89411 Geothermal Resource Risk in Indonesia A Statistical Inquiry Subir K. Sanyal, James W. Morrow, Migara S. Jayawardena, Noureddine Berrah, JUNE 2014 Shawna Fei Li, and Suryadarma ABSTRACT T his paper presents a statistical study of the geothermal resource risk in Indonesia, specifically, that the resource base and well productiv- ity are adequate and that the drilling cost per well is reasonable. This paper is timely because the Government of Indonesia is now embarking on an ambitious plan to rapidly scale up geothermal power capacity and eventually achieve a longer-term target of 9,500 MW by 2025. This study relies on the resource base estimates made by the Indonesian Government-owned enterprise P.T. Pertamina (now subsidiary Pertamina Geothermal Energy, or “PGE”) for nearly 80 sites, and productivity data on 215 wells in the country in the GeothermEx archives; these wells comprise some 80% of the total production wells drilled in Indonesia. Such a thorough national inventory of the geothermal resource base is available from very few countries. The geothermal resource base (proved-plus-probable-plus-possible) at a given site ranges from 10 MW to 800 MW with a log-normal distribution. More than 70% of the known Indonesian fields have a resource base greater than 50 MW and at least half of the fields (about 40) offer a resource base of 100 MW or more. Photo courtesy of GeothermEX Inc. Commercial wells in Indonesia vary in capacity from 3 MW to more than 40 MW, with a median value of 9 MW; this range and median are larger than seen in most countries; we believe the most likely range of well productivity worldwide is 4 to 6 MW. The commercial wells The findings, interpretations, and conclusions in Indonesia tend to fall in one of four groups in terms of capacity: (a) 3 expressed in this paper are entirely those of the authors and should not be attributed in to 5 MW representing “tight” wells; (b) 7 to 9 MW representing “typical” any manner to the World Bank Group, to wells; (c) 15 to 19 MW representing wells that usually produce from members of its’ board of executive direc- tors or the countries they represent, or the Government of Indonesia. Subir K. Sanyal is a Senior Advisor at GeothermEx James W. Morrow is a Database Manager at GeothermEx Migara S. Jayawardena is a Senior Energy Specialist at the World Bank Noureddine Berrah was formerly a Lead Energy Specialist at the World Bank Shawna Fei Li was formerly a Junior Professional Associate at the World Bank Suryadarma was previously the Director of Operations at PGE 2 a 100% steam-saturated but otherwise moderate- is statistically less than seen in many countries, the temperature reservoir or from a “steam cap”; and most probable value being in the range of $300,000 (d) greater than 27 MW representing “ultra-high” to $400,000 per MW. The average drilling success temperature and/or highly permeable reservoirs. rate and well capacity in Indonesia show a clear The last two types of wells, which occur in relatively “learning curve” effect; both success rate and well few countries other than Indonesia, represent 40% capacity tend to increase with the number of wells of the 215 wells studied in this paper. drilled, eventually reaching a plateau. Geothermal wells in Indonesia are mostly in the Given that (a) more fields with large resource 1,000 to 2,800 m  depth range and the drilling bases are encountered in Indonesia than in most success rate ranges from 63% to 73%, which are other countries, (b) well capacity in  Indonesia also typical in most countries. Using an existing typically is larger than in other countries, and (c) correlation of drilling cost versus well depth from drilling cost per well in Indonesia is smaller than in several countries, and the defined statistics on well most countries, the data indicates that the overall depth and productivity in Indonesia, we estimate resource risk in geothermal projects in Indonesia that the cost per MW well capacity in Indonesia should be lower than in other countries. Figure 1:  Indonesia’s Geothermal Prospects 95° 100° 105° 110° 115° 120° 125° 130° 135° 140° 15° 15° INDONESIA CAMBODIA MYANMAR INDONESIA VIETNAM ENERGY GEOTHERMAL POTENTIAL (MWe): SPECULATIVE = 9,533 MWe PHILIPPINES HYPOTHETICAL = 4,475 MWe PROBABLE = 10,317 MWe 10° 10° POSSIBLE = 728 MWe PROVEN = 2,305 MWe THAILAND Sulu TOTAL = 27,358 MWe Sea NATIONAL CAPITAL INTERNATIONAL BOUNDARIES A BRUNEI L Y DARUSSALAM 5° A S 5° M Natuna Besar I Celebes Talaud Is. Sea PACIFIC O CEAN A Simeulue Morotai SINGAPORE 1,961 534 Nias Halmahera 0° Lingga K A L I M A N TA N Waigeo 0° 13,778 Me Biak Siberut 50 Peleng Obi nt Bangka SULAWESI aw Misool Yapen SUMATRA Sula Is. 50 ai Ceram Is Belitung Buru PAPUA NEW GUINEA . PAPUA 5° Java Sea Muna Kai Banda Is. Enggano JAKARTA Sea Aru Is. Madura 9,304 JAWA 1,681 Alor Wetar Moa Babar Tanimbar Is. Bali Sumbawa Flores 0 200 400 Kilometers Lombok Sumba TIMOR-LESTE Arafura Sea 10° 10° Timor 0 100 200 300 400 Miles This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank INDIAN OCEAN IBRD 41049 Group, any judgment on the legal status of any territory, or any GSDPM AUSTRALIA JULY 2014 endorsement or acceptance of such boundaries. Map Design Unit 95° 100° 105° 110° 115° 120° 125° 130° 135° 140° Speculative Hypothetical Probable Possible Proven Total (MWe) (MWe) (MWe) (MWe) (MWe) (MWe) Sumatra 5,530 2,353 5,491 15 389 13,778 Kalimantan 50 — — — — 50 Sulawesi 900 125 761 110 65 1,961 Maluku 275 117 142 — — 534 Java 2,363 1,521 2,980 603 1,837 9,304 Papua 50 — — — — 50 Bali-Nusa Tenggara 365 359 943 — 14 1,681 Source: Authors based on information from Ministry of Energy and Mineral Resources, Indonesia. 3 INTRODUCTION Figure 2:  Histogram of total resource base in geothermal Indonesia has perhaps the largest num- fields in Indonesia ber of  known geothermal fields and 14 most ambitious plans for accelerated 12 geothermal power development in the 10 world: 9,500 MW  by 2025. Figure 1 Frequency 8 is a map of the country with estimated geothermal potential in various islands. 6 However, there is a perception among 4 many geothermal developers and finan- 2 ciers outside Indonesia that the resource 0 related risk in geothermal development 0 100 200 300 400 500 600 700 800 in Indonesia is substantial. Yet, we find Proved + Probable + Possible Reserves (MW) no statistical basis for this perception, Source: GeothermEx, Inc. even though there are some 100 known geothermal fields and prospects in the country, and nearly 300 deep wells. In this section we ana- a. adequate resource base; lyze the resource risk elements in Indonesia in a quanti- b. adequate well productivity; tative way and compare them to the same resource risk c. acceptable drilling cost per well; and elements in other countries with existing or imminent d. benign fluid chemistry. geothermal development. There are at least four fundamental requirements We will analyze here the uncertainty associated with that must be satisfied for any commercial geothermal the first three of the above requirements in Indonesia development project; these are: compared to that in other countries. Figure 3:  Frequency distribution of total geothermal ADEQUACY OF resource base in Indonesia fields on a log-normal RESOURCE BASE probability plot Before acquiring a  geother- 3.0 mal development concession 800 2.8 700 in a country a developer needs 600 500 to  have a  reasonable under- 2.6 400 300 standing of the resource base Log10 (Resource Base) 2.4 200 available at  the site. In  many Reserve Base, R 2.2 countries this information (MW) 2.0 100 is  incomplete or  unavailable, 90 1.8 80 70 whereas in Indonesia the govern- 60 50 ment developer PT Pertamina 1.6 40 30 had in the past systematically 1.4 20 explored most of the prospec- 1.2 tive areas, drilled deep wells 1.0 10 in several fields and assessed 0.01 0.05 0.1 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.8 99.9 99.99 the proved, probable and pos- Frequency (%) sible resource base in nearly all prospects. Figure 2 presents Source: GeothermEx, Inc. 4 a histogram of the total resource base (proved + probable + Figure 4:  Cumulative frequency of proven + probable + possible) in  Indonesian fields possible resource base in Indonesia as estimated by Pertamina. The 1.0 resource base displays a wide 0.9 Cumulative Frequency range, from 10 MW  to nearly 0.8 0.7 800 MW. 0.6 As expected in any nation- 0.5 wide assessment of this type, the 0.4 Proven + Probable + Possible Reserves 0.3 histogram indicates the occur- 0.2 rence of  progressively more 0.1 fields with smaller reserves, 0 0 100 200 300 400 500 600 700 800 that is, an approximately log- MW normal distribution of resource Source: GeothermEx, Inc. base. To  verify this outcome, in Figure 3 we have plotted the logarithm of the total resource base as  a  function of  the fre- Figure 5:  Histogram of well capacity, Indonesian fields quency of occurrence on a nor- (including non-commercial wells) mal probability scale. A  clear 40% linear trend of data points on this 35% plot establishes that the dis- 30% tribution of the resource base 25% Frequency is  log-normal. Therefore, Per- 20% tamina’s assessment of resource 15% base values appears consistent, 10% and so the prospect inventory 5% developed by  Pertamina has 0% significant value for develop- 0 to 2 2 to 4 4 to 6 6 to 8 8 to 10 10 to 12 12 to 14 14 to 16 16 to 18 18 to 20 20 to 22 22 to 24 24 to 26 26 to 28 28 to 30 30 to 32 32 to 34 34 to 36 ers looking for project sites in the country. Such a consistent MW national inventory of resource base is available from relatively Source: GeothermEx, Inc. few countries trying to attract geothermal developers. Figure 4 is  a  plot of  the cumulative frequency 102 known fields amounts to about 16,000 MW. It is likely of  resource base as  a  function of  the MW  capacity that further exploration and drilling will reveal even more in Indonesian fields. This figure indicates that there is exploitable geothermal sites and increase the inven- 50% cumulative probability that a geothermal field will tory of the resource base known today. Therefore, there have a resource base exceeding 100 MW; which is an is a sufficient resource base in the country to justify the attractively large scale of development for commercial development target of the Government of Indonesia. development. It is unlikely that a foreign developer would be  interested in  developing a  field in  Indo- nesia unless it offered the prospect for developing ADEQUACY OF WELL CAPACITY at least 50 MW; Figure 4 shows that more than 70% of the Indonesian fields (more than 50 fields) exceed We have accumulated statistics on 215 deep geothermal this threshold. The occurrence of so many potentially wells in Indonesia, which represent a large majority of all available commercial project sites in a country is rare. geothermal wells in the country; most of the individual The total estimated resource base in Indonesia in the well data remain confidential to the developer. This is an 5 adequate database for analyzing Figure 6:  Histogram of well capacity, Indonesian fields the well capacity risk. Figure 5 (commercial wells only) presents a histogram of per-well 25% MW capacity in Indonesia based on the data from the 215 wells. 20% A well capacity of less than 2 MW, and in some cases 3 MW, Frequency 15% is considered an unsuccessful 10% or  non-commercial well and 5% is usually turned into an injec- tion well or  observation well, 0% or is abandoned. Figure 5 indi- 3 to 5 5 to 7 7 to 9 9 to 11 11 to 13 13 to 15 15 to 17 17 to 19 19 to 21 21 to 23 23 to 25 25 to 27 27 to 29 29 to 31 31 to 33 33 to 35 35 to 37 37 to 39 cates that about 38% of wells are less than 2 MW in capacity, MW and as such, non-commercial. Source: GeothermEx, Inc. Therefore, the drilling success rate in Indonesia to date is about Figure 7:  Cumulative frequency of well capacity, 62%. However, this database includes exploration wells, con- Indonesian fields (commercial wells only) firmation wells, development 1.0 wells and make-up wells; the 0.9 drilling success rate generally Cumulative Frequency 0.8 0.7 becomes higher as  a  project 0.6 moves from exploration through 0.5 the confirmation, development 0.4 and operational stages in suc- 0.3 0.2 cession. The average well suc- 0.1 cess rate in the development 0 and operational stages of the 1.0 10.0 50.0 MW oldest operating field in Indone- Source: GeothermEx, Inc. sia (Kamojang) has been about 73%. This level of development drilling success is typical in most Figure 8:  Histogram of commercial well capacity, countries with geothermal Indonesian fields, (post 1990) potential. Figure 6 presents a  his- 20% togram of MW capacity of all 18% commercially successful wells, 16% 14% which are defined here as wells 12% with a capacity of at least 3 MW. Freqency 10% Figure 6 indicates an approxi- 8% mately log-normal distribution 6% 4% with a clear peak (at 3 to 5 MW). 2% Figure 7 shows a plot of the 0% cumulative frequency of  well 3 to 5 5 to 7 7 to 9 9 to 11 11 to 13 13 to 15 15 to 17 17 to 19 19 to 21 21 to 23 23 to 25 25 to 27 27 to 29 29 to 31 31 to 33 33 to 35 35 to 37 37 to 39 capacity of  the commercial wells in Indonesia; the wells vary Capacity (MW) in capacity from 3 MW to more Source: GeothermEx, Inc. than 40 MW, with a  median 6 log-normal but has multiple modes: Mode 1 ( 3 to 5 MW), Mode 2 (7 to 9 MW), Mode 3 (15 to 19 MW) and Mode 4 (27 to 31 MW). This multi-modal distribution can be  explained by referring to the several classes of geothermal fields discovered over the last two decades in Indonesia. Mode 1 typically represents “tight,” marginal wells that can randomly occur in any geothermal system, and as such, are not unique to Indonesia. Mode 2 (7 to 9 MW) wells are often referred to as the “most likely” in Indonesia. For example, JICA and WestJEC (2007) and JICA and WestJEC (2009) refer to 8 MW as the typical capacity for wells in Indonesia. Mode 2 wells represent those Photo courtesy of Migara Jayawardena / World Bank. value of 9 MW. The mean, Figure 9:  Maximum well temperatures vs. MW capacity, median and maximum values Indonesian fields of well capacity in Figure 7 are 60 all higher than seen in  other countries. Worldwide, a geo- 50 MW Capacity per Well thermal well capacity of 4 to 40 6 MW  is most common. It  is 30 worthwhile investigating the reason for this difference 20 between the well capacity dis- 10 tribution in Indonesia and most 0 other countries. 200 220 240 260 280 300 320 340 360 380 Maximum Well Temperature ( ˚c) Geothermal wells have been drilled in  Indonesia Source: GeothermEx, Inc. continuously since the 1970s. Given that over the last few decades the state-of-the-art and experience in geothermal Figure 10:  Histogram of maximum well temperature, well drilling and field develop- Indonesian fields ment have improved worldwide, let us  consider the statistics 25% of  wells drilled in  Indonesia 20% only since 1990, when the pace Frequency of  drilling activity accelerat- 15% ed in Indonesia. A histogram 10% of the MW capacity of all wells in  Indonesia since 1990 that 5% are at least 3 MW in capacity 0% is shown in Figure 8. 140 to 150 150 to 160 160 to 170 170 to 180 180 to 190 190 to 200 200 to 210 210 to 220 220 to 230 230 to 240 240 to 250 250 to 260 260 to 270 270 to 280 280 to 290 290 to 300 300 to 310 310 to 320 320 to 330 330 to 340 340 to 350 350 to 360 360 to 370 370 to 380 This figure reveals that when the pre-1990 wells are MW ignored, the histogram of well productivity is  no longer Source: GeothermEx, Inc. 7 producing from typical high-temperature liquid-dominated wells, as found in many countries. However, given that the Mode 3 and Mode 4 wells have much higher commercial value, the assumption of an 8 MW well capacity for the analysis of  the resource risk in  Indonesia significantly overestimates this risk. Let us examine what underlying resource conditions the Modes 3 and 4 wells may represent. Figure 9 is a plot of the MW capacity versus tem- perature of all 215 wells. This figure indicates two clus- ters of wells, one in the 230°C to 250°C range and the other in the 300°C to 340°C range. It can be shown that most of the Mode 3 wells falling in the 230°C to 250°C range typically represent wells producing from a 100% saturated steam reservoir, and the Mode 4 wells falling Photo courtesy of Migara Jayawardena / World Bank. in the 300°C to 340°C range Figure 11:  Histogram of well depth, Indonesian fields represent ultra-high tempera- 16% ture liquid-dominated wells, 14% sometimes producing from a 12% “steam cap”. Mode 3 and Mode 10% 4 wells occur in relatively few Frequency 8% countries besides Indonesia. 6% This diversity in  the 4% MW  characteristics of  wells 2% in Indonesia makes the assess- 0% ment of the resource risk more 0 to 200 200 to 400 400 to 600 600 to 800 800 to 1000 1000 to 1200 1200 to 1400 1400 to 1600 1600 to 1800 1800 to 2000 2000 to 2200 2200 to 2400 2400 to 2600 2600 to 2800 2800 to 3000 3000 to 3200 3200 to 3400 3400 to 3600 challenging. For example, Figure 10 is a histogram of the Depth (m) temperatures of the 215 wells; this figure shows only two Source: GeothermEx, Inc. modes rather than four as seen in Figure 8, but similar to what is seen in  Figure 9. Figures 8 and 9 imply that Modes 3 and 4 are uniquely defined and Figure 12:  Well depth vs. MW capacity, Indonesian fields represent well capacities more 35 attractive than are found in all 30 but a few countries, whereas MW Capacity per Well 25 Mode 2 represents wells with 20 a broad range of temperatures 15 and productivity found in many countries, and Mode 1 rep- 10 resents generally tight and/ 5 or relatively low temperature 0 500 1000 1500 2000 2500 3000 3500 wells. Given that the Mode 3 Well Depth (meters) and Mode 4 wells together represent nearly 40% of  the Source: GeothermEx, Inc. 8 215 wells, we  conclude that on a nationwide basis, Indone- Figure 13:  Histogram of drilling cost per MW for sia represents an unusually high successful wells, Indonesian fields occurrence of  commercially 30% attractive wells. 25% 20% Frequency WELL DEPTH AND 15% COST 10% Depth is the main determinant 5% of the drilling cost of a geother- mal well. Typically drilling cost 0% 100 200 300 400 500 600 700 800 900 1,000 1,100 1,200 1,300 1,400 1,500 1,600 1,700 1,800 1,900 2,000 2,100 2,200 in any country increases expo- nentially with depth; see, for Drilling Cost ($1,000/MW) example, GeothermEx (2004). Therefore, the depth distribu- Source: GeothermEx, Inc. tion of wells in Indonesia is an important issue in the assess- ment of resource risk in the country. Figure 11 shows m represent mostly Mode 3 wells while the points clustered a histogram of the depths of the 215 wells indicating that around 2,500 m depth represent Mode 4 wells. Figure 13 geothermal well depth in Indonesia ranges from about is a histogram of the drilling cost per MW of successful 1,000 to 2,800 m, which is  also representative of  the wells in  Indonesia, based on  the correlation in  Geo- range found in most countries with geothermal fields thermEx (2004) and the MW capacity versus depth data (GeothermEx, 2004). in  Figure 12. This figure shows that the drilling cost per Although there are no detailed statistics, we know MW well capacity in Indonesia ranges from US$100,000 from experience that drilling cost in Indonesia is some- to  nearly $2,000,000 the most probable range being what lower than seen in most countries. We have already $300,000 to $400,000 per MW; this is significantly lower shown that drilling success rate in Indonesia is compa- than in most other countries. rable to that in other countries. Given that (a) the well depth and drilling success rate in Indonesia are comparable to  those in  other countries, (b) the drilling cost per meter Figure 14:  Average drilling success rate vs. number is somewhat lower in Indonesia of wells drilled in Indonesia than seen in other countries, 80 and (c) that well capacity is gen- Average Drilling Success Rate (%) erally higher in Indonesia than 70 in other countries, we conclude 60 that the cost per MW of well 50 capacity in Indonesia is statisti- 40 cally lower than encountered 30 elsewhere. 20 Figure 12 presents a plot 10 of  the MW  capacity versus 0 0 50 100 150 200 depth of the 215 wells. It shows Number of Wells Drilled two clusters. The points clus- tered around a depth of 1,500 Source: GeothermEx, Inc. 9 of the wells on which we have Figure 15:  Average MW per well vs. number of wells no data) in Indonesia. drilled in Indonesia Clearly, the success rate 12 in  drilling has fluctuated but steadily increased as more wells 10 Successful Wells have been drilled, until a stable Capacity per Well (MW) 8 average success rate of about 62% was reached (after about 6 90 wells), when the learning 4 curve effect reached a plateau. All Wells As mentioned before, the data- 2 base in Figure 14 includes wells 0 in all stages of geothermal proj- 0 20 40 60 80 100 120 140 160 180 200 Number of Wells Drilled ects. Since the average drill- ing success rate progressively Source: GeothermEx, Inc. increases through the explora- tion, confirmation, develop- ment and operational stages THE LEARNING CURVE EFFECT of a project, a drilling success rate of 62% is an underesti- ON RESOURCE RISK mate for the development and operational stages, when the majority of the wells are drilled in a project. Geothermal resource development has been going on in Figure 15 is a plot of the average production capacity Indonesia for more than three decades. Over this long of all wells drilled and the average production capacity period the risk should have lessened due to improve- of only the successful wells drilled in Indonesia as a func- ments in technology and/or experience gained in avoid- tion of the total number of wells drilled. Figure 15 shows ing or mitigating such risk. Therefore, we examine to what that the average capacity of the successful wells reached extent the learning curve effect may have positively a plateau of 9.8 MW after about 200 wells were drilled. influenced some elements of resource risk in Indonesia. Figure 14 shows that the drilling success rate plateaued Figure 14 is a plot of the average drilling success versus at 62% after 200 wells were drilled. Therefore, the average the number of wells drilled (not including some 20% capacity of all wells drilled should have stabilized at 9.8 MW x 0.62, that is, 6.1 MW after 200 wells; this is confirmed by the lower curve in Figure 15, Figure 16:  Average drilling success rate vs. number implying that the statistics on of wells drilled in the Kamojang field, Indonesia the learning curve effect are internally consistent. 80 Figure 15 shows that, the Average Drilling Success Rate (%) 70 capacity per well increased until 60 it reached a plateau after about 50 40 wells were drilled. Then, after 40 about 140 wells were drilled, 30 a steady increase in the aver- 20 age well capacity ensued again, 10 even though the well success 0 rate remained nearly constant 0 10 20 30 40 50 60 70 80 at about 62% (Figure 14) as fur- Number of Wells Drilled ther wells were drilled. This Source: GeothermEx, Inc. apparent discrepancy stems 10 from the fact that over the last decade more and more countries. The resource base (proved-plus- probable- wells of Mode 2 and Mode 3 have been drilled. Figure 15 plus-possible) at a site ranges from 10 MW to 800 shows a peak in the average capacity of successful wells MW with a log-normal distribution. More than 70% of 9.8 MW, whereas the average capacity of  all wells of the known Indonesian fields have a resource base reached 6.1 MW, again reflecting an unchanged drilling greater than 50 MW, and at least half of the fields success rate of 62%. Therefore, although the learning (about 40) offer a resource base of 100 MW or more. curve effect on drilling success rate plateaued at 62%, the • Commercial wells in Indonesia vary in capacity from average MW capacity per well kept on increasing because 3 MW to more than 40 MW, with a median value of developers began tapping more Mode 3 and Mode 4 9 MW; this range and median are larger than seen wells. In other words, the assumption that a typical well in most countries while we believe the most likely in Indonesia is slightly better than the 4 to 6 MW seen range of well productivity worldwide is 4 to 6 MW. in other countries is unwarranted; it is substantially better • The commercial wells in Indonesia tend to fall in one (the average nationwide being about 10 MW) and is likely of four groups in terms of capacity: (a) 3 to 5 MW rep- to climb even higher as new projects begin tapping Mode resenting “tight” wells; (b) 7 to 9 MW representing 3 and Mode 4 wells. “typical” wells; (c) 15 to 19 MW representing wells The learning curve effect on the geothermal resource that usually produce from a 100% steam-saturated but risk is best illustrated by considering a specific geothermal otherwise moderate-temperature reservoir or from a field rather than Indonesia as a whole. The Kamojang “steam cap”; and (d) greater than 27 MW representing field in Indonesia is the best example for this illustration “ultra-high” temperature and/or highly permeable because it has the longest exploration, development and reservoirs. The last two types of wells, which occur production history in Indonesia (since the 1970s); see Sanyal in relatively few countries other than Indonesia, rep- et al (2000) and Suryadarma et al (2010). Figure 16 shows resent 40% of the 215 wells studied in this paper. the average drilling success rate at the Kamojang field • Geothermal wells in Indonesia are usually 1,000 to as a function of the number of wells drilled. Again, the 2,800 m in depth and the drilling success rate appears drilling success rate increases with drilling until a plateau to range from 63% to 73%, which are typical also at a drilling success rate of 73% is reached after about in most countries. Using a correlation of drilling cost 40 wells. versus well depth from several countries, and the statistics on well depth and productivity in Indone- sia, we estimate that the cost per MW well capacity CONCLUSIONS in  Indonesia is  statistically less than seen in  most countries, the most probable value being in the range • PGE’s geothermal resource base assessments are of $300,000 to $400,000 per MW. consistent; such a thorough national inventory of the • The average drilling success rate and well capacity geothermal resource base is available from very few in Indonesia show a clear “learning curve” effect; both success rate and well capacity tend to increase with the number of wells drilled, eventually reaching a plateau. • This analysis also implies that there is a sufficient geothermal resource base in Indonesia and that it is possible to scale-up as proposed in the GoI devel- opment targets. However, it will be essential to have Photo courtesy of GeothermEX Inc. an  adequate policy framework that will be  seen as credible by developers and enhance the investment climate, especially if the private sector is expected to play a major role in the proposed expansion. 11 REFERENCES GeothermEx, 2004. New Geothermal Site Identification and Quantification. Report prepared for Public Inter- est Energy Research (PIER) program of the California Energy Commission, April 2004. Available from: http:// www.energy.ca.gov/pier/project_reports/500–04–051. html. JICA and WestJEC, 2007. Master Plan Study for Geothermal Power Development in The Republic of Indonesia. Final Report, September 2007. JICA and WestJEC, 2009. Study on Fiscal and Non-Fiscal Incentives to Accelerate Geothermal Energy Develop- ment by Private Sector in The Republic of Indonesia. Interim Report, January 2009. Public-Private Infrastructure Advisory Facility (PPIAF), 2010. An Assessment of Geothermal Risks in Indonesia. Photo courtesy of GeothermEX Inc. Prepared by GeothermEx, Inc., for The World Bank, June 2010. Sanyal, S.K., A. Robertson-Tait, C.W. Klein, S.J. Butler, J.W. Lovekin, P.J. Brown, S. Sudarman and S. Sulaiman, Suryadarma, T. Dwikorianto, A.A. Zuhro and A. Yani, 2010. 2000. Assessment of Steam Supply for the Expan- Sustainable Development of the Kamojang Geother- sion of Generation Capacity from 140 to 200 MW, mal Field. Geothermics, Vol. 39, No. 4, December Kamojang Geothermal Field, West Java, Indonesia. 2010, pp. 391–399. Proceedings World Geothermal Conference 2000, pp. 2195–2200. The publication and dissemination of this paper is made possible due to the financial support from the Asia Sustainable and Alternative Energy Program (ASTAE), which is a multi-donor trust fund program administered by the World Bank. This paper was initially presented at Stanford University’s Workshop on Geothermal Reservoir Engineering, in 2011; and is based on a larger study conducted by GeothermEx for the World Bank (PPIAF, 2010) to assist the Gov- ernment of Indonesia with geothermal development. This work contributed to a program of support that is being provided by the World Bank Group to Indonesia where it is supporting the world’s largest geothermal expansion. The results of this work was also presented at an ASTAE-sponsored special session at The World Geothermal Congress 2010 held in Bali, Indonesia (27 April, 2010). The authors gratefully acknowledge the funding for this work by the Public- Private Infrastructure Advisory Facility (PPIAF) and ASTAE. The authors thank the Ministry of Energy and Mineral Resources of the Republic of Indonesia for their collaboration during the preparation of the study.