History of Geothermal Exploration in Indonesia From 1970 to 2000

Available online at www.sciencedirect.com

Geothermics 37 (2008) 220–266

History of geothermal exploration in Indonesia from 1970 to 2000 Manfred P. Hochstein a,∗ , Sayogi Sudarman b a b

Geology Department, University of Auckland, Private Bag 92019, Auckland, New Zealand Indonesian Geothermal Association, Jl Gatot Subroto Kav 18, Jakarta Sel. 12950, Indonesia Received 22 January 2008; accepted 22 January 2008 Available online 17 March 2008

Abstract Reconnaissance surveys undertaken since the 1960s show that more than 200 geothermal prospects with significant active surface manifestations occur throughout Indonesia. Some 70 of these were identified by the mid-1980s as potential high-temperature systems using geochemical criteria of discharged thermal fluids. Between 1970 and 1995, about 40 of these were explored using geological mapping, geochemical and detailed geophysical surveys. Almost half of the surveyed prospects were tested by deep (0.5–3 km) exploratory drilling, which led to the discovery of 15 productive high-temperature reservoirs. Several types of reservoirs were encountered: liquid-dominated, vapour-dominated, and a vapour layer/liquid-saturated substratum type. All three may be modified by upflows (plumes) containing magmatic fluid components (volcanic geothermal systems). Large, concealed outflows are a common feature of liquid-dominated systems in mountainous terrain. All explored prospects are hosted by Quaternary volcanic rocks, associated with arc volcanism, and half occur beneath the slopes of active or dormant stratovolcanoes. By 1995, five fields had been developed by drilling of production wells; three of them supplied steam to plants with a total installed capacity of 305 MWe . By 2000, with input from foreign investors, the installed capacity had reached 800 MWe in six fields, but geothermal developments had stalled because of the 1997–1998 financial crisis. © 2008 Elsevier Ltd. All rights reserved. Keywords: Geothermal exploration; High-temperature prospects; Selection criteria; Exploration surveys; Exploratory drilling; Java; Sumatra; Sulawesi; Bali; Flores; Indonesia

1. Introduction Geothermal exploration in Indonesia began in 1970 with the aim of finding and developing high-temperature geothermal systems. The developments between 1970 and 1990 (in many cases ∗

Corresponding author. Tel.: +64 9 373 7599; fax: +64 9 373 7436. E-mail addresses: [email protected], [email protected] (M.P. Hochstein).

0375-6505/$30.00 © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.geothermics.2008.01.001

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

221

until 1995) are not well documented. An attempt is made here to summarize the early surveys, referring to information in publications and reports written in English, mainly by Indonesian scientists and engineers, which are accessible in the public domain. Thus, the exploration of prospects is discussed where detailed geological, geochemical, and geophysical methods were combined to assess field characteristics of importance when siting exploration wells over inferred high-temperature reservoirs. Results of early geophysical surveys are discussed in more detail where they led to proper estimates of reservoir areas and, combined with important geochemical and geological findings, allowed a prediction of reservoir characteristics. Since most of the earlier exploration efforts are not listed in the scientific literature, theses and diploma reports of Indonesian geothermal graduate students attending the University of Auckland between 1979 and 2003 became an important source of information and were used for this paper. The geothermal terminology employed here is that adopted in Hochstein and Browne (2000). The description of a few prospects not covered by published work is based on observations and field notes collected by the authors. Descriptions of Indonesian geothermal resources probably started with the reconnaissance surveys described by Junghuhn over 150 years ago (Junghuhn, 1854), whose studies covered mainly active volcanoes and large thermal areas on Java. From around 1900 until the beginning of World War II, most of the Indonesian Quaternary volcanoes and their fumarole and solfatara fields were mapped by the Dutch colonial Geological Survey; the results were later published in the first volume of the Catalogue of the Active Volcanoes of the World (Neumann van Padang, 1951). A summary of documented thermal springs on Java, the Molucca Islands, and Sumatra can be found in the lists of global thermal springs by Waring (1965). After Indonesia gained independence, the Volcanological Survey of Indonesia (VSI) started work in the 1960s with reconnaissance-type surveys that led to the compilation of an inventory of sites with thermal manifestations. A map showing the location of these sites on Java and Bali was compiled by VSI in 1968 (Purbo-Hadiwidjojo, 1970). The studies were supported by the State Electricity Company (PLN) and the Bandung Institute of Technology (ITB). International and foreign missions (UNESCO, EURAFREP) visited several geothermal prospects at that time and, with reference to the size and type of manifestations, drew attention to prospects associated with hot spring discharges. A revised catalogue of volcanoes and fumarole fields in Indonesia published by VSI (Kusamadinata, 1979) provided important information now incorporated in a world-wide catalogue of volcanoes that can be accessed through the Smithsonian volcano website (see bottom of Table 1). All Indonesian geothermal systems associated with surface manifestations discharging fluids at boiling temperature occur in areas with Quaternary volcanism and active volcanoes along welldefined volcanic arcs. There are five active arc segments in Indonesia that define regions of interest for geothermal exploration (Fig. 1). Using plate tectonic concepts, all active Indonesian arcs can be interpreted as the result of sub-crustal melting induced by subducted lithosphere plates (Katili, 1975). The major plate tectonic structures shown in Fig. 1 had already been recognised during the 1970s (Hamilton, 1979). All young Quaternary volcanoes can be associated with cooling magma and igneous intrusions, which, in turn, are heat sources for active arc-type geothermal systems. The first inventory (in English) of Indonesian thermal areas and prospects, compiled by VSI as part of a New Zealand (NZ) Aid project in 1987 (NZMFA, 1987; Mahon, 1987), listed 215 sites. The inventory has been upgraded and about 245 thermal prospects are listed in its 1998 version, which is accessible through a VSI website (see bottom of Table 1). We have used the same names, numbering system, and coordinates of the geothermal sites shown in the 1998 VSI catalogue (with the exception of a few not yet given there). A list of 87 Indonesian geothermal

222

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

Fig. 1. Principal geographical features of Indonesia showing important plate tectonic structures and location of active volcanic arcs, based on information taken from Hamilton (1979), Simkin and Siebert (1994), and Hall (2002). Sites of explored geothermal prospects associated with arc segments are shown in Figs. 2–4.

prospects already covered by inventory/reconnaissance surveys was also presented by Manalu (1988). Another important registry of Indonesian geothermal prospects is that contained in an unpublished report by Kingston and Morrison (1997), which lists 204 sites and describes their state of exploration. The selection of Indonesian geothermal prospects for exploration studies was based on earlier reconnaissance surveys. The characteristics of the discharged thermal fluids, types of manifestation, and extent of thermal alteration at the surface, together with geothermometer data derived from chemical analyses, were taken into consideration for the selection. Initially, empirical (liquid and gas) geothermometers were used (Henley et al., 1984); later, theoretical based geothermometers (for example, Giggenbach, 1980, 1981) were often applied, using selected fluid samples. Between 1970 and 1995, about 70 sites were tentatively classified as high-temperature prospects where geothermometer data indicate deep fluid equilibrium temperatures of >220 ◦ C. Reconnaissance and more detailed exploration studies of most of the 70 prospects are discussed below. Geothermal exploration increased in 1994 when foreign and private investors were encouraged by the Indonesian Government to develop and to run so-called independent power projects (IPPs), which had to sell geothermal power under Energy Sales Contracts to the state electricity company PLN. This resulted in accelerated exploration and production drilling, which came to a halt as a result of a financial crisis in 1997–1998. The history of geothermal exploration in Indonesia between 1970 and 2000 has therefore been divided into three stages: (1) the ‘starting period’, covering 1970–1980; (2) a ‘diverse period’ from 1980 to 1995, and (3) an ‘accelerated development period’ from 1995 to 2000.

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

223

2. The first attempts (1918–1970) Exploration of geothermal resources associated with active fumarole and solfatara fields with the objective of generating electricity was first proposed in 1918. Initial exploration drilling was undertaken by the Volcanological Section (later to become the Volcanological Survey of Indonesia, or VSI) of the colonial Geological Survey of Indonesia (GSI), at Kawah1 Kamojang, on Java, in 1926. Several holes were drilled inside a large fumarole field. The third well (KMJ-3) was 66 m deep and produced steam. The last two holes (123 and 128 m deep) intermittently produced a two-phase mixture of steam and hot water. The shallow well KMJ-3 discharged continuously for another 50 years; the discharge rate was about 8 MW2 (about 10 t/h of steam) with a temperature of 140 ◦ C at the open lip when measured in February 1975. The two deeper wells stopped discharging some time after 1928 (Stehn, 1929). Historic photos of the first geothermal drilling efforts can be seen in Alzwar (1986). Another attempt to explore a solfatara field (K. Sikidang) was made at Dieng in 1928, sponsored by the Mines Department. A non-producing exploratory hole was drilled to 80 m depth, encountering a temperature of 145 ◦ C at the bottom (Radja, 1975). Further attempts to explore Indonesian geothermal fields by drilling were not made until 1972. The results of the earlier geological investigations were used to rank several prospects on Java for further investigation. The list included the volcanic complexes of Dieng, Gunung Tampomas, Gunung Salak and Gunung3 Perbakti, K. Kamojang, and the Cisolok prospect (Zen and Radja, 1970). In 1969, a PLN group (Power Research Institute) undertook a geothermal reconnaissance survey of Sulawesi (Radja, 1970). 3. Geothermal exploration (1970–1980) During the first PELITA (first 5-year development plan, 1969–1974), the Volcanological Survey group (VSI) completed a geothermal inventory of Sumatra, Sulawesi, and the Halmahera Islands (Radja, 1985; Soetantri, 1986). Geothermal exploration was supported by foreign aid projects. The Indonesian State Oil Company (Pertamina) entered geothermal exploration from 1974 onwards and became responsible for all geothermal exploration in Java and Bali, in line with Presidential Decree PD 16/1974. 3.1. Exploration of the Dieng prospect (Fig. 2) Between 1970 and 1972, the K. Sikidang sector of the Dieng volcanic complex was investigated under the auspices of a USAID program, involving US Geological Survey staff and VSI/ITB/PLN groups acting as counterpart. The prospect carries a certain volcanic risk, as indicated by its history of phreatic eruptions and gas hazards (Simkin and Siebert, 1994). All earth-science disciplines (geology, geochemistry, and geophysics) were used to assess the extent of the prospect and to locate drillsites. Several exploration holes were drilled in 1972, the deepest of which (DX2) reached 145 m, where it encountered 175 ◦ C; the hole was not productive (Radja, 1975). Most of the original objectives were not met, partly because of the inexperience of the drilling contractors.Pertamina took the project over in 1974 and repeated the geological, geochemical and 1

The abbreviation K. will be use for ‘Kawah’ (crater or a large crater-like depression) from here on. All heat discharge rates are listed in MW (i.e. MWt ); power plant capacity and inferred electricity generation rates are quoted in MWe . 3 The abbreviation G. will be used for ‘Gunung’ (mountain) from here on. 2

224

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

geophysical exploration studies, with the collaboration of a French contractor (BEICIP), and the aim of siting deep exploration holes. The resistivity surveys indicated the presence of a roughly 12 km2 area with shallow, low resistivities. In September 1977 the first well, DNG-1, was drilled near the acid manifestations of K. Sikidang to a depth of 1900 m. Dilute magmatic fluids were encountered, a reminder that phreatic and CO2 -driven eruptions were still a threat (Le Guern et al., 1982; Giggenbach et al., 1991). Acid alteration was dominant down to 1000 m and a maximum temperature of 326 ◦ C, with stable mineral equilibrium temperatures of 320–315 ◦ C, was measured between 1450 and 1600 m depth (Ganda, 1984; Fauzi, 1985). The well was productive but difficult to maintain and had to be abandoned after an accident (caused by a partly corroded valve). DNG-1 was the deepest geothermal well drilled in Indonesia during the 1970–1980 period. The second well (DNG-2, about 0.6 km south of DNG-1) was completed in August 1979 after reaching a total depth of 1660 m (Tmax ∼ 290 ◦ C). It was productive and could discharge about 80 t/h of steam with an initial, anomalously high, non-condensable gas (NCG) content of about 20% (by weight) (Bachrun et al., 1995). 3.2. The New Zealand geothermal aid program In 1971, the NZ consultant group Geothermal Energy Ltd. (GENZL) visited several geothermal prospects on Java and Bali and proposed reconnaissance studies of selected prospects using VSI inventory data. The sites to be studied were ranked according to the size of the area with manifestations, preliminary geochemical results, ease of access, and likely regional electricity demand. A proposal to investigate the K. Kamojang, Darajat, G. Salak, Cisolok and Bali prospects (see Fig. 2) was accepted and developed into a bilateral aid (Colombo Plan) project supported by the Indonesian and NZ Governments. The aim of the project was to use standard exploration techniques together with exploration drilling to demonstrate the feasibility of producing geothermal energy for electricity generation in at least one of the five selected prospects. The field surveys started in late 1972 and were supported by VSI, the first counterpart agency. Pertamina and PLN also participated from 1974 onwards. By 1974 the five prospects had been investigated and Kamojang and Darajat were selected for deep exploration drilling. 3.2.1. Exploration of the Kawah Kamojang prospect (Fig. 2) By 1974 the results of resistivity and shallow temperature gradient surveys had shown that the upper part of the Kamojang geothermal reservoir covers an area of at least 14 km2 . It was most likely capped by a thick layer saturated with steam condensates and contained electricconductive clay minerals; the natural heat loss rate was of the order of 100 MW (Hochstein, 1975). Geochemical data pointed to a vapour-dominated system (Kartokusumo et al., 1975); local tectonic structures and litho-stratigraphic sections were defined (Healy and Mahon, 1982). A mediumsize drilling rig was imported and the first deep well (KMJ-6) was sited near the centre of the low-resistivity anomaly, which delineated the extent of the thermally altered rocks. The well was started in late September 1974 and completed after 1 month to a depth of 615 m (Tmax = 239 ◦ C). It was discharged in late December 1974, producing about 6.5 t/h of steam through a 0.11 m (4.5 in.) diameter slotted liner. It confirmed that Kamojang is a vapour-dominated system, the fourth of its kind discovered worldwide (after Larderello in Italy, The Geysers in the USA, and Matsukawa in Japan). In rapid succession another productive (KMJ-7) and three non-productive (KMJ-8, 9, and 10) wells were drilled within a roughly 4-km2 large central area. By mid-August 1975 the last hole was completed to 760 m.

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

225

Fig. 2. Location of high-temperature geothermal prospects on Java and Bali explored between 1970 and 2000. Symbols used for each locality describe the type of system encountered as explained in the inset. The approximate and smoothed outline of Quaternary pyroclastics and lava flows are from geological maps published by UNESCO (1976) and the Geological Survey of Indonesia (1977).

The results led to an enlarged aid project with the aim of producing sufficient steam to generate electricity for a 30 MWe plant, all to be sponsored by NZ aid funds. The production drilling started with well KMJ-11 (September 1976), using a large drilling rig on loan from the NZ Government. The last well (KMJ-20) was completed in August 1979. Pertamina also joined the project with its own rig when drilling well KMJ-19. The depths of the wells varied between 935 m (KMJ-18 with an output of 125 t/h steam) and 1800 m (KMJ-15 with an output of only 5 t/h of steam). All production wells were vertical and produced through 0.18 m (7 in.) diameter slotted liners. Maximum temperatures were between 232 and 243 ◦ C (Grant, 1979a). At the end of 1979, all producing wells at Kamojang could deliver about 380 t/h of steam (NCG 30 g/kg.

232

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

resistivity area had been found. Three deep wells were drilled, the first of which (R-1) was the deepest (about 2740 m). Well R-1 encountered temperatures up to 307 ◦ C and rocks with significant acid alteration. It also penetrated the entire volcanic sequence and bottomed in sediments. Separated steam produced during testing (enough to generate 8 MWe ) contained a moderate amount of NCG (about 5% by wt.) which, together with trace gas and isotope characteristics, pointed to some deep input of magmatic constituents. Two more wells were drilled but had either zero or low outputs. Hence, exploration drilling shifted back to the Awibengkok field (western part) where three deep, large-diameter wells (0.3 m in the bottom section) were drilled during 1985. The first of these (AW-6) was successful; it was drilled to a depth of 1370 m, encountered fluids with a maximum temperature of 260 ◦ C and produced through a ∼0.25 m diameter slotted liner the equivalent of 20 MWe when tested in April 1985. It was the first large producing well in Indonesia. The other two wells (AW-7, AW-8) had similar characteristics with maximum depths of 1710 and 1830 m, maximum temperatures of 268 and 279 ◦ C, and equivalent outputs of 10 and 22 MWe , respectively. The entire project was now viable and a minimum power potential of around 145 MWe was predicted. Additional production drilling was stalled to wait for the outcome of the construction of a power plant by PLN, which took place during the following decade. There is no doubt that the exploration of the G. Salak (Awibengkok) prospect was a great success considering all the terrain obstacles that had to be overcome. The development also showed that a large, high-temperature reservoir with non-corrosive fluids can occur separately close to a volcanic geothermal system. Unfortunately, the earlier developments at G. Salak were not published and most of the important information listed here comes from short, later publications (Takhyan et al., 1990; Noor et al., 1992) and our own field notes. 4.2.2. Development of the Darajat field Between 1980 and 1983 a few minor follow-up exploration studies were undertaken by Pertamina at Darajat, consisting of MT soundings and the drilling of additional temperature–gradient holes to 200 m depth (Sudarman, 1983). After signing the JOC contract, the new developer (Amoseas Indonesia Co.) undertook additional follow-up surveys of the entire prospect between 1985 and 1986. These involved gravity, resistivity (using the CSAMT method), MT, airbornemagnetic, micro-earthquake, and soil–Hg surveys. An integrated interpretation model showed that the low-resistivity structure, thought to represent the propylitic zone, extends over an area of about 22 km2 , i.e. an increase of about 50% over that found previously. The demagnetised part of the reservoir showed an elongated structure covering ∼10 km2 (Whittome and Salveson, 1990). In 1987–1988 four deep exploratory wells (vertical depth between 1500 and 2300 m) were drilled within a ∼5 km2 area, approximately centred on the manifestations of K. Manuk. Two of the wells (DRJ-4 and DRJ-7) encountered maximum temperatures of 243 and 241 ◦ C and produced dry steam at rates of 81 and 88 t/h, respectively. The other two wells were almost non-productive. One of these (DRJ-6) confirmed the existence of the concealed, shallow outflow beneath the eastern margin of the field that had already been detected by well DRJ-1 ten years earlier. With a total steam output sufficient to produce about 24 MWe from three wells, it was predicted that the explored part of the reservoir would sustain steam production to run a 55 MWe power plant, as outlined by Dobbie (1991). Further production drilling was halted until financing of the power plant had been arranged by PLN.

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

233

4.3. Exploration of prospects with deep outflows 4.3.1. The Banten prospects The Citaman prospect on the souuthern slopes of G. Karang had been explored during the previous decade (Mulyadi, 1985). Additional MT studies, a micro-earthquake survey, and a temperature–gradient survey of 15 gradient holes were conducted in 1983. The surveys outlined a ∼6 km long, NNW-oriented target area exhibiting low resistivities at shallow depths and increasing temperature gradients towards G. Karang (Sudarman, 1985). At the toe of the target anomaly are the Citaman hot springs, discharging hot, neutral-pH bicarbonate water (Tmax = 67 ◦ C; heat discharge rate of ∼20 MW), which deposit travertine. Empirical geothermometers pointed to reservoir temperatures of up to 280 ◦ C. However, most of the characteristics listed are also features that occur at the toe of a concealed outflow of a hot water system in steep terrain (Hochstein, 1988). A deep exploratory well (BTN-1) was drilled to about 2100 m depth in 1985; the well site was about 3 km upstream of the thermal springs. It encountered a lateral outflow of hot water with temperatures between 100 and 120 ◦ C at 1000–2000 m depth (Hochstein, 1988). Exploration of the Batukuwung prospect in the Banten Caldera continued in 1986–1987 (Soemarinda, 1988). Fieldwork included dc-resistivity and MT studies, together with gravity, ground magnetic, temperature–gradient, and micro-earthquake surveys. Several temperature–gradient wells to 250 m depth encountered dilute neutral-pH bicarbonate water with temperatures of 55–65 ◦ C. It was inferred that the whole area is underlain by a concealed outflow of thermal water that has been diluted by groundwater infiltration. In view of the poor drilling results from the Citaman area, deep exploration drilling at Batukuwung was cancelled. 4.3.2. Cisolok–Cisukarame The Cisolok–Cisukarame prospect had been explored during the 1970s. Exploration was continued until 1983 with the drilling of about 20 shallow temperature–gradient holes; five reached depths between 100 and 150 m (Soetantri, 1986). Early resistivity surveys, performed in 1974, indicated that the hot springs at Cisolok and Cisukarame were discharge features of a concealed outflow of hot water travelling more than 9 km from an unknown source in steep terrain to the toe of the outflow near Cisolok. This model was still not accepted during the early 1980s. Instead it was postulated that a young intrusion outcropping near the Cisolok hot springs, which discharge boiling water and deposit massive travertine, was close to an inferred deep heat source. The ∼1200 m deep exploration well CIS-1 was drilled in late 1986 close to the Cisolok hot springs and encountered a ∼1000 m thick, concealed outflow of thermal water with a bottom temperature of about 120 ◦ C (Hochstein, 1988). 4.4. Exploration of systems associated with dormant volcanoes During the 1980s Pertamina explored a number of prospects associated with fumarole fields on the flanks of inactive Holocene volcanoes. Some of these prospects discharge dilute vapour condensates similar to those at Darajat and Kamojang, with SO4 contents derived from the oxidation of ascending H2 S gas. 4.4.1. Ungaran Two small, active fumarole fields occur at Gedong Sanga over the southern flank of the dormant Ungaran volcano, ∼2 km from its summit. Around the periphery, 5–10 km from the top, there are three other warm spring areas (Soetantri, 1986); dilute brine (∼15 g/kg TDS) is discharged

234

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

at Kaliulo, about 15 km to the east. The prospect was explored by Pertamina employing several geophysical methods (dc-resistivity, MT, and aeromagnetic surveys) between 1985 and 1990 (Budiardjo et al., 1989). Four temperature–gradient wells, up to 500 m deep, were drilled around the Gedong Sanga area; these showed slightly anomalous temperatures and temperature–gradient values near the bottom (47 ◦ C at 300 m depth). A deep-seated resistivity anomaly associated with a deep geothermal reservoir was inferred to occur beneath the summit area (Budiardjo et al., 1989). However, since about 90% of the prospect area was located inside protected forests and because of restricted access, further exploration was discontinued. 4.4.2. Wayang–Windu Gunung Wayang and G. Windu are two small lava domes with no historic eruption history. A fumarole field with surface acid alteration is found near the top of G. Wayang, within its crescentshaped crater. A smaller area with steaming ground occurs at G. Windu. Another smaller fumarole field lies ∼6 km north of G. Wayang extending over the southern flanks of the larger, also inactive G. Malabar stratovolcano. All thermal manifestations occur within an area of ∼30 km2 (Soetantri, 1986). A detailed exploration programme was launched by Pertamina in 1982, including geological, geochemical and geophysical surveys. The latter studies involved dc-resistivity surveys using Schlumberger arrays, ‘head-on’ resistivity profiling, MT and SP surveys, as well as gravity and temperature–gradient studies with temperatures measured in six gradient holes (down to 170 m depth) (Sudarman et al., 1986). The first phase of the geophysical surveys focussed on the Wayang–Windu prospect where an area of ∼25 km2 with low resistivity was outlined. Comparison of the constituents of the bicarbonate water discharged at Wayang–Windu with those from Kamojang and Darajat revealed a close affinity and it was inferred that the Wayang–Windu prospect might host a vapour-dominated reservoir (Sudarman et al., 1986). Follow-up MT-resistivity surveys during the next decade indicated that rocks with low resistivity extended beneath G. Malabar (Anderson et al., 1999, 2000), thus doubling the size of the prospective target area. The first deep exploration hole, WWD-1, was sited near the centre of the inferred Wayang–Windu anomaly. It was drilled at the beginning of 1991 to a depth of ∼1600 m, encountering 280 ◦ C at the bottom (Budiardjo, 1992). The hole penetrated a ∼900 m thick cover with a ∼350 m thick layer saturated with steam5 condensates at the bottom, underlain by a ∼600 m thick vapour-dominated layer. It bottomed in a liquid-saturated (∼20 g/kg TDS) deep reservoir. Another 0.6 km deep exploratory hole (MSH-1) was drilled by Pertamina at the end of 1993, about 5 km north of the discovery well WWD-1; it encountered a vapour-dominated (natural two-phase) zone at the bottom. At the end of the 1990s similar systems, but with a magmatic vapour5 core, were found at Patuha and Telaga Bodas. The Wayang–Windu prospect was rapidly developed when, at the end of 1994, a joint venture (initially by Indonesian and US companies) was established with the aim of conducting a ‘total project development’ (see Section 7). 4.4.3. Gunung Wilis About 10 km west of the centre of the G. Wilis stratovolcano lies an old crater (possibly from a hydrothermal eruption), now occupied by the 1.2 km2 Lake Ngebel. One kilometre south of the 5 The term ‘vapour’ is used in a narrow sense for the gas phase of water as it occurs underground, whereas the term ‘steam’ describes the same phase when discharged at the surface. However, ‘steam’ has been used in the literature for both settings, which explains the use of terms such as ‘steam condensate’ or ‘steam cap’ in a reservoir setting although strictly they should be identified as ‘vapour condensate’ or ‘vapour cap’.

M.P. Hochstein, S. Sudarman / Geothermics 37 (2008) 220–266

235

lake one finds a small fumarole field and a few hot springs (Tmax ∼ 47 ◦ C) discharging neutralpH chloride–bicarbonate water. The manifestations lie within a valley that drains the western slopes of G. Wilis and were mapped by VSI prior to 1983. Between 1983 and 1990 the Wilis prospect was explored by Pertamina using geological, geochemical, and geophysical surveys. A well-defined, low-resistivity target area was not found. A soil–Hg survey discovered a number of isolated, small anomalies over the western flank of the volcano (Mulyono, 1989). An airbornemagnetic survey located a few ill-defined possible targets (Rachman, 1990). Two slim holes were drilled to 500 m depth in the Ngebel area. The hole nearest to the lake (WSH-1) exhibited normal temperatures, whereas the other, close to the hot springs (WSH-2), found intermediate temperatures (Tmax = 149 ◦ C) and a small temperature inversion at the bottom. Some steam was produced from the 3 in. diameter hole but it was later abandoned; the second well intersected an outflow structure. It appears that the resource is small (Setyobudi, 1993). The project was later abandoned. 4.5. Exploration of active volcanic geothermal systems with significant acid surface manifestations The success at Dieng in 1977, when temperatures of up to 326 ◦ C were measured in the discovery well DNG-1, had a strong influence on geothermal exploration in Indonesia when it was assumed that fluids with similar high temperature could occur beneath the flanks of other active or dormant stratovolcanoes with active surface manifestations. Over 15 volcanic geothermal prospects were explored in Java during the 1980s, with seven showing significant acid surface alteration and manifestations associated with degassing intrusions. In these systems the discharged SO4 is derived from magmatic SO2 (Moore et al., 2002c). 4.5.1. Dieng Drilling of production wells was continued during the 1980s by Pertamina, mainly in the Sikidang area. By 1994, 24 deep wells had been drilled to depths between 1750 and 2500 m in an area of about 5 km2 around K. Sikidang. The wells encountered a liquid-dominated system at the bottom, saturated with very dilute brine (TDS of the order of 5–10 g/kg), a high-boron content (up to 10% of TDS), and different Cl/B ratios between adjacent wells (Suwana, 1986). The two-phase and vapour-dominated zones found at depths