Professional Documents
Culture Documents
Manuscript Draft
Title: Overview of the Wayang Windu Geothermal Field, West Java, Indonesia
Section/Category:
Keywords: Geothermal reservoirs; Vapor-dominated; Liquid dominated; Wayang Windu; Java; Indonesia
Abstract: The Wayang Windu geothermal field, West Java, Indonesia, is interpreted to be transitional
between vapor-dominated and liquid-dominated conditions with four coalesced fluid upwelling centers that
generally become younger and more liquid-dominated towards the south. Two of these centers are
associated with the large Gunung Malabar andesite stratovolcano and the other two with the smaller aligned
Gunung Wayang and Gunung Windu andesitic volcanoes to the south. The overall potential resource area
is of the order of 40 km2. Deep wells encounter a deep liquid reservoir whose top, which ranges from 0 to
400 m above sea level (m asl), becomes progressively deeper toward the south. As pre-exploitation
pressure versus elevation conditions were the same throughout the deep liquid-dominated reservoir it is
likely to be continuous across the field.
The deep liquid reservoir is overlain by three separate vapor-dominated reservoirs. The northernmost is the
largest as it is coalesced over two separate fluid upwelling centers. Its low gas content, size, prolonged
productivity and constant pressure at a given elevation preclude it from being a parasitic steam zone.
Mineralogical relationships demonstrate that the northern vapor zone was originally liquid-dominated with its
water level at an elevation as high as 1700 m asl. Subsequent boil off may reflect low recharge rates due to
hydrological isolation at depth. To the south, the vapor-dominated reservoirs decrease in thickness and are
characterized by progressively higher pressures, temperatures and gas contents, suggesting that the
southernmost vapor-dominated zone is the youngest and that these reservoirs become increasing older
toward the north.
Cover Letter
Dear Sue,
Mailing address:
Ian Bogie
Sinclair, Knight Merz Ltd
PO Box 9806, Newmarket
Auckland, New Zealand
Note that I have already sent an e-mail message and a letter to Ian
Bogie informing him of the acceptance.
Please let me know if you have any questions about the manuscript or
the attached files.
Best regards,
Marcelo
****************************************************
Marcelo J. Lippmann, Ph.D.
Editor, Geothermics Tel: 1-510-486-5035
Earth Sciences Division 90R1116 Fax: 1-510-486-5686
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8126 USA e-mail: mjlippmann@lbl.gov
* Manuscript
GEOT-D-08-Bogie
1. Sinclair, Knight Merz Ltd, PO Box 9806, Newmarket, Auckland, New Zealand.
2. Mandala Nusantara Ltd., Wisma Mulia 50th Floor, Jl. Jend. Gatot Subroto no. 42,
Jakarta 12710, Indonesia.
Abstract
coalesced fluid upwelling centers that generally become younger and more liquid-
dominated towards the south. Two of these centers are associated with the large Gunung
Malabar andesite stratovolcano and the other two with the smaller aligned Gunung
Wayang and Gunung Windu andesitic volcanoes to the south. The overall potential
resource area is of the order of 40 km2. Deep wells encounter a deep liquid reservoir
whose top, which ranges from 0 to 400 m above sea level (m asl), becomes progressively
deeper toward the south. As pre-exploitation pressure versus elevation conditions were
*
Corresponding author: Tel.: +64 9 913 8900; fax: +64 9 913 8901
E-mail address: ibogie@skm.co.nz (I. Bogie)
1
The deep liquid reservoir is overlain by three separate vapor-dominated reservoirs. The
northernmost is the largest as it is coalesced over two separate fluid upwelling centers. Its
low gas content, size, prolonged productivity and constant pressure at a given elevation
that the northern vapor zone was originally liquid-dominated with its water level at an
elevation as high as 1700 m asl. Subsequent boil off may reflect low recharge rates due
gas contents, suggesting that the southernmost vapor-dominated zone is the youngest and
Java; Indonesia
the provincial capital of West Java, Indonesia (Fig. 1). It is one of a cluster of
geothermal fields around Bandung that also includes Darajat (Hadi et al., 2003),
Kamojang (Utami, 2000), Karaha-Telaga Bodas (Moore et al., 2002, 2004), Papandayan
(Wibowo, 2006), Patuha (Layman and Soemarinda, 2003), Tampomas (Wibowo, 2006),
2
and Tangkuban Perahu (Wibowo, 2006). These fields lie within andesitic, volcanic
highlands formed by a concentration of volcanic centers in this part of the Sunda Arc.
The city of Bandung is located in a basin (Dam, 1994) near the center of the volcanic
highlands. That basin does not appear to be a back-arc basin, as it has arc volcanics on
either side, but may owe its origin to flexure from varying rates of subduction roll back
along the Sunda Arc. This arc has formed in response to the subduction of the Australian-
Indian Plate beneath the Eurasian Plate. It has been active since the Cretaceous
(Whittaker et al., 2007), but has undergone changes as increasing amounts of Australian
continental crust have become involved in the collision and it is undergoing roll back.
The dominant strike directions of the major faults in West Java are 40 and 340
(Wibowo, 2006; Fig.1), forming a conjugate pair of strike-slip faults, consistent with
2. Geothermal exploration
gravity and soil geochemistry surveys, as well as the drilling of temperature gradient
holes.
The first deep hole (then called WWD-1, now WWA-1ST, after being sidetracked to the
west, Fig. 2) was drilled by Pertamina to the west of the saddle between Gunung
3
Wayang1 and G. Windu in 1991 (Budiardjo, 1992; Ganda and Hantono, 1992; Ganda et
al., 1992). Well data showed that a perched steam-heated groundwater aquifer overlies a
reservoir. This was the discovery well for the Wayang Windu field and for transitional
liquid-vapor type geothermal systems. A 600-m deep slim hole (MSH-1) drilled by
The most intense surficial hydrothermal activity occurs adjacent to the small G. Wayang
and G. Windu volcanic centers (Fig. 3). Fumaroles, steaming and altered ground, and
acid-sulfate springs occur in the Wayang thermal area, which lies within a sector
collapse, with the current peak of G. Wayang representing an eastern remnant of a much
larger volcanic center, originally located to the west along the axis of alignment of other
small volcanic centers. The matrix of the debris flow from the sector collapse consists of
hydrothermal clay, suggesting slope failure related to the weakening of the volcanic
deposits by hydrothermal alteration and the increase of pore pressures with heating (Reid,
2004). A radiocarbon date of 7450 110 years was obtained from peat sampled from a
small swamp that had developed on top of the debris flow deposits at drill pad WWD.
This would represent a minimum age for the sector collapse itself.
1
The abbreviation G. will be used for Gunung (mountain) from here on.
4
Smaller areas of altered ground with acid-sulfate springs and weak fumarolic activity are
springs are present south of G. Malabar and to the south, west and east of the smaller
volcanic centers (Fig. 3). The springs have temperatures ranging from 25 to 66C and
are notable for their lack of Cl (Sudarman et al., 1986). The Cibolang spring in the south
has an elevated B content (16 ppm). As its other constituents indicate that it discharges
The northernmost area of hydrothermal alteration is exposed on the southwest rim of the
Malabar Caldera complex and there is strong alteration in the cirque of the Wayang
sector collapse. Small patches of altered ground are scattered around the area, although
the overall extent of the hydrothermal alteration is only apparent where deep cuts have
been made for roads and drill pads as much of the hydrothermal alteration is covered by a
sequence of young fresh ash beds. A paleosol between the overlying unaltered ash beds
and the underlying hydrothermally altered deposits in the vicinity of the WWQ pad has
yielded a radiocarbon age of 8700 90 years BP. This represents a minimum age for the
underlying alteration. Thus, only the alteration found in the immediate vicinity of
The northernmost known thermal activity in the Wayang Windu field occurs at a break in
slope southeast of Puncak Besar (Figs. 2 and 3). There are no known thermal
manifestations on the highest parts and northern slopes of G. Malabar despite the
5
evidence from the MT surveys for the field to extend beneath these areas. As the
prevailing weather during the rainy season is from the north, a rain curtain may be
3. Field development
The Wayang Windu field was developed by MNL (Magma Nusantara Ltd.) beginning in
1996 as a fast track development that started in the logistically easier areas in the south
and east with a combination of 1500-m deep slim holes (WWC-SH, WWJ-SH, WWL-SH
and WWR-SH; Nurruhliati, 1996; Thaysa, 2003) and deeper production drilling. The
existence of a large thermal anomaly was established, but initially productive wells were
restricted to sites immediately beneath and southwest of the young volcanic centers.
reservoir than the one found in the first well (WWA-1). These new wells confirmed the
results from slim hole MSH-1 in the north and indicated a vapor-dominated zone overlies
a deep liquid-dominated reservoir. Thus, a combination of deep and shallow wells was
drilled; success was greatest with the shallow wells. Three dry-steam wells with depths of
up to 1700 m, each produced steam equivalent to over 20 MWe 2. These wells were
drilled on the northernmost (at the time) MBD pad. In terms of resource potential, an
initial 220 MWe (gross) development was planned with a possible extension to 440 MWe
(gross), until the Asian financial crisis of 1997 intervened and the project was scaled back
2
All MWe values given are in terms of the power conversion capacity of the current Wayang Windu
power plant (i.e. 1.94 kg/s of steam per MWe).
6
The initial 110 MWe (gross) development obtained its main steam supply from the
northern two-phase reservoir, with some deep northern production and a combination of
shallow and deep production from wells further to the south, on the WWA pad upon
which the Pertamina discovery well was located. Two-phase fluid transmission pipelines
from these drill pads feed a central separator station with steam passing through a
scrubber before entering the dual inlet, 110 MWe Fuji turbine in the power plant. This
unit, which was installed in 1999, is one of the worlds largest operating geothermal
turbines (Murakami et al., 2000). As a result of this turbine installation, Wayang Windu
holds the distinction of being the most rapidly developed geothermal field of its size.
Both condensate and separated brine are reinjected by gravity in the southernmost part of
Unocal Indonesia became a 50% shareholder of the Wayang Windu project in 2001. The
poorly productive deep MBD-1 vertical well was sidetracked to intersect the shallow,
vapor-dominated zone and came to be the largest producer in the field at the time. That
was followed by workovers of other wells on the MBD drill pad, which reduced
production due to the installation of tiebacks (whereby the production casing is extended
to the surface reducing the internal diameter of the cased section of the well). WWQ-3,
another deep well, was also sidetracked, but with less success than MBD-1. Reservoir
pressure drawdown has reduced fluid production with time, but it has now stabilized in
wells tapping the vapor-dominated zones, indicating that they are major, sustainable
7
The Wayang Windu field was acquired by Star Energy Holdings Pty Ltd in 2004. A new
drilling program began in August 2006 to supply steam for a second 110 MWe Fuji
turbine at the existing power plant. The first well completed under this program (MBD-
5) produces the steam equivalent of 40 MWe making it at the time it was drilled the
largest dry steam well in the world. The whole eight-well program realized a total of 180
MWe and included make up wells for the existing turbine. The northern extension of the
field is currently being explored with the ultimate goal of obtaining steam to generate 440
MWe (gross), which on the results of reservoir modeling studies (Asrizal et al., 2006) is
eminently achievable.
Only the southern part of the western boundary of the Wayang Windu geothermal field is
well defined by drilling (Figs. 2 and 3). Anderson et al. (1999, 2000) found a good
relationship between the contoured elevation of the base of the conductive layer (i.e. the
surveys and the temperature distribution in the drilled portion of the field. Thus, the
continuation of elevated portions of the conductor outside the drilled area can be
Southwest of G. Windu, productive wells have been drilled on the basis of earlier
8
MT data indicated a deep base of the conductor with no doming or ridging. Since there
are indications that the southern area is the youngest part of the system (see below), the
position of the conductor is likely to have been dictated by earlier geothermal activity
when the deep liquid reservoir reached higher elevations. The conductor in the south now
appears to be too impermeable for the alteration mineralogy to re-equilibrate and allow
formation of a dome or ridge in its base, although the resistivity in the area of productive
wells southwest of G. Windu is slightly higher than where the non-productive WWE-1
Combining new well data with those of recent MT surveys, which have a greater station
density and have yielded better quality information, the bulk of the field is interpreted to
lie beneath G. Malabar, the andesite stratovolcano now centred at Puncak Besar (Fig. 3).
small volcanic centers, where it narrows down to approximately 2 km across. This gives
an overall potential resource area of approximately 40 km2. However, since core hole
drilling outside of the resource area encountered elevated temperatures at depth, the
The field owes its size to the presence of more than one fluid upwelling center, as will be
discussed below. This feature is found in other Javan geothermal fields. Layman et al.
(2002) identified three centers at Dieng, multiple centers (referred to as cells) were
9
suggested for Karaha-Telega Bodas (Nemok et al., 2007) and two geothermal centers
The stratigraphy of Wayang Windu has been discussed by Bogie and MacKenzie (1998)
who applied volcanic facies models to subdivide the various volcanic units at depth,
which define a series of overlapping andesitic piles. Their cross section has been
extended in Fig. 4; its the trace is shown in Fig. 2. Microdiorite, dolerite and diorite
porphyry dykes are found, but blind drilling and very limited coring have prevented the
clear recognition of any major intrusives. Andesitic lavas, pyroclastic and epiclastic
deposits predominate in the volcanic units with dacite only occurring at G. Gambung.
Quartz found in rocks of the other smaller volcanic centers is xenocrystic, and
Ash deposits of regional extent occur throughout these volcanic piles. Similar beds are
also found at Patuha (Layman and Soemarindo, 2003) and Awibengkok (Hulen and
Anderson, 1998), where they are referred to as paleosols. At Wayang Windu there are 14
different beds with thicknesses varying between 5 and 30 m that can be correlated
between wells, and numerous thinner ones with solitary occurrences. These beds have a
complex distribution suggesting draping over the existing topography with erosion in
steeper areas and ponding when deposited in valleys. Scanning electron microscopy and
XRD analyses indicates that they originally consisted of very fine-grained glass shards
and titanomagnetite, but the glass has altered to calcium smectite and in places to
10
interlayered smectite-illite. As these clays are usually found at temperatures below
200C (Anderson et al., 2000), we suggest the beds must have very low permeability,
since measured temperatures at the corresponding depths (> 300C in some instances) are
Gunung Malabar sits on the boundary fault of the Bandung Basin (Figs. 1 and 3; Dam,
1994). There is a multiphase summit caldera complex on G. Malabar. Rocks from the
volcano summit (east of the calderas), Puncak Besar (a prominent peak south of the
caldera complex directly above the Bandung Basin boundary fault) and G. Gambung (a
parasitic dacite dome to the southeast), all have K-Ar dates of 0.23 Ma and have bulk and
Gunung Bedil, the next volcanic center to the south, was dated at 0.19 Ma, and G. Windu,
the southernmost volcanic center at 0.10 Ma. A 0.49 Ma date for G. Wayang breaks the
trend of having younger centers towards the south. While the other young volcanic
centers are well preserved and samples for dating were obtained from their youngest
parts, G. Wayang has undergone sector collapse, and it is likely that the dated sample
from that eruptive center was taken from a much older part of the volcanic pile.
The geochemistry of the younger volcanic centers is variable, although G. Windu has
some trace element similarities to G. Malabar. As these rocks contain quartz xenocrysts
and diorite xenoliths, the chemical variation may be reflecting the degree of crustal
more extensive, but less intensive geochemical study of samples of older units collected
11
from the wells concluded that they are geochemically similar to the Malabar rocks
Structurally the field conforms best to regional patterns in the south, with faulting
exhibiting steep dips (> 80) and strikes of 30-40 and 330340. In the north, along the
southern boundary of the Bandung Basin, further deformation results from movement
along the boundary fault. Gunung Malabar is actively subsiding into the basin and is
deforming the basin fill, as can be seen by the presence of upthrust Tertiary sediments
The rocks penetrated by the wells have highly localized structural permeability; with the
most permeable geologic structures following the regional 40 strike. As these structures
have trends similar to regional faults, then it is likely they are strike-slip faults, which
would tend to have lower permeabilities than normal faults because of shearing and
consequent rock comminution (i.e. trituration). Structures that were reactivated as the
volcanic sequence was deposited will display less shearing at their upper ends than their
extensions into the basement. Thus, the same faults may have lower permeabilities at
depth due to prolonged shearing and comminution compared to the younger overlying
Bandyopadhay et al. (2006), however, calculated the direction of least principal stress in
the Wayang Windu area as far north as the WWQ pad, utilizing borehole breakouts (the
tendency for drill hole cross sections to become elongated in relationship to the local
12
stress field) as determined from the caliper measurements obtained from micro-resistivity
formation imaging logging. They found that the calculated stress field did not correspond
to the regional orientation, but had the least principal stress striking at 310, with an
overall normal faulting regime. The NE-striking faults are thus likely to have been
regional strike-slip faults reactivated as more permeable normal faults due to a change
extension may be even stronger at the boundary and inside the Bandung Basin.
deposits with more structurally limited alteration zones in the lava flows. Shallow
alteration (above and locally within the conductor), is marked by the presence of
kaolinite, alunite, natroalunite, and rare native sulfur. This alteration is associated with
base of the conductor. Pressure profiles from slim holes drilled to depths of up to 1500 m
that did not penetrate into the deep reservoir (Fig. 5) indicate perched aquifers that are
similar to those of perched steam-heated aquifers, it is likely that they feed the springs.
Possible condensate aquifers, not connected to the regional groundwaters, are indicated
tool; these aquifers are characterized by strong alteration to kaolinite, calcite, quartz and
13
by smectite along with near ubiquitous quartz, chlorite, calcite and pyrite with zeolites,
Kaolinite, calcite, anhydrite and quartz are found within parts of the conductor occurring
as an overprint.
With increasing depth, interlayered illite-smectite rather than smectite is found, until
there is a transition to a propylitic assemblage with its top marked by the presence of
corrensite and epidote. At greater depths, illite becomes the main sheet silicate.
assemblage are encountered still deeper. The formation of secondary amphibole appears
oligoclase and magnetite has been observed in well WWA-4. In the deep liquid
reservoir, wairakite and prehnite, along with epidote, are common as alteration and vein
Other than a generally prograde transition, with illite-smectite and corrensite detected
shallow in the vapor-dominated reservoirs, and prehnite in the lower parts of the deep
liquid reservoir, there is no clear difference in the hydrothermal alteration of the liquid-
reservoir, as has also been noted at Karaha-Telega Bodas (Moore et al., 2002).
Rare occurrences of advanced argillic alteration are found below the caprock of both the
14
pyrophyllite, diaspore, woodhouseite and dickite, with accompanying quartz, anhydrite
and pyrite. As these hydrothermal minerals are created under high-temperature, acid
conditions, and considering that the reservoir pH is now near neutral, and has
temperatures below those at which these minerals formed (Reyes et al., 1993), we
consider this deep advanced argillic alteration to be relict. This may possibly reflect the
In the northern part of the Wayang Windu geothermal field, hydrothermal epidote is
found at elevations up to 1330 m asl. The shallowest appearance of this mineral is above
the vapor-dominated reservoir, and all its first occurrences in well samples are above the
deep liquid reservoir (Figs. 4 and 6). Since generally in this type of geothermal reservoir
epidote forms at temperatures above 240C under near neutral pH conditions (Browne,
1978), we infer that the water level was previously higher. If the geothermal system is
related to recent volcanism, then it is no older than 0.23 Ma. In that case, the lowering of
the water level could not have been caused by tectonic uplift, which occurred much
The epidote occurs 400 m below the top of the smectite-bearing argillically altered rocks,
which marks the top of the conductor. If it is assumed that the temperatures followed
boiling point-to-depth conditions, the original water level would have been at an
elevation of 1730 m asl. To the south beneath the younger volcanic centers, the top of
15
the conductor is at 1400 m asl, and there is a concurrent deepening to the first appearance
of epidote (Fig. 4). Thus, the top of the conductor to the south can be considered to
The top of the epidote zone is very close to the tops of the Waringin volcanic unit (Fig. 4)
the northern part of the field, the Malabar volcanic unit consists mainly of lavas whose
average porosity is ~1% (Asrizal et al., 2006). The underlying Waringin unit consists
mainly of lapilli tuffs with an average porosity of ~8% (Asrizal et al., 2006). It is
possible that the original porosity of the pyroclastic deposits was higher prior to
alteration, but this would not be the case for the lava flows. Therefore, the thick
sequence of lava flows could have acted as the initial caprock of the geothermal system.
The permeability associated with vertical faults that cut these flows may have been
restricted by the presence of the regional ash deposits, which because of their high clay
content would tend to deform plastically rather than act as hosts for vertical conduits.
In the south, the Waringin volcanic unit contains a higher proportion of lava flows in its
upper parts. The overlying Pangalengan unit, which contains many ash deposits and fine-
grained sediments, could have formed the original cap in this part the field, and thus
might have influenced the distribution of epidote. However, a plot of the depth of the
first appearance of epidote has a shape similar to that of the 300C isotherm, suggesting
instead that the initial temperature distribution may outweigh the influence of porosity
16
In some places the original conductor has been overprinted by hydrothermal alteration
caused by perched steam-heated aquifers that, where topography allows, extend to higher
elevations and above topographic highs in the conductive layer. However, these aquifers
have higher resistivities (~ 5 m) than the main conductor (~ 2 m), due to the presence
of kaolinite, which is more resistive than smectite, as the predominant clay mineral.
The areas of the conductor that have slightly higher resistivity more clearly define the
four fluid upwelling centers at the Wayang Windu geothermal field (i.e. where productive
wells have been drilled; Fig. 3) than the elevation of the base of the conductive layer.
Presumably this is because the higher resistivity areas in the conductor reflect near
present conditions, whereas its base was defined during an early stage in the development
of the system. Similar higher resistivities in the conductor have been reported at Karaha-
Telaga Bodas (Raharjo et al., 2002), and may generally serve, in combination with the
geometry of the base of this conductive layer, as a pre-drilling indicator of the potential
North of the wells drilled at Wayang Windu the main conductor is found at elevations up
to 1800 m asl. In this part of the system it is possible that the water level may have
reached these highs during the history of the field. Moreover, the lower conductivities
indicative of perched steam-heated aquifers are found at even higher elevations. Since
the current deep liquid level is at much greater depth (~400 m asl in the north) and the
alteration mineralogy is that of a near pH-neutral fluid, most of the alteration in and
17
above the vapor-dominated zones is now above the water level and must be relict. That
is, the alteration, which includes the electrically conductive argillic zone, must have
formed early in the development of the system. This zone, which is characterized by a
caprock of the geothermal reservoir. On the margins of the field, where the base of the
conductor deepens, these clay-rich altered rocks form the lateral hydrological boundaries
of the vapor-dominated resource, at least in its upper parts. This may explain why this
steam zone is best developed in the north where there is the steepest drop off in the depth
of the margin of the conductor, forming a dome which encloses the vapor-dominated
reservoir.
above the deep water is only weakly developed. Epidote is partially replaced by calcite,
white clay (possibly kaolinite), pyrite and anhydrite, a further indication of its relict
nature. The pervasive calcite veining present in the rocks may have formed as the system
boiled off; but platy textures typical of boiling (Browne, 1978) are not commonly
condensate. Rarely, platy calcite and chalcedony occur in epidote-bearing rocks. The
suggestive of more intense boiling, which could result from localized pressure drops
(Moore et al., 2002, 2004), where its presence is interpreted as evidence of rapid boil off.
18
Fluid inclusion work has been limited by the amount of suitable sample material.
Abrenica (2007) reports homogenization temperatures between 228 and 255C, with the
mode at 235C, in primary fluid inclusions in quartz from a vein at 590 m asl in well
MBD-5; the current estimated temperature is 246C (the downhole logging tool did not
reach this depth). Melting point measurements for these inclusions indicate salinities of
0.53 to 1.05 wt% NaCl equivalent (Abrenica, 2007), reflecting both the dissolved salt and
gas contents of the trapped fluids. The present-day salinities, calculated on a gas-free
basis, of the deep liquid reservoir is ~2 wt% NaCl equivalent. Therefore, these inclusion
Secondary fluid inclusions in that quartz sample are vapor-rich and have higher
inclusions range from 241 to 334C, but since it is unlikely that a single-phase fluid was
trapped, they may not reflect temperatures prevailing at the time of their formation.
Primary inclusions in calcite from a sample collected two meters below the MBD-5
quartz sample (i.e., at 592 m asl) gave higher homogenization temperatures in the 264-
295C range (Abrenica, 2007), with the mode at 275C. Like the secondary inclusions in
quartz, these calcite-hosted inclusions are vapor-rich and have increasing vapor/liquid
trapping temperatures.
19
Epidote from 542 m asl in MBA-1 was observed to contain liquid-dominated fluid
obtained. As vapor-dominated conditions now prevail at that depth, it is inferred that the
liquid level was previously higher and that the epidote is relict.
The Wayang Windu geothermal resource has a deep, hot, neutral pH, liquid reservoir
(Fig. 6). Throughout the field within the deep liquid reservoir, pressures and temperatures
versus elevation are similar (Table 2; Fig. 5). In the north its top is at 400 m asl, and it
deepens towards the south where it is found near sea level elevation. Since under natural-
state (pre-exploitation) conditions the reservoir had almost the same vertical pressure
near pressure equilibrium there was little fluid flow. It is under-pressured with respect to
groundwater hydrostatic pressure and there is geochemical evidence for only limited
The deep liquid reservoir is possibly recharged from regions west of G. Bedil and G.
Wayang and southwest of G. Windu. In these areas, the decrease in the elevation of the
base of the conductive layer is less steep, and thus the early argillic alteration that may
more limited. In the recharge areas the shallowest well feed zones from the deep liquid
20
reservoir are cooler than elsewhere in that reservoir, but without actual temperature
inversions deeper in the well, indicating some limited ingress of meteoric waters into the
shallow zones.
reservoirs. The largest in the north appears to be coalesced over two fluid upwelling
centers (associated with Puncak Besar and G. Gambung), while the two farther south
(associated with G. Wayang and G. Windu) appear to be separate. In other words, the
that are located over four fluid upwelling centers. The characteristics of the vapor-
dominated zones change progressively towards the south; i.e. their pressures,
temperatures and gas contents increase, their thicknesses decrease, and are found at
greater depths.
However, if they were all hydraulically connected, the drop in fluid pressure for any
given elevation towards the north (i.e., the oldest part of the system if the ages of the
spatially related volcanic centers are also temporally related to the fluid upwelling
centers) may point to a northerly flow from the inferred youngest part of the system in the
south. The fact that the field can be divided up geochemically into a minimum of three
areas (Table 2), and that high temperatures are maintained at depth and are actually
highest beneath G. Wayang rather than in the Windu area, argue against a single heat
source in the south. Separation of these three vapor-dominated reservoirs may reflect the
21
distribution of impermeable, regional ash deposits that restrict vertical permeability (and
fluid flow), and the distribution of the more porous pyroclastic-rich deposits that host the
two-phase zones.
Interpretation of the MT surveys indicate that the Wayang Windu geothermal field may
extend approximately 7 km to the north of MBE-2 and that this well is located south of
two domed structures in the base of the conductor (Fig. 3), while MBE-2 and the major
producing wells on the MBA and MBD drill pads are associated with the southern dome.
In the north, a fourth area related to the northernmost domed structure in the base of the
conductor has now been drilled and found to be productive [well MBB-1 (see Figs. 2-4)
The four possible fluid upwelling centers areas in the field are spatially associated with
four eruptive centers (Puncak Besar, G. Gambung, G. Wayang and G. Windu, going from
north to south; Fig. 2), whose age decreases generally towards the south. When
considering these four particular areas, one finds that the deep liquid reservoir in the G.
Lovelock et al., 1982). The slightly lower temperatures at depth in the G. Windu area
when compared to those measured further north suggests that any concentration of
solutes by boiling (as is indicated by the downhole pressure profiles) is being offset by
22
The waters of the deep reservoirs in the G. Gambung3 and G. Wayang areas have similar
Cl/B ratios than those from the G. Windu region, but are much more saline and have
lower gas contents suggestive of boiling at high temperatures, with very limited mixing
with groundwaters (a process that requires continuous heat recharge to maintain reservoir
temperatures). Parts of the system where the deep liquid reservoir is more saline would
therefore have to be older than those at G. Windu in order to provide time for this extent
of boiling off to occur, which is consistent with the age of the fluid upwelling centers (i.e.
generally getting younger towards the south). The variation in gas content between the
areas (Table 2) may reflect a lower gas flux from the older centers, and may be
responsible for the deepening of two-phase conditions as the increase in gas content
7. Discussion
reservoir was the same throughout the drilled area and the deep reservoir was under-
pressured with respect to local cold-water aquifers, the hydrology of the system cannot be
interpreted in terms of liquid upflows and outflows typical of most geothermal fields
the low degree of under-pressure in the deep liquid reservoir, pre-exploitation pressures
and temperatures above that of the maximum enthalpy of steam in the vapor-dominated
reservoirs, and the limited depth range where vapor-dominated conditions prevail at
3
Fluid chemistry data on the deep liquid reservoir is lacking for the northernmost wells
as they are too shallow (i.e. they were completed above it). Thus, all the deep liquid
geochemistry data come from well MBE-2.
23
Wayang Windu also mean that the field cannot be strictly compared to the solely vapor-
dominated. This transition is most advanced in the northern parts of the field where the
drop in the deep water table has been largest. Towards the south, the vapor-dominated
zones are thinner and deeper and make up proportionality less of the resource. This
The Wayang Windu system is largely sealed off from surrounding and overlying cold-
water aquifers, but still receives deep heat and fluid recharge from possibly four fluid
upwelling centers that are inferred to be progressively younger to the south. Vapor-
dominated reservoirs have developed over the fluid upwelling centers, with a vapor-
dominated reservoir in the north coalesced over the fluid upwelling centers associated
with the Puncak Besar and G. Gambung volcanic centers. Two possibly separate vapor-
dominated reservoirs further south, lie above separate fluid upwelling centers associated
with G. Wayang and G. Windu. Allis (2000) concluded that stock-sized intrusives cannot
provide sufficient heat to maintain a long-lived hydrothermal system like the one at
Wayang Windu. Thus, it is probable that the field is underlain by a major multiphase
intrusion, which would be the ultimate heat source. The diorite xenoliths observed in the
24
The locations of the four fluid upwelling centers could overlie shallow apophyses of the
larger intrusive body, which have been fractured by a combination of secondary boiling
and thermal contraction, to provide conduits that channel steam and gas flow from a
much larger intrusive source at depth. If the ages of the eruptive centers can be related to
the fluid upwelling centers, it would explain how the geothermal system has been active
Early pulses of acidic, magmatic condensates produced from the intrusive apophyses may
be responsible for the formation of the rare advanced argillic alteration observed deep in
the Wayang Windu field. The deep liquid reservoir now has a neutral pH; thus there is
no evidence for the presence of acidic fluids produced by the condensation of magmatic
The deep magmatic degassing is now heating and partially recharging a large exploitable
reservoir. As this deep recharge is composed of water vapor and non-condensable gases,
it does not require a large mass, only enough to maintain the large volume of the system
under boiling-point versus depth curve conditions, and allow the upper part of the
original water-dominated reservoir to boil off. Whether or not there is sufficient heat
flow to dry out the entire reservoir is another problem. The volume of the deep
contiguous deep reservoir at Wayang Windu may be proportionately larger than that of
geothermal systems associated with a single or closely spaced multiple intrusive heat
25
The Wayang Windu field is part of a cluster of Indonesian high-temperature geothermal
fields, which include the vapor-dominated reservoirs of Darajat and Kamojang that may
Patuha (Layman and Soemarinda, 2003) and Karaha Telaga Bodas (Moore et al., 2002,
2004) exhibit magmatic vapor cores as defined by Reyes et al. (1993), and may represent
an earlier stage of transition than found in the Windu part of the Wayang Windu
geothermal field; the Wayang and Gambung-Puncak Besar areas being successively
A common feature of some of these Indonesian geothermal fields is that they underlie
sector collapses; this is most strongly the case at Darajat, where it includes most of the
field. This may possibly also be the situation at Kamojang, where a partially circular
collapse feature is reported, although this has been previously interpreted to be a caldera
(Healy and Mahon, 1982). Moore et al. (2002, 2004) consider that the sector collapse of
the G. Galunggung volcano triggered the drying out of the Karaha-Telaga Bodas system.
The collapse of G. Wayang may have also contributed to the boiling off of at least that
part of the Wayang Windu geothermal system. The radiocarbon dating (discussed in
Section 2.1) indicating that this sector collapse took place after the formation of the
shallow hydrothermal alteration further north, supports this notion. Other areas of major
slope failure that may also be interpreted as sector collapses are found on the western side
of G. Windu, north of G. Gambang, west of Puncak Besar and north of G. Malabar (Fig.
3). However, that to the north of G. Gambung and that to the west of Puncak Besar may
be more closely related to slippage on the Bandung Basin Boundary Fault. Hydrothermal
26
alteration and activity have not been reported from these areas and it is more difficult to
their apparent dearth in the rest of Java, or elsewhere in the world. Allis (2000)
deformation in the upper plate restricting the recharge from depth, but this applies all
along Java, not just to the Bandung region. This compression is enhanced by the
subduction of the Roo Rise (Whittaker et al., 2007), a thickened section of oceanic crust
of the down-going Australian Plate. However, there is a break in the Roo Rise in the
Australian Plate immediately adjacent to the Bandung area, which is marked by a large
bulk-sound velocity anomaly of the down going slab (Gorbatov and Kennett, 2003), with
Subduction rollback (Whittaker et al., 2007) is taking place all along the Javan part of the
Sunda Arc because of the subduction of the old (95 - 135 Ma) slab of the Australian
Plate. It is likely that this rollback is locally accentuated by the thinner and denser
oceanic crust adjacent to Bandung, and that this produces local extension in the overlying
plate, which may be responsible for the large number of volcanic centers and associated
geothermal fields in the Bandung area. The rollback may also be responsible for the
occurrence of multiple fluid upwelling centers in Javan geothermal fields, as the location
27
of the upper crustal plate and the deep magma source will vary with time as rollback
proceeds.
The fields around Bandung are in terrains of sufficient elevation for there to be room for
steam reservoirs to form above the general level of deep meteoric recharge. This
recharge may be limited by the poor permeability at depth as the faults in the basement
had an initial strike-slip movement that reduced fault permeability due to prolonged
shearing. The subsequent fault movement within the volcanic deposits was normal (at
least at Wayang Windu) resulting in higher rock mass permeabilities, thus favoring the
Acknowledgements
The authors wish to gratefully acknowledge the permission and support of the
management of Mandala Nusantara Ltd. to publish this paper and the help of Manfred
Hochstein in doing so. Constructive reviews by Rick Allis and Dick Henley are also
acknowledged, along with the editorial assistance of Greg Bignall, Sabodh Garg, Marcelo
Lippmann and Joe Moore. Thanks go to Mariano Gutierrez of Sinclair Knight Merz Ltd
and Iis Dian Indriani for their assistance with the figures.
References
Abrenica, A.B., 2007. Hydrothermal alteration and fluid inclusion studies in the northern
Wayang Windu geothermal field, West Java, Indonesia. MSc Thesis, Geological
28
Engineering Department, Faculty of Engineering, Gadjah Mada University, Yogyakarta,
Allis, R., 2000. Insights on the formation of vapor-dominated systems. In: Proceedings
Allis, R., Moore, J.N., McCulloch, J., Petty, S., DeRocher, T., 2000, "Karaha-Telaga
Alzwar, M, Akbar, N., Bachri, S., 1992. Geology of the Garut and Pameungpeuk
Anderson, E., Crosby, D., Ussher, G., 1999. As plain as the nose on your face:
geothermal systems revealed by deep resistivity. In: Proceedings of the Twenty-first New
Anderson, E., Crosby, D., Ussher, G., 2000. Bulls-Eye! Simple resistivity imaging to
reliably locate the geothermal resource. In: Proceedings of the World Geothermal
29
Asrizal, M., Hadi, J., Bahar, A., Sihombing, J.M., 2006. Uncertainty quantification by
using stochastic approach in pore volume calculation, Wayang Windu Geothermal Field,
Bandyopadhay, I., Juandi, D., Baroek, M., Hadi, J., Pramono, B., 2006. In situ stress
analysis in deviated wells using inversion method A case study from volcanic
pyroclastics of Wayang Windu Field, Java, Indonesia. In: Proceedings of the Jakarta
Bogie, I., Lawless, J.V., 2000. Application of mineral deposit concepts to geothermal
Bogie, I., Mackenzie, K.M., 1998. The application of a volcanic facies model to an
andesitic stratovolcano hosted geothermal system at Wayang Windu, Java, Indonesia. In:
Bogie, I., White, P.W., Lawless, J.V., 2003. Chalcedony within low-sulphidation
Browne, P.R.L., 1978. Hydrothermal alteration in active geothermal fields. Ann. Rev.
30
Budiardjo, B., 1992. Petrographic study of cores and cuttings from drillhole WWD-1.
Dam, M.A.C., 1994. The Late Quaternary evolution of the Bandung Basin, West-Java,
Ellis, A.J., Mahon, W.A.J., 1977. Chemistry and Geothermal Systems. Academic Press,
Fournier, R.O., Truesdell, A.H., 1973. An empirical Na-K-Ca geothermometer for natural
Ganda, S., Hantono, D., 1992. Alteration mineralogy of the Wayang Windu geothermal
field, West Java, Indonesia. In: Kharaka, Y.K. and Maest, A.S. (Eds.), Water-Rock
Ganda, S., Hantono, D., Sunaryo, D., 1992. Alteration mineralogy of the Wayang Windu
geothermal field, West Java, Indonesia. In: Proceedings of the Twenty-First Annual
Gorbatov, A., Kennett, B.L.N., 2003. Joint bulk-sound and shear tomography for
Western Pacific subduction zones. Earth and Planetary Science Letters 210, 527-543.
31
Hadi, J., Harrison, C., Keller, J., Rejeki, S., 2005. Overview of Darajat reservoir
Healy, J., Mahon, W.A.J., 1982. Kawah Kamojang Geothermal Field, West Java,
Indonesia. In: Proceedings of the. Pacific Geothermal Conference incorporating the 4th
Hulen, J.B., Anderson, T.D., 1998. The Awibengkok Indonesia, geothermal research
Layman, E.B., Soemarinda, S., 2003. The Patuha vapor-dominated resource, West Java,
Lovelock, B.G., Cope, D.M., Baltasar, A.J., 1982. Hydrogeochemical model of the
Tongonan geothermal field. In: Proceedings of the Fourth New Zealand Geothermal
32
Moore, J.N., Allis, R., Renner, J.L., Mildenhall, D., McCulloch, J., 2002. Petrological
Moore, J.N., Christenson, B., Browne, P.R.L., Lutz, S.J., 2004. The mineralogic
1499.
Murakami, H., Kato, Y., Akutsu, N., 2000. Construction of the largest geothermal power
plant for Wayang Windu project, Indonesia. In: Proceedings of the World Geothermal
Nemok, M., Moore, J.N., Christensen, C., Allis, R., Powell, T., Murray, B., Nash, G.,
Newcomb, K R., McCann, W.R., 1987. Seismic history and seismotectonics of the
33
Nurruhliati, R., 1996. Hydrothermal alteration of well WWR-SH, Wayang Windu
geothermal field, West Java, Indonesia. Geothermal Diploma Project Report 96.18,
Raharjo, I, Wannamaker, P., Allis, R., Chapman, D., 2002. Magnetotelluric interpretation
of the Karaha Bodas geothermal field, Indonesia. In: Proceedings of the Twenty-seventh
Reyes, A.G., Giggenbach W.F., Saleras, J.R.M., Salonga, N.D., Vergara, M.C., 1993.
Stoffregren R.E., Alpers C.N., 1987. Woodhousite and svanbergite in hydrothermal ore
Sudarman, S., Pujianto, R., Budiarjo, B., 1986. The Gunung Wayang Windu geothermal
area in West Java. In: Proceedings of the Fifteenth Annual Convention of the Indonesian
34
Suminar, A., Molling, P., Rohrs D., 2003. Geochemical contributions to a conceptual
Thaysa, A., 2003. Hydrothermal alteration and geochemistry of geothermal fluids study
of the Wayang Windu field, West Java, Indonesia. Final Project Report, Department of
Indonesia, 74 pp.
Utami, P., 2000. Characteristics of the Kamojang geothermal reservoir (West Java) as
Whittaker, J.M., Muller, R.D., Sdrolias, M., Heine, C., 2007. Sunda-Java trench
kinematics, slab window formation and overriding plate deformation since the
Wibowo, H., 2006. Spatial Data Analysis and Integration for Regional-Scale Geothermal
Prospectivity Mapping, West Java, Indonesia. Msc Thesis, International Institute for
35
Table 1
Alteration mineralogy (Possible relict phases shown in italics)
Amphibole
36
Table 2. Geochemical and pressure-temperature properties of Wayang Windu reservoir
areas prior to production
Notes
1
NaKCa geothermometer of Fournier and Truesdell (1973).
2
Non-condensable gases in weight percent.
3
Prior to production.
4
Well MBB-1 drilled in this area produces 20 MWe on initial discharge.
It did not penetrate into the deep liquid reservoir and fully stabilised gas,
temperature and pressure data are not yet available.
37
Figure captions
Fig. 2: Topographic map of the drilled portion of the Wayang Windu geothermal field;
contour interval: 100 meters; bold contour: 2000 m asl. Also shown are the locations of
drill pads, well tracks and of section A-B described in Figs. 4 and 6.
Fig. 3: Location of geothermal wells, thermal features, volcanic summits, calderas, and
sector collapses in the Wayang Windu geothermal field in relation to the base of the
conductor.
Fig. 4: Section across the Wayang Windu geothermal field showing well tracks,
geological units (extended north and south from that of Bogie and Mackenzie, 1998) and
geothermal wells versus elevation. Note that the casing in well WWE-1 has a hole
Fig. 6: Section across the Wayang Windu geothermal field showing well tracks,
isotherms and the known location of the tops of the vapor-dominated and deep liquid
38
Figure
Click here to download high resolution image
Figure
Click here to download high resolution image
Figure
Click here to download high resolution image
Figure
Click here to download high resolution image
Figure
Click here to download high resolution image
Figure
Click here to download high resolution image