teration mineralogy to their assessment. Bull. Volcanol. Res. 43, 1990, 279-349.
Engineering SGP-TR-209. Roedder, E., 1984. Fluid inclusions, reviews in mineralogy. Mineral. Soc. Am. 12,
Hedenquist, J.W., Henley, R.W., 1985. The importance of CO2 on freezing point mea- 644.
surements of fluid inclusions; evidence from active geothermal systems and im- Rosidi,, H.M.D., Tjokrosapoetro, S., Pendowo, B., Gafoer, S., Suharsono, 1996. The
plications for epithermal ore deposition. Econ. Geol. 80, 1379–1406. Geology of the Painan and Northeastern Part of the Muarasiberut, Quadrangles
Hedenquist, J.W., Reyes, A.G., Simmons, S.F., Taguchi, S., 1992. The thermal and geo- (0814-0714), Systematic Geological Map of Indonesia Scale 1:250000. Geological
chemical structure of geothermal and epithermal systems: a framework for inter- Research and Development Centre, Bandung.
preting fluid inclusion data. Eur. J. Mineral. 989–1016. Sasada, M., Roedder, E., Belkin, H.E., 1986. Fluid inclusion from drill hole DW-5, Hohi
Hickman, R.G., Dobson, P.F., van Gerven, M., Sagala, B., Gunderson, R.P., 2004. Tectonic geothermal area, Japan: evidence of boiling and procedure for estimating CO2 con-
and stratigraphic evolution of the Sarulla Graben region, North Sumatra, Indonesia. tent. J. Volcanol. Geotherm. Res. 30, 231–251.
J. Asian Earth Sci. 23, 435–448. Schmitt, A.K., Grove, M., Harrison, T.M., Lovera, O., Hulen, J., Walters, M., 2003. The
Hochstein, M.P., Sudarman, S., 1993. Geothermal resources of Sumatra. Geothermics 22, Geysers-Cobb Mountain Magma System, California (part 1): U-Pb zircon ages of
181–200. volcanic rocks, conditions of zircon crystallization and magma residence times.
Jones, C.G., Moore, J.N., 2018. Fluid Inclusion Microthermometric Analysis of Muara Geochim. Cosmochim. Acta 67 (18), 3423–3442.
Laboh Core Samples from Geothermal Wells ML-A1, ML-A2, ML-A3 and ML-H4 Stacey, J.T., Kramers, I., 1975. Approximation of terrestrial lead isotope evolution by a
Unpublished EGI Report to Supreme Energy dated Sept. 2018, 38p. two-stage model. Earth Planet. Sci. Lett. 26 (2), 207–221.
Milicich, S.D., Wilson, C.J.N., Bignall, G., Pezaro, B., Charlier, B.L.A., Wooden, J.L., Stimac, J., 2019. Update of Muara Laboh Stratigraphy. Unpublished Supreme Energy
Ireland, T.R., 2013. U-Pb dating of zircon in hydrothermally altered rocks of the Report.
Kawerau Geothermal Field, Taupo Volcanic Zone, New Zealand. J. Volcanol. Stimac, J., Mussofan, W., 2017. Update on Surface Stratigraphy and Possible Structures at
Geotherm. Res. 253, 97–113. Muara Laboh. Unpublished Supreme Report. 20 p.. .
Milicich, S.D., Chambefort, I., Wilson, C.J.N., Charlier, B.L.A., Tepley, F.J., 2018. The Stimac, J., Ganefianto, N., Baroek, M., Sihotang, A.M., Ramadhan, I., Mussofan, W., Sidik,
hydrothermal evolution of the Kawerau geothermal system. New Zealand. J. R., Alfiady, Dyaksa, D.A., Azis, H., Putra, A.P., Martikno, R., Irsamukhti, R., Santana,
Volcanol. Geotherm. Res. 353, 114–131. S., Matsuda, K., Hatanaka, H., Soeda, Y., Cariou, L., Egermann, P., 2019. An Overview
Moore, D.E., Hickman, S., Lockner, D.A., Dobson, P.F., 2001. Hydrothermal minerals and of the Muara Laboh Geothermal System, Sumatra. Accepted by Geothermics, xx,
microstructures in the Silangkitang geothermal field along the Great Sumatran fault xxx-xxx. .
zone, Sumatra, Indonesia. Geol. Soc. Am. Bull. 113, 1179–1192. Stimac, J., Goff, F., Janik, C., 2015. Encyclopedia of volcanoes, chapter 46. Intrusion
Moore, J.N., Allis, R., Renner, J.L., Mildenhall, D., McCulloch, J., 2002. Petrologic evi- Related Geothermal Systems, 2nd ed. Elsevier, pp. 779–822.
dence for boiling to dryness in the Karaha-Telaga Bodas Geothermal System, Sieh, K., Natawidjaja, D., 2000. Neotectonics of the Sumatran fault, Indonesia. J.
Indonesia. Proceedings 27th Workshop on Geothermal Energy. Geophys. Res. 105, 28295–28326.
Moore, J.N., Powell, T.S., Heizler, M.T., Norman, D.I., 2000. Mineralization and hydro- Ui, T., Takarada, S., Mitsuhiro, Y., 2000. Debris Avalanches. Encyclopedia of Volcanoes.
thermal history of the tiwi geothermal system, Philippines. Econ. Geol. 95, Elsevier, pp. 617–626 1417pp.
1001–1023. van Wyk de Vries, B., Davies, T., 2015. Landslides, debris avalanches, and volcanic
Muraoka, H., Takahashi, M., Sundhoro, H., Dwipa, S., Soeda, Y., Momita, M., Shimada, gravitational deformation. Chapter 38. Encyclopedia of Volcanoes. pp. 665–685.
K., 2010. Geothermal systems constrained by the sumatran fault and its pull-apart Wilkinson, J.J., 2001. Fluid inclusions in hydrothermal ore deposits. Lithos 55, 229–272.
basins in Sumatra, Western Indonesia. Proceedings World Geothermal Congress. Wilson, C.J.N., Charlier, B.L.A., Fagan, C.J., Spinks, K.D., Gravley, D.M., Simmons, S.F.,
Mussofan, W., Baroek, M.C., Stimac, J., Sidik, R.P., Ramadhan, I., Santana, S., 2018. Browne, P.R.L., 2008. U–Pb dating of zircon in hydrothermally altered rocks as a
Geothermal resource exploration alongGreat Sumatera fault segments in Muara correlation tool: application to the Mangakino geothermal field, New Zealand. J.
Laboh: perspectives from geology and structural play. Proceedings, 43rd Workshop Volcanol. Geotherm. Res. 176, 191–198.

120

You might also like