Unlocking Indonesias Geothermal Potential

UNLOCKING
INDONESIA'S
GEOTHERMAL
POTENTIAL
ASIAN DEVELOPMENT BANK
Unlocking
Indonesias
Geothermal
Potential
Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO)

2015 Asian Development Bank and The World Bank
The World Bank
1818 H Street, NW, US
Washington, DC 20433
Tel +202 473 1000; Fax +202 477 6391
www.worldbank.org; https://openknowledge.worldbank.org
Asian Development Bank

6 ADB Avenue, Mandaluyong City
1550 Metro Manila, Philippines
Tel +63 2 632 4444; Fax +63 2 636 2444
www.adb.org; https://openaccess.adb.org
Some rights reserved. Published in 2015.
Printed in the Philippines.
ISBN 978-92-9254-901-5 (Print), 978-92-9254-902-2 (e-ISBN)
Publication Stock No. RPT146821-2
Cataloging-In-Publication Data
Asian Development Bank and The World Bank.

Unlocking Indonesias geothermal potential
Mandaluyong City, Philippines.
1. Geothermal energy.
2. Indonesia.
3. Energy economics.
I. Asian Development Bank.
Asian Development Bank. The views expressed in this publication are those of the authors and do not necessarily reflect the views and
policies of the Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee
the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The mention of
specific companies or products does not imply that they are endorsed or recommended by ADB.
By making any designation of or reference to a particular territory or geographic area, or by using the term country in this document,
ADB does not intend to make any judgments as to the legal or other status of any territory or area.
The World Bank. This work is a joint product of the staff of The World Bank with external contributions. The findings, interpretations,
and conclusions expressed in this work do not necessarily reflect the views of the World Bank, its Board of Executive Directors, or
the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries,
colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank
concerning the legal status of any territory or the endorsement or acceptance of such boundaries.
Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank
or the Asian Development Bank all of which are specifically reserved.
Open Access. This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO)
https://creativecommons.org/licenses/by/3.0/igo/. By using the content of this publication, you agree to be bound by the terms of
said license as well as the Terms of Use of the ADB Open Access Repository at https://openaccess.adb.org/termsofuse and that of
The World Bank Open Access Repository at https://openknowledge.worldbank.org/terms-of-use
This CC license does not apply to non-ADB or non-World Bank copyright materials in this publication. If the material is attributed
to another source, please contact the copyright owner or publisher of that source for permission to reproduce. Neither ADB nor
The World Bank can be held liable for any claims that arise as a result of your use of the material.
AttributionIn acknowledging ADB and The World Bank as the source, please be sure to include all of the following information:
Asian Development Bank and The World Bank. 2015. Unlocking Indonesias Geothermal Potential. Asian Development Bank
and The World Bank. https://openaccess.adb.org; https://openknowledge.worldbank.org. Available under a CC BY 3.0 IGO license.
TranslationsAny translations you create should carry the following disclaimer:
Originally published by the Asian Development Bank and The World Bank in English under the title Unlocking Indonesias Geothermal
Potential 2015 Asian Development Bank and The World Bank. All rights reserved. The quality of this translation and its coherence with
the original text is the sole responsibility of the translator. The English original of this work is the only official version.
AdaptationsAny translations you create should carry the following disclaimer:
This is an adaptation of an original Work 2015 Asian Development Bank and The World Bank. The views expressed here are those
of the authors and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they represent
nor of the World Bank, its Board of Executive Directors, or the governments they represent. Neither ADB nor the World Bank endorse this
work or guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use.
Please contact OARsupport@adb.org or publications@adb.org if you have questions or comments with respect to content, or if you wish
to obtain copyright permission for your intended use that does not fall within these terms, or for permission to use the ADB logo. For use
of the World Bank logo, please contact The World Bank at pubrights@worldbank.org
Notes:
In this publication, $ refers to US dollars.
Cover photo of Wayang Windu geothermal power plant in West Java, Indonesia by Star Energy.
Contents
List of Tables, Figures, and Boxes
vi
List of Stakeholder Comments
ix
Abbreviationsx
Currency and Units
xii
Forewordxiii
Acknowledgmentsxiv
Executive Summary
1 Introduction
1.1 Background
1.2 Objectives
1.3
Defining the Geothermal Resource
1.4 Geothermal Targets
1.5 Scope
2 Tariff Design
2.1 General Principles of Tariff Design
2.2 Fixed Tariffs or Competitively Bid Tariffs
2.3 Ceilings
2.4 Production Costs or Benefits as a Basis for Ceiling Prices?
3 Implementation Issues and Procedures
3.1
3.2
3.3
3.4
3.5
Tariff Ceilings and Escalation

Transparency
Recovery of Incremental Cost
Tariff Setting as a Process
Conclusions on Implementation Procedure
4 Tariff Ceilings Based on Benefits

4.1
4.2
4.3
4.4
4.5
4.6
4.7
Alternatives to Geothermal
The Components of Avoided Costs
Macroeconomic and Global Energy Price Forecasts
Avoided Fixed Costs
Avoided Variable Costs
Avoided Global Externalities
Local Environmental Externalities
xv
1
1
2
2
3
9
10
10
11
13
15
18
18
18
19
21
21
22
22
24
24
25
26
29
33
iii
ivCONTENTS
4.8
4.9
4.10
4.11
5
Premium for Price Volatility

Avoided Cost of Transmission
Local Economic Development Benefits
Proposed Tariff Ceilings
Power Purchase Agreements
46
5.1
Escalation and Indexation
5.2 Renegotiation of Power Purchase Agreements
5.3 Adjustment for Project Size
5.4 Adjustment for Delay
5.5 Procedure
46
48
50
51
58
6 Recovery of Incremental Costs

6.1
6.2
6.3
Cost of Existing Projects

Incremental Costs of Future Projects
Impact on Ministry of Finance
7 Tendering
7.1
7.2
7.3
The Issues
Evaluation of Past Tenders
Options for Improving the Tender Process
8 The Tariff Impacts of Front-End De-risking

8.1
8.2
8.3
Cost Recovery Options for the Fund

Sensitivity Analysis
Conclusions and Recommendations
9 Institutional and Financing Issues
10
36
40
41
43
59
59
60
65
67
67
68
69
73
75
79
81
83
9.1 Structural Issues

9.2 The Role of the State
9.3 Pertamina Geothermal Energy Commercial Partnerships
9.4 Finance
9.5 Geothermal Risk Mitigation
9.6 Technology Paths to Faster Development
83
84
86
88
92
101
Unlocking the Sector: Study Team Recommendations
102
10.1 General Tariff Approach

10.2 The Tariff Process
10.3 Tariff Ceilings
10.4 Calculation of Benefits
10.5 De-risking of Projects
10.6 Power Purchase Agreements
10.7 The Ministry of Energy and Mineral Resources Geothermal
Production Cost Model
102
102
103
103
104
104
104
CONTENTSv
10.8
10.9
10.10
10.11
10.12
10.13
10.14
10.15
10.16
Power Purchase Agreement Renegotiation Policy

Transmission Connections
Improvements to the Tendering Process
Performance Bonds
The New Tender Entity
Measuring Resources
Binary Bottoming Plants
Other Required Supporting Actions of Government
Recommendation for Further Study
105
105
105
106
106
106
107
107
107
Appendixes
1 Reporting Codes
108
2 Transmission Interconnection Costs
113
116
Project Review
4 Tariff Structure and Project Finance
125
Background125
Financial Analysis at International Bank of Reconstruction and Development and ADB
126
Financial Analysis at Nominal Prices
128
Bankable Projects
132
Conclusions133
Front-Loaded Tariffs in Other Countries
134
5
Possible Technology Paths to Faster Development
136
Use of Ancillary Plant

Bottoming Binary Plants
Opportunities for Technical Improvements in Planned Developments
Using Condensing Steam Plant
136
140
144
6 Cost of Exploration
146
7 The Geothermal Tariff Issuance of June 2014
148
Key Provisions of Ministry of Energy and Mineral Resources Regulation No. 17/2014
Key Differences Between Ministry of Energy and Mineral Resources Regulation
No. 17/2014 and ADBWorld Bank Recommendations
8 The New Geothermal Law 2014
148
149
151
List of Tables, Figures, and Boxes

Tables
1.1
Castlerock Reassessment of Geothermal Work Area Potentials
2.1 The 2012 Geothermal FIT
2.2 FITs for Geothermal Energy
2.3 Bid Prices and Ceilings in Peru Renewable Energy Auctions
2.4 Size Classifications
2.5 Comparison of Approaches
4.1 World Energy Outlook 2013, Fuel Price Forecasts (at Constant 2012 Prices)
4.2 Macroeconomic and Fuel Price Forecasts
4.3 Fixed Costs (Investment Recovery and Fixed Operations and Management)
4.4 Variable Costs
4.5 Impact of Greenhouse Gas Valuations
4.6 Estimates of Local Health Damage Costs in Indonesia as Compared with the
Peoples Republic of China
4.7 Damage Cost of Local Air Pollutants
4.8 Forecast Errors
4.9 Local Multiplier Impacts (for Java and Sumatra)
4.10 Proposed Tariff Ceilings
4.11 Sample Calculation, Tariff Ceiling Adjustment
5.1
Classification of Costs
5.2 Operation and Maintenance Shares as Function of Discount Rate
5.3 Escalation of Cost Categories
5.4 Drilling Index Deflators (Relative to Current 2014 Prices)
5.5 Manufacture Unit Value Index Spreadsheet
5.6 Manufacture Unit Value Index (Index Values)
5.7 Adjusted Costs
5.8 Allowable Tariff Increase
6.1 Tariffs at Existing Geothermal Projects
6.2 Impact of Ceiling Prices on Ministry of Finance Subsidy (Java and Sumatra)
6.3 Impact of Subsidy on Assumptions (Java and Sumatra)
6.4 Subsidy Requirements to Meet Targets for Fast Track Program 2
8.1 Cash Flows for Recovery of Exploration Costs
8.2 Exploration Costs Funded by Geothermal Fund
8.3 Tariff Impact of $30 Million Exploration Program Funded by Developers
8.4 Cash Flows, Cost Recovery at Tender
8.5 Cash Flows, Cost Recovery at Financial Closure
8.6 Tariff Impact of $30 Million Exploration, 220 MW Project
8.7 Tariff Impact of $30 Million Exploration Program, Costs Recovered at
Financial Closure at 14% Weighted Average Cost of Capital
8.8 Impact of Developer Efficiency
8.9 Impact of Fee for Unsuccessful Exploration Programs
9.1 Summary of Geothermal Exploration Support Schemes Elsewhere
A2.1 Transmission Interconnections, Geothermal Projects
A2.2 Transmission Interconnections (500 kV), Large Coal Projects
A3.1 Project Status
vi
4
12
13
14
16
17
24
25
26
27
29
35
35
39
43
44
45
48
48
51
53
54
55
56
56
59
63
64
65
73
74
75
76
77
78
78
79
78
98
113
115
118
List of Tables, Figures, and Boxesvii
A3.2 Revised Targets

A4.1 Sample Project Financial Returns (at Constant Prices)
A4.2 Illustrative Weighted Average Cost of Capital Calculation
A4.3 Construction Disbursements at Nominal Prices
A4.4 Developer Cash Flows, at Base Tariff of 7.5 US/kWh and Equity Pari Passu
A4.5 Construction Period Disbursements, Equity Up Front
A4.6 Sensitivity of Tariffs to Financial Structure (Typical Project Financed with
International Finance Institution Concessionary Debt)
A4.7 Three-Tier Tariff Rates in Sri Lanka
A6.1 Expenditures Required through Financial Closure and Milestone MWs
A7.1 New Ceiling Prices
124
126
127
128
129
130
132
134
147
148
Figures
1.1
Geothermal Supply: Sumatra
1.2 Spatial Distribution of Geothermal Projects
1.3
Status of Geothermal Working Areas
1.4 Geothermal Projects by Tariff Status
2.1 Production Cost Estimates
2.2 Production Cost Based Tariff, Size Classifications
2.3 Recommended Form of the Production Cost Function
4.1 Nitrogen Oxide Damage Costs versus Per Capita Gross Domestic Product,
European Union
4.2 World Oil Price
4.3 Australian Coal Prices
4.4 Kalimantan Coal Prices
4.5 Heavy Fuel OilCoal Price Differentials
4.6 Gas Prices
4.7 Coal Price Forecasting Error
5.1
Geothermal Well Drilling Costs
5.2 Global Steel Price Trends
6.1 Castlerock Supply Curve, Java and Sumatra
6.2 Impact of a 12.5 US/kWh Ceiling Price
6.3 Adjusted Cost Curves
6.4 Estimates of Future Public Service Obligation Subsidy
7.1
Bid Price versus Project Size and Number of Bidders
9.1 Multilateral Investment Guarantee Agency Guarantee to Offshore Lenders:
Indonesia Hydro Project
A2.1 Transmission Cost in $/kW versus Installed Capacity
A4.1 Typical Indonesian Geothermal Tariff
A4.2 Equity Return versus Tariff
A4.3 Impact of Equity Disbursement Timing
A4.4 Impact of Debt/Equity Ratio (Equity Up Front)
A4.5 Bankable Project Definition
A4.6 Non-Concessionary Finance
A4.7 Three-Tier Front-Loaded Tariffs in Sri Lanka
A5.1 Back Pressure Steam TurbineStand Alone Operation
A5.2 San Jacinto Project, Nicaragua: 2 x 5 Megawatt Back Pressure Plant
A5.3 Schematic of Condensing Steam Turbine Plant
A5.4 McLachlan Plant, Wairakei, New Zealand: 55-Megawatt Condensing Steam
A5.5 Time from First Exploration Well to Commercial Operation Date
A5.6 Bottom Binary Plant Operating on Separated Water
A5.7 Bottoming 17-Megawatt Binary Plant at Wairakei, New Zealand
6
7
7
8
16
16
17
33
36
36
37
37
38
38
52
53
60
61
62
65
69
91
114
125
130
131
131
132
133
135
137
137
138
138
139
142
142
viiilist of tables, figures, and boxes

Boxes
1
Economic Quantity of Renewable Energy
2
Global Experience with FITs
3
Examples of Renewable Energy Tariff Transparency
4
Depletion Premium for Indonesian Coal
5
Global Social Cost of Carbon
6
Ministry of Finance Green Paper
7
Damage Cost Estimates for Indonesia
8
Impact of Public Funding of Up-Front Exploration in Mexico
9
Proposal for the Organizational Structure for Exploratory Drilling
10 The Geothermal Energy Development Program (Europe) and African Rift Geothermal
Development Program Schemes
11
Private Geothermal Insurance Schemes
A1.1 Specific Issues in Current Version of Indonesia Standard
3
11
20
28
30
32
34
80
94
96
97
111
List of Stakeholder Comments

1 Why Is Geothermal Energy Not Also Benchmarked Against Gas
Rather Than Only Coal?
23
Depletion Premium
27
Greenhouse Gas Valuation
31
Why Use Deterministic and Not Probabilistic Calculations?
43
The Tender Process
67
Use of the Geothermal Fund
91
A1.1 Use of International Code
108
ix
Abbreviations
3G
geology, geophysics, and geochemistry
ADB
Asian Development Bank
ARGeo
African Rift Geothermal Development Program
BBP
binary bottoming plant
BPP
back pressure plants
CV
calorific value
CCT
clean coal technologies
CER
Certified Emission Reduction
COD
commercial operation date
CTF
Clean Technology Fund
DSCR
debt service coverage ratio
FIRR
financial internal rate of return
FIT
feed-in tariff
fob
free on board
FTP
Fast Track Program
FTP1
Fast Track Program 1
FTP2
Fast Track Program 2
GDE
Geo Dipa Energi
GDP
gross domestic product
GeoFund
Geothermal Energy Development Program
GHG
greenhouse gas
GSCC
global social cost of carbon
HFO
heavy fuel oil
IEA
International Energy Agency
IFC
International Finance Corporation
IFI
international financial institution (Asian Development Bank, World Bank)
INAGA
Indonesia Geothermal Association
IPP
independent power producer
IUP
Izin Usaha Pertambangan (Mining Business Permit)
JBIC
Japan Bank for International Cooperation
Abbreviationsxi
JICA
Japan International Cooperation Agency
LNG
liquefied natural gas
MEMR
Ministry of Energy and Mineral Resources
MFF
multitranche financing facility
MIGA
Multilateral Investment Guarantee Agency
MoF
Ministry of Finance
MSOE
Ministry of State-Owned Enterprises
MUV
manufacture unit value (index)
NPV
net present value
O&M
operation and maintenance
OPEC
Organization of the Petroleum Exporting Countries
PGE
Pertamina Geothermal Energy
PIP
Pusat Investasi Pemerintah (Indonesia Investment Agency)
PLN
PT Perusahaan Listrik Negara (State Electricity Company)
PPA
power purchase agreement
PSO
public service obligation
RUPTL
Rencana Usaha Penyediaan Tenaga Listrik (Electricity Power Supply Business Plan)
SAGS
steam above ground system
SOE
State-Owned Enterprise
WACC
weighted average cost of capital
WEO
World Energy Outlook (of the IEA)
WKP
wilayah kerja pertambangan (geothermal work area)
Currency and Units

All references to dollars and $ refer to United States dollars
Except where noted otherwise, $1 = Rp11,500; 1 = Rp115
All references to tons refer to the metric ton (of 1,000 kg)
BTU
British thermal unit
GJ gigajoule
GW gigawatt
GWh gigawatt-hour
kJ kilojoule
kW kilowatt
kWh kilowatt-hour
MW megawatt
MWe
megawatt electric
MWh megawatt-hour
xii
Foreword
Geothermal energy represents one of the key options for Indonesia to achieve a comprehensive
approach to national energy development. The rapid increase in fossil-fuel based energy consumption,
which is subject to volatility in the world oil market, is the main challenge facing the countrys energy
supply. At the same time, growing greenhouse gas emissions from the use of fossil fuels imposes costs
on the economy and society. Geothermal energy provides one solution to these issues. It is a source
of clean, renewable and environmentally friendly energy for power generation. Furthermore, as an
indigenous and non-tradable energy source, it will enhance the countrys energy security by serving as
a natural hedge against the fluctuations of global fossil fuel prices. The Government of Indonesia has
recognized the role of geothermal energy and has put major efforts into promoting its development
with initiatives such as the Roadmap of Geothermal Development 20122025, the National Energy
Policy 2014, the issuance of a new geothermal tariff in 2014 and the Geothermal Law No. 21 of 2014.
Participation from all stakeholders, public sector as well as private sector, is essential to raise awareness
of the role of geothermal power in the national energy strategy.
This publication, Unlocking Indonesias Geothermal Potential, provides useful insights to policy
makers, investors, geothermal industry practitioners, and all geothermal development stakeholders.
The report identifies the main issues that are hindering geothermal power development and reviews
geothermal related policies and regulations in Indonesia. Its analyses and recommendations cover key
issues of the sectors development such as the geothermal tariff design, improvement in tendering
processes, Power Purchase Agreements and price renegotiation, institutional and other financing
issues.
I hope that all Indonesia geothermal stakeholders will benefit from the report. I believe that the report
provides valuable inputs toward acceleration of the future development of geothermal energy in
Indonesia.
Rida Mulyana
Directorate General of New, Renewable Energy and Energy Conservation
Ministry of Energy and Mineral Resources
xiii
Acknowledgments
This report is the result of a study by a joint team of the World Bank and the Asian Development
Bank (ADB). It is based on analysis and stakeholder consultations conducted from June 2013 to
June 2014. ADBs inputs were financed under ADB TA-7583-INO, Geothermal Power Development
Project, supported by the Government of Australias Department of Foreign Affairs and Trade. The
World Banks inputs were financed by the Energy Sector Management Assistance Program and the
Asia Sustainable and Alternative Energy Program. The work was initiated in response to a request
from the Government of Indonesias Ministry of Energy and Mineral Resources (MEMR) to revise its
geothermal pricing policy.
The authors of this report are Peter Meier, James Burke Randle, and James Vincent Lawless. The report
reflects the contributions of Muchsin Qadir (World Bank), Yuki Inoue (consultant, ADB), Maura Lillis
(consultant, ADB), Kazim Saeed (consultant, World Bank), Surya Darma (consultant, World Bank),
Olivia Tanujaya (World Bank), and Yerri Anullah (legal consultant, World Bank). The team would also
like to thank Richard Spencer (World Bank) for his helpful review. Pham Nguyet Anh (World Bank)
and Pradeep Tharakan (ADB) directed the study.
The team would also like to thank Rida Mulyana, Director General of New, Renewable Energy &
Energy Conservation (MEMR) for his guidance; Tisnaldi, Geothermal Director, and his team from the
Directorate of Geothermal (MEMR) for their cooperation during the study; and all other government
authorities, geothermal developers, and stakeholders who shared their perspectives and information
during the consultation meetings.
xiv
Executive Summary
This joint report of the Asian Development Bank (ADB) and World Bank presents a review of
Indonesias geothermal sector, prepared in the context of a request from the Government of
Indonesias Ministry of Energy and Mineral Resources (MEMR) for assistance with a planned revision.
The ministerial regulation for the new geothermal tariff was issued by MEMR in June 2014.
The 2012 feed-in tariff (FIT) was a first attempt to unlock the sector. But this FIT raised as many
new questions as it solved, and it is generally agreed that much more needed to be done to consult
with stakeholders than had been done before. For this reason, MEMR has engaged extensively with
stakeholders in the consultations for a new tariff issuance.
In our view, there are four main areas that need attention and that need to be addressed together.
Only concerted and coordinated action in all areas simultaneously will unlock the sector. The
underlying problem is really one of capital mobilization for a generating option that is unusually capital
intensive: just to achieve an additional 3,000 MW geothermal capacity in the foreseeable future will
require $4 billion in equity and $9.5 billion in debt finance (assuming $4,500/kW total cost, and 30%
equity). The problem of mobilizing equity is primarily one of the adequacy of tariffs to enable the upfront equity needed for explorationmuch more costly than in other countries where much of the
up-front exploration effort was funded as a pure public good.
A key problem for raising debt finance is that even the international financial institutions (IFIs)
(ADB, International Finance Corporation, World Bank/International Bank for Reconstruction and
Development [IBRD]) are reluctant to fund up-front exploration and typically will provide financing
only once 50% or more of the steam resource is proven. To date, targets for geothermal achievement
have not been set with full knowledge of the incremental costs of achieving them.
Clarifying the Role of the State

There are two major issues to be faced. The first is how best to resolve the competing interests of the
main state entities involved in the sector: the Ministry of Finance (MoF) is concerned about the
increasing size of the public service obligation (PSO) to PT Perusahaan Listrik Negara (PLN), Indonesias
state-owned electricity company, which it regards as increasingly unsustainable. Reducing the
magnitude of the PLN subsidy is its greatest concern in the power sector, which obviously conflicts with
the probable incremental costs of geothermal energy and the need to increase the subsidy to achieve
the geothermal targets. The Ministry of State-Owned Enterprise (MSOE) has as its main objective
the satisfactory commercial performance of its enterprises, which include both Pertamina and PLN.
Consequently the Pertamina Board of Directors (and in particular the director primarily responsible
for investment planning) are reluctant to allocate equity capital to PT Pertamina Geothermal Energy
(PGE) when compared to the much higher returns available in Pertaminas oil and gas plays. While an
objective comparison of the resource risk in oil and gas versus geothermal energy might well conclude
the risks are higher in oil and gas, these are more than offset by the very much higher returns. In short,
the regulated returns in the electricity sector are not commensurate with the risks of the geothermal
businesswith the result that PGE faces a continuous battle for resources from its parent company.
Finally, MEMR sees its role as the promoter of geothermal energy, and is responsible for supervising
the sectors development, including responsibility for implementing the Geothermal Law, and for tariff
setting. It is the entity primarily responsible for promoting geothermal energy, but whether its goals
can actually be methow to mobilize the equity and debt necessary to achieve successful project
developmenthas been left to others.
xv
xviexecutive summary
These differences in objectives have not been helped by poor communication between MEMR and
MoF in the matter of tariffs in the past. Indeed, basic principles of stakeholder consultation were
ignored in the issuance of the 2012 FIT. Fortunately, that lesson has been learned by MEMR, and its
efforts to consult with stakeholders in its new tariff issuance process has been exemplary.
The second issue is to clarify precisely the roles that each of the state entities currently active in
geothermal should play. Pertamina still owns many (legacy) concessions that remain undeveloped, yet
Pertamina is still bidding for new projects. Indeed, PGE was created by Pertamina to develop geothermal
energy projects, but as noted, is not provided with the equity funding necessary to successfully
develop the projects assigned to it. PT Geo Dipa Energi (GDE) was created in 2002 to develop some
specific projects, but it too was not funded with the adequate equity. PLN Geothermal has experience
with the power generation part of geothermal energy projects, but its role in developing particularly
the smaller geothermal energy projects in the eastern islands remains unclear. The management of
the Geothermal Fund has been given to an entity of the MOFbut it lacks the necessary technical
experience.
In this respect, we support the efforts of PGE to enter into partnerships with qualified private
developers as a way to bring in additional equity. However, this strategy will be successful only if the
previously negotiated prices for PGEs projects, set many years ago, and prior to the recent rises in
drilling costs, can be satisfactorily renegotiated.
Tendering
Competitive selection of private sector developers is mandated by the Geothermal Law. However,
notwithstanding the good intentions of the law in devolving the tender process to the provinces,
there is widespread concern that the tender process needs improvement. Particularly in the eastern
islands, the tender process has resulted in winning bids at prices that are so low that few believe
they can be implemented. In addition, many entities have won bids with insufficient technical and
financial capacity. The alleged deficiencies are many, including lack of technical capacity of the tender
committee (resulting in poor prequalification screening), bid bonds that are too small (so unqualified
bidders are not excluded), and performance bond requirements that are not imposed.
The principles that should apply to tendering have been well established in International Practice,
as exemplified by the procurement rules of ADB and the World Bank (Section 7). These should be
followed for Indonesia geothermal tenders. In particular, there should be a requirement to post a
significant bid bond (stipulated as a percentage of the total project cost rather than just a percentage
of the first year exploration program), of no less than $10 million. The winning bidders bid bond should
then be converted into a performance bond that can be released upon evidence of exploration drilling.
International experience demonstrates that the most effective way of improving the quality of a
tender process for exploiting a natural resource is to improve the quality of resource information
made available to bidders. No concession areawilayah kerja pertambangan (geothermal work
area)should be put to tender without a complete and independently certified package of geology,
geophysics, and geochemistry (3Gs). In the ideal case, and particularly in the eastern islands where
the larger developers have little interest in developing smaller projects, subsurface information should
also be provided, ideally with a minimum of three wells, with information presented to the standards of
an internally accepted resource code (see Appendix 1), and again, independently certified. Proposals
on how such a pre-tender exploration program should be organized, and the role of the Geothermal
Fund, are presented in Section 9 of the report.
The value of up-front de-risking as a public good is widely acknowledged in international geothermal
practice. Section 8 of the report presents an analysis of the impact of such de-risking on the tariff.
Executive Summaryxvii
We argue that if this up-front de-risking were provided by the Geothermal Fund, its costs should
be recovered from developers at the time of financial closure, at which point the weighted average
cost of capital is much less than that of privately provided risk equity in the early stages of project
development. Many of the principles that should govern an exploration program using public funds
have been presented before (by ADB, the Japan Bank for International Cooperation [JBIC], and
others). Here we provide further quantitative argument for doing so.
In the long term, Indonesia should establish a technically qualified, central tender entity to conduct
tenders on behalf of local governments. During the preparation of this report, Indonesias House of
Representative passed the Bill on Geothermal Energy as a revision to the previous Geothermal Law
No. 27 of 2003. One of the major changes in the new bill is that the geothermal concession tender and
issuance of geothermal license for power development will be carried out by the central government
(MEMR). The new bill assures the interests of the local governments through a production bonus
sharing when the power plant starts operation commercially as well as applicable local taxes. The key
point for such a centralized process is institutional longevity, which is essential to equitably regulate
a process where project gestation times can be 810 years. The central government will likely require
significant technical assistance, but the international finance institutions (IFIs) and bilateral donors
would certainly be interested in providing this. The question of how the size of the production
bonus will be determined by MEMR, and the mechanism for its recovery, will also require extensive
stakeholder consultation.
Tariff Reform
We recommend a return to the prior system of tender-determined tariffs. The 2012 FIT system
proposed in 2012 has the defect that with fixed prices, developers would then be selected on nonprice qualifications only (i.e., a beauty contest), which many developers oppose on grounds that
subjective evaluations are unreliable and unpredictable. If the improvements we recommend to
improve the tender process are adopted, then the system of competitively determined bid prices can
remain. However, the old ceiling price of 9.7 US/kWh needs revision. Indeed, any ceiling price set
for a tender bid today should not be based on what is an appropriate ceiling today, but what is an
appropriate ceiling for the date of commercial operation, which may be 79 years in the future.
We recommend that tariff setting be seen not so much as a one-time event, but as a process. This
again is proven by international best practice for the regulation of renewable energy tariffs: most
countries have a system of regular review of tariffs, based on a published methodology and stakeholder
consultation. These principles are elaborated in Sections 2 and 3.
Our recommendation for ceiling prices is provided in Section 4: they should be set on the basis of
the benefits of geothermal energy. Projects in which competitively bid costs exceed these benefits
should not proceed. We recommend a return to the avoided cost approach proposed by Castlerock in
2010, but with a more transparent process to translate these principles into a formal methodology for
forecasting a reasonable base price for projects whose commercial operation is 79 years away.
The benefits to Indonesia of increased geothermal energy are many. The first is the avoided costs of
PLN. But PLNs avoided costs are very different in the case of the big systems on Java and Sumatra
where the alternative is state-of-the-art coal (with high efficiency, state-of-the-art pollution controls),
than on the eastern islands, where the alternative is the diesel or small coal systems, sometimes less
than 25MW in size, but whose unit cost may be double that of an ultra-supercritical coal project on
Java. The second set of benefits relates to local regional economic development. It is one of the main
goals of government policy to encourage economic development in the eastern islands, for which
geothermal development provides an urgently needed contribution.
xviiiexecutive summary
The third set of benefits relate to the avoided externality costs of thermal generation, notably that
of avoided greenhouse gas (GHG) emissions. But this raises difficult questions of what value the
Government of Indonesia should place on avoided GHG emissions and the extent to which that value
may be higher than the current price in global carbon markets (an issue raised in the stakeholder
consultations). This highlights the need for agreement between the three ministries noted above, on
the incremental costs that Indonesia should be willing to pay.
In June 2014, MEMR issued a new geothermal tariff regulation (MEMR Regulation No. 17 of 2014)
based on the tariff study recommended by the World Bank/ADB. The discussion on the new tariff is
presented in Appendix 7.
Power Purchase Agreements and their Renegotiation

There are several aspects of the present power purchase agreement (PPA) process that could be
easily resolved. The present system of time-consuming, ad hoc, post tender negotiation of tariff
escalation terms should be dropped, and a single tariff escalation formula should be adopted for all
projects (consistent with international best practice for renewable energy projects). The prospective
PPA should be provided at the time of tender. All relevant terms and conditions, and particularly the
schedules relating to the tariff, should be fixed in advance.
In the short term, the question arises of how to unblock projects stalled for lack of an adequate tariff.
Developers argue that the inability to deliver a project on time and to the original tariff estimate has
often been the fault of the government, due to inordinate delays in permitting and resolution of land
issues. The proposed declassification of geothermal as a mining activity as one of the amendments to
the Geothermal Law will be helpful to new projects, but that will not help the currently stalled projects.
Therefore renegotiating some PPAs may be unavoidable if some of the currently stalled projects are
to move forward. Successful conclusion of the proposed PGE private partnerships will depend on
renegotiating prices established in some cases more than 10 years ago at levels that are no longer
reasonable in view of the rapid increases in drilling costs over the past 35 years.
Notwithstanding that we recognize that PPA renegotiation should be primarily a matter for the
contracting parties, i.e., between PLN and the developer, we believe there is benefit to MEMR in
issuing a policy statement that sets out the principles that should apply, and the circumstances under
which a renegotiation rather than cancellation should be consideredwhich should be limited to:
(i) delays attributable to the fault of government (to be established by an independent third party);
(ii) projects where delineation drilling after tender shows the project to be significantly larger, or
significantly smaller than estimated at tender, and (iii) projects for which capacity of individual units
was stipulated at time of tender, but where the developer subsequently wishes to install larger units
(e.g., build 1 x 110MW rather than 2 x 55MW as originally stipulated; choice of unit sizes should be
left to the developer at final design without penalty). Indeed, we argue that one of the reasons for the
new central tender entity is to facilitate any subsequent PPA renegotiation that may be attributable to
government delays.
The PPA should also clarify the arrangements for transmission connection. We recommend that, in
general, the developer build transmission line to the nearest PLN substation, recovering costs by a
non-escalating tariff adder (outside the bid tender price). These costs are a very small percentage
of total capital cost, and which are not material to selecting a qualified developer. PLN would take
over the line on the date of commercial operation of the generating project, and be responsible for its
maintenance.
Introduction
1.1 Background
Indonesia has abundant geothermal resources that can help meet the countrys rising electricity
demand and increase electrification rates. Indonesias estimated conventional hydrothermal
geothermal resource base is generally considered to be among the largest in the world. The
Government of Indonesia plans to achieve around 6,000 MW of installed geothermal power capacity
by 2020, a more than a fourfold increase of the end-2012 capacity of 1,335 MW. This ambitious plan
will require strong government support to materialize. Any shortfall in the expansion of geothermal
power generation capacity will most likely be met by additional coal-fired power plants.
Over the past decade, the government has intensified its efforts to scale-up and speed-up geothermal
power development.

In 2003, the Geothermal Law (Law 27/2003) was promulgated, making geothermal the only
renewable energy governed by its own law. The law mandated that future geothermal fields must
be transparently and competitively tendered for development. It also permitted operators of the
fields previously allocated to retain control of their assets. In 2004, the Ministry of Energy and
Mineral Resources (MEMR) issued the Blueprint for Geothermal Development in Indonesia,
which was intended as a roadmap to develop 6,000 MW of geothermal power capacity by
2020. In 2005, the Directorate of Geothermal Enterprise Supervision and Groundwater
Management were established by MEMR to strengthen sector management and support. This
became the Directorate of Geothermal Energy in November 2010. In 2006, MEMR initiated
the Master Plan Study for Geothermal Power Development in Indonesia funded by the Japan
International Cooperation Agency (JICA), further solidifying knowledge and understanding
about developing Indonesias geothermal resources.
In 2012, the MEMR issued a feed-in tariff (FIT) policy for geothermal electricity, based on
the analytic work supported by the World Bank and/or Global Environment Facility
Geothermal Power Generation Development Project.
In 2012, the Ministry of Finance (MoF) established a geothermal fund with more than
$200 million of initial capitalization to mitigate resource risks related to geothermal
development. The Asian Development Bank (ADB) provided early technical inputs on the
funds scope and design.1
Despite these efforts, progress in the last few years has been slow. The perception that the Indonesian
geothermal program has stalled is widespread, and exists among all stakeholders. From 20102013, just
135 MW was added, and best estimates suggest that by the end of 2016, no more than an additional
190 MW is likely.2 No power purchase agreements (PPAs) were signed under the 2012 FIT. A step
change in the pace of development for even 4,000 MW to be reached by 2020 is therefore required,
achievable only by a focused action program by government to resolve institutional, regulatory, and
tariff constraints.
AECOM, Geothermal Fund Report, Report to ADB, 2011; A. Wahjosoedibjo and M. Hasan. 2012. Geothermal Fund
for Hastening the Development of Indonesias Geothermal Resources. A paper presented to the 37th Workshop on
Geothermal Reservoir Engineering. Sanford University, California. January.
In 2014: Patuha, 55 MW. In 2015: Kamojang 5, 30 MW. In 2016: Ulubelu 3, 55 MW; Karaha, 30 MW; and Lahendong 5,
20 MW.
2Unlocking Indonesias Geothermal Potential
1.2 Objectives
The government has recognized these problems, and has proposed a series of actions to unblock the
sector, including:

amendments of the Geothermal Law (to declassify geothermal as a mining activity);
a revised tariff issuance; and
a new regulation on tendering.
To support the development of a new tariff approach, MEMR has requested assistance from the
World Bank (IBRD) and ADB. This report describes the findings of the team of technical experts, and
presents their recommendations. Our comments on the new June 2014 Geothermal Tariff Regulation
are provided in Appendix 7.
1.3 Defining the Geothermal Resource

Estimates of the magnitude of the available Indonesian geothermal resource vary greatly. The 2007
West Japan Engineering Consultants study estimated the exploitable potential across 50 fields at
9,000MW.3 In 2011, MEMR revised the countrys geothermal potential to 29,215 MW from 27,000MW
a decade earlier4indeed, the 27,000 MW figure is cited in many World Bank reports,5 and appears to
be the basis for claims that Indonesia possesses 40% of the worlds geothermal resources.6
The basis for these various estimates is unclear, for it is sometimes not fully appreciated that a resource
is only that portion of a natural occurrence (whether of energy, petroleum, or minerals) that can
feasibly and economically be extracted. Without the basis for such an estimate being made explicit,
including assumptions as to the technology pathway and power prices, resource estimates are of very
limited value. There is widespread perception that the estimate is too large, but no better estimate has
been made, so no one knows by how much.
At the level of an individual geothermal resource, it is important to not only have a good grasp of the
size of the resource for planning the development, but equally important to have the reliability of that
estimate quantified so that risks can be assessed and financing issues identified.
In 2010, Castlerock Consultants reassessed the more significant geothermal resources in Indonesia
on a consistent basis.7 It is clear from their analysis that some of the resource capacities were not only
initially overestimated, but have also had a tendency to be increased over time without new data or
justification being available. A more rational, systematic, and transparent methodology is needed.
This issue is, of course, not unique to the geothermal energy sector: a number of estimating and
reporting methodologies and codes have been developed in the petroleum and mineral industries,
often probabilistically based. Only recently has a similar approach been taken in geothermal practice.
Appendix 1 discusses the problem of reporting codes in detail.
3
West Japan Engineering Consultants. 2007. Master Plan Study for Geothermal Development in the Republic of Indonesia.
The Geological Agency (under MEMR) issues an annual Geothermal Area Distribution Map and the Geothermal
Potential in Indonesia.
See, for example, World Bank. 2011. Project Appraisal Report: Geothermal Clean Energy Investment Project.
Washington,DC.
J. Wilcox. 2012. Indonesias Energy Transit: Struggle to Realize Renewable Potential. Renewable Energy World.com.
14 September.
Castlerock Consulting. 2010. Phase 1 Report: Review and Analysis of Prevailing Geothermal Policies, Regulations and Costs.
Jakarta: Ministry of Energy and Mineral Resources.
Introduction3
A complicating factor with geothermal is that the value of the commodity producedelectricity
does not have an agreed international value as is the case for oil or mineral commodities. Therefore,
country and even site-specific factors have to be taken into account, including the cost of transmission.
Moreover, while petroleum or mineral resource estimates can be made on the basis of simple depletion,
geothermal systems may be recharged by heat and fluids during the exploitation of a resource.
1.4 Geothermal Targets

From the standpoint of economic analysis, the optimum quantity of geothermal energy that should be
in the energy mix is given by the intersection of the geothermal supply curve, and the avoided social
costs of thermal energy. These principles are illustrated in Box 1.
The avoided social cost is defined by PT Perusahaan Listrik Negara (State Electricity Company) (PLN)
production cost (for which the fuel input, including coal, is now valued at border prices), plus the
relevant externality cost, of which the largest component is the avoided cost of the cost of greenhouse
gas (GHG) emissions. The local health damage costs associated with local air pollutants such as
particulate matter less than 10 microns in diameter, sulphur oxides and nitrogen oxides are also a
component of the economic cost of fossil generation, but are quite small compared to GHG emission
damages (Section 4.7).8
Box 1: Economic Quantity of Renewable Energy
The economic rationale for renewable energy is straightforward: the optimum amount of renewable energy
for grid-connected generationwhich should be the basis for any targetis given by the intersection of the
renewable energy supply curve with the avoided cost of thermal electricity generation. If the price of coal to
PT Perusahaan Listrik Negara (State Electricity Company) (PLN) is subsidized (as it was in Indonesia until
recently), with PLNs avoided cost PFIN, then the optimal quantity of geothermal energy is QFIN. If the fuel
subsidy is eliminated, and the true economic cost of coal generation is PECON, then the optimal quantity
increases to QECON.
PRICE
If one also adds to the cost of coal generation

the global environmental damage cost , then
the true social cost increases further to PENV,
and the optimal quantity increases to QENV.
Obviously, the higher the damage cost , the
higher is the quantity that is justified to avoid
these environmental damage costs.
Uncertainty in the costs of geothermal projects
results in uncertainty in the supply curve: if
costs are higher than expected, the supply curve
shifts up, and the optimal quantities decrease.
Thus, the Castlerock report shows three supply
curves, defined by the variation in cost and
resource quality estimates (Figure 1.2).
P
P
GEOTHERMAL
ENERGY SUPPLY
CURVE
ENV
THERMAL ENERGY COST

Including ENV. DAMAGE
COST
ENV. DAMAGE COST,
ECON
FUEL SUBSIDY,
FIN
THERMAL ENERGY
FINANCIAL PRICE
QFIN
QECON
QENV
QUANTITY
QFIN = quantity supplied at PLNs avoided financial cost of

thermal energy PFIN, QECON = quantity supplied at Indonesias
avoided economic cost of thermal energy PECON, QENV = optimal
quantity supplied at Indonesias economic avoided cost including
environmental damage cost PENV.
Source: Authors calculations.
The Castlerock report estimated the levelized production cost of coal generation in the JavaBali grid at 6.3 US/kWh; the
local damage cost at 0.1 US/kWh, and the greenhouse gas (GHG) emission damage cost, based on $20/ton, at 1.4 US, i.e.,
14 times greater than the local damage cost. Currently, most World Bank project appraisals for geothermal projects use a value
of around $30/ton CO2, at which level the relative importance of local externalities are correspondingly smaller. The $30/ton
can be taken to be the World Banks current estimate of the global social cost of carbon (GSCC) used in economic analysis.
This is unrelated to the estimates of market prices (the Clean Development Mechanism [CDU] and the EU emissions trading
system [ETS]), which is currently much lower (and which should be used only in financial analysis).

Coal is the least-cost generation option for base-load in Indonesia. That is true even when local
environmental externalities are taken into account (i.e., damage costs from local air pollutants). Gas
would only be least cost for base load generation at very high valuations of GHG emissions (Section 4.6).
Geothermal electricity is by its nature suited to continuous production and therefore would substitute
for coal in the large coal-dominated grids of JavaBali and Sumatra.
The Castlerock Assessment

No such economic analysis was prepared to support the governments geothermal targets. Indeed,
one of the first tasks in the Castlerock study was to reassess the field-by-field estimate of probable
potential. This analysis takes into account the probabilistic variations in input parameters, and the
revised potentials represent the expected value of commercial potential (Table 1.1).
Table 1.1: Castlerock Reassessment of Geothermal Work Area Potentials
GoI
Castlerock
Change
No
Change
Capacity
Increase
Capacity
Decrease
No
Potential
MWe
MWe
MWe
[]
[]
[]
[]
110
TangkubanPerahu1,
WestJava
110
Kamojang5&6, WestJava
100
60
40
Ijen, East Java
110
110
IyangArgopuro, EastJava
55
55
Wilis/Ngebel, EastJava
165
39
126
Rawa Dano (Kaldera

Danau Banten)
110
217
107
Cibuni, WestJava
10
59
49
Cisolok, Cisukarame,
WestJava
50
30
20
Darajat, WestJava
110
110
10
KarahaBodas, WestJava
140
103
37
11
Patuha, WestJava
180
94
86
12
Salak, WestJava
40
40
13
Tampomas, WestJava
45
45
14
TangkubanPerahu2,
WestJava
60
60
15
Wayang Windu, WestJava
240
180
60
16
Baturaden, CentralJava
220
220
17
Dieng, CentralJava
115
41
74
1
1
1
18
Guci, CentralJava
55
55
19
Ungaran, CentralJava
55
62
20
SeulawahAgam, North
Sumatra
55
24
31
21
Jaboi, NorthSumatra
22
Sarulla1(NamoraILangit),
North Sumatra
1
1
330
220
110
1
continued on next page
Introduction5
Table 1.1continued
GoI
Castlerock
Change
No
Change
Capacity
Increase
Capacity
Decrease
No
Potential
MWe
MWe
MWe
[]
[]
[]
[]
110
128
18
55
53
23
Sarulla2(Silangkitang),
North Sumatra
24
SorikMerapi,
NorthSumatra
25
Muaralaboh, WestSumatra
220
30
190
26
LumutBalai,
SouthSumatra
220
204
16
27
RantauDadap,
SouthSumatra
220
172
48
28
Rajabasa, South Sumatra
220
49
171
29
Ulubelu3&4, Lampung
110
146
36
30
Lahendong5&6,
NorthSulawesi
40
40
31
Bora, CentralSulawesi
32
Merana/Masaingi,
CentralSulawesi
20
20
33
Huu, Sumbawa
20
20
34
Atadei, Lembata
35
Sokoria, Flores
36
Jailolo, NorthMaluku
37
SongaWayaua,
NorthMaluku
38
SungaiPenuh, Sumatra
39
Hululais, Sumatra
40
1
1
1
10
10
110
66
44
110
137
27
Kotamobagu1&2,
Sulawesi
40
40
41
Kotamobagu3&4,
Sulawesi
40
34
42
Sembalun, Flores
20
20
43
Tulehu, Maluku
20
20
44
SuohSekincau,
South Sumatra
230
219
11
45
SipoholonRia,
North Sumatra
75
75
46
BukitKili, Sumatra
83
23
60
47
GunungTalang, Sumatra
36
36
48
Suwawa, Sulawesi
110
14
96
49
Bedugul, Bali
10
208
198
50
Ulumbu, Flores
10
10
51
Mataloko, Flores
4,524
2,774
1,750
10
Total
1
1
1
1
1
1
1
1
1
1
1
1
20
14
GoI = Government of Indonesia, MWe = megawatt electric.

Source: Castlerock Consulting. 2010. Phase 1 Report: Review and Analysis of Prevailing Geothermal Policies, Regulations and Costs. Jakarta: Ministry of Energy
and Mineral Resources. Exhibit 4.1.

Of 51 waliyah kerja pertambangan (WKP) (geothermal work areas) examined, only 10 show no change,
and seven show an increase; while 20 WKPs show a decline, and 14 (or 27% of the total number) show
zero potential. The total potential of these fields is therefore reduced from 4,554 MW to 2,774MW.
Castlerocks detailed field-by-field assessment estimates suggest cumulative additions of about
2,100MW by 2020, or a total of 3,435 MW when the existing projects are included.
Castlerock then went on to provide the first rigorous evaluation of the geothermal supply curve for
the main geothermal areas: Figure 1.1 shows this for Sumatra. As explained in Box 1, such supply curves
are fundamental to the rational formulation of targets. The Castlerock supply curves are grounded
in detailed probabilistic modelling of exploration drilling (and its likely probability of success), which
permitted calculations of upper and lower bounds.
Where these curves intersect, the cost of coal generation provides the targetso at 6 US/kWh, the
median supply curve intersects at around 1,000 MW. If the value of avoided GHG emission damage
costs were 2 US/kWh, the Sumatra target should be set at around 1,300 MW: the lower bound curve
(most optimistic) has a value of 8 US/kWh at the higher value of around 1,700 MW.
Figure 1.1: Geothermal Supply: Sumatra

24
LCOE Cents per kWh
20
Color Key
Gross Power > 55 MW =
Gross Power < 55 MW =
16
12
200
400
600
800
1,000
1,200
1,400
1,600
1,800
2,000
Cumulative Power Plant Capacity/MWe

Mean Supply Curve
Upper Bound of Supply Curve
Lower Bound of Supply Curve
kWH = kilowatt-hour, LCOE = levelized cost of energy, MW = megawatt, MWe = megawatt electric.
Note: The (upper) blue line is the upper bound of the cost, but provides, at any given cost, the lowest supply; the (lower) brown line is
the lower bound of cost, but provides, for any given price, the lowest level of supply.
Source: Castlerock Consulting. 2010. Phase 1 Report: Review and Analysis of Prevailing Geothermal Policies, Regulations and Costs. Jakarta:
Ministry of Energy and Mineral Resources.
2013 World Bank Assessment

In September 2013, the World Bank prepared a revised project status report, provided in Appendix 3.
Figure 1.2 shows the resulting distribution of geothermal projects by region.
Note that the bulk of the new capacity that can be expected by 2020 is in Sumatra where 1,820 MW
is expected, compared to just 570 MW in Java/Bali, and 240 MW in the eastern islands (where project
size is likely to be quite small, constrained by the relatively small loads). The 4,400 MW that can
Introduction7
Figure 1.2: Spatial Distribution of Geothermal Projects

5000
Total Installed Capacity, MW
East new
Sumatra new
4000
3000
2000
JavaBali, new
1000
JavaBali, existing
JavaBali, existing
JavaBali, new
Sumatra, existing
Sumatra, new
East, existing
East, new
Total
2013
2013
1,124
0
123
0
88
5
1,340
2014 2015 2016
2014
1,164
0
123
0
88
15
1,390
2015
1,164
0
123
0
88
18
1,392
2017
2016
1,164
30
123
190
88
20
1,615
2018
2019
2017 2018
1,164
1,164
170
400
123
123
730
1,115
88
88
70
110
2,345 3,000
2020 2021
2019
1,164
570
123
1,727
88
115
3,787
2020
1,164
970
123
1,822
88
240
4,407
2022 2023 2024 2025
2021
1,164
1,040
123
1,942
88
273
4,630
2022
1,164
1,040
123
1,942
88
273
4,630
2023
1,164
1,040
123
1,942
88
273
4,630
MW = megawatt.
Source: World Bank, 2013.
Figure 1.3: Status of Geothermal Working Areas
GWA = geothermal working areas, MW = megawatt.

Source: R. Mulyana. 2013. Governments Programs to Accelerate Geothermal Energy. Paper for the Indonesia
International Geothermal Convention and Exhibition. Jakarta. June.
2024
1,164
1,040
123
1,942
88
273
4,630
2025
1,164
1,040
123
1,942
88
273
4,630

reasonably be expected by 2020 is somewhat lower than the total inventory of projects in the more
optimistic MEMR forecast (5,816 MW).9
Figure 1.4 presents the same data, plus the potential addition of binary bottoming plants (BBPs), but
broken down by the status of the tariff. Only 415 MW of new projects expected by 2020 have yet to
be tendered (plus an additional 466 MW of potential BBP add-ons). Assuming no renegotiation of
existing PPAs, any new tariff issuance would affect only these new projects.
Figure 1.4: Geothermal Projects by Tariff Status

8,000
MV by tariff category
6,000
Binary
not yet Tendered
4,000
Tender
negotiating PPA
signed PPA
2,000
existing
0
2013
2014
2015
2016
2017
2018
2019
2020 2021
2022
2023
2024 2025
Existing
2013 2014 2015 2016

1,335 1,375 1,375 1,375
2017
1,375
2018 2019
1,375 1,375
2020 2021 2022 2023 2024 2025

1,375 1,375 1,375 1,375 1,375 1,375
PLN
Signed PPA
Negotiating PPA
Reneg PPA
Tender
Not yet Tendered
Binary
Total
FracNewTariff
FracNoSigned PPA
5
5
8
10
0
0
0
190
0
0
0
0
0
5
5
5
0
0
0
0
0
5
5
35
0
99
99
116
1,340 1,489 1,491 1,730
0%
7%
7%
9%
0%
7%
7%
9%
50
770
0
115
0
35
194
2,538
9%
14%
75
75
1,165 1,440
0
227
115
115
220
440
50
115
287
395
3,287 4,181
10%
12%
20%
31%
95
95
95
95
95
95
1,450 1,450 1,450 1,450 1,450 1,450
517
567
567
567
567
567
115
115
115
115
115
115
440
440
440
440
440
440
415
588
588
588
588
588
466
489
489
489
489
489
4,872 5,118 5,118 5,118 5,118 5,118
18%
21%
21%
21%
21%
21%
40% 43% 43% 43% 43% 43%
PLN PT = Perusahaan Listrik Negara, PPA = power purchase agreement.

The view that Indonesias geothermal targets can be reached at negligible incremental financial cost
is a major issue for the sectora perception encouraged by the relatively high quality of the first few
geothermal projects that are now operating (the low hanging fruit). The highest cost among the
currently operating projects is 9.8 US/kWh, while most have costs below 8.5 US/kWh. It is also
pointed out that whatever may be the difficulties, the 1,335 MW that is in place ranks third in the
world, behind only the US and the Philippines.
Nevertheless, by whatever target is used as the yardstick, the impression that the geothermal program
has run into serious difficulties is widespread. There is general agreement among all of the stakeholders
that reaching 4,000 MW will require major reforms in policy and the institutional framework.
9
Government of Indonesia, Ministry of Energy and Mineral Resources. 2014. Energy as the Strategic Resources to Fulfil
Nations Growth and Prosperity, and the Role and Prospect of Geothermal. Paper presented at the Indonesia International
Geothermal Convention and Exhibition. Jakarta. 46 June.
Introduction9
1.5 Scope
Section 2 of this report discusses tariff design, and the advantages and disadvantages of alternative
approaches. In particular we discuss the question of whether tariff ceilings should be based on
estimates of production costs, or on the basis of estimated benefits.
Implementation and procedure of the proposed methodology is discussed in Section 3. We discuss
how tariff ceilings should be applied (for example, if a tender price is offered for a base tariff valid at
the commercial operating date (COD) some 68 years from tender, one should not calculate a ceiling
based on todays conditions), and stress the importance of transparency and stakeholder consultation.
Section 4 presents the calculation of tariff ceilings based on benefits. The main potential pitfall of
such calculations is an attempt at great precision: given the uncertainty of all forecasts, such precision
would be entirely spurious. For example, in the case of the benefit of the avoided local health damage
costs from thermal generation, one should not pretend that these can be estimated exactly, especially
where these are based on unreliable extrapolations of health damages estimated in other countries.
Issues surrounding PPAs are discussed in Section 5. Many geothermal projects are currently stalled
because of problems with long delays to projects, or resources proving to be much smaller than
originally envisaged. Renegotiation of PPAs between PLN and developers should be subject to a
formal policy issued by MEMR. The main principles of tariff adjustments are presented here.
Section 6 presents our estimates of the incremental costs of the geothermal targets, and the
likely impact on the Ministry of Finance (MoF) subsidy to PLN, followed by a discussion of our
recommendations for improvements to the tendering process in Section 7.
A quantitative analysis of the impacts of funding early-stage exploration on the tariff follows in
Section8. While the concept has long been proposed as one of the options for the Geothermal Fund,
the calculations presented here further reinforce the previous recommendations of ADB and others
for doing so.
Other key constraints to be unlocked are presented in Section 9. This includes the need for a more
nuanced presentation of the targets, the potential for PGE commercial partnerships, financing issues
(how to mobilize the $10 billion needed to reach the target), geothermal risk mitigation (including
the optimal use of the Geothermal Fund), and technology paths to faster development (BBP, larger
unit sizes).
The report concludes in Section 10 with a summary of recommendations, as presented to the
stakeholder consultation meeting held on 28 January 2014.
The report is interspersed with stakeholder questions or comments posed to us during the
consultations leading up to this report, along with our responses to these concerns.
Seven appendixes are provided. Appendix 1 discusses reporting codes. Appendix 2 discusses the
connection costs of geothermal projects compared to the costs of connecting large fossil fuel projects.
Appendix 3 provides a detailed project status review (as of October 2013). Appendix 4 discusses
tariff structure and project finance (and addresses stakeholder consultation meeting concerns about
criteria for bankable projects). Appendix 5 presents a detailed technical discussion of technology
pathways to faster development (particularly the possibility of retrofitting some existing projects
with BBPs where steam conditions are suitable). Appendix 6 presents an estimate of the exploration
drilling costs for the next 3,000 MW of geothermal projects. In Appendix 7, we comment on the new
geothermal tariff regulation issued by MEMR in June 2014. Finally, in Appendix 8, we comment on the
2014 revision to Geothermal Law No. 27 of 2003.
Tariff Design
2.1 General Principles of Tariff Design
The design of renewable energy tariffs should be guided by the following principles:

A tariff should be rational, and in support of clearly defined objectives. This would ensure that
the resources are not developed for their own sake simply because they exist, and because it
is generally held to be desirable.
The tariff methodology should be transparent (and documented as part of a tariff issuance),
with clearly stated assumptions.
A tariff should promote economic efficiency.
Recovery of any incremental costs should be transparent, credible to lenders, and equitably
allocated.
A tariff should be consistent with legislative requirements (in the case of Indonesia, this
means compliant with the 2003 Geothermal Law and its 2014 revision, as well as subsequent
regulations).
A tariff should be adaptable to changing circumstances. This requires the methodology to

have a defined basis and provide for review and updating to a clearly stated timetable.
Stakeholders should be consulted. While consensus is not always achievable, concerns

should be addressed.
The tariff policy environment should be stable. While ceiling prices may require annual
updating, the methodology should not be changed at frequent intervals.
The important questions in the detailed design of a tariff for geothermal energy are:

Whether tariffs should be fixed and available to all (as in so-called feed-in tariffs), or
whether tariffs should be set competitively.
If tariffs are fixed, whether they should be based on production costs or on benefits.
If tariffs are set competitively, whether they should be subject to a ceiling.
Whether a ceiling should be based on an estimate of production costs, or on benefits.
How the incremental costs (i.e., the difference between the geothermal tariff and PLNs
avoided costs) are recovered.
The distinction between economic and financial analysis is worth noting. Financial analysis deals with
the cash flows among the various stakeholders, while economic analysis deals with economic flows
from the perspective of the economy as a whole and includes consideration of externalities (such as
damage costs from local air emissions and GHG emissions) that are not reflected in financial cash
flows. From the economic perspective (i.e., from the perspective of the optimization of resources in
an economy), the ideal tariffwhether at the wholesale level (such as PLNs purchases of geothermal
power from developers), or at the retail level (PLNs sales to consumers)should reflect the economic
costs of production (and for consumer tariffs, the additional economic costs of transmission and
10
Tariff Design11
distribution). This ideal tariff should also be determined as a first step, so that the economic costs
of other objectives can be quantified (for example, it might be desirable for lifeline consumer tariffs
to protect low-income consumers to reflect the equity objective, but the costs of doing so should
be made explicit, and ideally be covered not by cross-subsidies from other consumers but by direct
government subsidies).
2.2 Fixed Tariffs or Competitively Bid Tariffs

Definitions
Two types of fixed tariff are in general use: FITs and avoided cost tariffs. In international practice, a
FIT is generally understood as being based on the production costs of the technology in question, as
in the original German model. Malaysia and the Philippines have such FITs (though in neither country
is there such a tariff for geothermal). The distinguishing feature of FITs is that they are technology
specific, and often differentiated by project size and other technical characteristics (such as additional
bonus payable for projects that meet criteria for domestic content or other technical attributes seen
as desirable). Fixed FITs have led to overcapacity at high prices, abandonment of programs, and an
increasing backlash from consumers and governments (Box2).
Box 2: Global Experience with FITs
Although fixed feed-in tariffs (FITs) have been very successful in some countries at enabling large installed
renewable energy capacity, particularly in Europe, there is increasing concern about the growing impact on
consumers and government budgets, leading to some abrupt recent policy reversals (such as in Spain, and
in the United Kingdom in the case of solar FITs). In Germany, the consumer surcharge that funds the FIT
has now risen to over 5 US/kWh, and there are increasing calls for a shift to a more competitive system.a
In several developing countries, FITs were unsuccessful at achieving targets, even where tariffs were
seemingly quite high: both Brazil and South Africa have replaced FITs by renewable energy auctions, which
have proven much more successful at lowering prices and achieving quantity targets. In Brazil, the average
price for wind under the original FIT scheme was 10.9 US/kWh; in the subsequent auctions, the average
price has fallen from 8.5/kWh in 2009 to 6.41/kWh in the 2011 auction.b
Sri Lanka had a very successful avoided cost tariff in place from 19962009 that successfully enabled a
vibrant small hydropower industry (by 2012, 188 MW in 77 projects were in place). Prices were high because
the tariff was set on the basis of Sri Lankas high cost thermal generation, largely based on high speed diesel.
But in 2009, with a coal project finally under construction, avoided costs were expected to fall, so developers
lobbied successfully for a change to generous technology-specific FITs (for example, offering 19 US/kWh
for wind). However, there was no sustainable mechanism for funding the incremental costs in place, so after
a few memorandums of understanding had been signed (guaranteeing access at the FIT), the program came
to a halt, as invoices from the utility (for the incremental cost) remained unpaid. The main lesson from
this experience is that generous FITs are not sufficient even to secure physical targets in the absence of
agreement on how the incremental costs are to be funded.c
In 2012, German residential customers paid $0.25/kWh for electricity, of which the surcharge for the FIT levy accounted for $0.039/
kWh, or 13.9% of the average bill. This surcharge rose to $0.0528/kWh in 2013 (excluding VAT)! (Power intensive industrial consumers
and the railways benefit from various degrees of exemption). See, for example, K. Neuhoff et al. 2013. Distributional Effects of Energy
Transitions: Impacts of Renewable Electricity Support in Germany. Economics & Environmental Policy. 2 (1). pp. 4145.
b
See World Bank. 2014. Renewable Energy Support Mechanisms: An Economic Analysis of the Design and Incentives and their Sustainability.
Washington, DC. However, it remains to be seen whether some of the low prices bid in the Brazilian (and other Latin American)
auctions will actually be achieved.
c
Ibid. Chapter 5.
a

An avoided cost tariff is one based on the benefits of renewable energy, which can be defined as
the avoided costs of the buyer. In the case of geothermal, the benefits are represented by PLNs
(avoided) cost of the thermal generation technology that geothermal replaces. In most cases this is
coal, although on some small eastern islands it may be oil. Geothermal also avoids the environmental
damage associated with fossil generation, and the costs of the volatility of fossil fuel prices.
By definition, avoided cost tariffs are not specific to the renewable energy technology in question
they should apply to all renewable energy technologies.10 This is based on the proposition that from the
perspective of avoiding the environmental impacts of thermal generation, it does not matter whether
emission reductions are achieved by any specific technologyand that it is better for the marketplace
to decide which technologies are most cost-effective than for government to decide what technology
mix is appropriate. Viet Nam and Sri Lanka had such avoided cost tariffs for qualified renewable energy
generators from 1996 to 2009.
The distinguishing feature of fixed tariffs, whether based on estimated production costs or on avoided
costs, is that they are potentially available to all (who can meet the technical requirements of the grid
code and who are judged financially capable of completing the project in a timely fashion).
Although the 2012 Indonesia geothermal tariff was announced by MEMR as a FIT (Table2.1), in fact
it was based on estimates of the avoided costs in the Castlerock report.11
Table 2.1: The 2012 Geothermal FIT
No.
1
2
3
4
5
6
Region
Sumatra
Java, Madura, and Bali
South Sulawesi
North Sulawesi
NTB, NTT, Maluku, and Papua
Maluku and Papua
Tariff (US/kWh)
High Voltage
Medium Voltage
10
11.5
11
12.5
12
13.5
13
14.5
15
16.5
17
18.5
NTB = Nusa Tenggara Barat (West Nusa Tenggara), NTT = Nusa Tenggara Timur (East Nusa Tenggara), US/kWh = cents per kilowatt-hour.
Source: Government of Indonesia, Ministry of Energy and Mineral Resources. Ministerial Regulation No. 22/2012.
Economic Efficiency
Competitively determined tariffs are the best guarantee of economic efficiency. Fixed FITs, set by the
government on the basis of estimated production costs, are suitable only where the sole rationale for
additional renewable energy is the achievement of physical targets.
If the government of a developing country desires, as a good global citizen, to place a value on
environmental goals (and be prepared to cover the incremental costs), it becomes even more
important that only the most cost-effective projects are implemented. That is best achieved by
competitive tender.
In Indonesia, the implication of a fixed FIT available to all (as in the 2012 MEMR tariff issuance) is that
tender awards would be made without consideration of price (in so-called beauty contests). Good
10
However, in practice, governments generally stipulate some maximum sizefor example, in Viet Nam, the avoided cost
tariff is available only for renewable energy projects no greater than 30 MW. Thus, while small hydropower projects no
greater than 30 MW automatically benefit from the avoided cost tariff, larger hydro projects are subject to project specific
negotiated tariffs.
11
Castlerock Consulting. 2010. Phase 1 Report.
Tariff Design13
developers (and especially international ones) are discouraged by such a procedure because they see it
as subjective and unreliable.
It is worth noting that none of the worlds leading geothermal countries (the United States,
the Philippines, Mexico, Italy, New Zealand, and Iceland) have FITs for geothermal (Table 2.2).12
Countries that do have fixed FITs tend to set them at very high levels because their resources are very
small or low-grade (e.g., Germany at 0.25/kWh, 33.7 US/kWh)13 or because of special circumstances,
such as Japan (2741US/kWh) where the motivation is the acute energy crisis in the aftermath of
the shutdown of its nuclear plants following the accident at the Fukushima nuclear power plant.
Table 2.2: FITs for Geothermal Energy
Size
Germany
Japanb
Italya
Taipei,Chinaa
<15 MW
>15 MW
<1 MW
Currency/kWh
0.25
27.3 Y
42.0 Y
0.20
4.80 NT$
US/kWh
33.7
26.6
40.9
27.0
17.0
kWh = kilowatt-hour, MW = megawatt, NT$ = NT dollar.

Sources:
a
E. Bscher. 2012. Feed-in Tariffs Blessing or Curse for Geothermal Energy? Worldwide Background and Overview. Geothermal Resource
Council Transactions. Vol 36.
b
Government of Japan, Ministry of Economy, Trade and Industry. http://www.meti.go.jp/english/policy/energy_environment/renewable
Setting fixed FITs on the basis of production costs is also subject to the same problems as setting tariff
ceilings on the basis of production costs.
2.3 Ceilings
Indonesia introduced the concept of a ceiling price (9.7 US/kWh) in 2009,14 below which the winning
tender bid would automatically be accepted, but above which the bid was subject to negotiation with
PLN. It is unclear what are the principles that govern such ad hoc negotiations (other than PLNs
desire to minimize the cost). It appears that no PPAs above 9.7 US/kWh were negotiated with PLN
under this provision.
There are two reasons why competitive tenders should be subject to ceiling prices:

to ensure that the bid price is reasonable (which it might not be if there are defects in the
tender process due to insufficient competition, collusion among bidders, or unrealistic bids
offered by inexperienced bidders); and
to ensure that the bid price does not exceed the benefits of the project.
Ceilings on competitively bid prices for renewable energy are widely used in international practice (Brazil,
Peru, South Africa), and are also appropriate for Indonesia. Ceiling prices are also used in India for solar.
It may be supposed that in a tender subject to a ceiling price, winning bids will be very close to that
ceiling price. The international experience is unclear. For example, in the renewable energy auctions in
Peru, winning bids have been from 53% to 82% of the ceiling price (Table 2.3).
12
Italy does have a FIT for geothermal, but only for projects smaller than 1 MW.
13
The system is complex, with several bonuses above the basic rate and several size categories: the figure cited here is for a
typical project.
14
Indonesia. Ministerial Regulation (MEMR) 32/2009.

Table 2.3: Bid Prices and Ceilings in Peru Renewable Energy Auctions
Winning bid
Ceiling
$/MWh
$/MWh
Winning bid as
% of ceiling
%
Small hydro
2009
60.2
74
81%
Solar
2010
221.0
269
82%
Wind
2010
80.4
110
73%
Biomass
2010
63.5
120
53%
$/MWh = dollar per megawatt-hour, hydro = hydropower.

Sources: Osinergmin. http://www2.osinerg.gob.pe/EnergiasRenovables/contenido/Resultado1raSubasta.html; International Renewable Energy
Agency. 2013. Renewable Energy Auctions in Developing Countries. IRENA.
On the other hand, the experience in South Africa suggests there may be a benefit to an undisclosed
ceiling price. In the first auction round (for wind, solar photovoltaic, and concentrated solar power,
2011) the average contract prices were from 98% to 114% of the disclosed ceilings. In a second 2012
round, the ceilings were undisclosed, and average prices were much lower (11.2 US/kWh for wind, as
opposed to 14 US/kWh in the first round). But this may be as much a reflection of how the ceilings
were set, i.e., on the basis of estimated production costs, as of the disclosure of the ceiling price.
Most international auction experience has been with small hydro and wind, where auctions were
for a large number of sites that bid for the right to a long-term PPA at the bid price. For both these
technologies, establishing the size of the resource is easy compared to that required for geothermal
energy. The ceilings in most cases are based on estimates of production costs, which for small hydro,
wind, and solar power projects are straightforward. Tenders for Indonesian geothermal projects are of
an altogether different type, where one has (ideally) many bidders for a single site about which there
is much resource uncertainty.
One may note that from the point of economic efficiency, it would not matter if bid prices were
close to the ceiling, because it is only important that the cost does not exceed the benefit. However,
the question would then be whether the corresponding producer surplus, which derives from the
economic rents associated with sites that can be developed at lower cost, should indeed accrue to
producers, or to government (electricity consumers). If the principal objective is to develop the huge
geothermal potential, that can only be achieved by building up a healthy private sector geothermal
industry, and nothing would give a better incentive than making geothermal development profitable
(i.e., letting producers capture the surplus at lower-cost sites). In time, if the tender process is efficient,
these profits will be competed away.
Ceiling prices are not meaningful if they are negotiable on an individual basis after tender. However,
whatever the rationale, ceiling prices should be reassessed from time to time. A ceiling established
several years ago may no longer be reasonable, or reflect benefits, today. For example, production
costs may change because of inflation, or benefits may change because international fossil fuel
prices rise.
The Indonesian 9.7 US/kWh threshold established in 2009 would clearly no longer be appropriate
for new projects today, because costs have risen considerably since then as a result of general inflation.
Drilling costs in particular have increased faster than inflation. For this reason, international best
practice is for renewable energy tariffs and ceiling prices to be reviewed on an annual basis by the
relevant body, which are usually regulators or ministries of energy.
Tariff Design15
2.4 Production Costs or Benefits as a Basis for

Ceiling Prices?
If the reason for a ceiling is that projects should be economically efficient, namely that costs, even if
competitively determined, should be less than benefits, then the ceilings should be determined by the
governments valuation of the benefits of geothermal energy.
But even if the main rationale of a ceiling is to ensure that bid prices are technically reasonable,
attempts to calculate plausible ceilings on the basis of production costs have significant problems:

Even bid prices with technically reasonable costs, and by technically and financially capable
bidders, may exceed the benefits, so economic efficiency is not assured. Since the incremental
costs of geothermal energy to PLN are paid for by government through the MoFs Public
Service Obligation (PSO), the requirement for economic efficiency is paramount.
Government can never have as much reliable information about costs of projects as
developers. However, PLNs costs are well known and can be estimated without great
controversy about their validity. PLN and MEMR have close relationships at the technical
level, and few of the data required are confidential.
Production costs depend on a range of financial variables such as the developers return on
equity, which are difficult for governments to determine objectively (what is a fair rate of
return for a project with high risk: 12%? 16%? 20%?).
MEMR can estimate the production costs for a particular project size and probable resource
characteristics based on engineers best estimates of most likely, or average values. But
given the wide range of uncertainty for geothermal projects, using average values as a ceiling
excludes (by definition) 50% of projects in that category. There is no objective answer as to
whether the ceiling should be 10% or 20% or 50% above the average.
This last problem is illustrated in Figure 2.1. In project economic analysis, one deals with uncertainty
in project assumptions with a formal risk analysis, using a technique called a Monte Carlo simulation.
The financial internal rate of return (FIRR) given the tariff, or the required tariff to achieve a given
equity FIRR, is calculated 1,00010,000 times, at each iteration using a different set of values for each
input assumption (sampled from their probability distributions). The result is a probability distribution
of the FIRR (or tariff), as shown in Figure 2.1.
In this illustrative example, the average required tariff is 11.5 US/kWh, which corresponds to the
most likely set of assumptions. But is this a reasonable value for the ceiling, because that would
exclude 50% of the cases (p50)? Even the values above the average are still technically reasonable
under circumstances less favourable than the average (e.g., more than expected number of wells for
delineation drillinga circumstance even the best planned drilling program cannot avoid). p70, which
excludes 30% of cases, is 13.5 US/kWh. p100, which includes all cases, is 16 US/kWh. But how does
one decide where to draw the boundary? There is no rational basis to decide how much to the right of
the average the ceiling should lie.
There is a perverse incentive in the definition of a production cost-based ceiling tariff, as earlier
proposed by MEMR (Table 2.4) which is classified according to size categories, and calculated using
the MEMR production cost model.
Clearly, with specification as a step function, no projects will be proposed between 56 MW and around
65 MW (at which the tariff is 10.5 US/kWh), when a 55 MW sized project obtains a tariff of 12.5 US/
kWh. In any event, there are substantial scale economies beyond 55 MW, so the size classification
should be extended to at least 220 MW. Moreover, instead of a step function, the ceiling would better be

Figure 2.1: Production Cost Estimates
0.2
AVERAGE
p50
0.15
CEILING
p70
0.1
0.05
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0 14.0
15.0
16.0
17.0
18.0
19.0
Base Tariff p (o), US/kWh

US/kWh = cents per kilowatt-hour, p50 = 50% probable value, p70 = 70% probable value.
Table 2.4: Size Classifications
Figure 2.2: Production Cost Based Tariff,

Size Classifications
US/kWh
10.5
20<MW<55
12.5
10<MW<20
15.5
<10
19.0
MW = megawatt, US/kWh = cents per kilowatt-hour.

Source: Government of Indonesia, Ministry of Energy
and Mineral Resources. 2014. Geothermal Production
Cost Model. January.
20
18
US/kW h
>55 MW
16
57 MW
12
10
stated with linear interpolation between

the calculated points, as illustrated in
Figure 2.3.
54 MW
14
20
40
60
80
Installed capacity, MW
100
120

In short, there are no credible mechanisms for setting tariff ceilings on the basis of production costs.
However, the production cost function, as for example calculated with the MEMR production cost
model, does have several other useful and important applications; for example, we propose use of
such a model for PPA renegotiations.15
There are, of course, some issues in the calculation of the benefits of geothermal energy (Section 4).
Some benefits are very hard to value, and future fossil fuel prices are difficult to forecast. However,
many such assumptions can be drawn from credible international sources, such as the oil and coal
15
See Section 5.
Tariff Design17
price forecasts published annually

by the International Energy
Agency (IEA), and assessments
of the benefits of GHG emissions
from the Intergovernmental Panel
on Climate Change (IPCC).
Figure 2.3: Recommended Form of the

Production Cost Function
20
US/kWh
18
16
14
12
10
8
recommended!
0
100
200
300
MW = megawatt, US/kWh = cent per kilowatt-hour.
Concerns
expressed
by
stakeholders about the current
tender process are legitimate.
However, these should be tackled
directly by improving the technical
capacity of tender committees,
and using the Geothermal Fund
for up-front exploration to improve
the quality and value of the data
available to bidders, rather than
setting ceiling prices based on
arbitrary definitions of reasonable
production costs. Table 2.5
summarises the comparison of the
two approaches.
Table 2.5: Comparison of Approaches

Criterion
Production costs
Transparency
Benefits
(avoided costs)
Good, provided assumptions are clearly stated,

and the methodology published.
Economic
efficiency
Not assured, since production costs say

nothing about benefits.
Assured, since projects whose

costs>benefits will not be built.
Data problems
Government can never know as much

about projects as developers.
PLNs costs are easily established.
Project
variability
High variability of project conditions.

Actual capital costs may vary greatly
according to manufacture order books,
interest rate variability, etc.
Project cost variation irrelevant: for

a ceiling price, it is only important
that benefits>costs as estimated in
competitive tender.
PLN = PT Perusahaan Listrik Negara.

One of the arguments against the avoided cost approach is that there is a higher risk of project not
going ahead if there are increases in project costs. While that may well be true, if the avoided cost
tariff is based on a sound economic rationale, then if a project does not materialize because its costs
increase to the point that the published avoided cost tariff no longer provides the necessary revenue,
then that is a good outcome: projects with costs higher than the economic avoided cost should indeed
not proceed.
Implementation Issues
and Procedures
3.1 Tariff Ceilings and Escalation

It is sometimes assumed that tariff ceilings should be based on todays costs (if based on production
costs), or todays benefits (if based on PLNs avoided costs). But that is not reasonable if the intent
is to apply the ceilings to a tender for a project that will deliver benefits only 68 years from the
tender date.
At the time of tender, developers estimate a base price that applies to the COD. That date may be
some 68 years in the future. A project tendered today may only start operation in 20202022.
The developers cost estimates for power plant construction, for example, would be based on
the expectation of construction costs several years in the future, when costs may have changed
significantlywhether as a matter of changes in market conditions for imported equipment or
inflation.
This is true regardless of whether the ceiling is set on the basis of production costs, or on the basis of
avoided costs (benefits). The estimates of production cost proposed by MEMR (Section 2, Table2.4)
are todays estimates of reasonable costs, but it is obvious that these would not be reasonable as
ceilings for a tender bid price that only comes into effect 68 years in the future.
In the case of benefits-based ceiling prices, MEMR must make a forecast of what are the likely benefits
at the COD. This, of course, requires a methodology for making such forecaststhe benefits in 2020
will depend on a range of assumptions about future coal prices, future coal project construction costs,
and the future value of GHG emissionsjust as the developers estimate of the base tariff depends
on his forecasts of the variables that affect investment and operation and maintenance (O&M) costs
some years in the future.
One might object that such forecasting is a technically complex analytical task requiring many
judgements that may be beyond the experience of MEMR. However, many of the necessary judgements
are routinely made by entities that specialize in such forecasts, which are not only published on the
web but which have widespread credibility and international acceptance. To forecast future oil and
coal prices, one can reasonably rely on the World Energy Outlook reports published annually by
the IEA; and for escalation of manufactured goods (say for power plant construction) one may use
forecasts of the manufactured unit value (MUV) index published quarterly by the World Bank.
3.2 Transparency
International practice provides many lessons about the importance of transparency and accountability.
In 1996, Sri Lanka introduced Asias first avoided cost tariff for renewable energy, based on the avoided
costs of thermal electricity. At the time, Sri Lanka lacked an independent regulator or government
renewable energy agency, so responsibility for the annual calculation of avoided cost was given to the
18
Implementation Issues and Procedures 19
Ceylon Electricity Board, which was also the buyer of electricity. Lack of clarity and transparency in
these calculations led to court action against the Ceylon Electricity Board by some developers.
The lessons were heeded: after the responsibility for tariff-setting was put in the hands of the newly
created Sri Lanka Public Utilities Commission (SLPUC), it made available the calculation spreadsheet
for the new FITswhich now sets the standard for tariff transparency (Box 3).
3.3 Recovery of Incremental Cost

International experience shows that credible recovery of the incremental costs of renewable energy is
the single most important requirement for the success of a renewable energy tariff.
Malaysias newly introduced renewable energy FIT sets the standard for the Association of Southeast
Asian Nations (ASEAN) for transparent recovery of the incremental costs of renewable energy. To
finance the initial FIT payments, the government advanced $60.4 million to the renewable energy
fund. The amount is to be paid back. Subsequently, the fund will rely on income from the additional 1%
tariff on monthly electricity rates (only applicable to consumers of 300 kWh and above). In Thailand,
the funds to pay the renewable energy adders are derived from the Energy Conservation Fund, whose
source of revenue is a tax on petroleum products.16
Unfortunately, in Sri Lanka, although the tariff itself is transparent, the arrangements for recovery
of incremental costs were not assured. The utility pays only its avoided financial cost, which is
substantially below the cost of new projects, especially for wind. Invoices presented by the Ceylon
Electricity Board to the Renewable Energy Fund remained unpaid, with the result that very few
memorandums of understanding that guarantee access to the generous FIT were signedwithout
which projects are unbankable.
In Indonesia, the present mechanism to assure PLNs financial credibility is well understood (given a
low consumer price, PLNs revenue requirements are made up by subsidy from MoF), though the
history of defaults has not been forgotten by developers. The introduction of performance-based
regulation should further enhance PLNs ability to raise sufficient funds to meet its investment
requirements.17 The substantial geothermal capacity under Fast Track Program 2 (FTP2) qualifies
for sovereign guarantees, and other foreign investment and purchase power parity projects could
also be protected by partial risk guarantees from the Multilateral Investment Guarantee Agency (MIGA)
and the PT Penjaminan Infrastruktur Indonesia, Persero (Indonesia Infrastructure Guarantee Fund).
However, in view of the current arrangements, and the desire of MoF to reduce the level of subsidy
to PLN, it is important to demonstrate that geothermal projects that potentially increase the subsidy
requirement are justified by the economic benefits to Indonesia. Fixed FITs or ceilings based on
production costs do not assure such economic efficiency.
16
Renewable energy producers in Thailand get the market price for electricity, plus a fixed payment (the adder) dependent
on the renewable energy technology in question.
17
Performance-based regulation is an approach to rate-making that provides better incentives to operate more efficiently
than the traditional approach of setting tariffs based on costs plus a predetermined rate of return.
Box 3: Examples of Renewable Energy Tariff Transparency

In Sri Lanka, the regulator published this spreadsheet as a basis for public hearings on proposed revisions to
the tariff. All figures here relate to tariffs for wind-based electricity.
Tariff Calculation for Wind-Based Electricity
Capital Cost (SLRs million/MW)
Construction Time (Years)
Construction Year
Equity (SLRs million)
Debt (SLRs million)
IDC (SLRs million)
Capitalized Total (SLRs million)
Loan repayment Period
Yearly Maintenance Cost (% of Capital Cost)
Plant Factor (%)
Capacity (MW)
Year
ROE
Principle payment
Cumilative principle paid at start
Interest payment
O&M
Total Cost
Production
Escallable O&M included (SLRs/kWh)
Years
Tariff (SLRs/kWh)
Escallable Incentive (SLRs/kWh)
Escallable O&M included (SLRs/kWh)
212
2
1
40%
10%
2.76
231.29
6
4.00%
32.00%
1
2
0%
50%
16.54
1
20.35
23.13
0
18.04
8.48
70.00
2.80
2
20.35
23.13
23.13
15.03
8.48
67.00
2.80
3.03 @
Yr 1-8
20.80
0.00
3.03
Yr 9-15
10.29
0.00
3.03
Year
3
20.35
23.13
46.26
12.03
8.48
63.99
2.80
4
20.35
23.13
69.39
9.02
8.48
60.98
2.80
5
20.35
23.13
92.52
6.01
8.48
57.98
2.80
6
20.35
23.13
115.65
3.01
8.48
54.97
2.80
7.64%
Yr 16-20
5.09% Escallation a
7.64%
1.68 with
3.03 @
3
Tariff Under 3 Tier Scheme
20.80
21.04
21.28
21.55
21.84
22.15
Year
Net Cashflow
O&M
Levelized Tariff (SLRs/kWh)
-1
87.56
8.48
19.43
0
4.96
9.1275148
9.8244725
10.574648
11.382106
12.25122
kWh = kilowatt-hour, MW = megawatt, O&M = operation and maintenance, ROE = return on equity, SLRs = Sri Lanka rupees,
SLRs million = million Sri Lanka rupees.
Source: Sri Lanka Public Utilities Commission, 2008.
In the Philippines, the National Renewable Energy Board (NREB) issues feed-in tariffs (FITs) based on
representations from developers and the Department of Energy. Assumptions and NREB determinations
are published (all figures below relate to tariffs for wind-based electricity).
Technology
Representative size (MW)
Propesed by
Developers
Estimates by DOE
Latest Estimates
by NREB
30.0
30.0
30.0
Project Cost ($ per kW)
2,758
2,255
2,758
EPC Cost
1,983
1,586
1,983
25.0
25.0
25.0
O&M Cost ($ thousand/unit/year)
134
100
100
Equity IRR (%)
20.1
16.5
16.0
Net Capacity Factor (%)
After-Tax WACC (%)

FIT (Php per kWh)
11.2
12.0
10.0
11.29
9.27
10.05
EPC = engineering, procurement and construction, IRR = internal rate of return, MW = megawatt, O&M = operation and
maintenance, PHP = Philippine Peso, WACC = weighted average cost of capital.
Source: Government of the Philippines, National Renewable Energy Board.
In Viet Nam, the regulator (Electricity Regulatory Agency of Viet Nam) published the avoided cost tariff
methodology report on the web, following a series of stakeholder consultation meetings. The calculations
for the annual update are prepared by the National Load Dispatch Centre.
Implementation Issues and Procedures 21
3.4 Tariff Setting as a Process

Tariff setting is not a one-time exercise that can be set aside for a few years once the regulation is
published, and revisited if and when some problem arises. Rather, tariff-setting should be seen as an
ongoing process, with a dedicated team in MEMRs geothermal directorate assigned to maintain a
database, monitor the external conditions, update and maintain the MEMR production cost model,18
and be responsible for the necessary annual update, including stakeholder consultations.
Such units are widespread in international practice. Countries with an independent electricity regulator
or Public Utility Commission (e.g., Sri Lanka), or a special board charged with setting tariffs (e.g., New
and Renewable Energy Board, Philippines) have such a technical unit as a matter of course. There are
also precedents in other countries where the regulatory body is housed within an appropriate ministry
or other entity (for example, in Viet Nam the Electricity Regulatory Authority of Viet Nam, which is an
entity of the Ministry of Industry and Trade, has a tariff unit with these functions).
3.5 Conclusions on Implementation Procedure

Annual Issuance
Tariff ceilings should be issued by MEMR on an annual basis, no later than 15 December of each year,
with values which would apply to tenders conducted in the following calendar year.
MEMR should issue a draft ceiling no later than 15 October of each year, followed by a stakeholder
consultation meeting no later than 15 November of each year, said date being the deadline for
submission of any written comments for consideration by MEMR.
Non-Negotiable Ceilings
Tariff ceilings should be non-negotiable. If the best price offer at tender exceeds the ceiling, then the
tender shall be deemed to have failed on grounds that no economically efficient project has been
offered (though from the economic perspective, this is a good outcome insofar as an uneconomic
project should indeed not proceed). Indeed, a project area can always be re-tendered if increases in
future benefits permit.
Transparency
The methodology of the tariff ceiling calculation should be published by MEMR.
18
Although this model is not recommended for calculation of tariff ceilings, we believe it can play an important role in PPA
renegotiations (see Section 5).
Tariff Ceilings
Based on Benefits
The purpose of a tariff ceiling is to ensure that the only projects that are built are those in which costs
do not exceed the ceiling. As noted, the benefits of geothermal projects can be defined as the avoided
economic costs of thermal generation. Therefore the first step in defining such ceilings is to examine
what thermal generation option would in fact be displacedwhich will not be the same in all parts of
Indonesia.
4.1 Alternatives to Geothermal

Geothermal projects have high annual capacity factors, and therefore can provide base load. Three
types of thermal alternatives are relevant:

In the large interconnected grids of Sumatra and JavaBali, the alternative is large, modern
coal projects, operating at high efficiency (supercritical, and in Java, ultra-supercritical) and
fitted with state-of-the-art environmental controls for the mitigation of local air pollutants
(particulates, nitrogen oxides, and sulphur oxides).
In the smaller grids of the other islands, there are small coal projects (less than 50 MW), of
much lower efficiency and higher unit costs, with less effective emission controls for local air
pollutants. PLN has some doubts about the practicality of this alternative, given its limited
experience with such small coal projects to date.19
On islands for which small coal projects are not technically feasible or environmentally
acceptable, the only practical alternative for electrification is diesel.
This means that the benefits of geothermal generation are likely to be quite different across Indonesia,
and three sets of tariff ceiling are therefore required.
One might argue for a finer geographical differentiationfor example, one might differentiate between
ultra supercritical coal projects in JavaBali, and smaller subcritical coal and mine-mouth projects
in Sumatra. However, mindful that the tariff ceilings as applied to tender bid prices are forecasted
68 years into the future, the differences that may apply to these various project types are much
smaller than the uncertainty that applies to forecasting. Great precision in project differentiation
would be spurious.
19
22
PLNs most recent RUPTL includes plans for some 70 small-scale coal systems ranging in size from 3 MW25 MW.
Tariff Ceilings Based on Benefits 23
Stakeholder Comment 1: Why Is Geothermal Energy Not Also Benchmarked

Against Gas Rather Than Only Coal?
Comment:
Since gas is being used on Java for base load, the avoided cost calculation should be based on the mix
of coal and gas actually used, rather than 100% coal as proposed. Geothermal should be considered
as a substitute for liquefied natural gas (LNG) on Java and Sumatra in view of the huge investments
in LNG.
Reply:
Use of gas would increase the avoided cost component of the benefit calculation.
It is true that in West Java, a significant amount of gas is currently being used for base-load generation.
However, this is an artifact of the takeorpay provisions of the current gas supply contracts, which
are short-term contracts not expected to persist over the long term. Moreover, because of the delays
in the large Fast Track Program 1 coal projects, gas is being used to make up the base-load shortfall.
These conditions are also not expected to persist in the long term. PLNs Rencana Usaha Penyediaan
Tenaga Listrik (RUPTL) (Electricity Power Supply Business Plan) makes it clear that in the future, gas
will be used primarily for intermediate and peak loads.
More importantly, the carbon dioxide emissions per kilowatt-hour (kWh) generated in a gas combined
cycle gas turbine plant are 38%40% of the corresponding emissions from a modern coal project.a
Consequently, while the avoided cost of gas generation will be higher, the avoided externality cost will
be significantly smaller. Also, though not monetized here, gas projects have no emissions of sulphur
oxide and particular matter less than 10 microns in diameter.
Since the principal rationale for geothermal, and the need for additional subsidy to cover its incremental
cost, is to offset the environmental impact (and especially the emissions) of coal, it would be unwise to
dilute this rationale, simply to increase the ceiling price. Indeed, geothermal and natural gas (whether
pipeline or LNG) are both clean fuels, and are therefore complements, not substitutes.
Large coal,
subcritical
Size class
IPCC default
Kg/GJ
Large coal
USC
Small
coal
200
1000
50
96.1
96.1
96.1
CCGT
gas
56.1
Efficiency
From PLN RUPTL
0.34
0.40
0.26
0.56
Heat rate
kJ/kWh
10,588
9,000
14,063
6,429
GHG emissions
Kg/kWh
1.018
0.865
1.351
0.361
[]
0.35
0.42
0.27
Ratio gas:coal

CCGT = combined cycle gas turbine, GHG = greenhouse gas, GJ = gigajoule, IPCC = Intergovernmental Panel on Climate Change,
Kg = kilogram, kJ = kilojoule, kWh = kilowatt-hour, PLN = PT Perusahaan Listrik Negara (State Electricity Company), RUPTL = Rencana
Usaha Penyediaan Tenaga Listrik (Electricity Power Supply Business Plan), USC = ultra-supercritical.
Source: Authors calculations, based on data from IPCC and PLN.
From the table above, we see that GHG emissions per net kWh from gas combined cycle gas turbine are between 27% and 42% of the
emissions from a coal project.
Sources: Written comments of Indonesia Geothermal Association (INAGA), March 2014; Representation of Chevron at the stakeholder
meeting at MEMR on 12 March 2014; Comments provided by Chevron, 11 March 2014.
a
4.2 The Components of Avoided Costs

The total economic benefit of geothermal energy consists of seven potential components:
(i) the avoided variable cost of thermal generation;
(ii) the avoided fixed cost of thermal generation (capital recovery and fixed O&M);
(iii) the avoided global externality cost of thermal generation;
(iv) the avoided local externality cost of thermal generation;
(v) the avoided costs to MoF of fossil fuel price volatility;
(vi) the avoided costs of transmission interconnection of thermal projects; and
(vii) the incremental local regional economic benefits of geothermal.
4.3 Macroeconomic and Global Energy Price Forecasts

Many of the variables used in the calculation depend on macroeconomic forecasts, and forecasts of
the future price of fuels. In the illustrative calculations presented here, dollar inflation is taken at 2.5%,
and Indonesian inflation at 4.5% (and with the base year exchange rate adjusted accordingly).
A range of coal and oil price forecasts is available from the most recent 2013 World Energy Outlook
published by the IEA.20 Fuel price forecasts provided by the 2013 World Energy Outlook for crude oil
and coal are presented in Table 4.1. Prices of fossil fuels are seen to rise (in real terms) under current
policies, but over the long run are expected to rise much less under the assumption of effective worldwide policies to reduce GHG emissions (new policies), and even fall in the 450 Scenario in which
the Intergovernmental Panel on Climate Change (IPCC) goals to limit atmospheric CO2 to 450 parts
per million are assumed to succeed. In the absence of much evidence that global agreement on GHG
emission reductions are imminent, we propose the World Energy Outlook current policy scenarios
as the basis for fuel price forecasts for renewable energy tariff-setting in Indonesia.
Table 4.1: World Energy Outlook 2013, Fuel Price Forecasts (at Constant 2012 Prices)
2012
2020
2025
2030
2035
$/bbl
109
120
127
136
145
New policies
$/bbl
109
113
115
121
128
450 scenario
$/bbl
109
110
107
104
100
Current policies
$/ton
99
112
116
118
120
New policies
$/ton
99
106
109
110
110
450 scenario
$/ton
99
101
95
86
75
Crude Oil
Current policies
Coal
$/bbl = dollar per barrel, $/ton = dollar per ton.

Source: International Energy Agency. 2013. World Energy Outlook 2013. Paris. Table 1.4. Prices are for average imports to the Organisation for
Economic Co-operation and Development countries.
20
International Energy Agency. 2013. World Energy Outlook 2013. Paris.
The resulting macroeconomic and fuel price forecasts are as shown in Table 4.2.
Table 4.2: Macroeconomic and Fuel Price Forecasts
Price level
2012
[]
2013
2014
2015
2016
2017
2018
2019
2020
OECD inflation
0.025 0.025 0.025 0.025 0.025 0.025 0.025 0.025
OECD deflator
US PPI
[]
0.025 0.025 0.025 0.025 0.025 0.025 0.025 0.025
Indonesia Inflation
[]
0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045
Exchange rate
Domestic Rp deflator
Oil prices
IEA Oil
Annual growth rate
10
Indonesia
11
1.000
Rp/$
[]
2012 $/ton
109
1.025
1.051
11500
11724
11953
1.00
1.05
1.09
1.077
1.104
1.131
12186 12424 12667

1.14
1.19
1.25
1.160
1.189
12914
13166
1.30
1.36
110
112
113
114
116
117
119
120
[]
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
1.2%
2013 $/bbl
100
101
102
104
105
106
107
109
@nominal prices
nominal $/bbl
100
104
108
112
116
120
125
129
12
MFO
nominal $/bbl
90
93
97
100
104
108
112
116
13
Coal prices
14
IEA coal
2012 $/ton
100.5
102.1
103.7
105.3
106.9
108.6
110.3
112.0
15
Annual growth rate
growth [ ]
1.6%
1.6%
1.6%
1.6%
1.6%
1.6%
1.6%
1.6%
16
@2103 prices
2013 [$/ton]
100
102
103
105
106
108
110
111
17
Indonesian Coal Index
2013 [$/ton]
78
79
80
82
83
84
86
87
18
Nominal
nominal [$/ton]
78
81
85
88
92
95
99
103
99
$/bbl = dollar per barrel, $/ton = dollar per ton, IEA = International Energy Agency, MFO = marine fuel oil, OECD = Organisation for Economic Co-operation
and Development, Rp = Indonesian rupiah, US PPI = United States Producer Price Index.
Source: IEA coal and oil prices from the International Energy Agency. 2013. World Energy Outlook 2013. Paris.
Assumptions:
Indonesia crude oil forecast for 2012 as from last available Rencana Usaha Penyediaan Tenaga Listrik (RUPTL) (Electricity Power Supply Business Plan). This price
is escalated at the same rate as the IEA oil price forecast. The basis is assumed to be the average price of the Organization of the Petroleum Exporting Countries
(OPEC) Reference Barrel. At the time of writing (mid 2014) the actual OPEC reference barrel stands at around $100/bbl, so the assumed (Indonesia) price of $101/
bbl is reasonable.
Organisation for Economic Co-operation and Development (OECD) and US PPI inflation by assumption, reflecting consensus forecasts. Exchange rate adjust
according to the ratio of OECD and Indonesia inflation. Marine fuel oil price as 90% of world oil price from PLN RUPTL. Indonesia inflation is taken as the
midpoint of the current Bank of Indonesia inflation target for 2014 of 4.5+1%.
4.4 Avoided Fixed Costs

The calculation of fixed costs for coal and oil projects is shown in Table 4.3. Investment cost recovery
depends on assumptions about debt to equity ratio (30% equity), the cost of PLN equity (assumed
here at 14%), and the cost of debt (7%).21 The assumptions for overnight capital costs and fixed O&M
are from PLNs RUPTL.
21
Equivalent to a WACC of 9.1%.

Table 4.3: Fixed Costs (Investment Recovery and Fixed Operation and Management)
Java and
Sumatra,
large coal
Eastern
Islands,
small coal
Oil
Overnight cost
$/kW
1,400
1,760
700
Development costs
$/kW
150
200
50
Interest during
construction
$/kW
291
287
81
Total financial cost
$/kW
1841
2,247
831
Equity
$/kW
552.3
674.2
249.2
RoE
[]
0.14
0.14
0.14
0.3
$/kW
77.3
94.4
34.9
$/kW
1289
1573
581
$/kW/year
90.2
110.1
40.7
Fixed O&M cost
$/kW/year
35.0
61.3
55.0
11
Total fixed cost/year
$/kW/year
202.5
265.8
130.6
12
Assumed PLF
[]
0.80
0.60
0.60
13
Generationc
hours/year
7008
5256
5256
14
Total fixed cost
US/kWh
2.9
5.1
2.5
Debt
Cost of debt
10
0.07
$/kW = dollar per kilowatt-hour, O&M = operation and maintenance, PLF = plant load factor, RoE = return on equity, US/kWh = cents per
kilowatt-hour.
Notes:
a
Assuming 30% equity (70% debt)
b
Cost of debt (interest rate)
c
Equivalent to kWh per year per kW of installed capacity
Source: Authors calculations based on data from PT Perusahaan Listrik Negara (State Electricity Company) (PLN).
Actual costs of large new coal projects using supercritical and ultra-supercritical technology show
considerable variation, with completed costs ranging from $1,300/kW for the 600 MW Cirebon
supercritical project to $2,240/kW for the 1,000 MW ultra-supercritical project at Indramayu.22
4.5 Avoided Variable Costs

The variable fuel costs are shown in Table 4.4. Costs per kWh of small coal projects are 50% greater
than large projects in large part because of significantly higher transportation costs (30%50% greater
per ton), and much lower heat rates. A discussion of calculating depletion premiums for Indonesian
coal follows in Stakeholder Comment 2 and Box 4.
22
Indramayu is financed by JICA (http://www.mofa.go.jp/mofaj/gaiko/oda/data /zyoukyou/h23/y110818_1.html). The 1,000

MW ultra-supercritical Manjung 4 project in Malaysia (ASEANs first ultra-supercritical project) has a total project cost of
$1,600/kW (but this is a project at an existing site with much of the infrastructure in place).
Table 4.4: Variable Costs

1
2
3
4
5
6
7
Coal price
Growth rate
Reference CV
Cost, fob
Cost, fob
Large Coal
Transport to Java
Heat value
9
10
11
12
13
14
15
16
17
18
Cost per ton

$/ton
Cost per million BTU
$/million BTU
Java coal cost
$/million BTU
Heat rate
38.5%
BTU/kWh
Variable fuel cost
US/kWh
Variable O&M cost
US/kWh
Total variable cost
US/kWh
Small coal, isolated grids, small islands
Transport to typical
Rp/ton
island
Heat value
KCal/kg
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Cost per ton

Eastern coal cost
Heat rate
25.6%
Variable fuel cost
Variable O&M cost
Total variable cost
Oil, isolated grids, small islands
Fueloil, HSD
1.125
0.1667
Assumed heat rate
34.0%
Variable fuel cost
Variable O&M cost
Total variable cost
5900
$/ton
[]
KCal/kg
$/million KCal
$/million BTU
Rp/ton
KCal/kg
2013
78.0
2014
81.2
4.1%
2015
84.5
4.1%
2016
88.0
4.1%
2017
91.6
4.1%
2018
95.3
4.1%
2019
99.2
4.1%
2020
103.3
4.1%
2021
106.6
3.2%
2022
110.1
3.2%
2023
113.6
3.2%
2024
117.3
3.2%
13.22
3.32
13.76
3.46
14.32
3.60
14.91
3.75
15.52
3.90
16.16
4.06
16.82
4.23
17.51
4.40
18.07
4.54
18.65
4.69
19.25
4.84
19.87
4.99
130840
PLN
estimate
4200
PLN
estimate
13.35
0.80
4.12
8,862
3.65
0.200
3.85
13.95
0.83
4.29
8,862
3.80
0.200
4.00
14.58
0.87
4.47
8,862
3.96
0.200
4.16
15.24
0.91
4.66
8,862
4.13
0.200
4.33
15.92
0.95
4.85
8,862
4.30
0.200
4.50
16.64
1.00
5.05
8,862
4.48
0.200
4.68
17.39
1.04
5.27
8,862
4.67
0.200
4.87
18.17
1.09
5.49
8,862
4.86
0.200
5.06
18.99
1.14
5.68
8,862
5.03
0.200
5.23
19.84
1.19
5.87
8,862
5.21
0.200
5.41
20.73
1.24
6.08
8,862
5.39
0.200
5.59
21.67
1.30
6.29
8,862
5.57
0.200
5.77
250000
27.86
1.67
5.27
13,333
7.02
0.002
7.02
29.11
1.74
5.49
13,333
7.32
0.002
7.32
30.42
1.82
5.72
13,333
7.63
0.002
7.63
31.79
1.90
5.96
13,333
7.95
0.002
7.95
33.22
1.99
6.21
13,333
8.28
0.002
8.29
34.72
2.08
6.48
13,333
8.63
0.002
8.64
36.28
2.17
6.71
13,333
8.95
0.002
8.95
37.91
2.27
6.95
13,333
9.27
0.002
9.27
39.62
2.37
7.21
13,333
9.61
0.002
9.61
41.40
2.48
7.47
13,333
9.96
0.002
9.96
115
19.21
10,037
19.28
0.002
19.28
117
19.44
10,037
19.51
0.002
19.52
118
19.68
10,037
19.75
0.002
19.75
119
19.92
10,037
19.99
0.002
19.99
121
20.16
10,037
20.23
0.002
20.23
122
20.40
10,037
20.48
0.002
20.48
124
20.63
10,037
20.71
0.002
20.71
125
20.87
10,037
20.94
0.002
20.95
127
21.11
10,037
21.18
0.002
21.19
128
21.35
10,037
21.43
0.002
21.43
$/ton
$/million BTU
$/million BTU
BTU/kWh
US/kWh
US/kWh
US/kWh
25.51
1.53
4.85
13,333
6.464
0.200
6.66
PLN
estimate
PLN
estimate
26.66
1.59
5.05
13,333
6.74
0.002
6.74
dollar per barrel

$/million BTU
BTU/kWh
US/kWh
US/kWh
US/kWh
113
18.75
10,037
18.82
0.200
19.02
114
18.98
10,037
19.05
0.002
19.05
4200
$/bbl = dollars per barrel, $/KCal = dollars per kilocalorie, $/million BTU = dollars per million British thermal units, $/million KCal = dollars per million kilocalories, $/ton = dollars
per ton, BTU = British thermal unit, BTU/kWh = British thermal unit per kilowatt-hour, CV = calorific value, fob = free on board, HSD = high speed diesel, KCal/kg = kilocalorie
per kilogram, O&M = operation and maintenance, Rp/ton = rupiah per ton.
Notes:
(1) The coal transportation cost provided by PLN for eastern islands ranges from 185,900 Rp/ton (for the Ende project, Flores Island), 455 nautical miles (NM), to 292,300 Rp/ton
for other small eastern island projects (Ternate) where distances are around 935 NM. 250,000 Rp/ton is taken here as a representative figure.
(2) Heat rates for small coal projects are very poor. PLN data for the Ende project (2 x 7 MW) is 14,507 BTU/kWh (23.5% efficient); for 2x 25 MW, PLN estimates 25.6% efficiency,
compared to 37% for large coal modern projects.
Stakeholder Comment 2: Depletion Premium

Comment:
The calculation of avoided fuel cost does not include a depletion premium.a Given that Indonesias
coal resources are finite, might it not be better to keep our coal in the ground and use it for power
generation when its value is much higher?
Reply:
Depletion premiums are significant where a resource is priced at its current or historical domestic
extraction cost (as its long run marginal cost), and where the so-determined domestic price is significantly
below the international price. This a major issue in many ASEAN countries in the case of gas: in VietNam,

Stakeholder Comment 2 continued
for example, domestic gas for power generation is priced at around $3.5/million British thermal
units (BTU), and the known domestic resource will be exhausted in another 1520 years, absent
new discoveriesat which point Viet Nam would have to face a gas price at international levels of
$12/million BTU or more. In such a case the current domestic price should indeed be adjusted to
include a depletion premium.
However, in the case of Indonesian supplies to PLN, coal is now already priced at its international
level and there are significant exports. Therefore, whether Indonesian coal is better left in the ground
today in the expectation of a higher price tomorrow depends on an assessment of future prices
of internationally traded coal, and how much coal there is in Indonesia. Indeed, in one scenario
considered in the most recent International Energy Agency, World Energy Outlook, the coal price
declines (seeTable 4.1)in which case the rational policy would be to export as much coal at the
currently higher price as possible!
Box 4 discusses the problems of practical calculation of the depletion premium for Indonesian coal.
a
For a more technical explanation of the depletion premium for an exhaustible resource, see, e.g., ADB. 1997. Guidelines
for the Economic Analysis of Projects. Manila. Appendix 6.
Source: Stakeholder Discussions.
Box 4: Depletion Premium for Indonesian Coal

The depletion premium is the amount equivalent to the opportunity cost of extracting the resource at some
time in the future, above its economic price today (and should be added to the economic cost of production
today). It is defined as follows:a
DPt =
(PST CSt)(1 + r)t

(1 + r)T
where
t
= year
PST
= price of the substitute (internationally traded coal)

at the time of complete exhaustion.
CSt
price of the domestic resource in year t
discount rate
year to complete exhaustion
The main problem in calculating the value of the premium is the uncertainty about when the resource is
exhausted, because the exploitable size of a resource is a function of its market value and the cost (and
technology) of extraction. Assessment of reserves can change very rapidlyas illustrated by the dramatic
developments in gas and oil extraction technology in the US (fracking). If the international cost of coal
increased, then doubtless additional resources would be discovered (or become economic) in Indonesia.
Indonesia is the worlds number one coal exporter, having overtaken Australian exports in 2005, so the
question of when Indonesian coal reserves will be depleted is controversial. According to Ministry of Energy
and Mineral Resources data, current remaining reserves (of all rank) are 21 billion tons, which at 2014
production of 342 million tons, would be depleted in 61 years. For high rank coals (>6,100 Kcal/kg), reserves
are only 2.6 billion tons but 2014 production is expected at 114 million tons, so depletion in 22years.b A more
pessimistic assessment is the 2013 BP Statistical Review of World Energy,c which assesses Indonesian coal
reserves at 5.5 billion tons, which could be exhausted in as little as 1415 years. However, total global reserves
are 861 billion tons, enough for more than 100 years.
See, ADB. 1997. Guidelines for the Economic Analysis of Projects. Manila. Appendix 6, Depletion Premium.
Government of Indonesia, Ministry of Energy and Mines. 2011. Domestic Market Obligation of Coal Policy in Indonesia. Jakarta.
Directorate of Coal Business Enterprise.
c
British Petroleum. 2013. BP Statistical Review of World Energy. June 2013. www.bp.com/statisticalreview
a
4.6 Avoided Global Externalities

The basis for the avoided global externality benefit should not be the current market price for CO2
in global carbon markets (which is a financial price that is also highly volatile), but should be the
economic price, namely the so-called global social cost of carbon (GSCC) (Box 5).
Since most geothermal projects in Indonesia benefit from concessionary finance offered by the global
community, it is reasonable that the value of avoided GHG used in the tariff calculation is consistent
with typical valuations used by the World Bank and ADB, currently around $30/ton CO2. This idea is
not entirely new to the Government of Indonesia, as discussed in Box 6.
As shown in Table 4.5, the valuation in US/kWh will depend on the fuel and the heat rate. At $30/ton
CO2, this ranges from 3.64 US/kWh in the inefficient small coal projects to 2 US/kWh for oil.
Table 4.5: Impact of Greenhouse Gas Valuations
IPCC default
Heat rate
Large coal
Small coal
Oil
US/kWh
US/kWh
US/kWh
kg/GJ
96.1
96.1
74.1
efficiency
0.39
0.26
0.34
kJ/kWh
9,351
14,068
10,590
kg/kWh
0.899
1.352
0.785
0.00
0.00
0.00
10
0.90
1.35
0.78
20
1.80
2.70
1.57
$/ton
30
2.70
4.06
2.35
40
3.59
5.41
3.14
50
4.49
6.76
3.92
60
5.39
8.11
4.71
70
6.29
9.46
5.49
80
7.19
10.82
6.28
90
8.09
12.17
7.06
14
12
US/kWh
10
8
6
4
2
0
10
20
30
40
50
60
70
80
90
100
GHG emission valuation, US$/tonCO2

CO2 = carbon dioxide, GHG = greenhouse gas, GJ = gigajoule, IPCC = Intergovernmental Panel on
Climate Change, kg = kilograms, kJ = kilojoules, US/kWh = cents per kilowatt-hour.

Box 5: Global Social Cost of Carbon
The literature on the global social cost of carbon (GSCC) is growing, with estimates ranging from
a small net benefit to costs of several hundred dollars a ton. Thus almost any estimate would
find some support. Tols 2008 meta-analysis of the peer-reviewed literature,a which updated
an earlier 2005 meta analysis,b cites 211 studies, and found an average estimate of $120/ton
carbon ($33/ton carbon dioxide [CO2]) for studies published in 19962001, and $88/ton carbon
($24/ton CO2) for studies published since 2001.
Much of the economics literature on the subject is highly technical, particularly with respect to
the choice of discount rate and assumptions about future global economic growth and income
inequalities: in general one can say that the lower the discount rate, the higher the social cost of
carbon (a value that may also change over time). The high valuation of the Stern Review (the
current social cost of carbon might be around $85/ton CO2)c is largely a consequence of the use of a
very low discount rate.d The 2007 Intergovernmental Panel on Climate Change report highlighted
the wide range of values of the GSCC in the literature as being in the range of $4/ton CO2
$95/ton CO2.
Carbon Valuations in World Bank Studies and Project Appraisals
Country
$/ton
CO2
Study
Reference
India
32
Policy study (2010)
G. Sargsyan et al (2011) Unleashing the Potential of

Renewable Energy in India, World Bank (2011)
Viet Nam
30
Trung Son hydro project
World Bank Project Appraisal Document (2010)
South Africa
29
Medupi coal project
Morocco
30
Ourzazate I CSP
In the United States, regulatory impact analysis requires consideration of the social cost of
carbon,e using a range of discount rates (from 2.5% to 5.0%), with values that increase over
time. For example, at a 5.0% discount rate the valuation is $12/ton in 2015, rising to $27/ton by
2050; at a 2.5% discount rate the valuation rises from $58/ton to $98/ton by 2050. In 2007, the
Government of the United Kingdoms Department of the Environment recommended a value
of 25/tonCO2 ($37/ton);f this was subsequently updated to a time-dependent system ranging
from 23/ton CO2 in 2015 rising to 48/ton by 2025 ($36/ton CO2$76/ton CO2).
Notes:
a
R. Tol. 2008. The Social Cost of Carbon: Trends, Outiers and Catastrophes. Economics E-journal. 200825. 12 August.
b
R. Tol. 2004. The Marginal Damage Costs of Carbon Dioxide Emissions: An Assessment of the Uncertainties. Energy Policy. 33 (2005),
pp. 2064-2074.
c
Government of the United States, Interagency Working Group on Social Cost of Carbon. 2013. Technical Update of the Social Cost of
Carbon for Regulatory Impact Analysis - Under Executive Order 12866. Washington, DC.
d
R. Price, S. Thornton, and S. Nelson. 2007. The Social Cost of Carbon and the Shadow Price of Carbon. Defra Evidence and Analysis
Series. Government of the United Kingdom, Department for Environment, Food and Rural Affairs (DEFRA); Government of the United
Kingdom, Department of Energy & Climate Change. 2009. Carbon Valuation in UK Policy Appraisal: A Revised Approach. London.
e
N. Stern. 2007. Stern Review on the Economics of Climate Change. Cambridge University Press. p. 304.
f
For a good discussion of these issues, and a review of the assumptions in the Stern Review, see, for example, C. Hope and D. Newbery.
2007. Calculating the Social Cost of Carbon. Cambridge University Electricity Policy Research Group; M. Grubb, T. Jamasb, and M. Pollitt
(eds). 2008. Delivering a Low Carbon Electricity System: Technologies, Economics and Policy. Cambridge University Press.
Stakeholder Comment 3: Greenhouse Gas Valuation

Comment:
Why should Indonesia bear the cost of greenhouse gas (GHG) emission reductions? Under the
Kyoto Protocol, Indonesia is not an Annex 1 country, and is not therefore obliged under the treaty
to reduce its GHG emissions.
Reply:
That is correct. But there are several arguments why it may be wise to show goodwill in this matter,
even if not mandated by international treaty.

Recognizing the strong increase in GHG emissions due to increased coal use, and as a
responsible global citizen, Indonesia has made public commitments to reduce its GHG
emissions.
When funding geothermal projects, global climate funds and the multilateral
development banks through which they are usually routed generally require
commitments to reduce GHG emissions, and an implicit or explicit valuation of these
benefits.
Even if there are other reasons for Indonesia to develop its geothermal resources,
particularly if these are difficult to value, reducing GHG emissions serves as a useful
general proxy indicator for environmental quality.
Comment:
Even if it were true that GHG emission reduction did constitute a benefit to Indonesia, the
proposed $30/ton CO2 is very high. Why should Indonesia value GHG emissions at $30/ton CO2
if the carbon price on global carbon markets is currently just $5/ton?
Reply:
It is indeed for the Government of Indonesia, and not a technical consultant or international
financial institutions, to determine what value Indonesia should assign to the avoidance of GHG
emissions. Table 4.5 is offered as guidance to show the relationship between that assumption in
$/ton CO2, and the potential impact on the tariff ceiling.
Prior to the final issuance of the tariff, it is suggested that the Ministry of Finance and the Ministry
of Energy and Mineral Resources, in consultation with the designated national authority under
the Kyoto Protocol, discuss the matter to decide on the final value.
Note that this valuation of the GSCC is unrelated to any financial benefit that may accrue to the developer
from the sale of carbon credits under the Clean Development Mechanism (or any successor to it). The
GSCC is included in the calculation of the avoided cost ceiling regardless of whether the developer
benefits from any carbon revenuewhich should be to his benefit (although subject to whatever taxes
are levied by the designated national authority on sales of Certified Emission Reduction[CER)] and any
income tax levied on the additional profit derived from their sale). Most countries with standardized PPAs
for renewable energy stipulate that any carbon sales revenue that may be collected by the developer are
for the developer to keep, and does not reduce the tendered price.

It may be supposed that this raises the potential issue of double counting: since the avoided cost of
GHG emissions is already included in the tariff ceiling (and paid for by government in the higher tariff),
should not any CER credits revenue accrue to the government rather than the developer? However,
there are several reasons why CER revenue should accrue to the developer:
(i) If the CER revenue accrues to government, there is no incentive for a developer to incur the
significant transaction costs of Clean Development Mechanism registration.
(ii) At the time of tender it is hard to gauge what CER revenue would actually be realized, so
many years in advance.
(iii) Even if the ceiling price includes the avoided cost of GHG emissions, this is only the ceiling
pricebid prices may be significantly lower.
In any event, even if CER revenue were viewed as a windfall, government recaptures a share of that
revenue through the corporate income tax. Indeed, under the proposed sharing of CER revenue in the
current PPA template (50% to the developer, 50% to PLN), in fact (at the margin) the government
collects another 30% of the developers 50% share (corporate income tax), so the total that accrues
to the developer is just 35%, from which must be deducted the substantial transaction costs of Clean
Development Mechanism registration. Given that, at least under present conditions in global carbon
markets, a forecasted revenue stream for future CER sales is unlikely to be bankable, it is unclear that
a 50/50 sharing of such revenues would provide much incentive for developers to actively pursue
carbon sales.
Box 6: Ministry of Finance Green Paper

The proposition that Indonesia should be willing to pay the true avoided cost for geothermal energy is not
new. The 2009 Ministry of Finance Green Paper on strategies for climate change mitigation proposed that
a reasonable and conservative estimate of the actual true cost of electricity incurred by the Government
of Indonesia is $0.13/ kilowatt-hour (kWh), derived on the basis of adding the cost of local air pollution
damage costs and a carbon price to the book cost paid by PT Perusahaan Listrik Negara (State Electricity
Company).
Green Paper Estimate of True Cost of Electricity

$/kWh
Capital cost
$0.054
Operating cost
$0.007
Fuel cost
$0.029
Fuel price cost risk
$0.010
Carbon cost
$0.020
Air pollution and other externalities
$0.010
Total
$0.130
Source: Government of Indonesia, Ministry of Finance, Green Paper:

Economic and Fiscal Policy Strategies. Table 4.
The basis for the carbon price component was a value of $20/ton CO2. The report also argues for a carbon tax
set at Rp80,000/ton (about $8/ton CO2) to be levied across the entire economyfor which macroeconomic
modeling showed a slight increase in gross domestic product. The report notes the commitment made by
the President of Indonesia at the 2009 Group of 20 meeting to reduce emissions by 26% by 2020, and up
to 41% with international help. The Green Paper analysis shows that a $30/ton CO2 levy would achieve such
a target. In short, the $30/ton CO2 valuation proposed for the geothermal tariff is consistent with previous
Ministry of Finance assessments.
Source: Government of Indonesia, Ministry of Finance. 2009. Green Paper: Economic and Fiscal Policy Strategies for Climate Change
Mitigation in Indonesia, Ministry of Finance and Australia Indonesia Partnership, Jakarta.
4.7 Local Environmental Externalities

It is widely acknowledged that fossil fuel combustion also results in local environmental impacts,
particularly from the local air emissions (NOx, SOx, and particulates). However, the valuation of
the damages is controversial, because the main share of the impact is human health, which in turn
depends on monetization of mortality and morbidity. The Castlerock Report has argued for inclusion
of these costs in any avoided cost tariff.
Many estimates are based on aggregate damage costs per kWh. But damage costs are a function of location
and affected population, of pollution control technology in place, fuel characteristics, weather patterns
(particularly on islands), and stack height. Aggregate $/kWh assessments therefore have little credibility.
Most comprehensive studies are for European Union and/or Europe and the US, which for a number
of reasons are difficult to extrapolate to developing countries. The most common method is to scale
impacts by per capita gross domestic product (GDP) (the so-called benefit-transfer method).
However, this method does not work within the European Union, so there is even less reason to think
it would work across the entire globe. Figure 4.1 shows the relationship between per capita GDP and
one component of damage costs (in /kg NOx emissions).
The lack of correlation is self-evident, testimony to the fact that damage costs are a function of a wide
range of factors, including local climate and weather patterns (in a country such as Indonesia, with many
islands subject to monsoonal climate, much of the local pollution is blown out to sea), lifestyles (in the
US, people spend a much greater proportion of their time indoors than in developing countries), and
health status (poor populations in developing countries tend to have a lower level of health than their
counterparts in the US and Europe, and are therefore more susceptible to pollution-related illness).
Notwithstanding these problems, several such estimates have been used in Indonesian renewable
energy project appraisals. For example, the 2011 Project Appraisal Document for the Geothermal Clean
Energy Investment Project used a value of health damage costs of $0.00546/kWh in its economic
analysis (Table 4.6). A recent assessment of the local environmental damages from Indonesian thermal
generation projects is further discussed in Box 7.
2005 Damage Cost (1,000 Euro/Kg NOx)
Figure 4.1: Nitrogen Oxide Damage Costs versus Per Capita Gross Domestic Product,
European Union
15
Germany
Austria
Hungary
Slovakia
10
Romania Croatia
Moldova
Bosnia and
Herzegovina
Ukraine
Bulgaria
Belarus
Czech Republic
Italy
Netherlands
United Kingdom
Ireland
Denmark
Spain
Latvia
Turkey
Estonia
Portugal
Malta
Belgium
Poland
Lithuania
Albania
France
Slovenia
20
Sweden
Greece
Finland
Cyprus
40
60
Per Capita GDP ($1,000)

GDP = gross domestic product, NOx = nitrogen oxides.
Source: Data from European Environment Agency. 2011. Revealing the Costs of Air Pollution from Industrial Facilities in Europe.
EEA Technical Report No. 15/2011. Luxembourg.
Box 7: Damage Cost Estimates for Indonesia

The most recent assessment of the local environmental damage costs from Indonesian thermal generation
projects appears to be Kusumawati et al.,a who studied damage costs at the Paiton coal project, the Gresik
gas project, and the Muara Karang oil project. These projects have emission factors for the main criteria
pollutants as follows:
Emission Factors (Gram/kWh)
Project
Paiton Coal
Muara Karang Oil
Gresik Gas
SO2
4.34
11.7
0
NO2
4.56
2.32
1.79
PM10
0.67
0.29
0
kWh = kilowatt-hour, NO2 = nitrogen dioxide, SO2 = sulphur dioxide, PM10 = particulate matter less than 10micrometers in diameter.
Source: W. Kusumawati, A. Sugiyono, and J. Bongaerts. 2010. Using the QUERI Model-AirPacts Program to Assess the External Costs of
Three Power Plants in Indonesia with Three Different Energy Sources. IMRE Journal. 4 (1).
Using the SIMPACT model, damage costs per kWh were estimated as follows:
Damage Costs, US/kWh (at 2010 Price Levels)
Gresik
Gas
0
PM10
SO2
NO2
Sulfates
Muara Karang
HFO
1.301
Paiton
Coal
0.207
0.517
0.016
0.051
0.063
0.008
0.148
0.042
Nitrates
0.036
0.173
0.045
Total
0.087
2.202
0.318
HFO = heavy fuel oil, kWh = kilowatt-hour, NO2 = nitrogen dioxide, SO2 = sulphur dioxide, PM10 = particulate matter less than 10micrometers
in diameter.
Source: W. Kusumawati, A. Sugiyono, and J. Bongaerts. 2010. Using the QUERI Model-AirPacts Program to Assess the External Costs of
Three Power Plants in Indonesia with Three Different Energy Sources. IMRE Journal. 4 (1).
These damage cost estimates differ slightly to those estimated by Liun et al.b who use the same SIMPACT
model as Kusumawati.
Damage Costs, US/kWh (at 2010 Price Levels) By Data Source
Kusumawati (2010 prices)

Liun et al (2007)
Gresik
Gas
0.087
0.074
Muara
Karang
HFO
2.202
Paiton
Coal
0.318
Surabaya
Coal
Tanjung Jati
Coal
0.097
0.646
HFO = heavy fuel oil, kWh = kilowatt-hour.
Both Kusumawati et al (2010) and Liun et al. (2007) use US damage cost estimates adjusted by purchase
power parity-adjusted per capita GDP. Thus, none of these estimates can be considered reliable,c and at best
are indicative of order of magnitude.
Notes:
W. Kusumawati, A. Sugiyono, and J. Bongaerts. 2010. Using the QUERI Model-AirPacts Program to Assess the External Costs of Three
Power Plants in Indonesia with Three Different Energy Sources. IMRE Journal. 4 (1).
b
E. Liun, A. Kuncoro, and E. Sartono. 2007. Environmental Impacts Assessment of Javas Electricity Generation Using SimPacts Model.
International Conference on Advances in Nuclear Science and Engineering. pp. 379384.
c
Indeed, none of these various studies appear in the peer-reviewed literature.
a
Table 4.6: Estimates of Local Health Damage Costs in Indonesia as Compared with
the Peoples Republic of China (2000 US/kWh)
Suralaya
N/a
N/a
N/a
0.200.65
TSP
SOx
NOx
Total
Peoples Republic of China

0.103
0.018
0.000
0.121
N/a = not available, NOx = nitrogen oxides, SOx = sulphur oxides, TSP = total suspended particulates, US/kWh = cents per kilowatt-hour.
Source: World Bank. 2011. Geothermal Clean Energy Investment Project. Project Appraisal Document. Washington, DC.
However, both the Peoples Republic of China and Suralaya studies cited in this table used benefit
transfer estimates from the US. The Castlerock report also quotes a study at Paitonbut this again
suffers from the same problem.
For project appraisal the recommended approach is to use the methodology proposed by the World Bank
Environment Department, where damage costs are related to kg of emissions and stack height (Table4.7).
Table 4.7: Damage Cost of Local Air Pollutants (expressed as $/ton
of pollutant per $1,000 of per capita GDP per million population)
PM10
Utility, High Stack

42
Medium Stack, Large

Industry
214
SO2
33
487
NOx
123
Self Generation
3,114
$/ton = dollars per ton, NOx = mono-nitrogen oxide, PM10 = particulate matter less than 10 microns in diameter, SO2 = sulfur dioxide.
Source: K. Lvovsky et al. 2000. Environmental Costs of Fossil Fuels: A Rapid Assessment Method with Application to Six Cities. Washington, DC:
World Bank.
However, while this approach is suitable for assessing a single proposed project, it is difficult to apply
to derive a credible average tariff premium estimate. In systems with high, unserved energy demand,
the main local health impact is from diesel self-generation sets, with uncontrolled emissions in densely
populated areas at ground leveltwo orders of magnitude greater than a utility project with modern
pollution controls, high stack, and relatively remote locations.
Conclusions
Based on this discussion, we find as follows:

Until there is a credible health damage assessment conducted for Indonesia, which is based
on local epidemiological and health data, valuations of the local environmental impact based
on the benefit transfer method are unreliable and not credible.
A de minimus charge of $0.001/kWh may be included, in recognition that the avoided

environmental impacts are not zero, and as a placeholder for possible future inclusion, but
$0.005/kWh as proposed by Castlerock has no rational basis.23
For a 2020 geothermal target of 4,000 MW, the potential impact of such a de minimus charge
on PLNs purchase costs is negligible.24
23
The Viet Nam regulator rejected a local environmental impact charge for the avoided cost tariff for renewable energy on
similar grounds: until a peer-reviewed Viet Nam-specific health damage study was available, any estimate was deemed to
be arbitrary and lack credibility.
24
This follows from the relative magnitude compared to GHG valuations. At $30/ton, the tariff impact is 2.4 US/kWh,
compared to 0.1 US/kWh for local environmental impacts (about 4%): it follows that the impact of local externalities on
the PLN subsidy would be no more than a few million US dollars.
4.8 Premium for Price Volatility

The Castlerock report first argued that fossil fuel price volatility should be factored into an avoided cost
tariff for geothermal energy.25 However, the calculations in that report reflect not price volatility (as that
term is normally understood) as much as a hedge against long-term price increases. But the avoided
production costs will already reflect the impact of higher fossil prices, if they occur. Fuel price volatility as
used in our report means fluctuations around the long-term trends.
Fuel price volatility is of particular concern to MoF, because fuel costs account for a major portion of
the total PSO subsidy provided to PLN, and MoF must budget these funds. Sudden and unforeseen
price increases can create great difficulty in fiscal management as additional funds then require
mobilization. On the other hand, a sudden decrease in price can result in unused funds, which might
have been spent on other projects (as occurred in the 2008/2009 price collapse). It is worth noting,
however, that the biggest shock to the PSO occurred as a consequence of the change in Indonesian
domestic coal pricing (coal to PLN is now priced at international market levels, rather than at the
previously subsidized price)a (desirable) policy change unrelated to international market conditions.
Oil
Global bubble
Crude oil price (Dubai), $/bbl, nominal
Figure 4.2 shows the nominal price

of crude oil (Dubai) since 1960.
One observes short-lived price
spikes (as observed in 1990/1991
at the time of the first Gulf War);
major step changes (in 1973, first
oil embargo); prolonged deviations
from trends, such as the period of
high prices between 1979 (Iranian
revolution) and the subsequent
collapse in 1986; a global speculative
bubble and its dramatic collapse in
2007/2008; and, most importantly,
the steady increase in price since
2000 that largely reflects the rise
of the Peoples Republic of China.
Smaller regular fluctuations are
superimposed around the longerterm trends.
Figure 4.2: World Oil Price, $/bbl

150
100
50
Indonesian coal is also traded on

the international markets, with
prices that are strongly correlated
25
Castlerock. 2010. Phase 1 Report.
Gulf War
Global crash
1973 Oil crisis

0
1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012
$/bbl = dollars per barrel.

Source: Organization of the Petroleum Exporting Countries (OPEC). 2014.
Annual Statistical Bulletin 2014. Vienna.
Figure 4.3: Australian Coal Prices, $/ton

200
Coal, fob Newcastle, $/ton
Coal
Short-term price volatility for coal is
less than for oil, but still subject to
the same general long-term trends
in international oil markets, as shown
in Figure 4.3.
Second oil crisis
150
100
50
1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012
$/ton = dollars per ton, fob = free on board.

Source: Platts International Coal Reports.
to Australian export coal that dominates the AsiaPacific coal trade (Figure 4.4A). Only in 2008
did Kalimantan coal prices diverge (fall below) significantly from the generally stable relationship, as
evident from Figure 4.4B. Indonesia coal futures are offered on a number of international commodity
markets, which could be used to hedge short-term volatility.
Figure 4.4: Kalimantan Coal Prices

200
120
100
Jul2008
Kalimantan, $/ton
$/ton
150
100
201 1
80
201 2
2009
60
201 0
2008
Apr2008
40
50
20
2007
2008
2009
2010
2011
2012
2007
50
100
150
200
Newcastle, $/ton
$/ton = dollars per ton.

Source: Platts International Coal Reports.
Gas
Figure 4.5: Heavy Fuel OilCoal Price Differentials

($/million BTU)
15
HFO-coal differential, $/mBTU
The urgency for PLN to reduce its

reliance on fuel oil is clear from Figure4.5,
which shows the differential between
coal and fuel oil on a BTU basis. Until the
1973 oil crisis, heavy fuel oil (HFO) was
only slightly more expensive than coal.
In the 1980s, the HFO (and crude oil)
price was propped up by Organization
of the Petroleum Exporting Countries
(OPEC) quotas, but the differential
narrowed again in 1986. However, since
2000, the Peoples Republic of China
has become a major importer of oil yet
an exporter of coal, so the differential
has widened.
10
-5
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000 2004 2008 2012
$/million BTU = dollars per million British thermal units, HFO = heavy fuel oil.
Sources: Calculated from data in Platts. 1981-2013. Platts International
Coal Report. McGraw-Hill Companies Inc. Issues 1-213. London; and
Organization of the Petroleum Exporting Countries. 19692013. Bulletin
Vols. IXLV. Vienna.
Gas prices, especially in North America, are even more volatile than oil. US Henry Hub gas prices are
especially volatile, but Asian LNG prices are less so, a consequence of the past predominance of long
term contractsthough the spot market now accounts for a growing share. The potential impact of
fracking shale gas in the US has been clearly visible since 2009, with sharp falls in the gas price. Gas
prices for the period 1960 to 2012 are shown in Figure 4.6. Most forecasts (IEA, World Bank) also see
LNG prices falling from present levels to around $8/million BTU by 2025, as the US joins the ranks of
LNG exporters.
Figure 4.6: Gas Prices, $/million BTU

20
LNG
$/million BTU
15
Europe
10
US
0
1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012
$/million BTU = $/million British thermal units, LNG = liquefied natural gas, US = United States.
Sources: World Bank, US Energy Information Administration.
Forecast Errors
By looking at the historical level of annual price fluctuations, one can calculate the shortfall or surplus
to budget for the PSO subsidy from MoF to PLN. There are, obviously, many different forecasting
rules, so for illustrative purposes we assume the forecast for year n is based on the average price of the
past 3 years (n-3, n-2, n-1). Figure 4.7 shows the forecasting error using this procedure.
Figure 4.7: Coal Price Forecasting Error, $/ton
80
Underestimate, $/ton
60
40
20
0
20
40
1974
1980
$/ton = dollars per ton.

1986
1992
1998
2004
2010
The calculations are shown in Table 4.8. The cost of the forecast error (as an absolute value, with
overestimate being equally undesirable as underestimate) is shown in column [4]. With Newcastle
coal at 6,300 Kcal/kg and a 2,200 Kcal/kWh heat rate, 0.35 kg of coal are needed per kWh. Therefore,
for example, if the error is $25/ton coal (2004), this would translate to 0.87 US/kWh. Since the cost
of coal is 1.85 US/kWh, in 2004 the error is almost 50% of the total.
Table 4.8: Forecast Errors
Forecast
Forecast
Error
=[1]-[2]
Error as
absolute
value
Forecast
Error
Geothermal
Benefit
$/ton
$/ton
$/ton
$/ton
US/kWh
US/kWh
[1]
[2]
[3]
[4]
[5]
[6]
2004
52.9
27.9
25.0
25.0
0.87
1.85
2005
47.6
34.8
12.8
12.8
0.45
1.66
2006
49.1
42.2
6.9
6.9
0.24
1.71
2007
65.7
49.9
15.8
15.8
0.55
2.30
2008
127.1
54.1
73.0
73.0
2.55
4.44
2009
71.8
80.6
8.8
8.8
0.31
2.51
2010
99.0
88.2
10.7
10.7
0.38
3.46
Year
Actual
Newcastle,
fob
2011
121.4
99.3
22.1
22.1
0.77
4.24
2012
96.4
97.4
1.1
1.1
0.04
3.37
2013
84.8
105.6
20.8
20.8
0.73
2.96
19.7
0.68
2.85
Average
$/ton = dollars per ton, fob = free on board, US/kWh = cents per kilowatt-hour.
Conclusions
This permits the following conclusions:

Since coal is the least volatile of the three main internationally traded fossil fuels, the most
important step in reducing Indonesias exposure to fuel price volatility is to complete the
planned shift from HFO to coal as soon as possible.
Annual coal price volatility has doubled in the last 10 years.
The 10-year average of the forecast error per ton coal increased from $9.9/ton (in the
previous years) to $19.7/ton (Table 4.8), though this latter average includes the 20082009
commodity price bubble and its collapse.
Forecast error per kWh of geothermal energy increased from 0.346 US/kWh in the previous
years to 0.68 US/kWh.
If the cost to MoF is limited to short-term borrowing, then the tariff premium is much smaller.
Note that when the price of coal increases as a matter of long-term trend, Indonesia (MoF)
already gets the benefit of the higher avoided cost.
4.9 Avoided Cost of Transmission

Opinions vary about who should be responsible for the transmission interconnection, and international
best practice suggests no clear answers. Some renewable energy independent power producers (IPPs)
would rather be responsible for the interconnection themselves on grounds that this ensures that the
line will be built on time and ready for the COD of the generating project.26 At the same time, given the
many delays experienced by geothermal projects, the buyer (i.e., in Indonesia, PLN) is often reluctant
to invest its resources in a transmission line only to discover the IPP is not ready.
Transmission interconnection costs are small. As shown in Appendix 2, average costs for geothermal
projects (as shown in PLNs RUPTL 2012) are around $50/kW, so with total capital investment of
$4,500/kW for typical geothermal generation projects, this will have little impact on the tariff. The
same is true for the interconnection costs of thermal projects: for large coal projects in Java, the
average is $27/MW.
On balance, we recommend that geothermal project developers be responsible for the construction
of the transmission line, but that the tender bid is exclusive of transmission costs. The tariff ceiling
would therefore also exclude the avoided costs of connecting fossil fuel projects.
If PLN agrees that the developer incurs the transmission costs, then those costs would be recovered
by a nonescalating transmission adder to the tariff to provide recovery of the capital costs over a
10-year period.27 Given that this will be a very small amount compared to the generation cost, bidders
need not compete on this at time of tender. In general we propose that the transmission connection
Sample Calculation: Transmission Adders

Assumptions: 110 MW, 90% annual capacity factor, 867 GWh per year. Developers transmission line outlays
$3 million in year 5, $2 million in year 6. Commercial operation date is 1 January of year 7. Weighted average
cost of capital at financial closure is 8.8%.
Net present value at weighted average cost of capital, of the transmission line outlays is $3.2 million. The
tariff adder that makes the NPV of the transmission recovery payments equal to this sum is 0.094 US/kWh
(10 annual payments of $0.81 million).
NPV
Transmission line
($ million)
3.2
Annual generation
(GWh)
Tariff adjustment
(US/kWh)
Recovery of costs
($ million)
3.2
Transmission cost
($/kW)
45.5
14
3.00
2.00
10
11
12
13
14
15
16
867
867
867
867
867
867
867
867
867
867
0.094 0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.81
0.81
0.81
0.81
0.81
0.81
0.81
0.81
0.00
0.81
0.81
17
$/kW = dollars per kilowatt, GWh = gigawatt-hour, MW = megawatt, NPV = net present value, US/kWh = cents per kilowatt-hour,
WACC = weighted average cost of capital.
26
That this is of real concern to developers is shown by the experience of the Peoples Republic of China, where 20% of
all large wind farms currently stand idle for lack of transmission evacuation capacity. Moreover in Indonesia, there were
serious delays in providing the transmission lines for the Wayang Windu and Darajat II projects.
27
Under the assumption that PLN is responsible for maintaining the transmission line.
be handed over to PLN on the COD, and that PLN be responsible for transmission O&M (given that
the meter is likely to be at the generating plant). Given that these costs will, in general, be quite small
compared to the generating project and steam field development, what is important is simply that the
proposed transmission arrangements be clearly specified at time of tender.
The calculation is straightforward: the adder is that value that makes the NPV of the stream of
transmission line outlays equal to the NPV of the 10 years of tariff recovery payments, that yearly
payment being the value of the adder multiplied by the expected annual generation.
4.10 Local Economic Development Benefits

There is much evidence that geothermal project development brings local and regional economic
benefits, given that some fraction of the construction and operating cost is spent locally, which boosts
local economic activity through multiplier effects. This is recognized in the economic literature in the
form of regional income and employment multipliers.
Although we know of no Indonesia-specific employment and regional income multipliers for
geothermal and coal projects, there are many studies in the international literature that suggest
renewable energy in general, and geothermal energy in particular, is more employment intensive than
thermal energy.28 Typical regional expenditure multipliers are in the range of 1.5 to 2.5, meaning that
for every Rp1,000 spent in the local economy, another Rp1,500Rp2,500 of expenditure is induced
(since those additional Rp are in turn spent by the first recipient in making further local purchases).
Thermal projects would have similar multiplier impacts, if not at the power plant site itself, then in coal
transportation and mining.29 However, the literature suggests that these multipliers are much lower
than for renewable energy.
We therefore propose that the tariff include a premium for the differential local provincial economic
development benefit that accrues from local expenditures on geothermalso for an average multiplier
of 1.5, a differential multiplier of 0.75 for geothermal (meaning that for every Rp1,000 spent locally
in a province, another Rp750 of local economic activity will be induced). A discussion about why
to use deterministic and not probabilistic calculations is included in the response to Stakeholder
Comment 4.
Table 4.9 shows such a calculation. The key assumptions are as follows:

Total expenditures are divided into three categories: exploration and construction (with a
local Rp share of 25%), routine O&M (with a local Rp share of 75%), and make-up well drilling
during operation (with a local Rp share of 25%).
The proportion of Rp costs spent in the local province is assumed at 75% for Java and Sumatra,
40% for eastern islands. The smaller share for the latter simply reflects the fact that small
islands may not have any qualified local firms that can offer the required goods and services,
and that these would be sourced from other parts of Indonesia.
The total local expenditures are taken from a World Bank project as an example of a typical
geothermal project.
28
R. Bacon and M. Kojima. 2001. Issues in Estimating the Employment Generated by Energy Sector Activities. World Bank,
Sustainable Energy Department; M. Wei, S. Patadia, and D. Kammen. 2010. Putting Renewables and Energy Efficiency to
Work: How Many Jobs Can the Clean Energy Industry Generate in the US? Energy Policy. 38. pp. 919931.
29
For example, the study by Wei, Patadia, and Kammen (2010), calculates average job creation for coal of 0.11 job-year per
GWh, but 0.25 job-years per GWh for geothermal.

Stakeholder Comment 4: Why Use Deterministic and Not Probabilistic Calculations?
Comment:
Setting the expected value of benefits as the ceiling at 50% probable value (P50) forces bidders (through
competitive tendering) from the expected value and into a failed project zone.a Ceiling tariffs should
be set at P90. Using a probabilistic method, when the P50 value (for 2020, against large coal) is 14.53
US/kWh, the P90 value is 16.24 US/kWh.
Reply:
Our calculation for 2020 is 14.7 US/kWh, which is very close to a probabilistic calculation. However, if the
recommendations we propose for improvements to the tendering process are implemented, it is doubtful
that use of any ceiling, however derived, would result in more failed projects. For example, suppose
qualified bidders assess the cost for some project at around 16 US/kWh, when the ceiling is 14.7 US/
kWh. They will not bid 14.6 US/kWh, just below the ceiling, since a qualified bidder will know that at the
lower price, the project cannot succeed. It may well be the case that a speculator or unqualified party
bids 14.6 US/kWh (or just below)and at this price the project is not viable, so the project fails. But if
indeed 14.7 US/kWh is the governments best estimate of the benefits, then the appropriate mitigant is
not to raise the ceiling to a P90 value of 16.24 US/kWh (as suggested)at which point indeed a good
developer would bid and develop a successful projectbut to exclude the unqualified bidders by rigorous
prequalification screening and a material bid bond. The point is not to set the tariff at a level to permit
successful development by a qualified bidder, but to limit geothermal development to those projects for
which costs are at or below the governments best estimate of the benefits.
Probabilistic (Monte Carlo) analysis is a useful technique to generate a range of answers based on
statistical variations of the inputs. However, it does require a reasonable understanding of the likely
variation of the input parameters, with the actual calculation then being undertaken a great many times
(e.g., 1,000 independent calculations would be a minimum) to produce a statistically meaningful series
of possible output results.
In the case of calculating the geothermal ceiling price, where this is based on the avoided cost of nongeothermal generation, the majority of the input parameters are not subject to variations that lend
themselves to statistical descriptions. Some are based on actual values provided from PT Perusahaan Listrik
Negara (e.g., fixed costs of thermal generation), some are simply fixed values selected to be representative
of additional loadings that are being applied to the calculation (e.g., value of carbon) and some are forward
projections taken from very limited published data (e.g., forward cost of coal). The actual value and the
fixed value inputs are not subject to statistical variation. The forward projection numbers are subject to
some statistical variation, but analyzing that variation is more easily undertaken by the simple sensitivity
technique of running the calculation with a known variation on the selected input parameter (e.g., using
alternative International Energy Agency forecasts for future energy prices, as in Table 4.1).
For the calculation of a ceiling price, a single value is required. It is not helpful to provide a range of
possible values, and hence, particularly given the nature of the input variables, probabilistic analysis is
not appropriate for this purpose. A simple sensitivity analysis will highlight where emphasis should be
placed in the selection of the representative value of the relevant input variable (as we have done, for
example, in the case of the valuation of greenhouse gas emissions in Table 4.5), but the end result still
needs to be presented as a single tariff number representing the benefit of geothermal, above which the
costs of geothermal outweigh the benefits. It may be noted that the most important assumption in the
benefits calculation is the valuation of carbon, which is a decision that must be directly confronted by
the government (see stakeholder comments on greenhouse gas valuation)and which should not be
buried among the many other variables in a probabilistic analysis.
Comments provided by Chevron, 11 March 2014. Similar comments were received from Indonesia Geothermal Association.
The tariff benefit is calculated such that the NPV of the induced benefits (row [10] of
Table 4.9, $75.5 million) is the same as the NPV of the tariff recovery (row [13]). This
calculated value is 1.8 US/kWh.
Table 4.9: Local Multiplier Impacts (for Java and Sumatra)
NPV 2014 2015
Rp portion
Exploration,
construction
0.25
$ million
Routine O&M
0.75
Make-up wells
Local province
shares
2016
2017 2018 2019
2020
2025
Rp share
6.3
0.0
0.0
$ million
8.0
8.4
8.8
11.3
0.25
$ million
0.0
0.0
4.2
0.0
Exploration,
construction
0.75
$ million
10.5
18.0
13.9
16.2
4.8
0.0
0.0
0.0
Operation
0.75
$ million
0.0
0.0
0.0
0.0
6.0
6.3
6.6
8.4
Make-up wells
0.75
$ million
0.0
0.0
0.0
0.0
0.0
0.0
3.2
0.0
Total provincial
expenditures
$ million
10.5
18.0
13.9
16.2
10.8
6.3
9.8
8.4
10
Multiplier
7.9
13.5
10.4
12.1
8.1
4.7
7.3
6.3
11
Energy sold
GWh
887.0
887.0
887.0
12
Tariff
increment
US/
kWh
1.80
1.80
1.80
13
Tariff recovery
$ million
75.5
16.0
16.0
16.0
14
Net cash flows
$ million
0.0
11.3
8.7
9.6
0.75
$ million
14.0
75.5
7.9
24.0
13.5
18.5
10.4
21.6
12.1
8.1
GWh = gigawatt-hour, NPV = net present value, O&M = operation and maintenance, Rp = rupiah, US/kWh = cents per kilowatt-hour.
The corresponding calculation for Indonesias eastern islands results in a tariff benefit equivalent of
0.97 US/kWh. These values are included in the total tariff ceiling build-up (Table4.10).
4.11 Proposed Tariff Ceilings

Table 4.10 shows the proposed tariff ceilings. These should be revised and published annually, and
would apply to all tenders issued during the year in question.30 MEMR would determine on a case-bycase basis which table applies (large grid, small island coal, small islands oil) to a particular tender, and
the tender entity determines the target COD. For example, if this table were to apply to a tender, say,
in November 2014, and the target COD were 2021, and the relevant region were eastern island, oil;
then the tariff ceiling would be 28.8 US/kWh.
30
In other words, if this table were published in mid-June 2014, then it would apply to all tenders issued between 1 July 2014
to 30 June 2015.

Table 4.10: Proposed Tariff Ceilings
2013
2014
2015
2.89
2.96
3.04
3.11
3.19
3.27
3.35
3.44
3.52
3.61
3.70
3.79
3.89
3.85
4.00
4.16
4.33
4.50
4.68
4.87
5.06
5.23
5.41
5.59
5.77
5.97
US/kWh
2.70
2.76
2.83
2.90
2.98
3.05
3.13
3.20
3.28
3.37
3.45
3.54
3.63
Local environmental
premium
US/kWh
0.10
0.10
0.11
0.11
0.11
0.11
0.12
0.12
0.12
0.12
0.13
0.13
0.13
Energy security
premium
US/kWh
0.68
0.71
0.74
0.77
0.80
0.83
0.87
0.90
0.93
0.96
0.99
1.02
1.06
Local eocnomic
development
US/kWh
1.80
1.85
1.89
1.94
1.99
2.04
2.09
2.14
2.19
2.25
2.30
2.36
2.42
Total benefit, coal,

large grids
US/kWh
12.0
12.4
12.8
13.2
13.6
14.0
14.4
14.9
15.3
15.7
16.2
16.6
17.1
Isolated grids/Small islands
10
Avoided fixed cost
US/kWh
5.06
5.18
5.31
5.45
5.58
5.72
5.87
6.01
6.16
6.32
6.47
6.64
6.80
11
Avoided variable cost
US/kWh
6.66
6.74
7.02
7.32
7.63
7.95
8.29
8.64
8.95
9.27
9.61
9.96 10.32
12
GHG emission
premium
US/kWh
4.06
4.16
4.26
4.37
4.48
4.59
4.70
4.82
4.94
5.07
5.19
5.32
5.45
13
Local environmental
premium
US/kWh
0.10
0.10
0.11
0.11
0.11
0.11
0.12
0.12
0.12
0.12
0.13
0.13
0.13
14
Energy security
premium
US/kWh
0.68
0.71
0.74
0.77
0.80
0.83
0.87
0.90
0.93
0.96
0.99
1.02
1.06
15
Local economic
development
US/kWh
0.97
0.99
1.02
1.04
1.07
1.10
1.12
1.15
1.18
1.21
1.24
1.27
1.30
16
Total benefit, isolated

grids
US/kWh
17.5
17.9
18.5
19.1
19.7
20.3
21.0
21.6
22.3
23.0
23.6
24.3
25.1
17
Eastern islands, oil
18
Avoided fixed cost
US/kWh
2.48
2.55
2.61
2.68
2.74
2.81
2.88
2.95
3.03
3.10
3.18
3.26
3.34
19
US/kWh
19.02 19.05
19.99 20.23 20.48 20.71 20.95
Large grids
Avoided fixed cost
US/kWh
US/kWh
GHG emission
premium
2016
2017
2018
2019 2020 2021
2022 2023 2024 2025
19.28
19.52
19.75
21.19
21.43
21.67
20 GHG emission
premium
US/kWh
2.35
2.41
2.47
2.53
2.60
2.66
2.73
2.80
2.87
2.94
3.01
3.09
3.17
21
US/kWh
0.10
0.10
0.11
0.11
0.11
0.11
0.12
0.12
0.12
0.12
0.13
0.13
0.13
22 Energy security
premium
US/kWh
0.68
0.71
0.74
0.77
0.80
0.83
0.87
0.90
0.93
0.96
0.99
1.02
1.06
23
Local economic
development
US/kWh
0.97
0.99
1.02
1.04
1.07
1.10
1.12
1.15
1.18
1.21
1.24
1.27
1.30
24 Total benefit, Eastern

Islands oil
US/kWh
25.6
25.8
26.2
26.6
27.1
27.5
27.9
28.4
28.8
29.3
29.7
30.2
30.7
Local environmental
premium
US/kWh = cents per kilowat-hour.

Adjusting Tariff Ceilings for Government-Funded Exploration Costs

In the event that government has paid for exploration drilling as a public good (for example under the
auspices of the Geothermal Fund) prior to tender, then the tariff ceiling shall be adjusted downward by
an amount that makes the NPV of the exploration expenditures equal to the NPV of the tariff ceiling
adjustment, where NPVs are calculated at the governments opportunity cost of capital.
If applicable, the ceiling tariffs shall be adjusted downward by the amount , in $/kWh, such that the
following equation is satisfied:
NPV(E) = NPV(R, )
where:
NPV(E)
NPV of the stream of annual expenditures on pre-tender exploration

drilling, evaluated at the discount rate corresponding to the
governments opportunity cost of capital
NPV(R, )
NPV of the tariff recovery payments, evaluated at the discount rate

corresponding to the governments opportunity cost of capital
NPV(R, ) is to be calculated as:
NPV (R , ) =
i
g i
(1 + r )i
where
gi = Average annual electricity sold in year i, in kWh per year
R = Discount rate corresponding to the governments opportunity cost of capital
The calculation is illustrated in Table 4.11. It is assumed that up-front exploration expenditures are
$5million in Year 1, $15 million in Year 2, and $10 million in Year 3. At the opportunity cost of capital
of 7%, the NPV is $25.9 million. For a 110 MW project at 0.9 load factor, the cost is recovered across
867 GWh/year. A value of 0.64 US/kWh makes the NPV of the cost recovery stream exactly equal to
the NPV of exploration. The tariff ceiling would therefore be adjusted downward by 0.64 US/kWh.
Table 4.11: Sample Calculation, Tariff Ceiling Adjustment
Exploration expenditures ($ million)
NPV
25.9
15
10
Annual generation (GWh)

Tariff adjustment (US/kWh)
Recovery of costs ($ million)
10
11
12
13
14
15
16
867
867
867
867
867
867
867
867
867
867
0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64 0.64
25.9
0.0
0.0
0.0
0.0
0.0
0.0
5.2
5.2
5.2
5.2
5.2
5.2
5.2
5.2
5.2
GWh = gigawatt-hour, US/kWh = cents per kilowatt-hour.

However, if the government recovers the up-front de-risking from the developer at the time of financial
closure (see Section 8), then no adjustment to the ceiling is required.
5.2
17
Power Purchase Agreements

The prospective PPA should be provided at the time of tender. All relevant terms and conditions, and
in particular the schedules relating to the tariff, should be fixed in advance. The present system of ad
hoc, post tender negotiation of tariff escalation terms should be dropped: a single tariff escalation
formula should be adopted as part of the tender, and should be binding on both parties. However,
there is a strong case to renegotiate the tendered price when there are delays outside the developers
control, and when the resource as revealed by delineation drilling proves to be significantly lower than
that expected at tender.
5.1 Escalation and Indexation

Current Practice
The most recent formulae for tariff escalation, as negotiated between PLN and developers, take the
form31
P(t) = [(1 )P(0)] + [P(0)USPPI(t)/USPPI(0)]
where
P(t)
P(0)
USPPI(t)
USPPI(0)

=
=
=
=
=
tariff in year t, in US/kWh

(base) tariff at COD = Tariff bid at time of tender, in year (0)
United States Producer Price Index for year (t)
United States Producer Price Index for year (0)
coefficient
is a coefficient that has been set by ad hoc negotiation and appears to be in the range of
0.25 to 0.4. As a matter of logic, should be the proportion of the cost structure attributable
to post-commissioning O&M (including the cost of make-up wells), but it is unclear that this
rationale is indeed the basis for its negotiated value in the past.
However, provided the value is set at the time of tender, and not subject to post-tender negotiation,
one might argue that the value of (or its precise rationale) does not really matter, for it can then be
factored into the cash flow forecast used by the developer to derive the bid price. All other things
equal, to produce the same equity return, a lower value of will result in a higher bid for the base price,
and would therefore matter only for bids close to the tariff ceiling. Therefore, we are of the view that a
rational evaluation of is in fact desirable.
31
46
Template of Geothermal PPA, published on the PLNs website.
Power Purchase Agreements47
International Practice
International practice reveals much variation, including:

Tariffs fixed at the time of PPA signature, with no escalation (an option offered in the Sri
Lanka FITs);
Tariffs fixed, and declining over time (as in the case of the Malaysian FIT) called tariff
degression in the German Modelthe idea being to offer relatively generous tariffs initially,
and then reduced over time so as to incentivize early introduction;32
Tariff escalation formulae of the same type as Indonesia, but where is based on the
regulators estimate of O&M costs for the technology in question, and escalated on an index
of domestic inflation (Sri Lanka FIT);
Tiered tariffs for recovery of investment costs, with a higher tariff in the first few years, and
a much lower tariff in subsequent years (Sri Lanka FIT).33 This is important where renewable
energy projects are mainly small (in Sri Lanka the fixed FITs apply only to projects less than
10 MW), and financed locally through the commercial banking system, so interest rates are high
and tenors are 57 years. Under these circumstances, a fixed constant tariff for investment cost
recovery would result in unacceptable debt service cover ratios (DSCR);34 and
Front-loaded geothermal tariffs, with which Indonesia does have some experience. (For
example, in 1993 Unocal signed an agreement with PLN for Sarulla, albeit later cancelled by
mutual agreement, which stipulated a price of 7.6 US/kWh for the first 14 years, 5.75 US/
kWh for years 15 to 22, and 5.21 US/kWh until year 30.35 At that time the cost of coal-fired
generation was around 4.0 US/kWh.)
Such options apply only to production-cost based FITs or to purely negotiated tariffs. For avoided cost
tariffs, most are contemporaneous tariffs, under which the applicable tariff is not fixed at time of PPA
signature, but is published each year by the regulator (avoided cost tariff in Sri Lanka from 19962009,
Viet Nam). Sri Lanka has a tariff collar where the amount payable cannot be less that 90% of the
published tariff in the year of PPA signature. In Viet Nam, there is both collar and cap whereby a developer
who opts for the 90% protection of his downside gives up the potential of the corresponding upside,
under which option a developer cannot receive more than 110% of the tariff at PPA signature.
Methodology for Estimating the Escalable Fraction

Using a financial model for a World Bank geothermal project, one can disaggregate the developers
total cash flow streams in the bid tariff determination by those components that are fixed at time of
financial closure (equity, debt service payments),36 and those that occur after COD (make-up wells,
routine O&M)costs which one would expect to escalate in nominal terms. Table 5.1 shows the NPVs
for these cost streams using the projects 11.8% calculated equity returns as the discount rate.37 The
share of costs after COD is 37.6%.
32
This proposition is widely asserted by the advocates of FITswhile nominally plausible, reliable evidence that degression
actually incentivizes early adoption is scant.
33
A tiered tariff was also proposed by the winning bidder for the Malitbog plant, part of the big Leyte geothermal project in
the Philippines.
34
See further discussion of DSCR in Appendix 4.
35
GeothermEx, Inc. 2010. Assessment of Geothermal Resource Risks in Indonesia. Washington, DC: PublicPrivate Infrastructure
Advisory Faculty and the World Bank.
36
Of course, while principal repayments are fixed at time of financial closure, depending on the type of financing involved
interest rates may be variable unless hedged with an interest rate swap.
37
Note that income tax is excluded from the calculation.

Table 5.1: Classification of Costs
NPV
Discount rate for NPV
0.118
Equity ($ million)
127.0 55.3 93.0

55.4
0.0
4.3
0.0
0.0
10
Debt service ($ million)
0.0
0.0
0.0
5.0
5.4 10.5
14.6 14.4 14.1
Total construction related

($ million)
182.3 55.3 93.0
4.3
0.0
5.0
5.4 10.5
14.6 14.4 14.1
O&M, make-up wells

($ million)
110.1
0.0
0.0
0.0 10.4 10.8 26.2
Total ($ million)
292.4 55.3 93.0
4.3
0.0 15.4 16.2 36.7 26.5 42.6 27.0
O&M share, as NPV
0.376
0.0
11.8 28.2
12.9
NPV = net present value, O&M = operation and maintenance.

However, the so-calculated O&M share is a function of the discount rate: at the projects overall
weighted average cost of capital (WACC) of 8% the fraction is 0.45; at the developer equity return
target of 14% the fraction is 0.34 (Table 5.2). Thus even the lowest share of 34% is considerably above
the 25% escalable share encountered in some PPAs.
Table 5.2: Operation and Maintenance Shares as Function of Discount Rate
Discount rate
O&M share
8.0%
0.45
10.0%
0.41
11.8%
0.38
12.0%
0.37
14.0%
0.34
WACC
Post tax nominal IRR
Developer equity target
IRR = internal rate of return, O&M = operation and maintenance, WACC = weighted average cost of capital.
Conclusions on Escalation
Provided the escalation formula is known at the time of tender, and applied to all projects rather than
set by ad hoc negotiation, we see no compelling reason why current indexation practice should be
changed, but based on such estimates as are available, for the time being, we recommend a value for
of 0.375.
5.2 Renegotiation of Power Purchase Agreements

The current reality is that many projects are stalled or delayed for a variety of reasons. MEMR cannot
meet its geothermal targets, and therefore a renegotiation of a significant number of projects is
inevitable. That being said, it is better that renegotiation be subject to a formal policy applicable to
all, rather than through ad hoc negotiations, which have high transaction costs and may well add yet
further delays.
Of course it is true that where a signed PPA exists, it is for the parties concerned (PLN and the
developers) to negotiate. Nevertheless, that does not preclude MEMR from issuing guidelines that
stipulate the principles that should be applied to both when a renegotiation is warranted, and what is
subject to negotiation. This is particularly true given that PLN is in principle ill-disposed to renegotiating
tariffs. It argues that the regulations (and the law) stipulate that the tariff shall be determined by
tender, and that if the bid price cannot be achieved, then the correct remedy is to return the WKP to
the government to be re-tendered.
MEMR may be reluctant to issue an explicit policy on renegotiation for fear that it would open a
floodgate of requests to take advantage of any tariff increase. There is also the potential danger that
unsuccessful bidders may protest if the revised base price is higher than their earlier rejected tender
offers. But with so many projects stalled, PPA renegotiations cannot be avoided, and if that process
lacks transparency, it would be even more likely to lead to protests.
The biggest potential issue is that all the unserious developers will seek to renegotiate. The difficulty is
that one cannot use exploration (or delineation) drilling expenditure as the yardstick for seriousness if
drilling cannot even begin for lack of the necessary permits, which is a common cause of delay.
However, serious developers can easily be defined as those who:

post the $10 million performance bond at the time of the original tender bid;
in the absence of a bond, show evidence of logistical and drilling expenditures of at least
$10million; and/or
in the absence of past exploration expenditure of at least $10 million, are willing to post such
a bond as condition precedent for a revised PPA (and would be obliged to show evidence of
funds before a renegotiation procedure is commenced, to avoid wasting PLNs time).38
A tariff should be renegotiated after tender under the following circumstances:

Delineation drilling shows that the resource as credibly defined using a recognized
international standard is significantly different from the reference power capacity stipulated
at time of tender (significant is defined as +25%).
Financial closure cannot be reached within the time frame estimated by the developer at the
time of tender for reasons outside the control of the developer.
Projects that have reached financial closure should not be open to renegotiation.
Any renegotiated tariff shall be subject to the additional ceiling that the increase may not be greater
than 50%, and that any renegotiated base price is below the applicable tariff ceiling (for the revised
COD at which the renegotiated tariff applies).
38
Such as a standby letter of credit, that can be drawn immediately upon signature of the revised PPA, or cancelled if the
parties cannot reach agreement (in which case the business licence would in any event be withdrawn, and the project
retendered).
5.3 Adjustment for Project Size

In the case of a significantly different resource size, the adjusted tariff shall be subject to change
according to a formula of the type:
P(MWa) = P(MWx) Mwa

MWx
where
MWx
tender reference project size
P(MWx)
bid tariff for project of the tender reference

MW size, in US/kWh
P(MWa )
tariff for project of the final size
MWa
final project size
coefficient, typically in the range of 0.150.3
MEMR could calculate the applicable value of the coefficient based on its geothermal production
cost model, which coefficient should be stipulated at the time of any tender.
Alternatively, to avoid the need for statistical estimation of the above production function, one may
directly use the MEMR production cost model, and use the (linearly) interpolated values as shown in
Section 2, Figure 2.3.
It could be argued that a developer would be discouraged from developing a slightly larger project in
the knowledge that the tariff would decrease. However, the relevant criterion is not MW for the sake
of MW, but rather developing the geothermal resource with an equitable distribution of the benefits
of the countrys resource endowment.
Sample Calculation: Variation in Project Size
Suppose the following at time of tender:

Estimated size of project: 55 MW
Base price established at tender: 9.5 US/kWh
Estimated COD: 2015
Ceiling price (Section 4, Table 4.10) of 12.8 US/kWh (large grids)
Subsequent delineation drilling shows the commercially viable size to be 40 MW. What should
be the revised tariff?
According to the MEMR production cost model, a 55 MW project has a typical tariff of 10.5 US/
kWh (Section 2, Figure 2.3). The interpolated value for a 40 MW project is 11.64 US/kWh, an
increase of 10.9% over the tariff for the original 55 MW project cost.
This percentage increase is then applied to the original bid price of 9.5 US/kWh, so the adjusted
tariff would be 9.5 x (1.109) = 11.65 US/kWh.
The increase is less than 50%, and the new tariff is lower than the 12.8 US/kWh tariff ceiling
applicable to a 2015 COD.
5.4 Adjustment for Delay

In the case of delay caused by permitting or other problems outside the control of the developer, the
allowable tariff escalation should be defined by the following procedure:

Let the original tender price be P*.
Calculate the tariff for the project in question using the MEMR production cost model, using
todays most likely overnight costs for drilling, steam above ground system (SAGS) and power
plant construction, say P(o).
If the agreed delay attributable to parties other than the developer is N years, recalculate the
tariff, using the same model, and the same set of technical assumptions (number of wells and
their success rate, etc.), but with cost assumptions corresponding to best estimates N years
ago. This recalculated base price is P(N).
The allowable rate of increase is P(o)/P(N), which is then applied to the original tender
price P*.
This procedure requires estimates of what prices were N years ago. The historical record shows that
the costs of different major components escalate at different rates: the cost of drilling in particular has
risen much faster than that of power plant construction, and faster than the general rate of inflation.
Geothermal energy project costs would be classified according to the categories shown in Table 5.3,
each deflated using its own index.
Table 5.3: Escalation of Cost Categories
Source
Drilling
Average drilling costs Average of indexes noted in Table 5.4
SAGS
Steel price index

(hot-rolled coil)
www.steelprices.com
Power plant
MUV index
World Bank. www.worldbank.org

(updated 34 times/year)
US costs not included above (contingencies,

management)
US GDP deflator
Government of the United States, US

Department of Commerce, Bureau of
Economic Analysis. http://www.bea.
gov/iTable/iTable.cfm?ReqID=9&step
=1#reqid=9&step=1&isuri=1
(updated 34 times/year)
Indonesian costs not included above (land,

permits, etc.)
Indonesian GDP
deflator
World Bank. www.worldbank.org;

International Monetary Fund.
www.imf.org
GDP = gross domestic product, MUV = manufacture unit value, SAGS = steam above ground system, US = United States.
Drilling
Over the last decade, geothermal drilling costs have increased at unprecedented rates, which cannot
be explained merely by any increase in depths or general inflation. Rather, the driving force has been
the demand for drilling in the oil and gas sector, which has been reactivated by the increase in global
oil prices and (in the US) by drilling for shale oil and gas (fracking). Figure 5.1 shows geothermal drilling
costs in New Zealand.

Figure 5.1: Geothermal Well Drilling Costs
Geothermal Well Drilling Cost
14.00
Well Cost
Well Cost (NZ$ million)
12.00
Mean Well
Cost
2.7% Inflation
10.00
8.00
6.00
4.00
2.00
1970
1980
1990
2000
2010
2020
NZ$ million = million New Zealand dollars.

Note: NZ$10 million = $8.2 million.
Source: H. Hole. 2013. Geothermal DrillingKeep it Simple. Proceedings of the New Zealand Geothermal Workshop. Rotorua.
There are no published indexes for drilling geothermal wells, but a number of cost indices are available
for the oil and gas industry, dominated by US costs. These include:

Spears Drilling and Completion Services Index (proprietary);
IHS39 Drilling Index (proprietary); and
US Bureau of Labor Statistics.40
Oil and gas drilling costs are strongly correlated with oil prices and numbers of drilling rigs in operation.
Well prices increased sharply in the global boom of 20062008, and then declined as the commodities
bubble and financial markets collapsed in 2008/2009. What matters is not the absolute value of costs
or their index values, but the relative changes over time, because under the proposed procedure the
level of prices is benchmarked against the current estimate in the MEMR production cost model.
Table 5.4 shows a comparison of these various indexes, and the deflators that result.
39
The company IHS was previously known as Information Handling Services, Inc.
40
The material can be accessed on the US Bureau of Labor Statistics website. The Industry/Product code is:
PCU21311121311101. See US Bureau of Labor Statistics. Producer Price Index (PPI) - Drilling Oil and Gas Wells Industry
Drilling Oil, Gas, Dry, or Service Wells. https://www.quandl.com/BLS/PCU21311121311101-Producer-Price-Index-PPI
-Drilling-oil-and-gas-wells-Industry-Drilling-oil-gas-dry-or-service-wells
Table 5.4: Drilling Index Deflators (Relative to Current 2014 Prices)

2003 2004 2005 2006 2007 2008 2009
100
IHSa
Deflators
New Zealand Geothermalb
2014
(est)
152
169
171
169
173.2
6.0%
15.1%
4.1%
7.1% 17.6% 13.8% 10.1%
11.2%
1.2% 1.2%
2.5%
0.58
0.61
0.70
0.73
0.79
0.92
0.80
0.88
0.98
0.99
0.98
1.00
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.3
10.0
10.3
10.6
11.1% 10.0%
9.1%
8.3%
7.7%
7.1% 10.7% 20.5%
3.0%
2.5%
0.47
0.57
0.62
0.66
0.71
0.95
0.98
1.00
383
0.38
Deflators
2013
127
12.5%
0.43
0.52
138
2012
122
Change
160
2011
106
Change
136
2010
US BLS PPIc
0.79
353
429
440
8.5% 12.0%
2.5%
0.80
0.87
0.98
1.00
1.08
1.1
1.08
1.025
Change
Deflators
Spearsd
1.16
0.8
Change
31.0%
1.13
Deflators
0.78
0.87
8.7% 24.1%
1.9% 1.8% 5.1%
0.85
1.07
1.05
1.05
IHS = company formerly known as Information Handling Services, Inc.; US BLS PPI = United States Bureau of Labor Statistics producer price index.
Sources:
a
IHS. www.ihs.com
b
See Figure 5.1.
c
Government of the United States, Department of Labor. Bureau of Labor Statistics website. www.bis.gov
d
Spears and Associates, Inc. January 2014. Drilling and Completion Services Cost Index. 4th Quarter 2013. Tulsa, OK.
Steam Above Ground System

We propose to index SAGS costs with international steel prices (Figure 5.2). Since steel pipe is
generally made from hot rolled coil, this is an appropriate index to use for this purpose.
Figure 5.2: Global Steel Price Trends
1200
1000
$/ton
800
600
400
200
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
cr coil
cr = cold rolled, hr = hot rolled.
Source: Steelprices.com. www.steelprices.com
hr coil
wire
rebar
1.00
Power Plant Costs

Power plant costs are proposed to be indexed by the MUV Index, published by the World Bank (which
also provides values for the US GDP deflator). This is a measure of the price of developing country
imports of manufactured goods in US dollar terms, and can be downloaded from the World Bank
website (Table 5.5).41
Table 5.5: Manufacture Unit Value Index
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
Manufactures U nit Value (MU V) Index

Release date: Jan 6, 2014
expressed in U.S. dollar terms
History 1960-2012; projections 2013-2025
email: gcm@worldbank .org
MUV Index
U .S. GDP deflator
Index 2010 = 100 ch%
Index 2010 = 100 ch%
83.32
3.8
66.05
82.49
-1
68.24
83.89
1.7
69.8
86.62
3.3
71.46
83.83
-3.2
72.98
91.84
9.6
74.5
90.15
-1.8
75.86
85.91
-4.7
77.16
82.17
-4.4
78
80.6
-1.9
79.11
79.56
-1.3
80.91
76.58
-3.8
82.76
75.68
-1.2
84.04
79.62
5.2
85.72
85.03
6.8
88.07
87.71
3.2
90.89
89.93
2.5
93.68
95.44
6.1
96.17
102.84
7.8
98.05
96.46
-6.2
98.81
100
3.7
100
108.89
8.9
101.96
107.53
-1.2
103.75
106.06
-1.4
105.93
107.73
1.6
108.31
108.96
1.1
110.98
110.47
1.4
113.22
112.01
1.4
115.51
113.65
1.5
117.85
115.38
1.5
120.23
117.21
1.6
122.65
119.1
1.6
125.13
121.05
1.6
127.66
123.05
1.7
130.24
125.09
1.7
132.87
127.17
1.7
135.55
3.7
3.3
2.3
2.4
2.1
2.1
1.8
1.7
1.1
1.4
2.3
2.3
1.5
2
2.7
3.2
3.1
2.7
2
0.8
1.2
2
1.8
2.1
2.2
2.5
2
2
2
2
2
2
2
2
2
2
ch% = percentage change, GDP = gross domestic product, U.S. = United States.
Source: World Bank. Manufactures Unit Value Index. www.worldbank.org. Accessed 6 January 2014.
41
MUV is a composite index of prices for manufactured exports from the fifteen major developed and emerging economies
to low- and middle-income economies, valued in US dollars. For the MUV (15) index, unit value indexes in local currency
for each country are converted to US dollars using market exchange rates and are combined using weights determined by
the share of each countrys exports in Group of 15 (G-15) exports to low- and middle-income countries.
The resulting index values are shown in Table 5.6. Rows [1] to [6] show the actual index values,
and rows [7] to [12] show the deflatorsi.e., the past values of the index when the 2014 value is
setat1. For example, if a project were delayed by 5 years, then the index values of 2009 would apply
(as highlighted in the table)so for example, drilling costs would be taken at 0.67 of the 2014 cost
estimate; power plant costs at 0.93 of the 2014 cost estimate, and so on.
Table 5.6: Manufacture Unit Value Index (Index Values)
Index Values
Land and
Permits
Indo
GDP
defl.
Drilling
Average
costs
SAGS
2005
2006
2007
2008 2009
6
2010
2011
2012
2013
2014
0.76
0.78
0.81
0.83
0.86
0.89
0.91
0.94
0.97
1.00
5.0
5.5
6.0
6.5
7.0
7.5
8.3
10.0
10.3
10.5
Steel
prices
52.5
53.4
56.0
59.6
56.7
58.3
61.4
60.4
60.7
62.3
Power Plant MUV
100.0
101.2
106.0
112.8
106.4
109.1
114.5
110.8
111.1 114.0
Other
US GDP 100.0
deflator
103.3
106.3
108.6
109.6
110.7
112.3
114.8
117.4
119.8
Deflators
Land and
Permits
0.76
0.78
0.81
0.83
0.86
0.89
0.91
0.94
0.97
1.00
Drilling
0.48
0.52
0.57
0.62
0.67
0.71
0.79
0.95
0.98
1.00
10
SAGS
0.84
0.86
0.90
0.96
0.91
0.94
0.99
0.97
0.98
1.00
11
Power Plant
0.88
0.89
0.93
0.99
0.93
0.96
1.00
0.97
0.97
1.00
12
Other
0.83
0.86
0.89
0.91
0.91
0.92
0.94
0.96
0.98
1.00
Indo GDP defl. = Indonesia gross domestic product deflation, MUV = manufacture unit value, SAGS = steam above ground system, US GDP =
United States gross domestic product.
Source: World Bank. Manufactures Unit Value Index. www.worldbank.org. Accessed 6 January 2014.
The illustrative calculations presented here use the data from a World Bank geothermal project and a
(simplified) financial model.
Rows 26 of Table 5.7 contain the investment costs over a 5-year period. The total is $359.1 million,
assuming that these are the 2014 costs. Assume further that the original bid (or negotiated) price
using these costs was 8.5 US/ kWh, and that the applicable tariff ceiling at the time of the bid was
10.4 US/kWh. Using a production-cost-based financial model and 2014 costs (i.e., the costs shown
in rows 27), the estimated tariff calculates to 9.47 US/kWh.
Table 5.8 shows the calculation for a range of delays, all calculated with the same production cost
model. For no delay, the cost is $359 million. For a 1-year delay, using the deflators for a 1-year delay as
shown in Table 5.6, the cost is only $351 million and tariff calculates to 9.29 US/kWh. The difference
in tariff is 0.17 US/kWh, or 1.87% of the 2013 price. Therefore, this increase is allowed on the original
bid price, and the new base price for the revised PPA would be 8.5 US/kWh X 1.0187 = 8.66 US/kWh.

Table 5.7: Adjusted Costs, $ million
$ million
1
Current Estimates
Land and Permits
Drilling
SAGS
Power Plant
Other
Total Initial Investment
Make-up Wells
Adjusted
10
Delay, years
11
Land and Permits

Drilling
13
SAGS
14
Power Plant
15
Other
16
Total Initial Investment
17
Make-up Wells
13.9
0.0
0.0
Total
Cost
1.0
0.475
0.475
139.2
69.6
55.7
31.8
0.0
6.8
8.7
11.6
4.7
152.4
0.0
36.7
43.4
57.9
14.5
35.7
8.0
10.9
7.0
7.4
2.4
77.6
110.1
73.1
76.8
21.6
359.1
18.0
5
12
Deflator
0.8
0.1
0.86
0.4
0.4
0.0
0.0
0.0
92.8
46.4
0.67
46.4
37.1
9.3
0.0
0.0
29.0
2.8
0.91
0.0
6.2
7.9
10.5
4.3
142.3
10.1
0.93
0.0
34.2
40.5
54.0
13.5
32.6
3.0
0.91
7.3
10.0
6.4
6.8
2.2
296.7
62.4
53.7
87.5
64.2
71.3
20.0
0.66
12
SAGS = steam above ground system.

Table 5.8: Allowable Tariff Increase
2014
Delay
Increase
Tariff
Years
$ million
Change in Tariff
Change
Total Cost
US/kWh
$ million
9.47
359
2013
8.0
9.29
0.17
1.87%
351
2012
13.4
9.15
0.31
3.31%
346
2011
31.2
8.60
0.87
9.15%
328
2010
51.1
8.12
1.34
14.21%
308
2009
62.4
7.84
1.63
17.19%
297
2008
59.3
7.81
1.66
17.55%
300
2007
77.6
7.41
2.06
21.73%
282
2006
92.8
7.06
2.41
25.43%
266
US/kWh = cents per kilowatt-hour, US = United States.

Similarly, for a 5-year delay, the allowable tariff increase would be 1.63 US/kWh, or 17.19% of the price
had it been calculated 5 years ago. The new base price would therefore be 8.5 US/kWh X 1.1719 =
10.26 US/kWh. The increase is below the threshold of 50%, and the new tariff is below the applicable
tariff ceiling of 10.4 US/kWh. Note that the developer does not need to reveal his estimates of capital
costs: the increase is applied to his original tender price.
Are there perverse (and unintended) incentives in such a procedure? Obviously, the longer the delay,
the greater is the tariff increase that is allowable, so it might appear that a developer has an incentive
to delay, and particularly so in the case of a unserious developer who has neither posted a bond, nor
incurred any exploration expenses. Since the delay interval is fixed only once, the reason for the force
majeure has been cured, it may be thought that there is a disincentive to cureat least as long as it
takes for the renegotiated price to reach the tariff ceiling.
However, we offer the following rejoinders to such concerns:

Under our recommendations, developers have to post a $10 million bond as a precondition for
any PPA renegotiation. This they recover only once expenditures for exploration or delineation
drilling are actually incurred: that sum should be sufficient to discourage speculators.
The delay interval is only that which can be shown to be outside the control of the developer,
as certified by an independent expert engaged by the tender entity42a precondition of
which is that the force majeure is cured so that project development can resume.43 A
developer cannot know for sure what the independent determination of the delay will be.
Moreover, since declaration of force majeure requires a declaration of what best efforts to
cure are being proposed, this provides additional evidence to the evaluator to judge whether
or not good faith efforts to cure can be shown: and if a developer has not acted according
to this plan, or the plan to cure is unreasonably protracted, then the share of delay that is
attributable to government can be adjusted downward.
PPA renegotiations can drag on for years, which cannot be in the interest of government or
developers. The proposed procedure reduces the negotiation to confirmation of the period
and cause of delay, after which the actual tariff adjustment is then mechanistic and not
subject to protracted debate.
Adjustment of Power Purchase Agreements for Proposed Pertamina

Geothermal Energy Partnerships
PGE is presently negotiating commercial partnerships for some of its old projects, but these are
unlikely to be viable at the old, previously agreed tariffs.
We recommend that the tariffs in these cases be recalculated in two stages. In stage one, the originally
agreed price would first be adjusted for the new WACC, which may be significantly different to the
original estimate with the entry of private partners who would contribute significant equity up front.
The adjustment factor is simply the new WACC (with the commercial partner) divided by the original
WACC at the time the first tariff was negotiated. The WACC-adjusted base tariff therefore follows as:
WACC-adjusted base tariff = BT*(1-)*WACCAF+ BT
where
a
proportion of the net present value of all project costs attributable to

O&M and make-up well expenses (see Section 5.1).
BT
original agreed tariff, in US/kWh.
In the second stage of the adjustment, the WACC-adjusted base tariff would be adjusted for delay,
using the same methodology as the competitively tendered projects (see above), and subject to the
same tariff ceiling.
42
The independence of the expert so engaged by the tender entity is an issue, and may need to be an individual international
expert, rather than local Indonesia expert.
43
In this context, force majeure means any government force majeure, whether formally enumerated in the PPA or not, that
is outside the control of PLN (as the counterparty in the PPA). For example, PLN has no control over the timely issuance of
environmental and forestry permits, and unreasonable delays in such issuance, not attributable to the developer, would be
grounds for an adjustment in price under this provision.
5.5 Procedure
We recommend that in the future, tenders should be conducted by a new central entity on behalf of
central and provincial governments.
The recommended procedure in the case of delay is as follows:

The developer notifies the buyer, and the independent tender entity, of force majeure as
required by the PPA, and that therefore he cannot meet the agreed COD. This requires the
developer to declare what efforts he is taking to cure the problem.
The delay in question must be at least 1 year.
The new tender entity44 appoints an independent technical expert to assess the validity of
a claim that the delay has occurred for reasons outside the control of the developer, and to
recommend the number of years that can reasonably be attributed to government.
The final value of the period of delay may only be known once the condition of force majeure
is cured.
Once the period of delay is agreed, and the project moves forward again, the tender entity
calculates the recommended value of the revised base price, using its production cost model,
as described above, but in any event no greater than the applicable tariff ceiling, and no
greater than 50% of the original tender price. If either of these two ceilings are exceeded, the
project would need re-tendering unless the developer accepted the ceiling price.
It is recognized that the establishment of this new tender entity may take some time, during which
time MEMR should assume the duties as described in the previous paragraphs, in collaboration with
local governments (who own the site) and MoF.
44
See Section 7 on the proposal to create a new central tender entity.
Recovery of Incremental Costs

In Indonesia, the mechanism for recovering the incremental cost of geothermal energy is the existing
subsidy mechanism from MoF to PLNin effect, the incremental costs are borne by government,
not by consumers. At some point it may be that the subsidy is eliminated and the consumer tariff
becomes fully cost-reflective, but that is unlikely in the short term. Consequently the focus here is on
the magnitude of additional subsidy that will be required from MoF if the pace of geothermal energy
development accelerates, as called for by the FTP2 program.
6.1 Cost of Existing Projects

There is general agreement that many of the low cost geothermal fields have already been developed:
many of these now operating projects have costs below that of coal. Tariffs at the existing projects are
shown in Table 6.1.
Table 6.1: Tariffs at Existing Geothermal Projects
COD
MW
US/kWh
Muara Laboh
2017/2018
220
9.40 (PPA)
Sarulla 1
2017/2018
330
6.79 (PPA)
Rajabasa
2020/2021
220
9.50 (PPA)
Rantau Dedap
2019
220
8.86 (PPA)
Blawan Ijen
2019
110
8.58 (PPA)
Atadei
2016
9.50 (PPA)
Ungaran
2019
55
8.09 (PPA)
Sorik Marapi
2019/2020
240
8.10 (tender price)
Suoh Sekincau
2020/2021
220
6.90 (tender price)
Cisolok Cisukarame
2019
50
Rp630 (tender price)
Jaboi
2019
10
Rp1,705 (tender price)
Tangkuban Perahu
2019
110
Rp533.6 (tender price)
Jailolo
2017
10
Rp1,727 (tender price)
Sokoria
2017/2019
Rp1,250 (Tender price)
Rawa Dano
2019
110
8.39 (tender price)
Tampomas
2019
45
6.50 (tender price)
Batu Raden
2018/2019
110
9.47 (tender price)
Ngebel/Wilis
2019/2020
165
7.55 (tender price)
Ciremai
2019
110
9.70 (tender price)
Guci
2019
55
9.09 (tender price)
Huu Daha
2021
20
9.65 (tender price)
Seulawah Agam
2018
110
6.90 (tender price)
COD = commercial operating date, MW = megawatt, PPA = power purchase agreement, Rp = rupiah, US/kWh = cents per kilowatt-hour.
Sources: Government of Indonesia, Ministry of Energy and Mineral Resources. Jakarta; various published information.
59

With the change in coal pricing to international levels, in 2013 PLNs coal generation costs were around
6.7 US/kWh (significantly above earlier costs when coal price to PLN was at a low administered
price). The generation-weighted average cost is 6.3 US/kWh.
6.2 Incremental Costs of Future Projects

Methodology
To calculate the subsidy requires estimates of the costs of future projects. This can be displayed in the form
of a supply curve, which plots cumulative capacity as a function of cost. Figure 6.1 shows such a supply
curve for the geothermal projects in Java and Sumatra, as calculated in 2010 by the Castlerock report. This
showed a total capacity of geothermal projects in the two main islands of 2,432 MW. Also shown in this
figure are the estimated PLN avoided costs (2013) and the former 9.7 US/kWh ceiling price.
Figure 6.1: Castlerock Supply Curve, Java and Sumatra

16
Production cost, US/kWh
14
12
9.7 US/kWh ceiling
10
A
PLN avoided cost (6.7 US/kWh)
500
1,000
1,500
2,000
2,500
Source: Authors calculations based on data in Castlerock. 2010. Phase 1 Report.
If only the projects are built for which costs are below the 9.7 US/kWh ceiling, then 1,949 MW
would be built. The other 483 MW of geothermal projects in the supply curve exceed the ceiling and
would not be built. The incremental costs associated with this level of geothermal development are
represented by the area A under the curve. This area represents the additional subsidy that must be
paid to PLN by MoF. For the costs as shown here, this calculates to $142 million per year once all
1,949MW have been builtassuming the bid tender prices were at the levelized cost of energy as
reflected in the supply curve.
Figure 6.2 shows the potential impact of raising the ceiling to 12.5 US/kWh, which intersects the
supply curve at 2,362 MW. Now the incremental costs increase by the additional amount represented
by the areas, B and C ($125 million), for a total subsidy of $268 million per year once all 2,362 MW
Recovery of Incremental Costs61
Figure 6.2: Impact of a 12.5 US/kWh Ceiling Price

16
14
Proposed new ceiling, 12.5 US/kWh
12
9.7 US/kWh ceiling
10
8
500
1,000
1,500
2,000
2,500
MW = megawatt, PLN = PT Perusahaan Listrik Negara, US/kWh = cents per kilowatt-hour.
have been built. Note that this is the subsidy for just the existing projects (1,335 MW) plus 1,027 MW
of new projects.45
The same methodology can be used for the eastern islands, where PLNs avoided financial cost is
11.5US/kWh. Of the 362 MW in the eastern island supply curve, 235 MW have costs below 11.5 US/
kWh, and so require no subsidy. There are 127 MW between 11.5 US/kWh and the new eastern island
ceiling of 20 US/kWh, which would require a subsidy of $17 million.
In other words, based on the original Castlerock supply curve data, the total subsidy required to
develop such geothermal capacity as lies below the new ceiling prices (2,342 MW in Java and Sumatra
at 12.5 US/kWh and 362 MW in the eastern islands at 20 US/kWh) is $285 million per year. 46
The Castlerock supply curve reflected 2010 costs, under the presumption of constant prices. But for
many projects under development, exploration and delineation drilling has barely begun, but drilling
costs in particular have increased significantly since then. The Castlerock estimates of the levelized
cost of energy clearly no longer apply.
On the other hand, PLNs avoided costs will also not stay constant. According to the IEA forecast
of coal prices (in their current policies scenario), at constant prices international coal prices will rise
from $100/ton in 2013 to $110/ton in 2020. As shown in Table 6.3, by 2020 PLNs avoided cost of
coal generation in the JavaBali and Sumatra grids will increase to 8.5 US/kWh, and to 9.9 US/kWh
by 2025.
45
That is, 1,335 MW + 1,987 MW = 2,432 MW.
46
The Castlerock report estimated the total 2020 incremental cost at $376 million (under the same assumption that the
supply curve reflected tender prices). Castlerock. 2010. Phase 1 Report.

Thus, as shown in Figure 6.3, over time both the PLN avoided cost curve (blue) and the cost curve
for geothermal projects (green) shift upward as shown in Figure 6.3. But note that as the supply curve
shifts upward, the amount of geothermal capacity for any given tariff ceiling decreases (i.e., shifts to
the left in the figure, from 2,362MW to 1,949 MW).
Figure 6.3: Adjusted Cost Curves

25
20
15
Proposed new ceiling 12.5 US/kWh
10
500
1,000
1,500
2,000
2,500
MW = megawatt, PLN = PT Perusahaan Listrik Negara, US/kWh = cents per kilowatt-hour.
Baseline Subsidy Estimates

The impact on subsidy of these two adjustments is shown in Table 6.2. For the proposed ceiling of
12.5 US/kWh, the subsidy required is $214 million/year (up from $125 million in the static analysis).
However, as noted, at this ceiling price, only 1,949 MW of new geothermal energy is built, down from
2,342 MW in the static analysis.
Table 6.2: Impact of Ceiling Prices on Ministry of Finance Subsidy (Java and Sumatra)
Ceiling
Installed
Capacity
Incremental
Capacity
MoF
subsidy
Average
subsidy
Incremental
subsidy
US/kWh
$ million
$/kW/year
$/kW/year
MW
MW
PLN avoided cost (2014)
6.7
186
186
Old ceiling
9.7
1583
1397
120
76
86
Higher ceilings
11.0
1900
317
197
104
243
11.5
1900
197
104
12.5
1949
49
214
110
361
13.5
2028
79
248
122
428
14.0
2094
66
277
132
434
15.0
2156
62
305
141
450
16.0
2237
82
348
156
537
Proposed ceiling
17.0
2292
55
381
166
606
18.0
2332
40
407
175
642
19.0
2332
407
175
20.0
2362
30
430
182
753
$/kW/year = dollars per kilowatt per year, MoF = Ministry of Finance, MW = megawatt, PLN = PT Perusahaan Listrik Negara, US/kWh = cents
per kilowatt-hour.
Note that as we move up the supply curve to ever more expensive projects, the amount of subsidy
per additional kW increases. At the 12.5 US/kWh ceiling, the average subsidy is $110/kW/year. But at
14 US/kWh, the average subsidy is $277/kW/year.
There are many uncertainties in this analysis, the most important being the assumption that the tender
prices as bid correspond to the levelized cost of energy as in the (modified) Castlerock supply curve.
But as noted, declarations of ceiling prices may influence bid prices, and the winning bid could be
close to the ceiling price.47 On the other hand, with up-front Geothermal Fund de-risking, the bid tariff
should be correspondingly lower. As discussed in Section 8, the impact on the tariff can range from
1 to 3 US/kWh, depending on project size (the smaller the project, the greater the relative impact on
the tariff).
Table 6.3 shows the results of alternative assumptions. In column [4] we show the subsidy estimates
if bids are at the levelized cost of energy. Column [5] shows the impact of up-front Geothermal Fund
de-risking. Column [6] makes the most pessimistic assumption that the bid price will be at the ceiling
price, and column [7] at the ceiling price adjusted for Geothermal Fund re-risking. The subsidy impact
of the proposed ceiling is seen to be in the range of $149 million to $316 million per year.
47
See discussion at Section 2, Table 2.3.
Table 6.3: Impact of Subsidy on Assumptions (Java and Sumatra)

MoF subsidy if tender prices are at:
LCOE
LCOE with
de-risking
@tariff
ceiling
@tariff ceiling
adjusted for
de-risking
[3]
[4]
[5]
[6]
[7]
MW
$ million
$ million
$ million
$ million
Ceiling
Installed
Capacity
Incremental
Capacity
[1]
[2]
US/kWh
MW
PLN avoided
cost (2014)
6.7
186
186
Old ceiling
9.7
1583
1397
120
104
168
152
Higher ceilings
11.0
1900
317
197
141
298
242
11.5
1900
197
141
298
242
12.5
1949
49
214
149
316
251
13.5
2028
79
248
170
345
267
14.0
2094
66
277
188
368
279
15.0
2156
62
305
206
388
290
16.0
2237
82
348
234
413
299
17.0
2292
55
381
256
428
303
18.0
2332
40
407
274
438
305
19.0
2332
407
274
438
305
20.0
2362
30
430
291
445
306
Proposed
ceiling
LCOE = levelized cost of energy, MoF = Ministry of Finance, MW = megawatt, PLN = PT Perusahaan Listrik Negara, US/kWh = cents per kilowatt-hour.
There is additional uncertainty in the supply curve itself. The adjustments made here (increasing the
Castlerock cost estimates by 2% per year) are fairly simplistic, and we recommend that the supply
curve be updated in detail.48
Subsidy to Meet Fast Track Program 2 Targets

FTP2 calls for 4,925 MW of new geothermal projects, none of which are yet completed. The total in
the Castlerock supply curves is 2,362 MW in Java and Sumatra, and 362 MW in the east for a total of
2,794 MW, or 1,459 MW of new projects. It follows that to meet the FTP2 targets, another 3,466 MW
is required. The subsidy requirements to meet the targets of FTP2 are presented in Table 6.4.
Under the optimistic assumption that the average subsidy cost of these additional projects is the
same as that calculated for those in the Castlerock supply curve, then the additional annual subsidy
requirement is $601 million, or $428 million with the benefit of up-front de-risking. This brings the
total subsidy requirement to meet the FTP2 target to between $774 million and $1,085 million. If the
average cost of the additional projects is greateras would seem likelythe subsidy requirements will
be that much higher.
48
This will require a project-by-project review of the status of development, with updated costs and assessment of drilling
prospects, and with updated assumptions about financing structure. Most important, an attempt should be made to
extend the supply curve to all of the projects in FTP2. Such a study is outside the scope of this report.
Table 6.4: Subsidy Requirements to Meet Targets for Fast Track Program 2
No De-risking
With De-risking
Annual
Subsidy
MW
Annual
Subsidy
$ million
$/kW
$ million
$/kW
In Castlerock supply curve

of which <12.5 US
1,949
316
251
483
129
56
2,432
445
183
306
126
362
39
108
39
108
Total
2,794
484
173
345
124
Existing
1,335
>12.5 US
Java and Sumatra
Eastern islands
New projects
1,459
Additional projects
3,466
601
173
428
124
FTP2 target new projects
4,925
1,085
173
774
124
$/kW = dollars per kilowatt, FTP2 = Fast Track Program 2, MW = megawatt.

6.3 Impact on Ministry of Finance

These incremental costs and benefits should be placed in the context of the overall level of subsidy
provided by MoF to PLN. Figure 6.4 presents the World Banks estimates of the PSO derived on
the basis of PLNs financial model with a 30-year time horizon. In 2013, PLNs estimated revenue
requirements were Rp230 trillion ($23 billion), of which Rp127 trillion was covered by tariff revenue,
and Rp103 trillion (or 44%, $10 billion) by the PSO subsidy.49
Figure 6.4: Estimates of Future Public Service Obligation Subsidy

800
Rp trillion
600
400
200
206
161
58
93
230
103
103 113
127
2010
2012
254
101
153
288
299
113
77
175
329
50
279
379
419
58
53 361
46
326
303
349
496
527
39
55 488
53 441
399
452
595
570
624
44
42
44 553 580
526
222
2014
Tariff Revenue
2016
2018
2020
2022
2024
Subsidy from MoF (PSO)
MoF = Ministry of Finance, PSO = public service obligation, Rp = Indonesian rupiah.

Source: World Bank estimates based on Perusahaan Listrik Negaras 2013 Rencana Usaha Penyediaan Tenaga Listrik.
49
The actual figures from PLNs 2013 Annual Report show revenue of Rp257 trillion, with tariff revenue of Rp153 billion and
Rp101 trillion in PSO subsidy.

The actual level of a future PSO subsidy is subject to several key uncertainties, notably assumptions
about future tariff increases,50 and future coal prices. However, there is general agreement that the
level of PSO subsidy will decline over the next few years, and by 2020, the PSO subsidy should have
fallen to Rp53 trillion ($5.5 billion).51
Nevertheless, despite the uncertainty over future PSO subsidy payments, the incremental cost of
geothermal subsidies could well reach $1 billion per year if the FTP2 targets are to be metthough
this would fall in the event that international energy prices were to increase over the levels envisaged
by IEA (Section 4, Table 4.1).
50
Here we assume 15% tariff increases each in January 2015 and 2016, and 2.75% in each of years 20192023.
51
In the PLN financial model, the level of geothermal generation is explicit only for its own projects. Purchases of geothermal
energy from IPPs is included in the category purchased power.
Tendering
7.1 The Issues
Revised tariff ceilings as proposed in this report are important, but will not alone unlock geothermal
development in Indonesia. The geothermal law requires competition in the selection of geothermal
developers, the main features of which are as follows:

The tender process is conducted under the two-envelope system in which the price
envelope is opened only after meeting the technical and administrative requirements in the
technical proposal envelope.
In the previous Geothermal Law No. 27/2003, where a WKP or wilayah kerja pertambangan
(geothermal work area) falls entirely within a single province, the Geothermal Law 2003
devolves responsibility for conducting the tender to the provincial governments concerned.
Only where a WKP covers more than a single province was the responsibility assigned to
the Ministry of Energy and Mineral Resources (MEMR). Under the new bill on geothermal in
2014, the central government will take over the responsibility of geothermal tender from local
governments. While this shift is likely a major step forward, the successful implementation
will depend on the capacity of the government tender committee and the regulation on the
new tender mechanism.
The current tendering process has several deficiencies. The first set of issues relates to the procedures
of tendering, which include:

poor prequalification standards, which have encouraged unqualified bidders to offer

unrealistically low prices that cannot be achieved in practice;
that the technical capacity of the government tender committees will likely require
strengthening;
that the requirement for the winner of the tender to post a $10 million performance bond has
not been enforced; and
that the presently required bid bond of $100,000$200,000 is too small, and does not
discourage companies from submitting an unrealistic price.52
The second set of issues relates to the information available to bidders at the time of tender, including:
52
that heretofore the PPA, and its tariff schedules that govern escalation and indexation, have
required negotiation after tendering; and
that the information on the geothermal resource often lacks any subsurface information,
making it difficult for bidders to reliably estimate costs. In many cases even basic geology,
geochemistry, and geophysics (3G) information is incomplete.
The current requirement is that the bid bond is 2.5% of the estimated first-year exploration program. There is no minimum
first-year requirement. Even if the first-year program were $5 million, then 2.5% is just $125,000.
67

These issues were often mentioned in our discussions with developers, both informally and as
expressed at the formal stakeholder consultation meetings (Stakeholder Comment 5). However,
these discussions were limited to the prominent developers with a presence in Jakarta, and do not
necessarily reflect the views of smaller developers who have won tenders for small projects in the
eastern islands.
Stakeholder Comment 5: The Tender Process

Comment:
Chevron has stated that a meaningful and substantial bid bond is required, and failed projects
should have meaningful consequences.
Other developers raised similar issues in informal discussions. They are often discouraged from
bidding in the smaller projects because they see the tender process as unpredictable; for that
reason, they are strongly opposed to beauty contests as envisaged by the 2012 FIT under which
the wilayah kerja pertambangan (geothermal work area) (WKP) would be awarded on grounds
other than price. Tender committees in remote provinces are seen as lacking adequate technical
credentials to be able to make sound judgements about technical capacity to deliver geothermal
projects on time and on budget.
Reply:
We are in general agreement with the comments raised. Our recommendations that the tender
process should follow international best practice for competitive procurement address the
concerns directly (including better qualified tender committees, and a significant bid bond that
will discourage speculators and poorly qualified bidders).
Sources: Written comments from Chevron, received 11 March 2014; informal discussions with developers.
7.2 Evaluation of Past Tenders

The actual number of successful tenders is small. According to data provided by MEMR, in 14 of
21tenders, there was only one bidder who passed the first stage in the two-envelope system that is
used. Again according to MEMR, in all of the 14 cases reporting only one bidder (whose price envelopes
were opened), PPA negotiations are underway or already signed. Only one tender is listed as tender
failed. Under the Indonesian rules for tendering, a tender is considered failed only if the number of
bidders is less than three: if two of the three bidders fail the test of the first envelope, the fact that only
one price bid is examined (in the second envelope) does not invalidate the tender.
Figure 7.1 shows the bid price as a function of project size and number of bidders.
The sample size is small, so inferences drawn from this analysis require caution. Potential challenges
to consider include
53
that there is little evidence of scale economies, which could be evidence that bidders have
placed little reliance on MEMR estimates of resource capacity;53
that projects with two bidders are not significantly different from those with one;
One possible explanation, particularly for smaller projects, is that bidders have simply acted to secure the concessions in
the expectation that they will later decide how much can actually be built.
Tendering69
Figure 7.1: Bid Price versus Project Size and Number of Bidders
US/kWh
20
15
10
50
100
150
200
250
300
Project size, MW
1 bidder
2 bidders
3 bidders

Source: Data provided by Ministry of Energy and Mineral Resources, January 2014.
that projects with three bidders have significantly lower prices than those with one or two;
however, whether these projects, with bid prices from 5.8 to 6.85 US/kWh, can actually be
delivered at this price is unclear, and
that only two of the twenty bid prices are higher than the 9.7 US/kWh ceiling.
These findings do support some of the criticisms of the developers. The large number of bids for
which just a single price envelope is opened suggests that the system is open to manipulation. With
so many obviously unqualified bidders submitting bids to be rejected at the first-envelope stage, the
field becomes clear for the one remaining bidder. Moreover, it is fairly clear that the very low bids (even
where there are three bidders whose price bids were examined) are very unlikely to be realized.
7.3 Options for Improving the Tender Process

Because the devolution of tendering to the provinces is enshrined in the geothermal law, fundamental
changes to the procedure may be difficult, and it is therefore useful to distinguish between the longer
term objectives (which may require changes in law), and those that can be implemented in the shortterm by regulations issued at the MEMR level.
The Long-Term Objective: Central Tendering

In principle, the question of how to resolve the problems with the tender process is straight-forward
to resolve with appropriate technical assistance to develop a set of standards for prequalification of
bidders, and for the composition and qualification of tender committees. Such technical assistance
has, in fact, been provided by Fichtner under Global Environment Facility funding to the MEMR:54
54
Fichtner, Consulting Services for Design and Preparation for Geothermal Investment Transaction, MEMR. Various reports,
20102012.

However, in the long-term we believe that efforts should be made to establish a new professional
entity for tendering at the center. The rationale for such a new body includes:

that is much easier to provide technical assistance from multilateral development banks and
bilateral donors to an established body with some institutional longevity, rather than to many
provincial or municipal and/or regency tender committees that are constituted on an ad hoc
basis for short time periods;
that such a tender committee could also be involved in tendering other renewable projects,
particularly hydropower, which has similar problems; and
that an entity with institutional longevity is in a much better position to deal with post-tender
issues as may be associated with adjustments in bid prices warranted for delays that are not
the fault of the developer.
An important question would be how the interests of the local governments would be secured (which
may be at the provincial, municipal, or regency level). One option would be for the local government
to be formally represented on the tender committee.
It may be noted that the present system is not necessarily in the interests of local government, and
that there may well be good reasons for them to approve reform of the current system. Simply put,
the local benefits of geothermal energy, from direct fees and taxes, as well as from the stimulus to
local economic development, are only realized in practice if development succeeds. Unrealistic bids
by unqualified developers are not in the interests of local government if these projects fail. They may
well be better off if a capable developer is selected, even though they may lose total control in the
selection process.
Another long-term goal that may take some time to achieve is to improve the quality of
information available at time of tender, for example by using the Geothermal Fund for this purpose
(see Section9.4). In the ideal case at the time of tender:

a complete and independently verified package of 3G information should be provided;
the resource should be measured and resource capacity estimated using an internationally
accepted method, and
at least three successfully tested wells should be provided, funded by the state, the costs of
which are to be reimbursed at the time of financial closure.55
Improving the quality of information available to bidders is particularly important for the smaller
projects in the eastern islands.
During the preparation of this report, Indonesias House of Representative passed the Bill on
Geothermal Energy as a revision to the previous Geothermal Law No. 27 of 2003. One of the major
changes in the new bill is that the geothermal concession tender and issuance of geothermal license
for power development will be carried out by the central government (MEMR). See Appendix 8 for
a fuller discussion of the new law.
Options for the Short Term

In general, tendering should follow the practices and procedures of international best practice, as
exemplified by ADB and World Bank procurement guidelines. Beyond the two-envelope system
already in use, three elements would be vital for Indonesia to adopt in geothermal tendering:
55
See detailed discussion in Section 8 (which includes provision for recovery of the costs of unsuccessful exploration drilling
by the Fund).
Tendering71
rigorous prequalification;
requirement for a substantive bid bond (1%2% of the project cost, but preferably not below
$10 million);56 and
post-qualification review to ensure compliance of the lowest evaluated bidder with all
requirements before winner is announced.
The bid bonds for bidders who do not pass the first-envelope evaluation can be released immediately;
the bid bonds for unsuccessful bidders whose second envelope is opened would only be released
once the contract with the winner has actually been signed.
This procedure does not necessarily exclude small companies, but the consortia must be in place for
prequalificationfor which purpose memorandums of understanding will not be sufficient, but would
require a credible legal agreement to document the joint venture.
Post-Award Audits
In the Philippines, the winner of any tender by the Power Sector Assets and Liabilities Management
Corporation (involving the sale of state assets, including geothermal assets), is subject to a postaward audit by the Privatization Bids Awards Committee, which ensures that all of the tender terms,
including the posting of any bond, has been met before declaring a final winner. A similar audit system
should be established in Indonesia. Such a review is also part of the normal procedure under the
procurement regulations of the IFIs.
Performance Bonds
There is currently a requirement for the winning bidder to post a $10 million performance bond.
However, this has apparently never been enforced. There is, obviously, no point in such a requirement
if it is not enforced.
It has been argued that one of the reasons why the $10 million bond requirement has not been
enforced is the high uncertainty about the actual prospects of a commercial project. For example, the
recent KfW report for Bappenas argues the following:57
Consider the situation in which, after an IUP (Izin Usaha Pertambangan) holder conducts
verification exploration, the site potential has been overestimated. In that case, the IUP holder
may be justified in making the commercial decision not to drill, in which case it should not
forfeit the performance guarantee.
However, the problem with this argument is self-evident: the result has been winning bids by
unqualified entities. A bidder not prepared to post the recommended bid bond is unlikely to be
serious. A complete 3G package (or, better yet, up-front drilling by the government as a public good)
mitigates the problem of inadequate information at time of tender.
Developers dislike performance bonds because they tie up (expensive) equity capital. However,
a bond requirement is not unreasonable as a pledge of performance in developing the project, but
that purpose is surely demonstrated once exploration (or delineation) drilling has commenced.
Consequently, it seems reasonable that the bond could be paid down over a 2-year period upon
evidence that the equivalent funds have been expended on exploration drilling (in the case where
Geothermal Fund has not already provided this), or for delineation drilling.
56
The lower bound is for larger projects.
57
Partnership International, for BAPPENAS. 2013. The Indonesia Geothermal Handbook. Jakarta.

Note that tying up equity capital in a performance bond is not without its consequences on the buyer:
the cost of raising that additional equity will eventually be recovered by the developer in his tariff.
There is no free lunch.
Under our proposed recommendation of a substantive bid bond, the winning bidder would convert his
bid bond into a performance bond, to be drawn down as noted above.
Power Purchase Agreement Issues

The tariff schedules of the PPA should be fixed at time of tender, with the same escalation and
indexation provisions applicable to all (rather than being the subject of ad hoc post-tender
negotiations).
In some cases (notably in the eastern islands), project size is subject to an upper bound because of
demand or network constraints. Experience shows that one cannot always reliably predict what the
best commercial project size will be, yet this is crucial to make cash-flow projections, because the
projects revenue requirements must be recovered across the available kWh sales. Projects that are
smaller than expected at tender, for whatever reason, may not therefore be commercially viable at the
tariff bid. While one could argue that under these circumstances such a project should be re-bid, the
difficulty of valuing any work that has already been done (and what compensation for same, if indeed
any, should be provided to the previous incumbent) makes such re-bids difficult.
A preferable solution is to agree, in advance at the time of tender, a formula that would automatically
adjust the base tariff price to the actually achievable project size. Section 5 outlines proposals for the
procedures by which such adjustments could be made.
PLN may object that its planningboth to ensure the requisite transmission expansion, and to ensure
that local loads are metis made more difficult if project sizes are not fixed in advance. But experience
shows that projects that are stalled for inadequate tariffs because projects are smaller than originally
planned are an even greater problemand re-bids imply even greater delays, which are in the interest
of neither the developers nor PLN.
Dealing with Delay

Government Regulation No. 70/2010 stipulates that if developers of concessions issued prior to
the issuance of Government Regulation No. 59/2007 do not exploit the field by 31 December 2014
(extended from the previous deadline 31 October 2010), they shall have to return the concession
to the government. In practice, this applies to Pertamina/PGE and PLN (Ulumbu, Mataloko, and
Tulehu), who are the only such concession holders. Deadlines from 31 August to 31 December 2014,
for existing holders of mining permitsIzin Usaha Pertambangan (mining business permits) (IUP) and
tender winners were given by MEMR Regulation No. 17 of 2014. It remains to be seen whether the new
deadline will in fact be enforced.
Even if drilling some minimum number of wells within a specified time could be stipulated, good
and financially credible developers point out that in many instances, lack of progress is the fault of
government, not of developersan argument well exemplified by the Sarulla project, where sorting
out legal issues surrounding asset ownership has taken 10 years. This kind of problem can only be
mitigated by resolving the necessary regulatory permits before bidding. This may involve significant
preparatory work by the government (and MEMR), but in the absence of precleared sites, there
will be no easy solution to enforcing rules about lack of developer progress. The possibility of using
Geothermal Fund resources for improving the quality of information available to bidders at the time of
tender is available to bidders at the time of tender is set out in Section 9.5.
The Tariff Impacts of

Front-End De-risking
In the past, the cost of initial exploration in Indonesia has been assumed by the developer, for which
only equity is possible since debt finance requires a substantial (and in recent years, increasing)
fraction of the required steam resource to be proven in terms of well deliverability. For investors to put
up such up-front high-risk equity requires high rates of return, which must necessarily be recovered
though the tariff. The objective of this section is to present quantitative estimates of the tariff impact
of such front-end exploration costs, since such estimates have not been quantified in previous
discussions of front-end de-risking.58 We assume here that this up-front exploration is undertaken by
the Geothermal Fund, as proposed in Section 7.
The costs of exploration drilling to establish the existence of a geothermal resource depend little on
the ultimate size of a project. For the sake of illustration, we assume this up-front exploration costs
$30 million, including support infrastructure for drilling. The tariff impact of its recovery will depend
on two main factorshow many kWh of energy will ultimately be produced, and what is the time lag
between exploration and the commercial operation.
Table 8.1 illustrates the necessary calculations. In rows [1][11] are shown the developers cash flows for
a $30 million outlay for exploration in the case of a 220 MW project, with COD in Year 8. Assuming a
90% plant capacity factor, 1,754 GWh/year of electricity is available for cost recovery. To achieve the
Table 8.1: Cash Flows for Recovery of Exploration Costs
NPV
1
Developer
Exploration
outlay
$ million
220
MW
Capacity
COD date
Energy
0.9
GWh
Tariff Impact
1.30
US/kWh
Equity IRR to
developer
21.1
$ million
21.1
$ million
[ ]
24.0%
10 PLN
11
10
20
10
15
20
30
22.8
22.8
22.8
22.8
22.8
22.8
22.8
22.8
22.8
22.8
22.8
22.8
7
Net cash
flows
r=24%
1
Cost to PLN
12 Net cost to
Indonesia
10
20
1,754 1,754 1,754 1,754 1,754 1,754
r=12%
$ million
83.2
22.8 22.8 22.8 22.8 22.8 22.8
$ million
83.2
22.8 22.8 22.8 22.8 22.8 22.8
COD = commercial operation date, GWh = gigawatt-hour, IRR = internal rate of return, MW = megawatt, NPV = net present value,
PLN = PT Perusahaan Listrik Negara, r = discount rate, US/kWh = cents per kilowatt-hour.
58
De-risking is perhaps an imprecise term, because the risks of drilling are whatever they are. In this section the term is used
in the context of transferring the exploration risk from the developer to other parties, such that the cost to the buyer (PLN)
may be reduced. Of course, better 3G data can reduce subsequent drilling risk, but we here assume that a comprehensive
3G data package has been prepared prior to tender.
73

assumed target 24% internal rate of return (IRR) on such up-front capital requires 1.3 US/kWh. The
corresponding cost to PLN is $22.8 million per year, or $83.2 million in NPV terms when using the
assumed 12% opportunity cost of capital (to the Government of Indonesia) as the discount rate.
Now suppose that the $30 million of up-front exploration expenditure is covered by the Geothermal
Fund as a pure public good, with no recovery from the developer. In such a case, the benefit to PLN is
a tariff that is 1.30 US/kWh lower than in the base case. So the return to government as a whole is also
24%: for every US dollar of up-front geothermal exploration outlay, and assuming that the exploration
is successful and leads to a commercially viable power generation project, the implied rate of return
on exploration outlays (i.e., the avoided tariff impact if recovered by the developer) is 24%. The net
impact (cost) to Indonesia falls to $24.9 million, as reflected in Table 8.2.
Table 8.2: Exploration Costs Funded by Geothermal Fund
NPV
1
Discount rate
r=12%
Geothermal fundoutlays
($ million)
24.9
PLN
Incremental cost to PLN

($ million)
Net impact on
government ($million)
24.9
IRR to Govt on GF
expenditures
24.0%
Net cost to Indonesia

($million)
24.9
10
20
0.0
10
15 20 30
10.0
20.0
10
20
GF = Geothermal Fund, Govt. = government, IRR = internal rate of return, NPV = net present value, PLN = PT Perusahaan Listrik Negara,
r = discount rate.
Assumptions: 220 MW, COD in Year 9.
The total cost to Indonesia, under the same NPV assumptions as above, falls from $83.2 million to
$24.9 million. Such is the difference between private equity funding with recovery 9 years later, and
providing high-risk exploration as a public good. It explains why so many countries with substantial
geothermal energy development (Philippines, Mexico, New Zealand, Iceland, Kenya59) funded a
comprehensive geothermal exploration program from the state budget.60
As noted, the tariff impact will be a function of the size of the project, and the time lag to COD. As
shown in Table 8.3, for a 220 MW project the tariff impact is between 1 and 2 US/kWh; but for small
projects, the tariff impact is significantly higherfor a 10 MW project more than 10 US/kWh. Of
course, for very small projects three successful exploration wells probably means no further delineation
or even production drilling is necessary, so time and cost to COD will be much shorter.
59
Though in the cases of Iceland and Kenya, only recently have IPPs been allowed to participate in development.
60
However, the exploration activity was not undertaken as part of the development of a new power project, but as a separate
exercise to identify available resources for future development. That was also the situation in New Zealand before deregulation
of the power industry. Even in the US, government-funded surveys have been conducted by the US Geological Survey.
The Tariff Impacts of Front-End De-risking 75
Table 8.3: Tariff Impact of $30 Million Exploration Program Funded by Developers (US/kWh)
Installed Capacity, MW
Years to
COD
20
10
220
110
55
4.88
9.76
6.05
12.10
2.73
7.50
15.01
1.69
3.38
9.31
6
7
8
1.05
2.10
4.20
1.30
2.60
5.21
10
Tariff impact, US/kWh
20 MW
8
6
55 MW
4
110 MW
2
220 MW
Years to COD
COD = commercial operation date, MW = megawatt, US/kWh = cents per kilowatt-hour.
8.1 Cost Recovery Options for the Fund

The above calculations assume no cost recovery. Ultimately, the cost of geothermal exploration will fall
either on government, electricity consumers (through tariff increases), or providers of concessionary
finance.61 From the perspective of government as a whole, it matters not whether the costs fall on the
Geothermal Fund run by the governments investment unit or on subsidy provided by MoF to PLN.
However, as discussed below, there exist good reasons to recover the exploration costs where
commercial development is enabled. There are two obvious points in the development process at
which this might occur: at the time of tender, and at the time of financial closure.
Cost Recovery at Time of Tender

Under the assumption of a successful 2-year exploration program, the work area could be tendered in
Year 3, at which point the successful bidder would be required to refund the costs of the exploration to
the Geothermal Fund, plus interest. Although in the case of large projects that still require additional
delineation drilling this would require equity to do so, with a measured resource a considerable fraction
61
Note that in the case of concessional finance, where the effective interest rates are far below the governments opportunity
cost of capital, some part of the cost is in effect transferred to the global community that provides concessional finance.

of the project risk has been reduced, so the required target equity IRR would be much lower than before
the resource is measured. We may assume that at the point of tender, the target equity IRR falls from
the 24% assumed previously to 20%, which would be suitable where delineation drilling still needs to be
undertaken. For small projects, where the exploration drilling may already provide a significant fraction of
the total steam requirement, then a lower equity IRR target would be suitable.
As shown in Table 8.4, this would require a payment from the developer, at time of tender, of
$35million, which the developer recovers at 20% IRR (on his up-front equity) by an increment to
the tariff.62 Under the same conditions as shown in Table 8.1 (220 MW, COD in Year 8), this results
in a tariff increment of 0.84 US/kWh (compared to 1.3 US/kWh if the developer funds early stage
exploration himself entirely from equity, Table 8.1). To be sure, this is a substantial up-front payment,
but this should be compared with the level of expenditure he would have had to incur during the
first few years, had the tender been issued before initial exploration. This also eliminates unserious
bidders interested simply in holding the site (as has allegedly occurred in some cases in the past). The
net impact on government is $53.3 million, some $30 million less than if developers pay for up-front
exploration (Table 8.1, $83.2 million).
Table 8.4: Cash Flows, Cost Recovery at Tender
Recovery through repayment at time of tender
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Developer
Payment to
GF at tender
Installed
capacity
COD
Energy
Tariff impact
NPV
$ million 20.4
220
8
0.9
0.84
Tariff revenue
Developer
cash flows
Equity IRR to
developer
Geothermal
Fund
Geothermal
Fund outlays
Repayment at
tender time
Net impact
PLN
Cost to PLN
Net cost to
Indonesia
10
15
20
30
35.3
MW
GWh
US/
kWh
$ million
$ million
[ ]
0
20.4
0.0
0.0
0 1,754 1,754 1,754 1,754 1,754 1,754
0.0 0.0 0.0 0.0 0.0

0.0 35.3 0.0 0.0 0.0 0.0
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
20.0%
$ million 24.9 10.0 20.0

$ million
25.1
35.3
$ million
0.2 10.0 20.0
35.3 0.0 0.0
$ million
$ million
53.5
0.0
0.0
53.3 10.0 20.0
0.0 0.0 0.0 0.0 0.0 14.7 14.7 14.7 14.7 14.7 14.7
35.3 0.0 0.0 0.0 0.0 14.7 14.7 14.7 14.7 14.7 14.7
COD = commercial operation date, GF = Geothermal Fund, IRR = internal rate of return, MW = megawatt, NPV = net present value,
PLN = PT Perusahaan Listrik Negara, US/kWh = cents per kilowatt-hour.
The Sarulla project demonstrates that developers are prepared to make substantial payments
for access to proven resources at time of tender. Unocal held the original WKP and undertook
substantial exploration including drilling, proving two separate good resources (and possibly a third).
It then relinquished the concession, for which PLN paid them $60 million. The concession was then
62
It is assumed that the Geothermal Fund charges interest at 12%, rolled into the outstanding balance at 12% (just like
capitalized interest during construction). Therefore, even though the exploration expense was $30 million, the repayment
at time of tender would be $35.3 million.
Table 8.5: Cash Flows, Cost Recovery at Financial Closure

NPV
1
2
Developer
r=14%
Payment to
$ million 22.0
0
0
GF at FC
3 Installed
220
MW
capacity
4 COD
8
5 Date of FC
5
6 Energy
0.9
GWh
7 Tariff impact 0.45 US/kWh
8 Tariff
$ million
22.0
revenue
9 Developer
$ million
0.0
0
0
cash flows
10 IRR to
[ ]
14.0%
developer
11 Geothermal
r=12%
Fund
12 Geothermal
$ million 24.9 10.0 20.0
Fund outlays
13 Repayment
$ million
24.0
0.0
0.0
at financial
closure
14 Net impact
$ million 0.9 10.0 20.0
15 PLN
16 Cost to PLN
$ million 28.6
0.0
0.0
17 Net cost to
$ million 29.5 10.0 20.0
Indonesia
1
2
9
0
10
0
15
20
30
0 42.34
0 1,754 1,754 1,754 1,754 1,754 1,754
7.9
7.9
7.9
7.9
7.9
7.9
42.3
7.9
7.9
7.9
7.9
7.9
7.9
0.0
0.0
0.0
0.0
0.0
0.0
42.3
0.0
0.0
0.0
42.3
0.0
0.0
0.0
0.0
0.0
0.0
42.3
0.0
0.0
0.0
0.0
7.9
7.9
7.9
7.9
7.9
7.9
7.9
7.9
7.9
7.9
7.9
7.9
COD = commercial operation date, FC = financial closure, GF = Geothermal Fund, GWh = gigawatt-hour, IRR = internal rate of return,
MW = megawatt, NPV = net present value, PLN = PT Perusahaan Listrik Negara, US/kWh = cents per kilowatt-hour.
retendered and MedcoOrmat won the tender despite complications in the bidding process. One of
the conditions of tender was a cash payment to PLN of $70 million. The fact that there was then a
9-year delay for other reasons does not undermine this conclusion of willingness-to-pay for access to
proven resources.
Table 8.4 also shows the corresponding cash flows to PLN and the Geothermal Fund. With interest
included in the repayment to the Geothermal Fund, the net impact on the Fund is a positive (NPV) of
$0.3 million. The cost to PLN of the incremental tariff necessary to fund the tender payment at 20%
equity return to the developer is now only $14.7 million per year (down from $22.8 million per year if
the developer conducts the exploration program at an equity return target of 24%).
Cost Recovery at Time of Financial Closure

Table 8.5 illustrates the cash flows if cost recovery occurs at time of financial closure, assumed here at
3 years prior to COD. Now the costs of fund repayment can be rolled into the debt, and at which point
only a small amount of additional equity would be needed. The WACC at this pointpost delineation
drilling, with the resource provenwill unlikely be greater than 12%, so the tariff impact falls to just
0.45 US/kWh.
The net annual cost to PLN falls to $28.6 million, and the total impact on Indonesia as NPV is
$29.5million. If the WACC is based on largely concessional debt, that WACC may be much lessfor
example at 8%, the total of exploration costs to Indonesia falls to $16.8 million. Table 8.6 compares the
tariff impact for a 55 MW project under the various scenarios and as a function of the time to COD.

Table 8.6: Tariff Impact of $30 Million Exploration, 220 MW Project
Developer
%
Years to COD
Equity Return, %
Geothermal Fund
Recovery at Tender
Geothermal Fund Recovery

at Financial Close
24%
20%
14%
US/kWh
US/kWh
US/kWh
0.85
0.58
0.35
1.05
0.70
0.40
1.30
0.84
0.45
1.61
1.01
0.51
COD = commercial operation date, US/kWh = cents per kilowatt-hour.

The figure in Table 8.7, drawn to the same scale as Table 8.3, illustrates the tariff impact when
exploration costs are funded at time of financial closure. Even for 20 MW projects, the impact is less
than 4 US/kWh (compared to 6 US8 US/kWh if the developer must pay). It explains why so many
tenders in small eastern islands have failed or have only had one bidder.
Table 8.7: Tariff Impact of $30 Million Exploration Program, Costs Recovered at
Financial Closure at 14% Weighted Average Cost of Capital
Installed Capacity, MW
Years to COD
220
110
55
20
10
2.59
5.18
1.07
2.96
5.91
1.23
3.37
6.75
0.70
1.40
3.85
6
7
0.40
0.79
1.58
0.45
0.90
1.79
0.51
1.01
Tariff impact, US/kWh
10
20 MW
55 MW
110 MW
220 MW
3
Years to COD
COD = commercial operation date, MW = megawatt, US/kWh = cents per kilowatt-hour.
In small projects typical of eastern islands, a successful $30 million exploration program may well mean
that most (if not all) of the necessary steam resource has been established: this expenditure exceeds likely
bid and performance bond requirements, so the level of remaining risk may be quite small. Nevertheless,
unserious bidders would still be discouraged by the requirement to post a significant bid bond.
8.2 Sensitivity Analysis

The above analysis makes several assumptions about which there is uncertainty. First, there is a
presumption that the cost (and success) of the exploration program as conducted by developers
at their own risk is no different to one conducted by the Geothermal Fund. However, it might be
argued that an exploration program conducted by a developer at his own risk would likely be more
cost-effective than a program conducted under contract for fee by a drilling company, and managed
by a fundwhich, as noted elsewhere, lacks geothermal technical expertise.63 In short, what is
accomplished by the fund for $30 million might be achieved by a competent and experienced private
developer for 5% to 30% less.
However, as shown in Table 8.8, even at 30% more efficiently, the cost to Indonesia is still more than
double that of Geothermal Fund exploration drilling, and with a tariff impact of 0.91 US/kWh.
Table 8.8: Impact of Developer Efficiency (220 MW, Year 8)
Ratio of Developer Cost: Fund Cost
Tariff Impact
Cost to Indonesia
US/kWh
$ million
1.0 = $30 million
1.30
83.2
0.9 = $27 million
1.17
74.9
0.8 = $24 million
1.04
66.0
0.7 = $21 million
0.91
58.2
GF 100% funding
0.00
24.9
GF = Geothermal Fund, MW = megawatt, US/kWh = cents per kilowatt-hour.

Second, the above calculations are all with respect to a single exploration program that is assumed
to result in a commercially viable electricity generation program. That is by no means assured.
Consequently, the net impact on Indonesia needs to account for some unsuccessful projects. This
has one of two consequenceseither the resources of the fund are drawn down (or replenished from
the state budget, or even by contributions from donors), or they can be replenished by charging the
successful projects a fee.
Such a fee could therefore be levied on all (successful) projects to cover the costs of failed exploration
schemes. However, if that fee is payable only at financial closure, at which point there is very little
remaining uncertainty in the project, the fee is effectively a pass-through, borne either by the sources
of concessionary finance, or by MoF through the PLN PSO.
Table 8.9 shows the impact of the fee on the tariff, under the assumption that it is payable at time of
financial closure. The impact is smallest, obviously, for the large 220 MW projecteven at 50%, the
fee raises the tariff impact from 0.33 US/kWh to 0.46 US/kWh. For a small 20 MW project, the
impact is much greater: a 50% fee raises the cost from 3.65 US/kWh to 5.11 US/kWh.
63
Even if expert consultants are engaged by the fund to oversee the drilling program, there is no substitute for a powerful
profit motivation.

Table 8.9: Impact of Fee for Unsuccessful Exploration Programs, US/kWh
(Exploration Cost Plus Fee Paid at Time of Financial Closure)
Installed Capacity
0%
220 MW
55 MW
20 MW
0.33
1.33
3.65
10%
0.36
1.43
3.94
20%
0.38
1.54
4.23
30%
0.41
1.65
4.53
40%
0.44
1.75
4.82
50%
0.46
1.86
5.11

Assumption: weighted average closing costs at financial closure = 12%.
However, what is important is the predictability of the fee, so it can be factored into the developers
cash flow forecasts at time tender. If indeed it is only recovered at the time of financial closure, there
is little to no impact on the developer.
Box 8: Impact of Public Funding of Up-Front Exploration in Mexico

A recent study of the impact of up-front exploration funded by the public sector, and improvements to
the concession regime, is reported for Mexico. As shown in the figure, the total impact on the tariff was a
reduction of 19%, or 1.8 US/kWh, comparable to the estimates presented here for Indonesia.
US/kWh
100
92
14
74
80
60
40
20
0
Current
Cost
(NPV)
Amendments
to the
Concession
Up-Front
Exploration
Drilling
NPV = net present value, US/kWh = cents per kilowatt-hour.

Source: SENER. 2013. Initiative for the Development of Geothermal Regime in Mexico.
Potential
Cost
Nevertheless, there remain the usual questions of moral hazard as applicable to all insurance and
quasi-insurance schemes, which applies in this case not to developers (as when occurs for traditional
insurance drilling schemes), but to the Geothermal Fund itself. Since the cost of failures can be
recovered by the fee, what incentive is there for efficiency in running an effective exploration program?
If a drilling contractor is hired by the fund to do the exploration for a fixed fee, he also has no incentive:
consequently, there needs to be some bonus system to incentivize the drilling contractor. There are
many examples of how to do this within a commercial drilling contract.
8.3 Conclusions and Recommendations

The desirability of up-front exploration being funded by governments has long been noted. The
analysis presented here shows just how large is the cost difference between government and private
exploration where tariff recovery follows only many years later and explains why so many countries
have funded comprehensive geothermal exploration programs from the state budget.
The analysis also shows that by far the lowest cost option is for the fund to recover exploration
expenditures at financial closure. This is because at financial closure the repayment to the fund can
be rolled into the debt, at which point the relevant cost is the incremental WACC. Depending on what
proportion of equity remains to be contributed by the developer, the incremental cost may be just the
cost of debt, likely to be significantly lower than the developers equity return. Indeed, where this cost
of capital is below the governments opportunity cost of capital (as will almost certainly be the case
if a significant portion of the debt is from IFIs), the net cost of exploration to Indonesia will be lower
than if cost recovery is not attempted. Even if additional equity contributions are required, at financial
closure the incremental return on equity contributed at that point will be significantly lower than the
rate required for additional equity at the point of tender.
Moreover the analysis shows that even if the cost recovery from the development includes a
premium for the fund to cover the costs of unsuccessful exploration programs, which do not lead to a
commercially viable project and hence to cost recovery, the failure rate could be as high as 60% and
the fund could still maintain its nominal starting balance.
The main objective of the Government of Indonesia should be to fund as much of the incremental
cost of renewable energy from the IFIs under concessionary financial terms as possible. This is not an
unreasonable strategic policy imperative insofar as the main beneficiary of GHG emissions reductions
is also the global community.
However, the reality remains that the private sector operations of the International Finance
Corporation or of ADB will not lend to projects until the resource is largely proven. With funds from
the Clean Technology Fund (CTF), ADB approved a loan to the Rantau Dedap IPP project to support
drilling costs on a nonrecourse basis in June 2014. Sovereign operation funds of ADB and IBRD can
be used for exploration since these loans benefit from a sovereign guarantee. However, the headroom
for such borrowing is limited and MoF is increasingly reluctant to provide these guarantees. Therefore
the use of the governments Geothermal Fund as a means to progress project development to the
stage wherein upon resource confirmation, IFI debt instruments become available, is an excellent
option.
There remains the question of what impact such an exploration program, and cost recovery at time
of financial closure, would have on unserious bidders, particularly for small projects in eastern islands.
It is true that the recommended scheme would substantially de-risk small projects if a significant
part of the steam resource can be established at time of tender, in principle making it easier for
unserious bidders to bid (particularly if the costs are to be recovered at financial close rather than at

time of tender). However, the best mitigants to eliminate unserious bidders is (i) rigorous technical
prequalification; (ii) rigorous evaluation of technical capacity at the first-envelope stage of tender
evaluation; and (iii) the requirement for a substantial bid bond.
The recommendations that follow confirm, and further support with quantitative estimates of the
benefits, earlier recommendations made by ADB for the Geothermal Fund:64

The resources of the Geothermal Fund should indeed be used to establish a geothermal
resource prior to tender, with at least three successful wells.
The recovery of exploration costs from the commercial developer should be at financial
closure, and not at tender: at this point in the development process, a substantial fraction
(if not all) of the cost recovery can be rolled into the debt.
The repayment obligation should include an appropriate interest charge reflecting the
governments actual borrowing costs. However, the extent to which an additional fee is
necessary to cover the costs of exploration efforts that do not result in commercially viable
projects may need further study.
The repayment that the fund will require at the time of financial closure should be stipulated
in the tender document. Since this will be dependent upon the time that elapses to financial
closure, the payment should be presented in table form. To provide maximum certainty for
the bidders cash flow projections, a fixed interest rate should be used.
In short, over wide ranges of assumptions, funding of initial exploration by the government, with
recovery of some or all of the cost from the developer at the time of financial closure, is the optimum
strategy from the perspective of Indonesia. This is true regardless of who in Indonesia bears the
ultimate costwhether PLN and its consumers (once cost-reflective tariffs are attained), or the
government (if incremental costs are absorbed by MoF as part of the PSO).
64
See, e.g., the 2011 AECOM report to ADB that made many of these same points (AECOM, Geothermal Fund Report,
Report to ADB, 2011).
Institutional and
Financing Issues
9.1 Structural Issues

An important problem within the Indonesian geothermal sector is the conflicting objectives of the
main government stakeholders in the face of the perceived incremental coststhough as noted, there
is much uncertainty about what the actual incremental costs really are. These objectives could be
briefly described as:

Ministry of Finance (MoF) is concerned about the size of the PSO to PLN, which it regards
as increasingly unsustainable. Reducing the magnitude of the subsidy to PLN is its greatest
concern in the power sector, which obviously conflicts with the probable incremental costs of
geothermal energy, and the need to increase the subsidy to achieve the geothermal targets.
The Ministry of State-Owned Enterprises (MSOE) and state-owned enterprises (SOEs)

have as their main objective the satisfactory commercial performance of the enterprises,
which include both Pertamina and PLN. Consequently, the Pertamina Board of Directors
(and in particular the director primarily responsible for investment planning) are reluctant
to allocate equity capital to PGE when compared to the much higher returns available in
Pertaminas oil and gas plays. While an objective comparison of the resource risk in oil and gas
versus geothermal might well conclude that risks are higher in oil and gas, this can be more
than offset by the much higher returns. In short, the regulated returns in the electricity sector
are not commensurate with the risks of the geothermal businesswith the result that PGE
faces a continual battle for resources from its parent company.
The Ministry of Energy and Mineral Resources (MEMR) sees its role as the promoter of
geothermal energy, and is responsible for supervising the sectors development, including
responsibility for implementing the Geothermal Law and for tariff setting. It is the entity
primarily responsible for promoting geothermal energy, but whether its goals can actually be
met is determined by others.
A more nuanced view of MoFs position is that it is not so much the magnitude of the tariff that
is of concern; indeed, noted in Box 7 (Section 4), MoF is not averse to green objectives and even
carbon taxes. Rather, in the case of the Fast Track Program (FTP) projects that benefit from sovereign
guarantees, the concern is that in the event that the guarantee is invoked, MoF officials might stand
accused of corruption for causing a government loss, if the selection of the developer cannot be shown
to have been properly tendered. In addition, since MoF must provide the necessary subsidy to cover
the incremental costs of geothermal, there is an obligation to ensure economic efficiency.
These differences in objectives have not been helped by communication problems between MEMR
and MoF in the matter of tariffs in the past. Indeed, basic principles of stakeholder consultation were
not followed in the issuance of the 2012 FIT. The result was paralysis: as noted in the introduction,
very few additional projects will be added in the next few years. Wait and see is the openly admitted
posture of Pertamina, PGE, and other significant sectors of the developer community, including the
potential commercial partners.
83
9.2 The Role of the State

In most countries with a large degree of geothermal development, the primary role of the state in
the geothermal sector is exploration drilling, then often leaving the development of power generation
projects to the private sector at a stage when much of the remaining cost can be raised as debt. But
Indonesia has taken a different path, with its Geothermal Law assigning exploration drilling to the
private sector as part of the competitive system for bidding WKPs. Indeed, the extent to which the
Geothermal Fund can itself undertake drilling is one of the unresolved questions (Section 9.4).
Consistent with the view that geothermal development was a mining activity, Pertamina was given
the largest share of geothermal resources and the responsibility for developing them. PGE was
subsequently created as the vehicle for accomplishing this objectivethough it was Pertamina
who recently bid for a concession for Seulawah Agam in Aceh, not PGE. In addition, PT Geo Dipa
Energi was established in 2002 to develop fields in Dieng and Patuha, though without being provided
adequate financial resources. The reality of geothermal project development is that it requires large
amounts of up-front equity, so there is little point in a state-owned geothermal development entity if
a clear commitment to provide the necessary equity is lacking.
Ministry of Finance
At present, MoF is not permitted to provide guarantee to PLN payments (a prohibition that is not
unique to geothermal projects). Payment guarantees under the FTP1 and FTP2 projects require a
waiver through presidential decree. The mechanism for provision of guarantee to PLN is stipulated
in MoF Regulation No. 139/2011, 225/2013, and 173/2014. However, given the history of past PLN
defaults, the difficulty of obtaining payment guarantee of PLN obligations is an important barrier to
geothermal energy development, and needs to be addressed by a new regulation.
A Strategic Vision for the Sector

A coherent government strategic vision for the sector is needed. MEMR generally focuses only on the
technical aspects of the sector and not its financial and investment problems. The targets announced
for geothermal development, with a list of WKPs and projects to be developed by a particular date, is
not always consistent with actual financial and equity investment constraints. Resolving the underlying
conflicts existing between MoF, MSOE and/or SOEs and MEMR requires agreeing on the incremental
costs of different levels of geothermal energy and how these costs are to be funded. Only when such
agreement has been reached will be it reasonable for the government to decide what should be the
balance of the state and private sector, and then allocate the necessary investment resources to
Pertamina (and PGE), and Geo Dipa Energi, to enable that strategy to be carried out.
Pertamina Geothermal Energy

PGEs fundamental problem is that it is not in reality an independent company, adequately
and independently capitalized to undertake the mission entrusted to it. Although it has its own
independently audited accounts, and functions by law as a corporation, in reality it is better described
as an independent division of Pertamina. It is entirely dependent upon Pertamina for its equity
requirements to develop projects. This is perhaps well illustrated by the recent problems at Ulubelu,
where it took 18 months to obtain Pertamina approval for additional equity to complete the drilling
program after the initial phase encountered a lower success ratio than projected.
It seems evident that Pertamina is reluctant to commit equity capital to PGE to develop geothermal
projects when much higher returns can be obtained from the oil and gas sectors. If that is indeed the
case, there should be some time limit on Pertaminas legacy concessions, and Pertamina should face
limits on bidding on new concessions in the absence of progress on its existing fields.
Institutional and Financing Issues 85
Nevertheless, if the government wishes Pertamina and PGE to play a meaningful role in the countrys
geothermal development, then the MSOE should consider setting up a separate benchmark for the
equity returns for geothermal projects, and treat geothermal projects as a separate line of business
that should not be seen as competing with oil and gas. The rationale for such a benchmark is that
comparable equity returns to oil and gas projects would simply increase the MoF tariff subsidy to
PLN, and that government should recognize the overall economic returns to country of geothermal
investment, rather than the more narrow interest of SOE equity returns.
Geo Dipa Energi

GDE was established on 5 July 2002 to develop geothermal fields in Dieng (Central Java) and
Patuha (West Java). GDE currently owns a 60 MW power plant at Dieng (Unit 1) which is inoperable
at the moment but under rehabilitation to restore it to about 40 MW. GDE is also planning to
commission a 55 MW plant at Patuha (Unit 1) in the third quarter of 2014. GDE became a stateowned enterprise on 29 December 2011, following the passing of Government Regulation
No. 62/2011.
Following its new status as a state-owned enterprise, GDE planned a very ambitious expansion plan.
In the near term, GDE plans to commission additional units at Dieng and Patuha. Each of these sites is
expected to support steam resources that can support a maximum of 400 MW. Since 2013, GDE has
received capacity development grant assistance from ADB and Agence Franaise de Dveloppement
(AfD). ADBs assistance65 focused on enhancing GDEs capacity in safeguards and procurement
practices and preparing them to engage with international development banks. AfDs assistance
targeted reviewing steam resource assessments at their two locations and preparing an expansion
plan. Through this technical assistance, GDE is currently in discussions with ADB and AfD to finance
for expansions.
Perusahaan Listrik Negara Geothermal (State Electricity Company)

PLN has a long history in Indonesian geothermal development. The first geothermal power plant which
was commissioned in 1983, Kamojang 1, was developed and owned by PLN. PLN owns and operates
a number of geothermal power plants with the steam supplied from PGEs geothermal areas. PLN
Geothermal was established in 2008 to develop and maintain the upstream activities for geothermal
concessions.
PLN and PLN Geothermal hold concessions in the eastern area of Indonesia: Tulehu (Maluku),
Ulumbu and Mataloko (East Nusa Tenggara), and Sembalun (West Nusa Tenggara). An integrated
upstream and downstream geothermal development for those projects resulted in the power plants
at Mataloko (1 x 3 MW) and Ulumbu (2 x 2.5 MW). The projects have provided PLN valuable
experience in the geothermal development. However, in 2013 PLN decided to suspend the activities
of PLN Geothermal because there are already two other SOEs (Pertamina/PGE and Geo Dipa Energi)
engaged in geothermal development on behalf of the state.66 While PLN Geothermal is still a legally
established entity and holds the concession, PLN appears to have plans to continue its geothermal
program under PLNs own Renewable Energy Division.
65
ADB. 2011. Technical Assistance for Geothermal Power Development Project. Manila.
66
PT PLN (Persero). 2013. Annual Report 2013. Jakarta.
9.3 Pertamina Geothermal Energy

Commercial Partnerships
Objectives
Largely at the behest of Pertamina, PGE appears to be negotiating a number of proposed commercial
partnerships67 as a way to raise the necessary additional equity that it cannot get from Pertamina in the
present fiscal environment. The basic strategy is to establish new project companies in which PGEs
equity takes the form of contributed assets rather than new cashi.e., the value of the preparatory
work of site development is already completed. This constraint means that the new entities will be
majority-owned (and controlled) by the new partners, with PGE holding only a small minority share.
A second objective is to bring to PGEs operations the partners presumably superior technical and
management capabilities, since the operational management of the projects will be taken over by the
commercial partners.
Tariffs
Whether the commercial partnerships will proceed will depend, among other factors, on whether the
tariff permits the desired equity return objectives of the commercial partners. It seems likely that these
are higher than PGEs equity return targets that were the basis for past agreements on tariff with PLN.
PLN and PGE have jointly commissioned a study by Sinclair Knight Merz Limited to reassess the
tariffs necessary for, among others, the proposed Tompaso 40 MW and Karaha 30 MW projects to
succeed. After reviewing the tariff assessment proposal, PLN and PGE signed a Heads of Agreement
with tariffs in the range of 8.4 US/kWh11 US/kWh,68 as compared to the previously agreed tariffs of
less than the old 6 US/kWh8 US/kWh ceiling. Such results may well reflect negotiating positions,
and depend on key assumptions: one should bear in mind that developers will always claim a regulated
tariff is too low, regardless of what level is proposed. The relatively small sizes of the projects analyzed
may have some bearing on the high tariffs. Unfortunately, as of the time of writing, we have not been
given access to this report, and therefore cannot comment on its assumptions or the extent to which
the results are reasonable.
Availability of Commercial Partners

While the progress or the outcome of the negotiations is unclear, an interesting question is whether
there are suitable partners available. To be suitable, partners need to have a strong balance sheet and
the ability to raise up to $100 million in equity or nonrecourse debt for drilling; a mandate to participate
in overseas projects (which precludes, for example, Comisin Federal de Electricidad of Mexico and
some other SOEs); good expertise and preferably a proven track record in geothermal projects;
and preferably (but not necessarily) some Indonesian experience. Possible additional candidates
could include:

Origin (Australia): Origin is suitable given its ownership of Contact Energy, which is New
Zealands largest geothermal generator, and its part-ownership (along with Tata Power
of India) of OTP Geothermal, which is undertaking the Sorik Merapi project in Sumatra.
67
Think Geoenergy. 2013. Pertamina to Partner with Chevron and Star Energy on a Project Development. 20 November.
http://thinkgeoenergy.com/archives/17192
68
PLN press release. The Signing of the Head of Agreement (HoA) PLN PGE about the Basic Price of Geothermal Steam
and Electricity Power. 24 April 2014. Jakarta. http://www.pln.co.id/eng/?p=2886
However, it declined to participate in the first round of commercial partnership negotiations,

and appears to be giving geothermal a low priority overall.
69
Supreme Energy (Indonesia, with overseas backers): Supreme Energy appears not to have
been formally invited to participate in the first round. Supreme Energy is in fact an umbrella
for three separate project-specific companies, each with different shareholding, and as such it
is not a single potential partnership company. The participants may be already fully occupied
with their own three concessions.
Ormat (US): Ormat is one of the worlds largest geothermal generators as well as equipment
suppliers. It is participating in the Sarulla project along with Medco. It has the financial
resources and downstream track record to partner with PGE, but perhaps lack upstream
credibility. It may also consider that it is already sufficiently exposed in Indonesia already with
Sarulla.
Mighty River Power (NZ): Mighty River Power is well qualified, being New Zealands second
largest geothermal generator, and has in recent years taken on several overseas projects.
However, not all of those projects have gone well and the company has previously mentioned
that Indonesia was not in its preferred portfolio.
ENEL (Italy): ENEL is suitable, being the worlds second or third largest geothermal company,
and it has taken on overseas projects in several countries including Chile, El Salvador, and
Nicaragua, but lacks Indonesian experience and has not previously expressed an interest in
Indonesia.
Calpine (US): Calpine is suitably qualified, being the worlds largest geothermal company.
It has previous experience in Indonesia, but it exited post-1997 and has not expressed any
interest in returning, presumably because it regards the country risk as too high.
Other US and Canadian geothermal companies: The other US and Canadian geothermal
companies are either too small or financially under-resourced to enter the Indonesian market
(e.g., Ram Power, Alterra, US Geothermal), or, like Calpine, suffered bad experiences with
the PPA defaults in 1997 and are unlikely to return (e.g., California Energy, Florida Light, and
Power).
Icelandic companies and consortia: There are several Icelandic geothermal companies,
some of which are SOEs, some of which are private, and some of which have overseas
investors. In the past, Icelandic companies have shown an ability to work together on an ad
hoc basis so they are best considered as a group. Icelandic geothermal expertise is world class
and companies have aggressively looked for both consulting and investing opportunities in
other countries, with strong support from their government. After the 2008 collapse of the
Icelandic banking sector they have lacked financial credibility, but more recently have for
example started on a project in the Philippines. As a whole, they should be considered serious
contenders for partnerships in Indonesia.
Marubeni (Japan): Marubeni is an engineering, procurement and construction contractor

and generator. It has previously expressed an interest in geothermal projects in Indonesia,
has partnered with Supreme Energy for one project, and has adequate financial resources. It
lacks in-house upstream expertise but has shown a willingness to contract that in. It could be
a serious contender.
Mitsubishi (Japan): Mitsubishi has entered the geothermal business by acquiring 20% of Star
Energy shares.69
Sumitomo (Japan): Sumitomo is partnered with Supreme Energy on two of its projects.
This is Sumitomos first foray into geothermal energy development as distinct from
Mitsubishi Corporation. 2012. Mitsubishi Corporation Enters Geothermal Business in Indonesia. http://www.mitsubishicorp.
com/jp/en/pr/archive/2012/html/0000017168.html

engineering, procurement, and construction contracting. Through its links with Fuji it has
good downstream expertise, but lacks upstream expertise. It is probably waiting to see
the results of their participation in Supreme Energy before entering into other geothermal
projects in Indonesia.

GDF-Suez (France): GDF-Suez is also partnered with Supreme Energy, but in all three of the
Supreme Energy projects. It has good downstream expertise in fossil fueled projects, but also
lacks upstream expertise. It is also probably waiting to see the results of its participation in
Supreme Energy before entering into other geothermal projects in Indonesia.
9.4 Finance
Geothermal energy production is capital intensive, and will require billions of dollars in debt and equity
to realize the targets set by MEMR. The cost of exploration alone, for the next 3,000 MW is estimated
at $2.8 billion (see Appendix 6) for details. With little of this likely to come from Pertamina, and the
present resources of the Geothermal Fund limited (at present) to just $300 million, the bulk of this will
need to be raised from the private sector. It remains to be seen whether this can in fact be mobilized.
Financing the downstream power generation project once the resource is confirmed poses few
issues unique to geothermal. Off-take risk, readiness of the transmission connection, and guarantee
of payment by PLN are common to all IPPs. The technology of power generation using geothermal
steam is well established and cost overruns of the type associated with hydropower projects (due to
geotechnical and tunneling risks) would be rare.
What is difficult is debt finance of the exploration stage. As noted above, if loans from the Geothermal
Fund require 100% collateral, then such loans are equivalent to equity.70 It is not unreasonable to
conclude from global experience that the best prospects for successfully developing the geothermal
resource is for government to take the upfront exploration risk, and simply tender as IPPs the
downstream partwhich is the route followed by the Philippines. The Indonesian approach that
places that risk on the developer makes the funding of the exploration stage the critical problem.
As long as geothermal development is subject to unpredictable institutional constraints (one again
thinks of the Sarulla project), the lack of lender enthusiasm with financing high-risk geothermal
projects is understandable. The harsh reality is that what the geothermal sector needs most to facilitate
financing is a stable tariff regime, some form of payment guarantee for PLNs off-take obligations, and
a predictable regulatory environment.
Our consultations with developers suggests that the larger, established developers are content
to assume the exploration risk (i.e., finance with their own equity) for larger projects, but have less
interest in the smaller projects in the eastern islands, even if they were de-risked by the Geothermal
Fund (as suggested in previous sections). Large developers are only interested in large projects.
Nevertheless, in addition to the $3 billion for exploration, and assuming that the results of exploration
will indeed result in an additional 3,000 MW of capacity, with the power generation portion likely to
cost at least $2,500/kW, the total financing requirement will be close to $10 billion. Many put the total
cost at $4,500/kW, which for 3,000 MW comes to $13.5 billion.
In principle, there is no shortage of potential sources of concessionary finance: the World Bank Group,
ADB, IFI, and JICA and JBIC have ambitious plans to provide support, including CTF funds.
70
It is unclear to what extent this requirement could be replaced by a Bank Guaranteethough such a guarantee would also
require a corresponding security.
International Bank for Reconstruction and Development

IBRD can lend to public developers (and publicprivate partnerships projects) with sovereign
guarantees, so there is in principle no reason why IBRD loans could not be used for exploration drilling
under the protection of the guarantees.71 But the question is the extent to which the MoF would
be prepared to issue guarantees if it had doubts about the risks for a given project (though all FTP2
projects supposedly qualify for MoF guarantees).
In the case of Ulubelu and Lahendong projects, IBRD is not providing funds for drilling, only for steamfield development and the power plant: funding for drilling has to be provided by Pertamina, and delays
in securing such funds have had significant impacts on the projected COD. In this project, MoF is the
borrower, who on-lends to Pertamina, who in turn provides (debt) funding for PGE. At present, MoF is
not permitted to provide guarantees to funds directly lent to entities such as PGE, though a discussion
is currently underway for a mechanism which would permit direct loans to SOEs with MoF guarantees.
This would require a new government regulation. Of course, this is not a matter only for IBRD, but one
that affects lending by ADB and JBIC as well.
ADB
ADBs experience in financing geothermal projects in Indonesia dates back to 2002. ADB financed
PLNs Lahendong Geothermal Power Plant Units 2 and 4 with the governments sovereign guarantee.72
Building on this experience, ADB prepared a multitranche financing facility (MFF) for multiple projects
to PGE in 2011. An MFF requires extensive initial due diligence, but in return it allows additional funds
to be disbursed quickly with simplified administrative processes. The MFF was also designed to
incorporate the use of CTF to further improve the financial viability of the target projects. However,
the disbursement planning of these sovereign loans to SOEs have become subject to parliamentary
approval in Indonesia since 2009, and this caused delays in securing financial resources for Pertamina
for PGE projects. As a result, the government of Indonesia and Pertamina decided not to pursue any
further use of sovereign-backed financing for PGEs geothermal projects. The preparation of ADBs
MFF was therefore cancelled in 2012.
ADB later received a request from GDE for technical assistance for its resource assessment and
corporate planning in 2012. Through this assistance, GDE and ADB are currently planning a new
sovereign-backed loan for its expansion plans. Despite the needs for large financing for its two working
areas with a total potential of 800 MW, GDE, as a new SOE, has limited borrowing capacity. Therefore,
loans from multilateral or bilateral financing agencies with sovereign guarantees have become the
most sensible solution. Since the loans would be sovereign guaranteed, ADB is also able to finance
exploration drilling activities, which commercial loans cannot cover.
ADBs Private Sector Operations Department has also been involved in the sector, and the signing
of financing agreements between lenders and the developer of the Sarulla project was reached in
March 2014. This is the first new geothermal IPP project73 in the country for over a decade. The
funds from CTF are used as mezzanine financing, making the debt package more attractive to allow
further acceleration of the development. The involvement of ADB and JBIC (see below) also enabled
cofinancing by six commercial banks. Since the MFF was cancelled, Pertamina also approached and
mandated ADBs Private Sector Operations Department for a direct corporate loan with CTF for their
geothermal expansions in March 2014. ADBs Private Sector Operations Department is in discussions
with other IPPs regarding financing with CTF at the various stages of project development, including
exploration.
71
This has been done in some other countries (such as Djibouti in the early 1980s), but it has not always worked well.
72
Both power plants were built under ADB Loan 1982-INO: Renewable Energy Development Sector Project.
73
This excludes expansions such as Wayang Windu 2, completed in 2009.

In addition to the role of sovereign loans, this illustrates another key role of IFI funding in leveraging
private lending to geothermal projects. There is only a limited number of commercial financing deals
for geothermal energy in the world. More successful cases are required before private lenders can
actively provide financing for the sector, and these IFI-backed private projects will lead the way. As
long as the regulatory framework, including processes for land acquisition and permissions, allows
developers to invest, IFIs are also equipped with financing resources such as CTF to support these
projects, as seen in the case of Sarulla.
Japan Bank for International Cooperation

JBIC provided the largest portion of the debt (42%) for Sarulla. JBIC has played an important role
in large-scale project financing in Indonesia and will continue to do so for IPP-based geothermal
development. It further facilitates the participation of commercial lenders by providing political risk
guarantees to cofinanciers. JBIC will likely be the lead bank for any geothermal IPP projects with a
significant participation of Japanese developers. However, JBIC normally requires a government
guarantee on the off-take agreement with PLN and is likely unable to finance exploration drilling.
With its capacity to finance a large portion of the required debt, however, it plays a pivotal role in
enabling the financial closure of a project. Therefore, a balanced mixture of JBIC loans and financing
for exploration through facilities like CTF would likely be the most effective financing sources for some
geothermal IPPs in the short- to mid-term.
Japan International Cooperation Agency

JICA is the public sector peer to JBIC in Japans bilateral partnership with Indonesia. It has also played
an important role in the Indonesian geothermal sector by providing sovereign loans to PLN and PGE.
The total commitment made by JICA for Indonesian geothermal projects is over $1.1 billion, the largest
among development partners. Similar to IBRD and ADB, JICAs loans have not yet financed drilling
activities. Its loans for the Lumut Balai (PGE) and Tulehu (PLN) projects, however, contain portions for
exploration drilling. These may be the first drilling activities to be financed by IFIs in Indonesia. JICAs
case will, therefore, likely provide important lessons for IFI-financed drilling activities, which will likely
be more complicated than sponsor-financed drilling owing to the requirements of IFI procurement
guidelines based on international or national competitive bidding. If the Government of Indonesia and
Pertamina resume the use of sovereign loans for Pertamina projects, JICA will likely resume its leading
role in geothermal financing for SOEs in Indonesia as well.
Stand-By Financing
PGEs recent problems with obtaining the necessary additional capital to fund Ulubelu and Tompaso
drilling highlights the need for flexible financing arrangements. There are good models from hydro
development that apply here. Hydro IPPs also face considerable uncertainty, albeit not from
resource uncertainty but from construction cost uncertainty, especially tunnelling risk. Indeed, some
commercial banks will often declare we take no tunnelling risk. Therefore it is not uncommon,
especially in Latin America, for hydro IPPs to arrange for standby financing arrangements triggered
by certain geological outcomes that cannot be predicted prior to actual tunneling or dam foundation
work.74 But that can only be arranged if any additional standby equity requirement has also been
arranged in advance. It also emphasizes the need for internationally credible resource estimation and
reporting. The current arrangements for PGE obtaining additional equity from Pertamina are a case in
point: any hope for flexible financing is academic under such conditions for equity.
74
World Bank. 2010. Peru Overcoming the Barriers to Hydropower. Energy Sector Management Assistance Program Report.
53719-PE. Washington, DC.
Payment Guarantees and the Role of the Multilateral Investment

Guarantee Agency
Indonesian IPP development suffered a severe blow in the aftermath of the 1997 financial crisis, as
PLN defaulted on some of its payment obligations. Consequently, it is entirely understandable that
geothermal IPPs have sought guarantees for PLNs PPA obligations.
Where equity is provided to geothermal projects by foreign investors, there is an opportunity for MIGA
to provide assistance. MIGA is designed to insure foreign investors (say Chevron) against default by
a state-owned enterprise (such as PLN).75 While MIGA has not yet done a recent geothermal deal,
36% of its global fiscal year 2013 portfolio is in the power sector, and will be providing guarantees for
an Indonesian hydropower project (Figure 9.1). Tenor is up to 15 years, with 20 years on an exceptional
basis, and up to 90% of equity, and 95% of debt can be insured. One of the conditions for a MIGA
guarantee is a No Objection Letter by MoF to such guarantee. MoF should be ready to provide such
a letter.
The cost of a MIGA guarantee varies according to country, sector, transaction, and type of risk. A typical
premium for a power project would be 125150 basis points per year (1.25%1.5%) for insuring the SOE
payment obligation. However, on the strength of MIGAs AAA rating and zero risk classification by the
Bank for International Settlements, a MIGA guarantee can lower the cost of debt. With a guarantee in
place, a typical financing at the London Interbank Offered Rate (LIBOR) + 500 basis points could be
obtained at LIBOR + 350400 basis points, so the effective net cost may be quite small or negative.76
However, because MIGA is limited in the volume that it can provide in the coming year (that applies
not just to geothermal IPPs), the guarantee by MoF is vital for new geothermal projects. Until the PLN
tariff reaches cost recovery level, lenders will require some sort of guarantee to PLN payments, which
Figure 9.1: Multilateral Investment Guarantee Agency Guarantee to Offshore Lenders:
Indonesia Hydro Project
PT PLN
(Indonesia)
Foreign
Investor
Owns
(100%)
Lenders
Lenders
Indonesia Power (IP)

Owns
(51%)
Owns
(49%)
Equity
EPC
Contract
EPC Contractor
Project
Enterprise
Project
Enterprise
(Indonesia)
(Indonesia)
Guarantee
Debt
O&M
Agreement
PPA
PT PLN
IP
EPC = engineering, procurement, and construction; IP = PT Indonesia Power; MIGA = Multilateral Investment Guarantee Agency;
O&M = operation and maintenance; PPA = power purchase agreement; PT PLN = PT Perusahaan Listrik Negara.
Source: Multilateral Investment Guarantee Agency, Project Brief. Rajamandala Hydropower Project. http://www.miga.org/projects/index
.cfm?pid=1367 (accessed 10 September 2014).
75
MIGA has two distinct products that potentially apply here: non-honoring of SOE financial obligations, and breach of
contract by an SOE.
76
From MIGA.

MoF needs to explicitly include in its overall support to the sector. As with the case of direct loans to
SOEs, a new regulation is required to allow MoF to do so (and to protect MoF officials to make those
decisions).
9.5 Geothermal Risk Mitigation

Geothermal Fund
In 2011 the government established a Geothermal Fund, now funded to $300 million. However, a
satisfactory model for use of these funds has yet to be developed, and no disbursements have yet
been made. A viable model is urgently required, which will require a change in the current regulations
of the fund.
One of the proposed disbursement models is based on providing loan funds to developers for upfront exploration drilling. As noted above, such lending conditions are considered onerous, since they
are reported to require 100% collateral, in effect requiring equity (which if available, might just as well
be used directly for exploration rather than pledged as collateral). For the loans to holders of IUPs
to be effective, it is essential that this collateral requirement be revised.77 However, there is already
another financing scheme for government-led exploration before tender in the current regulation for
the operation details of the Geothermal Fund. This mechanism is much better structured, and rather
than improving the loan scheme for IUP holders, MoF should focus on operationalizing the funding
scheme for exploration before tender. This requires careful attention to the shift of tender authority
from local governments to the central government under the new bill on geothermal.
A major concern is related to the use (and waste) of government funds for unsuccessful projects:
according to the Law of State Finance 17/2003, actions by government officials causing losses to
the state, even if merely by negligence, are considered corruption. This principle might be applied to
exploration whose costs might ultimately be unrecoverable if a WKP proves not to be commercially
viable.
This concern can be mitigated. As suggested by MoF, decisions about which resource areas are to be
selected for exploration drilling can be vetted by an independent panel of international technical advisors,
who would confirm that, based on the best available surface data, exploration drilling in a particular area
was prudent. That would absolve government officials from the charge of negligence. In addition, one
may impose a fee on successful projects to make up any potential losses of unsuccessful projects.
Ideally, tenders should only proceed on the basis of measured resources (say a minimum of three wells),
for which we recommend use of the Geothermal Fund. The fund would finance exploration drilling
as a public good, albeit with recovery of costs at time of financial closure of a project that proves to
be commercially viable. Private capital for up-front exploration expenditures is expensive, which will
ultimately be reflected in the tariff. Up-front de-risking will not only improve the quality and number
of bidders, but result in a lower equity return required by bidders, reducing the tariff. Calculations in
Section 8 show that the smaller the project, the greater the likely impact on the tariff (ranging from
1 US/kWh for large projects to 3 US4 US/kWh for small projects in the eastern islands).
We recognize, however, that it may take some time to find a satisfactory mechanism for using the
Geothermal Fund for this purpose, since at present the fund lacks the technical capacity to manage
an exploration program. We accept the recommendation of the Indonesia Geothermal Association
(INAGA) that the fund resources should be directed in the first instance to the eastern part of
77
Also as noted previously, there is the question of whether a bank guarantee in the same amount would suffice.
Stakeholder Comment 6: Use of the Geothermal Fund

Comment by the Indonesia Geothermal Association, March 2014:
Relating to the utilization of the Geothermal Fund, the Indonesia Geothermal Association
(INAGA) can understand that the Geothermal Fund is used to reduce exploration risk. However,
we also need to consider that the energy crisis has to be immediately resolved and geothermal
power plant is the part of the Acceleration Program. If the government has to conduct the
exploration drilling before tendering out, it will produce better data quality. However, it will also
take longer process since the exploration program takes 35 years to complete while many wilayah
kerja pertambangan (geothermal work areas or WKPs) are to be explored, so the acceleration
program will be delayed.
Therefore, we suggest using the Geothermal Fund more effectively. The Geothermal Fund should
be used for geothermal development in the eastern parts of Indonesia where there is limited
investor interest, and therefore government support is required to develop geothermal resources
in the area. For other areas where many investors are interested to develop, the Geothermal Fund
is not needed since many bidders will participate in the competitive tender process as long as the
ceiling tariff is attractive to them.
Reply:
We agree that the first priority should be to improve resource data in the eastern islands, where
indeed there would be little interest in the big developers to develop projects below 30 MW.
However, the funds resources should be used to prepare a comprehensive geology, geophysics,
and geochemistry package (3G) for all projects prior to tender, regardless of location. And
because resources are scarce, a prioritization scheme is needed.
We also agree that it will take time for the Geothermal Fund to establish a workable model for
exploration, to establish a new central tender entity, and to establish its technical expertise to
be able to manage an exploration program (for which international technical assistance will be
required). However, with assistance from the Japan International Cooperation Agency (JICA),
a consensus is emerging on the best way forward, in which the exploration program would
be managed under a long-term contract to a suitable consulting company on behalf of Pusat
Investasi Pemerintah (Indonesia Investment Agency) (PIP) and Badan Geologi, with appropriate
safeguards to avoid conflicts of interest, and independent certification by international experts.
Box 9 presents the JICA-proposed scheme.
Source: Personal communication from INAGA. 11 March 2014.
Indonesia where the interest of the large developers is limited (Stakeholder Comment 6).78 We also
accept the suggestion of Chevron that in the first instance the fund should be directed to ensure the
completeness of the governments geological data and to ensure the adequacy and completeness of
the 3G information made available at time of tender. However, even this less ambitious activity will
require technical assistance, and expert advisors to assist in the interpretation of the data.
Box 9 shows the JICA proposal for the organizational structure for exploratory drilling, with much
emphasis on the avoidance of conflicts of interest and the need for independent review to certify
results. The details of some of the individual elements still require further elaboration, but the study
team supports this general approach.
78
Personal communication from INAGA. 11 March 2014.

Box 9: Proposal for the Organizational Structure for Exploratory Drilling
A viable model for the effective use of the resources of the Geothermal Fund must recognize the following
realities:

Pusat Investasi Pemerintah (Indonesia Investment Agency) (PIP), as an arm of the Ministry of Finance,
does not have adequate technical capacity to undertake an exploratory drilling program.
Much of Indonesias technical expertise in geothermal resource assessment resides in its Badan Geologi.
For the information to be of real value as a basis for tendering, the geology, geophysics, and geochemistry
(3G) and exploratory drilling information must adhere to an internationally recognized reporting code
(see Appendix 1), and be independently certified.
Japan International Cooperation Agency-supported technical assistance is preparing a concept for an

organizational structure, as shown here:
Independent
Certifier
Data
Independent certier is typically a consulting company, and responsible for

certifying the data prepared by the PMT.
Selected from the qualied pool of companies, and allocated to a project.
To ensure neutrality, the independent certier who is selected will have no
Data
Certication capital tie with IPM contractors and PMT.
Drilling Sponsor
Certied Data
(PIP +Badan Geologi)
Local/Central/
Government
RIP with
Certied Data
IUP Tender
Participants
Reporting Drilling progress, provide analyzed data
PMT is typically a consulting company, employed on long-term contract and

responsible for multiple projects during the period.
Management Team) PMT is responsible for (i) supervising detailed survey (incl. additional management
team survey), (ii) selection of drilling sites, (iii) preparation of drilling planning,
Drilling
Reporting
(iv) supervising exploratory drilling, and (v) resource evaluation.
Plan
Drilling progress
PMT(Project
Field Supervisor
Day to day
Instruction
Reporting
Drilling
progress
IPM (Integrated
Project Management)
On-site representative of the drilling sponsor and is in overall charge of the

drilling and associated activities, and responsible for giving day-to-day
instructions about the drilling operation to the drilling contractor at site.
Monitoring progress of works towards drilling completion, obtaining log and
preparing daily activity report.
Compilation of exploratory drilling mission according to sponsors spec.
Procure sub-contractors and supervising their activities.
IPM contractor must bear the operational risk of exploration activities, but does
not need to bear the results of exploration activities.
Legend:
Contract
Service Flow
Key elements of such a structure include:

a suitably qualified consulting company, under long-term contract to PIP and Badan Geologi, to manage
the program and be responsible for multiple projects during the contract period;
an independent entity to certify results (who may be an individual international expert), drawn form a
pool of certified experts, and free of potential conflicts of interest before and after tender; and
experts drawn from this pool also advise PIP and Badan Geologi on the prudence of committing
resources to particular projects, which are then passed to the project management team for execution.
Source: PWC/KRI/Maxeed, Proposed Organization Structure for Geothermal Fund Facility, Progress Report to JICA, January 2014.
Support Mechanisms in International Practice

A variety of support mechanisms are found in international practice, which might be adopted by the
fund. These fall into three broad categories: (i) drilling insurance, (ii) direct grants, and (iii) revolving
funds.
The objectives of such schemes vary. In some cases, such as many projects in Europe and perhaps to
a lesser extent in the US, the objective has been to stimulate commercial geothermal development
in a known geological setting with readily quantifiable risks. Therefore, the emphasis is on successful
production rather than exploration or innovation. Similarly in Germany, the proposed scheme will
initially be applied to well-known areas to gain more confidence in geology and scheme practice.
In a later step, new geologies (e.g., North German Basin) and new technologies (e.g., Enhanced or
Engineered Geothermal System) will be covered by the scheme. In contrast, in Australia and the
African Rift Geothermal Development Program (ARGeo) program, the emphasis was more on
exploration and proof of technical concepts, so projects that test a range of geological settings were
favored. Most schemes other than the wide-ranging whole-project loans, are designed to operate on
a well-by-well basis. The ARGeo scheme has a rollover provision under which the maximum payment
reduces from the initial to subsequent wells.
Drilling Insurance
A number of drilling insurance schemes have been tried or proposed worldwide, including the World
Bank-supported Geothermal Energy Development Program (GeoFund) for Europe, and ARGeo in
Africa. Under the GeoFund, the Geological Risk Insurance program, which is the cornerstone of both
programs, is designed to absorb a major portion of that risk by providing insurance against the failure
of the reservoir confirmation drilling program to confirm the existence of an economically exploitable
geothermal resource. Such economic viability is to be determined on the basis of the temperature,
flow and/or pressure, and chemical suitability of the geothermal fluids to allow for power production
or in the case of the GeoFund, significant direct use projects (e.g., district energy or industrial process).
Criteria for success or failure are to be negotiated prior to the initiation of the drilling activities.
Some level of partial compensation can be awarded for wells that are below the level of anticipated
production, but cannot be considered to be eligible for full compensation. For example, if a project
is based on an estimated well productivity of 5 MW per well, but in fact proves capable of producing
only 3 MW, a partial payment may be made if such terms have been negotiated at the time the
agreement was entered into. The coverage applies only to eligible cost factors; any failure is based
solely on geological parameters and no coverage is available to cover so called drilling risk for which
conventional drilling insurance may be available (Box 10).
One major feature of the rollover provision is that coverage will be available only on a declining basis.
For example, only 80%, 60%, and 40% of eligible expenses would be eligible for coverage for each of
the three wells. Eligible expenses have not been fully defined at this point in time, but could exclude,
for example, excessive mobilization and demobilization costs, infrastructure development costs
(roads, pipe lines, etc.). A cap may also be applied to the eligible expenses in as much as the entire
budget for the ARGeo Geologic Risk Insurance program is only $12.3 million. This should go a long
way in helping to confirm viable geothermal resources and provide the needed incentive for attracting
capital for project development.
Some schemes have been funded by governments or development agencies, which effectively provides
free insurance subject only to an administrative fee. In other cases such as the unique Munich Re
scheme in Germany, which is offered by a private commercial entity rather than a bank or international
agency, a substantial premium is charged to fully reflect the risk to the insurance company. In no case
is full coverage provided. There is always some excess to the developer (Box 11).

Box 10: The Geothermal Energy Development Program (Europe) and African Rift
Geothermal Development Program Schemes
The insurance program is in the form of a contingent grant with no money made available to the recipient
of the insurance coverage until such time that the well is deemed to be a failure. To protect both the World
Bank and the developer, it is critical that the well is instrumented and all drilling parameters are recorded so
that disputes are minimized should there be a call upon the insurance.
In the case of the Geothermal Energy Development Program (or GeoFund),a only one project developer
has so far received insurance coverage. In this case the primary project sponsor was the Hungarian oil and
gas company, MOL. The coverage amount was 85% of eligible expenses and the developer only asked
that the coverage be against failure to encounter sufficient fluids to make the project economically viable.
The sponsor felt there was no risk that the temperature would not be adequate based on prior drilling for
hydrocarbons in the area. After completion of the drilling and well testing activities, the well was deemed a
failure based on the lack of flow and the World Bank paid out an amount in excess of $3 million.
The African Rift Geothermal Development Program (ARGeo program),b designed to initially serve six East
Africa Rift Zone countries, was initiated in 2006. The ARGeo program will provide many of the same types
of assistance as the GeoFund. However, there is no direct investment funding window provided in the
ARGeo program. A major component of ARGeo is directed toward technical assistance and the focus is
upon capacity building, removal of legal and intuitional barriers, and conducting exploration activities.
Under the ARGeo program, the United Nations Environment Programme (UNEP) is in charge of all predrilling technical assistance, while the World Bank retains management of the Geologic Risk Insurance
Window and also has the ability to provide limited technical assistance to address post drilling issues. Some
post drilling issues, however, must receive some attention prior to the initiation of the drilling phase. These
include negotiation of the concession or lease agreement and, where outside financing is required, the basis
for a power purchase agreement (PPA) which provides project participants with an assurance that there will
be a market for the power that may be produced if the drilling phase is successful. Other post drilling technical
assistance can be provided for developing business plans and financing packages, and final negotiation of
the PPA. A recent addition is an internship scheme where project developers can nominate individuals to
receive internship positions with geothermal developers and/or operators in order to gain needed expertise
prior to the project coming on line. The cost for the internship is to be borne by the World Bank.
A major new feature of the ARGeo Risk Insurance Window is the so-called rollover feature. Upon applying
for coverage, applicants can request that the insurance coverage be eligible to be rolled over to as many as
three wells if each preceding well is successful. The argument was made successfully that one successful well
would not provide sufficient resource information to allow the developer to finance the project or attract
additional equity partners. Three successful wells would provide much greater assurance that a viable
geothermal resource would be developed for power generation.
World Bank. Geothermal Energy Development Program (GeoFund). http://www.worldbank.org/projects/P075046/geothermal-energydevelopment-program-geofund-1st-tranche?lang=en
b
World Bank. African Rift Geothermal Development Program Project (ARGeo). http://www.worldbank.org/projects/P100203/africanrift-geothermal-development-program-argeo?lang=en
a
Box 11: Private Geothermal Insurance Schemes

The first insurance to enter the market was Munich Re in 2003. The insurance covered the exploration risk of
the Unterhaching project. The insurance was established with support and pressure from the state of Bavaria.
The premium was more than 1 million (about $1.25 million) and also included a considerable nonclaim
refund. Subsequently, insurance brokers (i.e., Marsh and Willis) and insurance companies like Axa, Gothaer,
Swiss Re and R&V entered the market. There are both insurance companies acting as a direct, unique insurer
of the risk and insurance brokers distributing the risk between one leading and several contributing insurance
companies on the market.
Technical drilling risks can be insured on a standard basis. The conditions for an offer of discovery risk
insurance on the private market are a mature project preparation and a substantiated geological-technical
exploration and development concept of the geothermal project.
As no standards have been established for this kind of insurance yet, the cooperation between project
developer and insurer is of major importance. The clear definition of scenarios, best- and worst-cases,
measures and procedures is crucial to produce a reliable and transparent policy. Both the drilling concept
and the layout of the test program for the certification of results should be specified in advance and form
part of the insurance policy. The implementation and continuous revision of those concepts in the light of
the actual geological conditions is the responsibility of the project developer and is realized in dialogue with
the insurer.
The general concept of the private insurance solutions is to let the customer choose the desired insurance
sum according to the expected investment costs. Currently, a premium of at least 1 million is required for an
insured sum of about 10 million (about $12.5 million) plus an own risk share (deductible) of about 1 million.
The major downside of such an approach is the lack of such insurance providers worldwide and the very high
premiums that any insurance provider would require, especially in new exploration areas.
Outside of Western Europe (France, Germany, Switzerland) the uptake of the schemes has generally
been disappointing and in some cases most of the funds have not been used. There are various reasons
why that might be so, but the following general issues need to be taken into account for such schemes.

If the full risk of the insurance cost is to be covered, in most cases that can only be poorly
quantified, which makes the schemes unrealistically expensive. The larger developers would
prefer to take a portfolio approach and, in effect, self-insure.
The legitimate cost of wells needs to be carefully established in advance. That can be done
by having an agreed schedule of anticipated well costs. A particular issue is what constitutes
a reasonable rate for the developers own input as, for example, if it owns the drilling rig or
undertakes the well testing using its own equipment.
Drilling plans need to be flexible to cope with unexpected subsurface conditions.
Failures or cost overruns due to mechanical failures or poor drilling practice need to be
excluded. That, together with the previous point, implies that a degree of technical supervision
be provided by the funding agency.
Success and failure of wells must be carefully defined beforehand. That can be covered
by careful testing protocols, the cost of which is included within the well cost (as it can be as
high as several hundred thousand dollars). There nevertheless remains a potential for abuse.
There are anecdotal instances in the US of successful wells being cemented up to claim the
insurance, with the developer having nevertheless benefitted by locating the geological targets
and proving the resource. One way around that, which was proposed for Chile (but not yet
implemented), was that no payout would be made unless the concession was relinquished.

The above four points mean that the funding agency must have a good deal of technical
expertise, either in-house or contracted, and a fair degree of involvement overseeing the
drilling process. That all adds to the cost of the scheme. There is also a cost to companies in
preparing an application. In the case of the GeoFund and some of the US schemes, this cost
could itself be subsidized by the fund, but that is uncommon.
It is in the national interest that resource data collected during drilling be filed with some
central agency such as the national Geological Survey or Ministry of Energy. Developers may
wish to keep this information confidential for an embargoed period.
Fundamentally, there are conceptual difficulties with a scheme that rewards failure (a socalled moral hazard, an issue common to all insurance schemes). When such a scheme was
first raised in Indonesia, the larger and more successful developers were opposed to it on the
grounds that it would encourage less able companies to enter the industry.
Table 9.1: Summary of Geothermal Exploration Support Schemes Elsewhere

Country/Scheme
Premium
Required
Nature
Scope
Cap/project
$ million
% Cover
Criteria for
Payment
Germany
Federal
Loan
No
Drilling and
heat plant
3.5 per well

80
and 2.8 for
the heat plant
Completion
Drilling Risk
Cover
No
Drilling
1.8
50% of
originally
planned
costs
Failure
Exploration
Risk Cover
Higher interest
rate while drilling,
10%20%
Drilling
None
80% of
drilling
costs
Failure
State
Loan
No
Drilling
1.3
25
Failure
Commercial
Insurance
Yes
Drilling
Partial
Failure
Switzerland
Risk cover
No
Drilling and
testing
5080
Failure
France
Risk cover
Yes
Whole
project
Up to 90
Failure
Australia
Grant
No
Drilling
50
Completion
UCCDP
Loan
No
Drilling
2090
Failure
GRCP
Loan
No
Drilling and
geoscience
5090
Failure
Iceland
Loan
No
Drilling and
geoscience
60
Failure
GeoFund
Risk cover
No
Drilling
Up to 80
Failure
ARGeo
Risk cover
Yes
Drilling
Up to 80
Failure
4.7
US
3
World Bank
ARGeo = African Rift Geothermal Development Program (World Bank), GeoFund = Geothermal Energy Development Program, GRCP = Geothermal
Reservoir Confirmation Program (US), UCCDP = user-coupled confirmation drilling program, US = United States.
Notes: Only the more relevant schemes are included in the table.
Drilling usually includes well testing and stimulation (if used).
Direct Grants
The simplest form of support is direct grants for drilling to the developers, regardless of well success.
A scheme of this nature was set up in Australia, while a second example is found in the Geothermal
Risk Mitigation Fund (GRMF) funded by KfW in East Africa. Issues with this type of scheme include
the following.

No such scheme has covered the full cost of drilling; the level has varied from 30% to 80%.
That can pose an obstacle. For example, very little of the Australian scheme was used
because developers could not raise the 50% matching funding, in part because of a steep
rise in drilling costs there since the scheme was set up. That was exacerbated by a lack of
geothermal technical drilling standards in some states of Australia, meaning that petroleum
regulations were (inappropriately) applied, adding to the cost.
What is included in the well costs needs to be defined: for example, infrastructure costs such
as roads and water supply may be included. Mobilization and demobilization costs may be
substantial and may be spread over several wells.
As with an insurance scheme, there needs to be a very careful technical and financial
evaluation of the applications, which will usually be on a competitive basis. The GRMF
scheme may well provide a good benchmark for this process: The first round of application
has been evaluated and contract negotiations for the successful projects are underway at the
time of writing.
While payments are not contingent on the success or failure of the wells, there are still criteria
and milestones to be met and evaluated.
Revolving Funds
The basic concept of a revolving fund is that a portfolio of projects is supported by way of a loan,
and that successful projects repay the funds provided for future revenue, whereas those that are
unsuccessful do not.79 It therefore shares the characteristics of an insurance scheme in that what
constitutes success and failure must be defined. Other issues are as follows:

The risk of nonrepayment is contingent not only on technical success but also commercial
success. It is not uncommon for some productive wells to be drilled but the project does not
proceed for various reasons, such as a low tariff or permitting difficulties, and therefore there
may be no revenue.
There may be a very long gap between the initial drilling and revenue resulting. Either the
funding agency has to cover that gap or interest rates need to be set so low as to remain
attractive to the developer.
Implications of Centrally Funded Exploration Drilling

Castlerock suggested that a government-controlled exploration body undertake geothermal
exploration in Indonesia, for Indonesias national benefit and to encourage participation by IPPs.
Castlerock proposed funding through concession fees, whereas others have suggested this could be
provided by the Geothermal Fund. Another option is to assign one of the existing companies (Geo
Dipa or PGE) to undertake this role. In the case of Geo Dipa, since it is already SOE, there is no legal
barrier; but there are barriers in terms of adequate capacity and experience and also potential conflict
79
This makes it a declining balance fund, unless those that are successful cover the costs of those that fail, which is expensive
and creates a moral hazard.

of interest if the company is also allowed to bid for these projects. For PGE, there is a legal barrier since
it is not considered an SOE. But all agree that however such a body is set up, it needs to be technically
competent and will require expert supervision.
To be beneficial, the body would require a good system for data collection, collation, and archiving
something that has been notably absent in Indonesia so far. Donor agencies have in the past made
proposals as to how to improve the situation with a national database.
The question of access to that database is critical: Would it be free, freely available to anyone subject to
a fee, or available only to those who prequalified for concessional bidding? And if not free, how would
it be priced? In New Zealand there is a similar situation with legacy geothermal data, which is now
held by SOE technical agencies that have, in some cases, priced data access so highly that developers
have found it simpler and cheaper to repeat the surveys. Australia has taken the opposite approach,
whereby all natural resource data collected by government agencies is made freely available, as is that
by exploration companies after an embargo period.
Standardization of resource capacity estimates is also important (see Appendix 1). To be of most value,
exploration should include the preparation of such estimates. If, as proposed, exploration includes
drilling, further issues arise because geothermal wells are not only a source of resource data, but are in
themselves both assets and potential environmental liabilities.
Geothermal wells drilled for exploration may, and hopefully should, be usable either for production or
re-injection. They therefore have a tangible value, and if projects where such wells exist are to be taken
up by a developer, it is likely that the developer would wish to purchase those wells. But there are also
good arguments that they should be priced at less than the cost of replacement, because:

They may not be in optimal or even feasible locations for ready interconnection to a coherent
power schemethey may for example be on the wrong side of a large river valley or other
terrain obstacle, or there may be unresolved land access issues for pipelines.
They may have been drilled at a smaller diameter, or with a sub-optimal casing configuration
than a developer would use on its own behalf.
They may have been sitting unused for many years and undergone corrosion or other
degradation, which reduces their working life, especially if not well maintained.
Wells not commercially productive may still have some value for reservoir monitoring. Once again,
though, that value is significantly less than the cost of replacement.
Wells also represent a liability. Both productive and commercially unusable wells potentially have the
risk of leaking or blowing out, causing damage. For example, a well may be hot and have a large flow
rate, but be unusable because the fluid is too acidic or has too high a gas content. In any event, there
needs to be regular monitoring and preventive maintenance. Maintenance includes not only the well
and its site, but also road access to it, which can be a significant cost in remote locations.
One approach is to require that unsuccessful wells be permanently cemented up and abandoned, but
that precludes subsequent use for monitoring. It also precludes the possibility of future conversion
of an unsuccessful well to a successful one through stimulation techniques, or simply because an
alternative generation technology such as binary becomes economic.
Another approach is to require that the developer purchases and assumes the liabilities for all
exploration wells in a field, whether successful or not. This is a simple concept, but cannot prevent the
developer from walking away from it altogether in the absence of some kind of performance guarantee
or surety.
Finally, some consideration needs to be given as to how the exploration body sets priorities for drilling.
Would potential developers be allowed to lobby the exploration body to prioritize areas of most
interest to them?
9.6 Technology Paths to Faster Development

In Appendix 5, we discuss a range of technological options to improve cost-effectiveness and achieve
faster development. These include:

Retrofitting existing projects with binary bottoming plants (BBP): The economics
depend on resource conditions, but in Appendix 3 (and in particular Appendix 3, TableA3.1)
we estimate that an additional 490 MW of BBPs could potentially be installed. Since this would
not be suitable for competitive tendering, this would be an exception to our recommendation
against a technology specific FIT. Such a FIT should be issued by MEMR. This would not
apply to new projects, for which a developer is free to propose any technology mix he or she
deems suitable for the site.
Use of larger units: Most of the developments, and in particular those planned by PGE, are
based around unit sizes of 55 MW. In some cases, there are good reasons why that should
be so. However, in other cases such as Ulubelu 3 and 4, where two 55-MW units are to be
installed on the same resource simultaneously, there appears to be no good reason to do
so instead of installing a single 110 MW unit. Indeed we recommend that in future, tenders
should simply reference some total capacity, e.g., 110 MW instead of 2 x 55 MW.
Use of units other than new ones: The biggest obstacle to the use of such plants in
Indonesia so far is believed to have been that, in some cases, government and IFI procurement
processes, and lenders requirements, preclude the use of anything other than a newly
manufactured plant.
Flexibility in development: There are several instances in the world where plants have
been ordered for a particular project and then switched to a different project by the same
owner when, for some reason, the original project became stalled or delayed. This included
some of the Philippine National Oil Company Energy Development Corporation plants in
the Philippines, and others in New Zealand and Nicaragua. Such opportunities will arise
opportunistically but may be a means of keeping the overall program moving when there are
unexpected difficulties. However, this may also fall foul of IFI procurement and/or financing
rules, which require the use of funds for the purposes intended. If a turbine is ordered, say,
for Ulubelu, but then used somewhere else, it is no longer eligible to be financed by the
World Bank.
10
nlocking the Sector:

U
Study Team Recommendations
10.1General Tariff Approach

With regard to general tariff approach, we recommend the following approach:

Tariffs should continue to be set by tender, but with improvements to the tendering process
and PPAs.
MEMR should issue non-negotiable tariff ceilings based on benefits (which are PLNs avoided
costs adjusted for positive and negative externalities).
We recommend against fixed FITs based on production costs on the grounds of their lack of
economic efficiency, except in the special case of retrofitted binary plants.
On the following grounds, we also recommend against tariff ceilings based on estimates of production
costs:

Information asymmetry, since government can never have the same knowledge of costs and
technical parameters as developersand indeed, for precisely this reason, no countries with
large geothermal resources have tariffs based on government estimates of production costs.
Lack of economic efficiency.
Conceptual difficulty of calculating a ceiling based on assessment of likely project costs.
10.2The Tariff Process

With regard to the tariff process, we recommend the following:
102
Tariff ceilings must be calculated according to a published methodology, so that developers

and lenders can assess factors that are likely to change in the future.
Tariff issuance should not be a one-time measure. Regardless of the basis for tariff ceilings,
the tariff should be reviewed annually, no later than 15 December for the following calendar
year.
Ideally this review would be preceded by a draft issuance and a stakeholder consultation
meeting to permit stakeholder comments prior to final issuance (consistent with international
best practice).
The annual tariff ceiling review should be based on updated information for PLNs avoided
costs.
The tariff ceilings shall apply only to new projects tendered after the date of issuance of the
tariff ceilings.
Unlocking the Sector: Study Team Recommendations 103
10.3Tariff Ceilings
With regard to tariff ceilings, we recommend the following:

Tariff ceilings should be based on the benefits of geothermal energy, as assessed by the
Government of Indonesia.
Making such ceilings non-negotiable ensures that only projects for which costs are less than
benefits are undertaken. This assures economic efficiency and that the subsidy provided by
MoF to PLN is for economic projects.
The economic benefits of geothermal energy are defined by PLNs avoided economic costs
of thermal generation, i.e., with all fuel costs assessed at international prices.
Government will never have as much information on geothermal production costs as

developers, while PLNs costs are transparent and can be assessed easily and reliably by
MEMR. For this reason, an avoided cost tariff (or tariff ceiling) is to be preferred over one
based on estimated production costs.
10.4Calculation of Benefits
The benefits of geothermal energy should be based on a calculation of benefits (avoided costs) for
three general applications:

Projects that connect to the interconnection systems of JavaBali and Sumatra, for which the
relevant benefit is the avoided cost of coal generation (in large projects);
Projects on small islands (or isolated grids on large islands), where the alternative for PLN to
meet base-load requirements is small coal projects (typically less than 50 MW); and
Projects on small islands where small coal projects are not practical, usually for logistical or
environmental grounds, and where the only practical alternative is diesel generation.
The applicable tariff ceiling category will be published by the tender agency for every tender.
The total tariff ceiling will consist of the following elements of avoided cost (benefit):

The avoided variable cost of thermal generation;
The avoided fixed cost of thermal generation (including capital and fixed O&M);
The avoided global externality cost of thermal generation;
The avoided local externality cost of thermal generation;
The avoided costs to MoF of fossil fuel price volatility; and
The incremental local economic development benefits.
MEMR should assess these costs on an annual basis using data from PLN for the avoided variable
cost of thermal generation, the avoided fixed cost of thermal generation, and the incremental local
economic development benefits, and from its Geothermal Directorate for the rest.
10.5De-risking of Projects
We recommend that the de-risking of prospective geothermal work areas prior to tender should be
undertaken with the resources of the Geothermal Fund, with revisions to the funds rules to make them
practicable (such as revising the requirement for 100% cash collateral). This would be particularly
important for small projects in the eastern islands, where the interest of the larger developers is likely
to be small. More specifically, we make the following recommendations:

At least three successful wells should be drilled and tested prior to tender.
The costs of such drilling and evaluation shall be recovered from the successful bidder at time
of financial closure of the project.
The amount payable at financial closure shall include an interest charge and a cost recovery
charge (the rates for both to be set at the time of tender).
The costs of the de-risking as incurred by the government shall be deducted from the
applicable tariff ceiling for the tender according to the published tariff ceiling methodology.
The ownership of the wells, including any unsuccessful wells, and responsibility and/or liability
for same, shall be transferred to the successful tenderer.
10.6Power Purchase Agreements

With respect to PPAs, we recommend the following:

The present system of time-consuming, ad hoc, post tender negotiation of tariff escalation
terms should be dropped.
A single tariff escalation formula should be adopted for all projects (consistent with
international best practice for renewable energy project).
The prospective PPA should be provided at the time of tender. All relevant terms and
conditions, and particularly the schedules relating to the tariff, should be fixed in advance.
10.7The Ministry of Energy and Mineral Resources

Geothermal Production Cost Model
MEMR has developed a useful geothermal production cost model. Even though we recommend that it
not be used to calculate ceiling tariffs for reasons that we have detailed in this report, we recommend that:

The model be documented and published, and be made available to stakeholders as a

yardstick to assess the reliability of more detailed models (such as the Sinclair Knight Merz
Limited model).
The model be used in any renegotiation of PPAs (as recommended in this report).
10.8Power Purchase Agreement Renegotiation Policy

Although we recognize that PPA renegotiation should be primarily a matter for the contracting parties,
i.e., between PLN and the developer, we believe there is benefit to MEMR issuing a policy statement
that sets out the principles that should apply, and the circumstances under which a renegotiation
rather than cancellation should be considered. This would be limited to the following cases:

Delays attributable to the fault of government;
Projects where delineation drilling after tender shows the project to be significantly larger or
significantly smaller than estimated at tender; and
Projects for which capacity of individual units was stipulated, but where the developer
subsequently wishes to install larger units (e.g., build 1 x 110 MW rather than 2 x 55 MW as
originally stipulated). The choice of unit sizes should be left to the developer at final design
without penalty.
We also recommend that renegotiation should not be permitted in the absence of payment of the
performance (bid) bond or actual exploration costs of at least $10 million.
10.9Transmission Connections
With regard to transmission connections, we make the following recommendations:

There must be clarity in the treatment of transmission connection costs at the time of tender.
The published ceiling tariff excludes the avoided cost of transmission for thermal projects.
In general, the transmission connection should be funded and built by the developer, then
handed over to PLN at the time of commissioning for subsequent maintenance.
Recovery of the incremental cost of the transmission connection, if paid by the developer,
should be through a nonescalating tariff adder over 10 years.
The metering point should be at the generating facility switchyard.
Transmission losses in the transmission connection shall be the responsibility of PLN. The
technical specifications for the connection that may affect the loss-rate shall be stipulated at
time of tender. In any event, the line must meet the technical requirements of PLNs grid code.
10.10Improvements to the Tendering Process

The principles that should apply to tendering have been well established in international practice,
as exemplified by the procurement rules of ADB and the World Bank. These should be followed for
Indonesia geothermal tenders. In particular there should be a requirement to post a significant bid
bond stipulated as a percentage of the total project cost (rather than just a percentage of the first year
exploration program).
We recommend that, as a long-term objective, tenders should only be undertaken if at least three
wells have been drilled to measure the resource. In addition:

The PPA shall be provided at time of tender, and its tariff clauses not subject to subsequent
renegotiation, except as expressly provided under the PPA renegotiation policy;
All bids shall be denominated in US dollars; and
A significant bid bond shall be imposed.

Indonesia should establish a technically qualified central tender entity to conduct tenders on
behalf of local governments. The interests of the local governments can be assured by appointing a
representative to the tender committee. Such an entity would require significant technical assistance,
but the IFIs and bilateral donors would certainly be interested in providing this.
10.11Performance Bonds
There is currently a requirement for the winning bidder to post a $10 million bond. However, this
has apparently never been enforced. There is, obviously, no point in such a requirement if it is not
enforced. Developers dislike performance bonds because they tie up (expensive) equity capital, which
is not without its consequences on the buyer. The cost of raising that additional equity will eventually
be recovered by the developer in his tariff. We recommend as follows:

The performance bond requirement should be enforced.
The bond should be drawn down over a 2-year period upon presentation of evidence that the
equivalent funds have been expended on exploration drilling (in the case where Geothermal
Fund has not already provided this), or expended on delineation drilling where the project has
benefitted from the Geothermal Fund de-risking program.
However, were the winning developer to repay the Geothermal Fund for the de-risking cost at
the time of tender award, then there should be no additional performance bond requirement
(since the cost of exploration drilling would likely be significantly higher than $10 million).
10.12The New Tender Entity

It is clear that the present procedure for tendering is often unsatisfactory. This could be due largely to
inadequate technical qualifications of tender committees, which has sometimes led to poor quality
control of short lists and unrealistic bids by entities which have neither the technical nor financial
capacity to deliver. We consequently recommend the establishment of a new central entity that will
conduct tenders on behalf of provincial governments and local entities. During the preparation of
this report, Indonesias House of Representative passed the Bill on Geothermal Energy as a revision
to the previous Geothermal Law No. 27 of 2003. One of the major changes in the new bill is that
the geothermal concession tender and issuance of geothermal license for power development will be
carried out by the central government (MEMR). The new bill assures the interests of local government
through a production bonus when the power plant starts commercial operation, in addition to
applicable local taxes.
10.13Measuring Resources
Some developers complain about the poor quality of data being offered in the geothermal concession
tendering process.80 INAGA and developers should make use of or adapt the existing Australian or
Canadian geothermal reporting codes (which are essentially identical to each other). These have been
extensively peer reviewed and are endorsed by the International Geothermal Association (IGA). They
have built up a track record of use and validation.
Alternatively, the existing Indonesian Geothermal Standard should be improved to the point where
it can serve the same purpose, recognizing that this process would take some time. Use of such a
code would ensure that data presented at time of tender meets generally accepted standards on what
constitutes a measured resource.
80
As noted at stakeholder consultation meetings and in discussions with individual developers.
10.14Binary Bottoming Plants

With respect to binary bottoming plants (BBP), we recommend the following:

MEMR should encourage the installation of BBPs at existing projects, where resource quality
is suitable.
It is not technically or commercially appropriate for such retrofits to be subject to a

competitive tender.
Therefore, MEMR should issue a separate production cost-based FIT, also to be reviewed
annually, and also subject to the tariff ceilings, which would apply only to such retrofitted
projects. This is an exception to our general recommendation against production-cost based
FITs in favor of competitive tendering, but this applies only to the expansion of existing
projects by BBP, for which competitive tendering is not suitable.
Proposals to increase the capacity of an existing project by the addition of BBP shall be
subject to the adequacy of the power evacuation capacity of the transmission line, and on
small island systems of the ability of the system to absorb the extra base-load power, and
hence be subject to approval of PLN. The developer shall have the option to pay for any
reasonable incremental network costs imposed on PLN, but such costs shall not affect the
offered BBP tariff.
10.15Other Required Supporting Actions of Government

Other required supporting actions from the government include:

If the government wishes Pertamina and PGE to play a meaningful role in the countrys
geothermal development, then the MSOE should consider setting up a separate benchmark
for the equity returns for geothermal projects, and treat geothermal projects as a separate
line of business within Pertamina that should not be seen as competing with oil and gas. The
rationale for such a benchmark is that comparable equity returns to oil and gas projects would
simply increase the MoF tariff subsidy to PLN, and that government should recognize the
overall economic returns to country of geothermal investment, rather than the more narrow
interest of SOE equity returns.
MOF is not permitted to provide guarantee to PLN payments (a prohibition that is not unique
to geothermal projects). Payment guarantees under FTP1 and FTP2 projects requires a waiver
through presidential decrees. However, given the history of past PLN defaults, the difficulty
of obtaining payment guarantee of PLN obligations is an important barrier to geothermal
development, and needs to be addressed by a new regulation.
10.16Recommendation for Further Study

This report argues that a key to the successful development of a geothermal development policy in
Indonesia is knowledge of the resource supply curve. The 2010 Castlerock study should be updated,
and include a detailed examination of the likely costs based on the current situation. This should
follow the probabilistic techniques used in the original Castlerock report. However, it should not be
limited to estimates of the levelized cost of energy, but also take into account the likely financial costs.
This work should be initiated as soon as possible.
Appendix 1
Reporting Codes
Many developers complain about the poor quality of data being offered in the geothermal concession
tendering process. Estimates of the magnitude of the available Indonesian geothermal resource vary
greatly. The 2007 WestJapan Engineering Consultants study estimated the exploitable potential
across 50 fields at 9,000 MW.1 As noted previously, in 2011 MEMRs Geological Agency revised
the countrys geothermal potential to 29,215 MW from 27,000 MW a decade earlierindeed, the
27,000MW figure is cited in many World Bank reports,2 and appears to be the basis for claims that
Indonesia possesses 40% of the worlds geothermal resources.3
The basis for these various estimates is unclear, for it is sometimes not fully appreciated that a resource
is only that portion of a natural occurrence (whether of energy, petroleum, or minerals) that can
feasibly and economically be extracted. Without the basis for such an estimate being made explicit,
including assumptions as to the technology pathway and power prices, it is essentially meaningless.
There is a widespread perception that the estimate is too large, but no better estimate has been made,
so no one knows by how much.
At the level of an individual geothermal resource, one needs not only to have a good grasp of the size
of the resource for planning the development, but it is equally important to have the reliability of that
estimate quantified, so that risks can be assessed and financing issues identified.
An Indonesia-specific geothermal estimation methodology was produced as an Indonesian standard
(SNI 13-6169-1999) in 1999 and has been subject to two subsequent revisions in 2000 and 2004.
It is not clear whether that standard was rigorously applied for producing the 27,000 MW estimate,
and it is certain that it has not been used for the inflated estimates on which some of the wilayah
kerja pertambangan (geothermal work areas) (WKP) tendering has taken place, as pointed out by
Castlerock.4
The Indonesia Geothermal Association (INAGA), which is affiliated to the International Geothermal
Association (IGA), has reported that they are considering developing a new Indonesia-specific
geothermal estimation standard, related to the P1, P2, P3 categories used to report the quality of a
petroleum resource. The intention of that is commendable, but it may not be particularly helpful in
that it will undoubtedly take significant time to develop and validate, and, at least initially, will not
have any international recognition, limiting value for resource certification as a basis for financing.
Furthermore, to be of much value, it would have to represent a substantial extension of the existing
standard, which lacks default parameters for estimation and is weak on standards for reporting (as
opposed to estimation methodology).
Rather, it is recommended that INAGA and developers be encouraged to make use of, or adapt, the
existing Australian or Canadian geothermal reporting codes (which are essentially identical to each
other). These have been extensively peer-reviewed and are endorsed by the International Geothermal
108
West Japan Engineering Consultants. 2007. Master Plan Study for Geothermal Development in the Republic of Indonesia.
World Bank. 2011. Project Appraisal Report.
J. Wilcox. 2012. Indonesias Energy Transit: Struggle to Realize Renewable Potential. Renewable Energy World.Com.
Castlerock. 2010. Phase 1 Report.
Reporting Codes109
Association (IGA). They have built up a track record of use and validation. That they are applicable
to the Indonesian situation is demonstrated by the fact that they have been used in three feasibility
studies sponsored by the World Bank,5 and at least one other significant feasibility study leading to an
important investment decision by an independent power producer (IPP).6 Stakeholder Comment A1.1
summarizes our replies to views expressed, and Box A1.1 enumerates the specific issues in the current
Indonesian standard.
Before rushing to judgment about the causes of the slow progress in reaching the announced targets,
one may well ask whether the problem lies in the targets themselves. If what is realistically available
(at a reasonable price and where there are no grid or market constraints) is only 2,000 MW, then the
existing achievement of 1,335 MW would be quite satisfactory. Indeed, with 1,335 MW of capacity
installed, Indonesia ranks third in the world, behind only the United States (3,400 MW) and the
Philippines (1,900 MW).7
At the end of 2012, the installed capacity of geothermal electricity generation in Indonesia was 1,335
MW. The national energy policy8 stipulates that by 2025, 5% of total primary energy supply shall be
from geothermal, translated by the governments Road Map for Development Planning of Geothermal
Energy as 9,500 MW. Ministerial regulation of MEMR 15/2010 lists the specific power plants to be
developed by the governments Fast Track program (FTP), which total 3,967 MW by 2014. Taking into
account other projects not listed in Ministerial regulation of MEMR 15/2010, but known to be under
development, the target increases to 5,710 MW.9 The current PT Perusahaan Listrik Negara (State
Electricity Company) Investment Plan (Rencana Usaha Penyediaan Tenaga Listrik 2012 [Electricity
Power Supply Business Plan] 2012) now anticipates total additions of 3,611 MW between 2013 and
2020, to bring the total by 2020 to 4,816 MW.10
Lumut Balai, Ulubelu, and Tompaso.
Star Energy at Wayang Windu.
Followed by Mexico (1.0 GW), Italy (0.9 GW), New Zealand (0.8 GW), Iceland (0.7 GW), and Japan (0.5 GW). REN21.
Renewables 2013. Global Status Report. Frankfurt.
Presidential Decree No. 5/2006.
Castlerock assessment of geothermal targets, December 2010.
10
PLN. 2012. Rencana Usaha Penyediaan Tenaga Listrik 20122021 (Electricity Power Supply Business Plan). Jakarta. Table 5.11.
110APPENDIX 1
Stakeholder Comment A1.1: Use of International Code
The following comments were made by the Indonesia Geothermal Association (INAGA) during
discussions.
(1)Indonesia already has its own code [implying that use of an international code is not
necessary].
Reply: Indonesia has a national standard produced in 1990 plus several updates, which is specific
as to terminology and methodology but which is not a code in the sense of defined minimum
requirements for reporting in terms of transparency and materiality. (There are requirements in
the 2005 revision and templates for reports about the resource characteristic, but not about the
assumptions and methodology used for the resource capacity estimate). It is currently under
revision and the revised version has not yet been made public, so there is uncertainty about
what it actual means and when that uncertainty will be resolved, nor has it been internationally
peer reviewed, nor ratified. We also note that the Indonesian standard does not appear to have
been used (or if it has been used, not in a transparent way) in resource capacity estimates on
which current tendering and forward planning is based. That issue is extensively covered in
the appendixes to the Castlerock report, where rigorous resource estimates using very similar
methodology to that in the Indonesia standard end up with much lower figures than the official
estimates.
(2)The Australian and/or Canadian code is based on mining practice.
Reply: The code has used some terminology from similar mineral codes but that is all. The
methodology for resource estimation is geothermal-specific and draws heavily on the Society of
Petroleum Engineers (SPE) practices. To quote the Australian code:
The Society of Petroleum Engineers (SPE) and the World Petroleum Congress
(WPC) have jointly proposed definitions of standard terms for booking petroleum
reserves. Their guidelines for the Evaluation of Petroleum Reserves and Resources
(SPE and WPC, 2001) are drawn upon significantly here for methodology, as is the
more recent Petroleum Resources Management System (SPE, WPC, American
Association of Petroleum Geologists [AAPG] 2007).a
(3)The Australian code has not been adopted by the International Geothermal
Association.
Reply: That is not quite correct. The Australian code has been reviewed by the International
Geothermal Association (IGA) and endorsed, if not yet accepted as an IGA standard. The IGA
Reserves and Resources Committee has also agreed that the Australian code will be used as the
basis for an international set of definitions on renewable energy to be developed in the very near
future for the United Nations Framework Classification (2009) renewables framework under
development by the United Nations Economic Commission for Europe. The work of alignment is
underway and is intended to be complete within the first half of 2014.
(4)It would therefore be better to adapt the present Indonesian code if in fact needed.
Reply: In principle we have no objection to the use of the Indonesian standard, and the World
Bank Group would be happy to collaborate with INAGA and/or MEMR to revise the standard to
make it suitable for this purpose.
As quoted in J. Lawless. 2010. Geothermal Lexicon for Resources and Reserves Definition and Reporting. Adelaide: Australian
Geothermal Energy Group. p. 3.
Reporting Codes111
Box A1.1: Specific Issues in Current Version of Indonesia Standard

Some specific issues with the current version of the Indonesian standard are as follows:

There is only a Bahasa Indonesia version. To be credible to international investors, there also needs to
be an official English version. That is particularly the case since some of the terminology does not align
well with international practice. For example, in the Indonesian standard there is a resource category
called: Cadangan Terduga which would usually be translated as: Contingent Reserves. But the
criteria for that category do not include drilling and well testing. The term reserve is usually only used
for the highest category of geothermal, petroleum, or mineral resources. For example, the Australian
code requires that well deliverability has been demonstrated to be able to declare a reserve. The
equivalent Joint Ore Reserves Committee mineral code requires in effect a bankable feasibility study be
conducted, and Society of Petroleum Engineers for petroleum restricts the use of reserve to resources
which can be brought into economic production within 5 years or less. So the Indonesian terminology
seems well out of step with world practice.
The Indonesian standard is essentially one-dimensional. It enumerates a progression of increasing

categories of geological and other knowledge about the resource, starting from regional scale surface
studies and culminating in drilling and well testing. In contrast, the Australian code and most other
resource reporting systems in geothermal, mineral, and petroleum are two- or three-dimensional. In the
Australian code, the first axis is geoscientific certainty, as it is in the Indonesian standard, whereas the
second axis relates to project feasibility through consideration of issues such as economic, marketing,
environmental, social, legal, and regulatory factors.
While the 2004 revision of the Indonesian standard does include a reasonably comprehensive list of
data to be collected, factors to be considered, and report outlines for an exploration program,a it does
not require explicit disclosure of the assumptions that go into a resource or reserve capacity estimates.
These assumptions include the technology pathway, the basis for selecting values for resource
parameters such as area and temperature, and the modifying factors that can render an otherwise
feasible project uneconomic such as concession tenure and environmental issues. In contrast, the
Australian code explicitly requires the disclosure of these assumptions through the concepts of
transparency and materiality. A fundamental principle of the Australian code is that a report on a
resource capacity estimate should contain sufficient information for an independent authority to
check the validity of the assumptions made and to replicate the estimate with a reasonable degree of
accuracy. That would not be possible for reports prepared under the Indonesian standard.
There is no requirement under the Indonesian code for the person producing an estimate to be
competent or qualifiednor accountable. In contrast, the Australian code is explicit in these matters,
including minimum years of experience in the type of geothermal system being assessed. The parent
association maintains a register of competent persons. Inclusion on the register requires agreement to
conform to a defined code of ethics. It would be possible (and desirable) for the Indonesia Geothermal
Association (INAGA) to take a similar role in Indonesia.
The Indonesian standard permits the use of the volumetric method (stored heat) and numerical
simulation to estimate reserves and resources, which is appropriate. For stored heat, while the 2000
revision of the standard lists preferred or default single values (within broad reservoir temperature
bands) for various parameters such as recovery and conversion factors at various resource and reserve
categories, and does allow the use of other values, it provides no guidance on how those values should
be selected. In contrast, for example, the Australian code lexicon provides default values for:
Electricity conversion factors as a function of reservoir and ambient temperature, along with
guidance on the relationship of conversion factors to technology pathway (in contrast, for
example, the default factor of 10% in the Indonesia standard for a 125C resource and 90C cut off
temperature is quite unrealistic).
112APPENDIX 1
Box A1.1. continued
Recovery factors as a function of rock type, porosity or fracture density based on international
peer reviewed research. Unlike the Australian code, the Indonesian standard does not distinguish
between the minimum reservoir cut-off grade for stored heat (i.e., the minimum temperature below
which a portion of the reservoir should not be included in the resource), and the base temperature
for energy extraction. As explained in the Australian code lexicon, those temperatures can differ by
100C or more, and perhaps more significantly, the difference between them can vary greatly from
project to project depending on the technology pathway chosen, so a single default value even for
the difference is not appropriate.
The Indonesian standard, while it gives a description of the numerical simulation process, gives
no guidance on how a numerical reservoir simulation model should be used to determine resource
capacity thereby through eventual failure criteria for the project (e.g., pressure or temperature).
See Clotworthy et al. (2010)b for a discussion of this issue.
The Indonesian standard does not address issues related to probabilistic estimates, which are
commonly used for stored heat estimates (and less commonly for numerical resource simulation).
Without consideration of how to apply a probabilistic approach, the Indonesian standard cannot
provide the basis for the World Banks recommendation for basing tenders on P90 resource estimates.
Most investors would expect the use of probabilistic estimates or at the very least declaration of the
uncertainty of the estimates. To quote the Australian code:
In most situations, rounding to the second significant figure should be sufficient. For
example: 31 thermal megawatts and 6.5 electrical megawatts. There will be occasions,
however, where rounding to the first significant figure may be necessary in order to convey
properly the uncertainties in estimation. This would usually be the case with Inferred
Geothermal Resources. To emphasize the imprecise nature of a Geothermal Resource
estimate, the final result should always be referred to as an estimate not a calculation.
Competent Persons are encouraged, where appropriate, to discuss the relative accuracy
and/or confidence of the Geothermal Resource estimates. Where a statement of the
relative accuracy and/or confidence is not possible, a qualitative discussion of the
uncertainties should be provided. Use of probabilistic estimates is encouraged.c
Notes:
Although there is a far more detailed list of criteria in the Australian code, Appendix D.
A. W. Clotworthy, J. V. Lawless, and G. Usher. 2010. What is the End Point for Geothermal Developments: Modelling Depletion of
Geothermal Fields. Proceedings of the World Geothermal Congress. Bali. 2529 April.
Australian Code for Reporting of Exploration Results, Geothermal Resources and Geothermal Reserves. 2010. The Geothermal
Reporting Code (Second Edition). Australia.
Appendix 2
Transmission
Interconnection Costs
What are the typical costs of transmission interconnection for geothermal projects? Table A2.1
shows transmission connection costs as estimated by PT Perusahaan Listrik Negara (State Electricity
Company) (PLN) in the 20122021 in its investment plan. We note considerable variation in distance
(from 1 km to 80 km) and in cost (from less than $1 million to $8million).
Table A2.1: Transmission Interconnections, Geothermal Projects
Conductor
From
Province
km
COD
Status
Cost
Cost
$1,000/
$/kW
km
$ million
Year
Wayang Windu
Jabar
220 150 kV
2 cct, 2xZebra
32
6.11
2014
Proposed
28
Lawu
East Java
165 150 kV
2, 2xHawk
32
2.44
2019
Planned
15
76
Wilis/Ngebel
East Java
165 150 kV
2 cct, 2xZebra
60
5.91
2018
Proposed
36
99
Arjuno Welirang East Java
110 150 kV
2 cct, 2xHawk
74
5.65
2019
Planned
51
76
Ijen
110 150 kV
2 cct, 2xZebra
60
5.91
2019
Proposed
54
99
East Java
MW
Cost
191
Baturaden
Jateng
110 150 kV
2 cct, 2xZebra
20
1.97
2018
Planned
18
99
Tangkuban
Perahu I
Jabar
110 150 kV
2 cct, 2xZebra
0.49
2018
Proposed
98
Rawadano
Banten
110 150 kV
2 cct,
2xTACSR410
30
4.5
2018
Proposed
41
150
Gunung Ciremai Jabar
110 150 kV
2 cct, 2xZebra
40
3.94
2019
Planned
36
99
Bedugul
Bali
100 150 kV
2 cct, 1xHawk
0.22
2017
Planned
55
Tangkuban
Perahu II
Jabar
60 150 kV
2 cct, 2xZebra
10
0.99
2017
Proposed
17
99
Umbul
Telomoyo
Jateng
55 150 kV
2 cct, 2xZebra
16
1.58
2019
Planned
29
99
197
Guci
Jateng
55 150 kV
4 cct, 2xZebra
20
3.94
2018
Planned
72
Iyang Argopuro
East Java
55 150 kV
2 cct, 2xZebra
60
5.91
2017
Proposed
107
99
Ungaran
Jateng
55 150 kV
2 cct, 2xZebra
30
2.96
2018
Planned
54
99
Gunung Endut
Jabar
55 150 kV
2 cct, 2xZebra
80
7.88
2019
Planned
143
99
Patuha
Jabar
55 150 kV
2 cct, 2xZebra
0.06
2014
Ongoing
60
Cisolok
Sukarame
Jabar
50 150 kV
2 cct, 2xZebra
60
5.91
2017
Proposed
118
99
Tampomas
Jabar
45 150 kV
2 cct, 2xZebra
35
3.45
2018
Planned
77
99
Karaha Bodas
Jabar
30 150 kV
2 cct, 2xZebra
20
1.97
2016
Committed
66
99
Kamojang
Jabar
30 150 kV
2 cct, HTLSC
(1xHawk)
0.15
2016
Proposed
150
Cibuni
Jabar
10 70 kV
2 cct, 1xHawk
50
2.77
2016
Proposed
277
55
$/kW = dollars per kilowatt, $1,000/km = thousand dollars per kilometer, cct = circuit, COD = commercial operating date, km = kilometer,
kV = kilovolt, MW = megawatt.
Note: The terms Hawk, Zebra, TACSR410, and HTLSC refer to conductor types.
Source: PT Perusahaan Listrik Negara (State Electricity Company of Indonesia). 2013. Rencana Usaha Penuyediaan Tenaga Listrik (RUPTL)
(Electricity Power Supply Business Plan 20122021). Jakarta.
113
114APPENDIX 2
With one or two exceptions, the cost per MW is very small, with an average of $46/kW (if one excludes
the outlier Cibuni, $277/kW). Compared to total project costs of $3,500/kW or more, the transmission
interconnection cost should not have a significant impact on the bid tariff, were responsibility for the
interconnection passed to the developer.
As shown in Figure A2.1, there is some evidence of scale economieslarger projects have somewhat
lower transmission costs (as $/kW). We note that most lines are double circuit Zebra, irrespective of
the intended MW capacity or distance, suggesting that the transmission configuration has not been
optimized for capacity, but probably more for commonality.
Figure A2.1: Transmission Cost in $/kW versus Installed Capacity

300
Transmission cost, $/kW
Line Length, km
200
80
60
60
100
35
20
20
30
16
10
60
74
30
40
60
20
32
0
0
50
100
150
32
200
250
$/kWh = dollar per kilowatt-hour, km = kilometer, MW = megawatt.
Transmission Interconnection Costs 115
Transmission Connections for Thermal Generation Projects

The average cost for interconnection of large coal projects is $27/kW, so significantly lower than for
geothermal projects (Table A2.2). This is probably due to the cost being spread over a greater line
capacity, with greater configuration optimization and generally easier terrain than for many of the
geothermal projects.
Table A2.2: Transmission Interconnections (500 kV), Large Coal Projects
Length
Cost
km
$ million
COD
Status
Cost
Cost
$/kW
$/km
4.2
415
MW
Conductor
Jawa-5
2,000
2 cct, 4xZebra
20
8.30
2018
Planned
Banten
625
4 cct, 4xDove
40
13.06
2016
Committed
20.9
327
Jawa-6
2,000
2 cct, 4xZebra
80
56.44
2021
Planned
28.2
706
Jawa-1
1,000
2 cct, 4xZebra
116
48.14
2017
Planned
48.1
415
Indramayu
1,000
4 cct, 4xDove
200
65.28
2017
Proposed
65.3
326
Jawa-3
1,320
2 cct, 4xZebra
20
8.30
2017
Planned
6.3
415
Adipala
660
2 cct, 4xZebra
28
11.62
2014
Ongoing
17.6
415
$/km = dollars per kilometer, $/kW = dollars per kilowatt, cct = circuit, COD = commercial operating date.
The impact on the tariff can be calculated in the usual way, that is: Given an assumed number of kWh
over which the transmission tariff can be recovered, what transmission tariff is necessary to meet the
opportunity cost of capital? Assuming a 2-year construction time immediately prior to the commercial
operation date of the generating station, and an average capacity factor for coal projects of 80%, the
incremental impact of the tariff calculates to a negligible 0.05 US/kWh. This is about 1.8% of the
avoided fixed costs of thermal generation at large projects (Section 4, Table 4.3).
Appendix 3
Project Review
In late 2010, Castlerock estimated a realistic target would be additions of 600 MW by the end of
2014, and 1,400 MW by the end of 2016. Four years have elapsed since then, so it is timely to review
those figures, as shown in Table A3.1. This review starts from the Government of Indonesias stated
geothermal development plan as of July 2013. Then new estimates of the probable capacity and
completion date for the stages shown are given. In terms of capacity, considerable weight has been
given to the Castlerock1 estimates, since those were based on a systematic re-examination of the
data on a consistent basis. In some cases the Castlerock estimates have been further modified here
on the basis of additional data that was not available to it at the time. This has led both to increases
and decreases. Where the estimated resource capacity is larger than the planned development stage
(e.g., for developments on remote islands with a small market), the planned development does not
represent the full resource capacity. It is assumed that all MW figures are net at the transmission
interface.
In terms of timing, where development is well underwayas for Ulubelu units 3 and 4a specific
estimate of commercial operations date (COD) can be given. In other cases the estimate is based
on the stage of development of the project and the regulatory process. Castlerock gave a detailed
breakdown for the timing of new projects, assuming that the overall time for a project would be
69years depending on the regulatory process followed. For the present purpose, if there is no other
more specific information available a slightly simpler approach has been taken, since the details of the
current status of some of the projects are not available, with the following assumptions as to time to
COD from the defined present status:

Significant number of wells drilled, power purchase agreement (PPA) finalized, financial
closure or close to it: 2 years. This is based on the assumption that a power plant cannot be
ordered until financial closure is achieved and from that point the time for construction and
commissioning will be about 2 years.
Some wells drilled, commercial arrangements advanced, source of funding for remaining
drilling identified, PPA finalized or under negotiation: 3 years.
A few wells drilled, concession tendered, surface exploration complete: 4 years.
No wells drilled and/or concession not tendered: 7 years.
Obviously if exploration so far has not been promising, longer times will apply.
The next two columns of the table list possible additional generation from bottoming binary plants
(Appendix 5). The following assumptions are made:
116
No bottoming binary is assumed at the dry steam fields (that are less suitable for bottoming
binary plants).
Only a small amount of bottoming binary is possible at the very high enthalpy fields such as
Wayang Windu and Dieng.
Castlerock Consulting. 2010. Phase 1 Report: Review and Analysis of Prevailing Geothermal Policies, Regulations and Costs.
Jakarta: Ministry of Energy and Mineral Resources.
Project Review117
Otherwise, for liquid dominated fields, a bottoming binary capacity of 15% of the condensing
steam generation is possible.
Where the binary capacity would be less than 5 MW the possible capacity is calculated but it
is not included in the total. Binary plants smaller than 5 MW are available, but it is considered
less likely that adding very small binary plants would be considered worthwhile.
Where the project is on an isolated island and the plant capacity is thought to be driven by the
market size rather than the resource size, no binary is added.
For existing plants or those well advanced in the engineering, procurement, and construction
process, it is assumed that an extra 12 months would be needed to add a binary plant. But
for plants which are only at the conceptual planning stage, it is assumed that binary could be
installed simultaneously with the main plant.
The binary additions listed are genuinely additional to the main plant capacity, and can be assumed to
operate through the lifetime of the project. Therefore they offer a real opportunity to accelerate the
program.
It is also possible to add early generation by back pressure plants (see below), which may be a good
opportunity to accelerate the program and add confidence in the resources, but the assumption is
that those would eventually be replaced by condensing plants, so they do not represent permanent
additions to the capacity and so are not included in the table. In terms of quantities, one might expect
to install such plants when one third to one half of the drilling is complete, and they produce half as
much electricity per unit quantity of steam, so the generation available might be in the range of 10 to
20% of the main plant.
The final column of the table gives a brief comment on the current status of the project so far as can
be determined, including the following definitions:

Surface exploration includes all geoscience prior to drilling. Note that this can cover a wide
range of activities from early reconnaissance through to detailed geophysics. Exploration that
requires the latter is designated as advanced exploration. Logically, surface exploration
should be complete before drilling commences, but this is not always the case.
Shallow drilling means shallow and/or small diameter drilling, sometimes called slim-hole
drilling, but excluding very shallow temperature gradient boreholes (which are occasionally
but not very often used in Indonesia). A slim-hole can add valuable resource information,
which may raise or lower the resource potential. But if a shallow and/or slim hole is not
productive, this does not necessarily mean the resource is not viable.
Exploration drilling means a program of one to four full diameter wells has been drilled,
which may or may not have been commercially successful.
Part drilled means that a significant proportion of the wells required for the full planned
development have been successfully completed.
Lahendong 4
Ulubelu
19
20
20
Lahendong 3
Ulumbu
17
18
117
Wayang Windu 2
16
60
110
20
13.3
Kamojang 4
20
Sibayak
Lahendong 2
13
110
14
Darajat 3
12
2.5
20
197
120
120
55
55
15
Mataloko
Darajat 2
11
110
Salak 4
Wayang Windu 1
Salak 3
Lahendong 1
Salak 1 & 2
Darajat 1
10
95
Kamojang 3
55
Kamojang 2
30
Kamojang 1
MW
Project
Probable
Capacity
by Date
2012
2012
2012
2009
2009
1998/
2008
2008
2007
2009
2004
2001
2000
2000
1997
1997
1994
1994
1987
1987
1983
Possible
COD
Date
1,335
1,225
1,205
1,200
1,180
1,063
1.049
989
969
859
857
837
727
632
435
315
195
140
85
30
MW
Cumulative
16.5
30
18
18
MW
Additional
Binary
Capacity
2014
2014
2014
2014
2014
2014
2014
2014
2014
Binary
Date
99
83
80
80
77
77
77
77
74
74
74
71
66
66
36
18
MW
Cumulative
Binary
Table A3.1: Project Status
1,434
1,307
1,284
1,279
1,256
1,139
1,126
1,066
1,043
933
930
907
792
697
471
333
195
140
85
30
MW
Total
Running at ~80 mw because of resource

constraints but those should be solvable
Drilling complete
Significant resource issues probably

preclude expansion
Status
118APPENDIX 3
Muara Laboh
Sarulla 1 (Namora-ILangit)
Sungai Penuh
Karaha
Wayang Windu 3
Lahendong 5 & 6
Jailolo
31
32
33
34
35
36
37
Sarulla 2 (Silangkitang)
27
Ulubelu 3 & 4
Lumut Balai 1 & 2
26
30
Ulumbu 4
25
Kamojang 5
Ulumbu
24
Ulumbu 5
Dieng
23
28
Ulumbu
22
29
Mataloko
21
Project
10
40
110
30
110
220
110
100
2.5
30
80
110
2.5
40
MW
Probable
Capacity
by Date
2017
2017
2017
2017
2017
2017
2017
2016/2017
2016
2014
2016
2016
2015
2014
2014
2013
2013
Possible
COD
Date
2,345
2,335
2,295
2,185
2,155
2,045
1,825
1,715
1,615
1,612
1,582
1,502
1,392
1,390
1,385
1,345
1,340
MW
Cumulative
4.5
16.5
33
16.5
15
16.5
MW
Additional
Binary
Capacity
2018
2017
2017
2017
2018
2017
2016
Binary
Date
213
213
207
207
203
186
153
137
122
122
122
116
99
99
99
99
99
MW
Cumulative
Binary
2,558
2,548
2,502
2,392
2,357
2,231
1,978
1,851
1,736
1,734
1,704
1,618
1,491
1,489
1,484
1,444
1,439
MW
Total
Not yet drilled, surface exploration

advanced, potential market may be larger
if mining goes ahead
Part drilled, low success, pertamina

threatening to stop
Part drilled, moderate success
Part drilled, moderate success, total

capacity probably larger than shown
Drilling just started; surface exploration

advanced; part in national park
Exploration drilled, some success,

but difficult terrain, high gas, acidity
Exploration drilled, some success
Part drilled, moderate success,

may be pushing limit of field capacity
so down-rated
Field drilled, but not necessarily this part
Drilling complete
Part drilled, successful but small field
Part drilled but low success
Part drilled
Not currently operating; may be repaired

to 80% capacity
Drilling complete (not confirmed)
Drilling complete (not confirmed)
Status
Project Review119
Lumut Balai 3 & 4
Margabayur
Hululais
Rajabasa
Cibuni
Wayang Windu 4
Patuha 1
Atadei
Huu Daha 1
Sokoria
Seulawah Agam
Sorik Merapi
Rantau Dedap
Suoh Sekincau
Jaboi
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
Project
220
220
110
55
15
20
110
110
10
110
110
55
110
MW
Probable
Capacity
by Date
2019
2019
2019
2019
2019
2018
2018
2018
2014
2018
2018
2018
2018
2018
2018
Possible
COD
Date
3,612
3,605
3,385
3,165
3,055
3,000
2,985
2,965
2,960
2,850
2,740
2,730
2,620
2,510
2,455
MW
Cumulative
33
33
16.5
8.25
16.5
16.5
8.25
16.5
MW
Additional
Binary
Capacity
2019
2019
2019
2019
2018
2018
2018
2018
Binary
Date
378
378
345
312
296
287
287
287
287
271
271
271
254
238
230
MW
Cumulative
Binary
3,990
3,983
3,730
3,477
3,350
3,287
3,272
3,252
3,247
3,120
3,010
3,000
2,874
2,747
2,684
MW
Total
Not yet drilled
Not yet drilled. Suoh and sekincau are

2 separate fields, may be technically difficult
Part exploration drilled, should be

a good field

advanced otp proceeding slowly,
may be serious technical issues and
some volcanic risk
Shallow drilled, may be smaller demand

than planned
Shallow drilled, surface exploration

advanced

advanced, market may be larger if mines
can be supplied
Shallow drilled, resource may be larger

than market
~50% drilled
Part drilled, moderate success
One exploration well. Next to patuha boundaries not clear
Not yet drilled, land access issues
Exploration drilled, drilling problems
Surface exploration advanced but not

drilled within lumut balai wkp
This part not drilled yet, surface exploration

advanced
Status
120APPENDIX 3
Rawa Dano
Kamojang 6
Mataloko 2
Way Ratai
Gunung Talang
Simbolon Samosir
Patuha 2 & 3
Tampomas
Ungaran 1
Ngebel / Wilis
Dieng
Ciremai
Gunung Endut
Guci
Ungaran 2
Umbul Telomoyo
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
Project
20
20
40
40
55
40
55
110
20
20
55
60
110
MW
Probable
Capacity
by Date
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
2020
2019
2019
2019
Possible
COD
Date
4,262
4,242
4,222
4,222
4,182
4,142
4,087
4,047
3,992
3,992
3,882
3,862
3,842
3,787
3,782
3,722
MW
Cumulative
8.25
8.25
8.25
16.5
MW
Additional
Binary
Capacity
2020
2020
2020
2020
2020
2020
2019
Binary
Date
454
454
454
454
448
442
434
428
419
419
403
403
403
395
395
395
MW
Cumulative
Binary
4,715
4,695
4,675
4,675
4,629
4,583
4,520
4,474
4,411
4,411
4,284
4,264
4,244
4,181
4,176
4,116
MW
Total
Not yet drilled, only basic exploration done
Not yet drilled
Not yet drilled
Not yet drilled
Not yet drilled, only basic exploration done
Not yet drilled, assuming this applies to

the other fields within the Dieng wkp

advanced but not encouraging
Not yet drilled
Not yet drilled, preliminary exploration

not encouraging
Part drilled, moderate success; this capacity

only achievable if it includes the cibuni
sector

not encouraging
Not yet drilled, volcanic risk, preliminary

exploration not promising; figure based on
assuming this heading refers to bukit kili
Not yet drilled
Shallow drilled
One well drilled, good success in field

overall
Shallow drilled, not successful but looks

promising
Status
Project Review121
Bonjol
Danau Ranau
Kepahiyang
Bedugul
Gunung Lawu
Mataloko 4
Oka Larantuka
Lainea
Sokoria 4
Tangkuban Perahu 2
78
79
80
81
82
83
84
85
86
87
Jailolo 2
Ulumbu 3
74
77
Mataloko 3
73
Kotamobagu 1
Bora
72
Kotamobagu 2
Songa Wayaua
71
75
Tulehu
70
76
Iyang Argopuro 1
69
Project
20
20
50
40
40
40
40
40
20
20
MW
Probable
Capacity
by Date
2022
2021
2021
2021
2021
2021
2021
2021
2021
2021
2020
2020
2020
2020
2020
2020
2020
2020
2020
Possible
COD
Date
4,630
4,630
4,625
4,605
4,602
4,597
4,577
4,527
4,487
4,447
4,407
4,402
4,362
4,322
4,317
4,312
4,307
4,302
4,282
MW
Cumulative
MW
Additional
Binary
Capacity
2021
2021
2021
2021
Binary
Date
489
489
489
489
489
489
489
484
478
472
466
466
460
454
454
454
454
454
454
MW
Cumulative
Binary
5,118
5,118
5,113
5,093
5,090
5,085
5,065
5,010
4,964
4,918
4,872
4,867
4,821
4,775
4,770
4,765
4,760
4,755
4,735
MW
Total
Not yet drilled, may not be technically

feasible
Not yet drilled
Not yet drilled, only very basic exploration
Shallow drilled
Not yet drilled, may be technical issues
Part drilled, moderate success, but stalled

by environmental issues, unlikely to be
resolved soon
Not yet drilled, only basic exploration

advanced
Drilled, some success
Drilled, some success
Shallow drilled
Not yet drilled
Not yet drilled, surface exploration not

promising; may only be ok for binary
Not yet drilled, may not be very large

done
Status
122APPENDIX 3
Cisolok Cisukarame
Tangkuban perahu 1
Cisolok Cisukarame
2&3
Batu Raden
Ijen
Arjuno Welirang
Iyang Argopuro 2
Marana/Masaingi
Huu Daha 2
Sembalun
89
90
91
92
93
94
95
96
97
98
MW
Possible
COD
Date
4,630
4,630
4,630
4,630
4,630
4,630
4,630
4,630
4,630
4,630
4,630
MW
Cumulative
MW
Additional
Binary
Capacity
Binary
Date
456
489
489
489
489
489
489
489
489
489
489
MW
Cumulative
Binary
5,118
5,118
5,118
5,118
5,118
5,118
5,118
5,118
5,118
5,118
5,118
MW
Total
Not yet drilled

advanced, but no evidence of large resource

not promising
Not yet drilled
Not yet drilled, may be serious technical

issues
Shallow drilled, may be serious issues

with acidity

advanced but not very encouraging
Not yet drilled

not encouraging
Drilled, unsuccessful

not encouraging
Status
Sources: Estimates of capacity are based on Castlerock Consulting. 2010. Phase 1 Report: Review and Analysis of Prevailing Geothermal Policies, Regulations and Costs. Jakarta: Ministry of Energy and Mineral Resources, as
modified by more recent information, and as assessed by the technical experts on the study team. The estimates of COD are based on PT Perusahaan Listrik Negara (State Electricity Company of Indonesia). 2013. Rencana
Usaha Penuyediaan Tenaga Listrik (RUPTL) (Electricity Power Supply Business Plan 2012-2021). Jakarta; Castlerock. 2010. Phase 1 Report.
COD = commercial operation date, MW = megawatt, WKP = wilayah kerja pertambangan (geothermal work area).
Sipoholon Ria-Ria
88
Project
Probable
Capacity
by Date
Project Review123
124APPENDIX 3
It can be concluded that the targets shown in Table A3.2 now appear achievable provided the
Pertamina Geothermal Energy (PGE) commercial partnerships go ahead.
Table A3.2: Revised Targets
Revised
Assessment (MW)
Including Binary
Additions (MW)
End 2013
1,345
1,345
2014
1,390
1,489
2016
1,615
1,736
2020
4,400
4,739
MW = megawatt.
Appendix 4
Tariff Structure and

Project Finance
Background
At the stakeholder consultation meeting of 19 December 2013, a question was raised about the
adequacy of tariffs under project finance. It was noted that the usual criterion of project viabilitythe
financial internal rate of return (FIRR) to investors equitywas not a reliable indicator of bankability,
because the primary measure of cash-flow adequacy in project finance used by lenders is not FIRR, but
debt service coverage ratio (DSCR). It is entirely possible that a project meeting a given equity return
target is not bankable because of poor DSCRs in the early years of debt service. Simple calculations of
FIRR may also be misleading because of other conditions imposed by lenders (such as the funding of
debt service and major maintenance escrow accounts before dividends can be paid).
This is related to the problem of calculations of the levelized cost of energy being used to indicate
the level of tariff required. Because levelization calculations are typically made over the economic
lifetime of a project (typically 2530 years), the levelized cost of energy is a reliable indicator only
if the cost of debt were also spread over the same time horizon. In a project-financed project with
a preponderance of locally financed debt, tenors of more than 10 years are seldom achievable. Only
with highly concessional finance with generous grace periods and long tenors is the levelized cost of
energy a useful indicator. In most cases, debt service obligations are front-loaded.
Figure A4.1: Typical Indonesian Geothermal Tariff

10.5
Tariff, US/kWh
10
9.5
9.5
9.6
9.7
9.7
9.8
9.9
10.0
10.1
10.2
10.2
Nominal prices
9.50
9.39
9.28
9.17
9.07
8.96
8.86
8.76
8.5
8.66
8.57
at 2014 prices
8
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
US/kWh = cents per kilowatt-hour.

Assumption: PPI index escalates at 2% per annum, applied to 40% of the base price.
125
126APPENDIX 4
Unfortunately, most renewable energy tariffs are back-loaded in consequence of escalation clauses,
and many decline over time in real terms if escalation is below the rate of inflation. Recent geothermal
tariffs in Indonesia have taken the form of a nonescalating base tariff (typically 60% of the total base
price), with the balance escalating according to the United States (US) Producer Price index (PPI). For
a hypothetical base price of 9.5 US/kWh, the tariff would escalate over time as shown in Figure A4.1.
At constant 2014 prices, the tariff decreases over time.
This raises the question of how such a tariff revenue stream matches the developers actual profile of
cash revenue requirement (composed of operation and maintenance [O&M] costs) (including major
outlays for make-up wells), debt service obligations, taxes, and equity returns. Our analysis shows that
the two match poorly.
Financial Analysis at International Bank of Reconstruction

and Development and ADB
The general practice in international financial institution (IFI) appraisal reports is to estimate the financial
viability of a project by calculation of a project financial return at constant prices, that is independent
of the actual financial structure of a project, and that is unrelated to actual cash flows in nominal terms
(as would be used by a developer in a project-finance context). This project financial return is then
compared to a calculation of the weighted average cost of capital (WACC): the project is declared
financially viable if the project return is greater than or equal to the WACC. Table A4.1 shows a typical
example of such a project financial return at constant prices.
Table A4.1: Sample Project Financial Returns (at Constant Prices)
NPV
1
Installed
capacity
Capital
$/kW
cost
Investment $ million 282.2
O&M
$ million
58.7
3
4
5
Make-up
wells
6 Taxes
7 Total cost
8 Levelized
cost
9 Revenue
10 Energy
11 Tariff
12 Tariff
13 Financial
flows
14 FIRR
73.1
76.8
21.6
8.4
8.4
8.4
10.5
9.4
10
15
77.6
110.1
$ million
18.0
$ million
$ million 390.4
$/kWh
0.068
GWh
$/kWh
$ million
$ million
5711
$ million
11.4%
77.6
110.1
73.1
76.8
30.0
8.4
26.4
8.4
18.0
10.5
27.4
8.4
8.4
10.5
18.0
0.0
26.4
10.5
887.0 887.0 887.0 887.0 887.0 887.0 887.0 887.0

0.075 0.075 0.075 0.075 0.075 0.075 0.075 0.075
0
0 0.0
0.0 66.5 66.5 66.5 66.5 66.5 66.5 66.5 66.5
77.6 110.1 73.1 76.8 36.6
58.1 40.1 56.0
39.1
58.1 40.1 56.0
FIRR = financial internal rate of return, GWh = gigawatt-hour, MW = megawatt, O&M = operation and maintenance, US/kWh = cents per
kilowatt-hour.
Note: Actual calculations to 30 years.
Source: World Bank.
20
MW
Tariff Structure and Project Finance 127
The corresponding WACC calculations are shown in Table A4.2.

Table A4.2: Illustrative Weighted Average Cost of Capital Calculation
Share
Nominal
Cost
Real Cost
Equity
0.460
0.1400
6.44%
0.1122
5.16%
CTF
0.227
0.0025
0.06%
0.0220
0.50%
IBRD
0.313
0.0502
1.57%
0.0246
0.77%
WACC
Inflation, $
8.01%
5.93%
0.025
CTF = Clean Technology Fund, IBRD = International Bank for Reconstruction and Development, WACC = weighted average cost of capital.
The conclusion which is drawn from this (under the particular assumption of the tariff as shown,
7.5US/kWh) is that the project is financially viable since project financial return equals 11.4%, which
is greater than the real (inflation adjusted) WACC, 5.93%.
Such a conclusion is not reliable, for several reasons:

The implications of the WACC calculation is that equity contributions are pari passu with
debt. Particularly in the case of geothermal projects, that is not reasonable: the up-front
exploration and delineation drilling will be front loaded with equity.
For the tariff to be shown as constant implies that in nominal terms, the entire tariff escalates
at the assumed inflation rate. Since the analysis is presented in US dollars, that would require
the entire tariff to escalate at 2% (US PPI)but in reality, most power purchase agreements
(PPAs) negotiated by PT Perusahaan Listrik Negara (State Electricity Company) (PLN) allow
PPI escalation only on 25%40% of the base price (as shown in Figure A4.1).
O&M costs are at constant prices, which implies that both foreign and domestic cost
components escalate at the same rate. That seems very unlikely.
The WACC calculation shows a post-tax equity return for developer of 14%, which is
consistent with (no) taxes appearing in Table A4.1. But since the calculation of corporate
income tax will be dependent upon the time pattern of interest payments (which will not be
constant through the life of the project), it is unclear that the line item for taxes is reliable.
Nowhere in the (large) spreadsheet model could we find any reference to DSCR. That may
not be unreasonable for the project in question, which is 100% financed by international
finance institutions (IFIs), but that is not always the case.
The announced rationales for such a procedure are (i) that at the time of the project appraisal report
(submitted to boards of IFIs for approval), the details of the financial structure are often not yet
finalized, and (ii) that WACC is a valid numeraire for assessing project financial returns.
That may be reasonable in the case of traditional IFI projects involving loans to a large state-owned
utility, for which the average cost of debt is indeed composed of a mix of loans of different tenors and
interest rates. It is less reasonable in the case of private sector project finance, for which the detail of
project cash flows is everything.
128APPENDIX 4
Financial Analysis at Nominal Prices

When one uses exactly the same assumptions, but with the actual financial structure used to calculate
the cash flows, and at nominal prices (we assume here 2.5% $ inflation, set equal to the US PPI), the
calculations show rather different results.
Table A4.3 shows construction period disbursements, assuming that 75% of the capital cost is in
dollars, and 25% in Indonesian rupiah (under a 4.5% Indonesian inflation rate assumption). Total
construction funding rises from the overnight cost of $359.1 million to $406.2 million when interest
during construction and construction period escalation is included. The equity and debt proportions
are identical to those shown in Table A4.2, and equity is disbursed pari passu with debt.
Table A4.3: Construction Disbursements at Nominal Prices
Year
Share
Construction: Uses of Funds
Overnight costs, constant $
$ share, constant $
0.75
Sum
$ million
359.1
77.6
110.1
73.1
76.8
21.6
$ million
269.3
58.2
82.5
54.8
57.6
16.2
$ million
287.3
59.7
86.7
59.0
63.6
18.3
$ million
89.8
19.4
27.5
18.3
19.2
5.4
$ million
101.6
20.3
30.2
21.0
23.2
6.8
$ million
0.0
0.0
0.0
0.0
$ million
388.8
79.9
inflation adjusted, nominal $
Rp share, as constant $
Rp share, at Rp inflation
VAT on Rp share
Total construction cost
IDC/service charge
10
CTF
0.0
$ million
0.7
0.0
0.1
11
IBRD
0.1
$ million
16.6
0.6
12
Total uses of funds
$ million
406.2
13
Construction: Sources of Funds
14
Equity
15
Balance for debt
16
Debt
17
CTF
18
IBRD
19
Total sources of funds
0.25
0
0.46
116.9 80.0
0.0
86.8
25.1
0.2
0.2
0.2
2.0
3.5
4.8
5.7
80.5
119.0
83.7
91.8
31.1
$ million
186.8
37.0
54.8
38.5
42.2
14.3
$ million
219.3
43.5
64.3
45.2
49.6
16.8
0.42
$ million
101.5
20.1
29.8
20.9
23.0
7.8
0.58
$ million
117.8
23.4
34.5
24.3
26.6
9.0
$ million
406.2
80.5
119.0
83.7
91.8
31.1
CTF = Clean Technology Fund, IBRD =International Bank for Reconstruction and Development, IDC = interest during construction,
Rp = Indonesian rupiah, VAT = value added tax.
Source: World Bank.
Table A4.4 shows the developer cash flows. Other assumptions here are:

that the tariff escalates according to the PLN PPA formula, with 60% of the base price
escalating at the US PPI;
that the foreign exchange share of O&M is 25%, the Rp share is 75%;
that make-up wells have the same foreign exchange share as the first investment cost; and
that the corporate tax rate is 34%.
Table A4.4: Developer Cash Flows, at Base Tariff of 7.5 US/kWh and Equity Pari Passu
Year
1
Shares
NPV
10
15
O&M Costs
O&M ($ share)
0.25 $ million
2.10
2.10
2.10
2.10
2.10
O&M (Rp share)
0.75
$ million
6.30
6.30
6.30
6.30
6.30
Make-up wells
($ million)
0.75
$ million
0.00
0.00 13.50
0.00
13.50
Make-up wells (Rp)
0.25 $ million
0.00
0.00
4.50
0.00
4.50
Cash Disbursements
Equity
O&M Costs
$ million
37.1
54.8
38.6
42.4
14.5
O&M ($ share)
$ million
2.38
2.44
2.50
2.69
3.04
10
O&M (Rp share)
$ million
8.00
8.40
8.82
10.21
13.03
11
Make-up wells ($)
$ million
0.00
0.00 16.05
0.00
19.55
12
Make-up wells (Rp)
$ million
0.00
0.00
6.30
0.00
9.31
13
Principal Repayments
14
CTF
$ million
15
IBRD
$ million
16
Interest
$ million
17
CTF
18
IBRD
19
Income tax
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.4
5.4
0.0
0.0
6.4
6.4
6.4
$ million
0.2
0.2
0.2
0.2
0.1
$ million
6.2
6.4
6.2
5.3
3.7
$ million
34.0
0.0
0.0
0.0
0.0
9.1
9.1
2.5
9.7
1.7
$ million
371.5
37.1
54.8
38.6
42.4
40.4
26.6
49.0
39.9
62.3
22 Energy
GWh
5382
23
$/kWh
20 Total cash out

21
Revenue
Tariff
24 Revenue
$ million
Total Cash Flows
$ million
26 Post-tax IRR, nominal
[ ]
25
27
Deflated Cash Flows
28 Real, post tax

29 DSCR
0.0
0.0
887.0
0.083
0.0
66.5
67.2
67.9
70.0
74.0
37.1 54.8 38.6 42.4
26.2
40.6
18.9
30.2
11.7
36.2 52.2 35.9 38.4
23.1
35.0
15.9
23.6
8.1
10.4%
$ million
[ ]
0.0
887.0 887.0 887.0 887.0

0.075 0.076 0.077 0.079
7.7%
37.1
54.8
38.6
42.4
14.5
$/ kWh = dollars per kilowatt-hour, CTF = Clean Technology Fund, DSCR = debt service coverage ratio, GWh = gigawatt-hour, IBRD = International
Bank for Reconstruction and Development, IRR = internal rate of return, Rp = Indonesian rupiah, O&M = operation and maintenance.
At the stated base tariff of 7.5 US/kWh, the post-tax, nominal FIRR is 10.4%, significantly below the
14% return assumed for developer in the project financial analysis. To achieve the stated equity target
requires a base year tariff of 8.6 US/kWh, not 7.5 US/kWh.
130APPENDIX 4
Figure A4.2: Equity Return versus Tariff
0.25
0.2
FIRR
0.15
Hurdle rate = 14.0%
0.1
0.05
0
0.04
0.06
0.08
0.1
0.12
Tariff, $/kWh
$/kWh = dollars per kilowatt-hour, FIRR = financial internal rate of return.

The debt/equity ratio and equity timing is critical. The above calculations assume a developer
equity share of 46%, and for simplicity it is generally assumed in project financial analysis that equity
contributions are pari passu with debt. For geothermal projects that is unlikely. Table A4.5 shows the
same calculation, but for equity up-front (i.e., disbursed ahead of debt).
Table A4.5: Construction Period Disbursements, Equity Up Front
Share
Sum
$ million
359.1
77.6
$ million
269.3
$ million
$ million
110.1
73.1
76.8
21.6
58.2
82.5
54.8
57.6
16.2
287.3
59.7
86.7
59.0
63.6
18.3
89.8
19.4
27.5
18.3
19.2
5.4
$ million
101.6
20.3
30.2
21.0
23.2
6.8
$ million
0.0
0.0
0.0
0.0
$ million
388.8
79.9
116.9
80.0
86.8
25.1
Construction: Uses of Funds
Overnight costs, constant $
$ share, constant $
inflation adjusted, nominal $
Rp share, as constant $
Rp share, Rp inflation
VAT on Rp share
Total construction cost
IDC/Service Charge
10
CTF
0.0
$ million
0.4
0.0
0.0
0.1
0.1
0.2
11
IBRD
0.1
$ million
11.7
0.0
0.2
1.6
4.1
5.8
12
Total uses of funds
$ million
400.9
79.9
117.1
81.7
91.1
31.2
13
Construction: Sources of Funds
14
Equity
$ million
184.4
79.9
104.5
0.0
0.0
0.0
15
Balance for debt
$ million
216.5
0.0
12.6
81.7
91.1
31.2
16
Debt
17
CTF
0.42
$ million
90.9
0.0
5.3
34.3
38.2
13.1
18
IBRD
0.58
$ million
125.6
0.0
7.3
47.4
52.8
18.1
19
Total sources of funds
$ million
400.9
79.9
117.1
81.7
91.1
31.2
0.75
0.25
0
0.46
0.0
CTF = Clean Technology Fund, IBRD = International Bank for Reconstruction and Development, IDC = interest during construction,
Rp = Indonesian rupiah, VAT = value added tax.
When equity is up front, several consequences arise. First, is that interest during construction
decreases, so the total completed financial cost decreases slightly from $406.2 million to
$400.9million. Second, at the old base tariff of 7.5 US/kWh, the FIRR falls from 10.4% to 9.4%. At the
14% FIRR target, the required tariff is 9.2 US/kWh. But when equity is provided up front, at least in the
private sector the aggregate FIRR target will also increase, so the gap between the achievable FIRR and
the target really increases by more than 1%. Figure A4.3 shows the differences over a range of tariffs.
Figure A4.3: Impact of Equity Disbursement Timing

0.2
0.15
Hurdle rate=14.0%
FIRR
0.1
0.05
0
Equity up front
0.05
0.04
0.06
0.08
0.1
0.12
Tariff, $/kWh
$/kWh = dollars per kilowatt-hour, FIRR = financial internal rate of return.
Figure A4.4 illustrates the impact of the debt to equity ratio. The lower the equity share, the higher lies
the curvewhich means that for a given FIRR, the higher the equity share, the lower the required tariff.
Figure A4.4: Impact of Debt/Equity Ratio (Equity Up Front)

0.25
30%
0.2
40%
46%
FIRR
0.15
0.1
0.05
0
-0.05
0.04
0.06
0.08
Tariff, $/kW h
$/ kWh = dollars per kilowatt-hour, FIRR = financial internal rate of return.
0.1
0.12
132APPENDIX 4
These results are summarized in Table A4.6.
Table A4.6: Sensitivity of Tariffs to Financial Structure (Typical Project Financed with
International Finance Institution Concessionary Debt)
Equity Share
Equity Contributed
Required Tariff for

14% Equity Return
US/kWh
46%
Pari Passu
8.6
46%
Up front
9.2
40%
Up front
8.7
30%
Up front
8.0
/kWh = cents per kilowatt-hour.

Bankable Projects
A bankable project in project finance is one that meets both the objective of investors (for example
as the desired return to equity) and the objective of lenderswhich translates to adequacy of cash
flows to meet debt service obligations, most frequently captures by the debt service coverage ratio
(DSCR). Both of these targets will vary from project to project depending on the relative perceptions
of project risk.
In fact, once a geothermal project reaches financial closure, much of the project risk has been
reduced. Lenders and equity holders in other types of renewable energy project face significant
resource risk after project completionannual variations in hydrology in small hydropower projects,
or annual variations in average wind speeds can vary by +30% over long term averages as established
in feasibility studieswhich may have significant impact on cash flows. One of the major advantages
of a geothermal project is that once constructed, annual plant factors will likely be fairly constant in
the 90%95% range (always assuming competent operation and maintenance).
Figure A4.5 illustrates this idea. Bankable projects are to be found in quadrant I, where the quadrants
are defined by the minimum requirements for DSCR (here shown as 1.8) and FIRR (here shown as
Figure A4.5: Bankable Project Definition
4
IV. Unbankable
FIRR<target
DSCR
I. Bankable
2
6.5
baseTariff US/kW h
7.0
7.5
8.0
8.5
9.0
9.5
11.0
10.5
10.0
1
III. Unbankable
FIRR, DSCR<target
II. Unbankable
DSCR<target
0
0
0.05
0.1
0.15
0.2
FIRR
US/kWh = cents per kilowatt-hour, DSCR = debt service coverage ratio, FIRR = financial internal rate of return.
Assumptions: 40% equity, equity up front.
14%). In all of the other quadrants, either or both requirements are not met. The curve in Figure A4.5
shows the relationship between FIRR and DSCR at each level of base tariff.
We see that above a tariff of 7.2 US/kWh, projects meet the DSCR requirement; but at that tariff, the
FIRR is an unsatisfactory 8%. Only at a tariff of about 9.1 US/kWh is the FIRR target achieved, and
the curve enters bankable territory. In this case it is clear that the FIRR requirement is indeed binding,
rather than DSCR.
This conclusion changes with nonconcessionary commercial finance. In Figure A4.6 we show the
result for the same project, again with 40% equity up front, but now financed at 7% over loan tenors of
8 years with just 2 years grace during construction of the power plant.
Under these conditions we see that the 14% hurdle rate for FIRR is reached at a base tariff of 11.5 US/
kWh, but at this level, the DSCR of 1.44 is still below the previous DSCR target of 1.8. To reach this
threshold requires a tariff of 14.2 US/kWh, at which point the FIRR is 18%.
Figure A4.6: Non-Concessionary Finance
3
IV. Unbankable
FIRR <target
I. Bankable
DSCR
10
11
12
III. Unbankable
FIRR, DSCR <target
0.05
13
14
15
16
17
II. Unbankable
DSCR < target
0.1
0.1 5
0.2
0.25
FIRR
DSCR = debt service coverage ratio, FIRR = financial internal rate of return.
Assumptions: 40% equity, equity up front.
Conclusions
The analysis permits a range of important conclusions relevant to the geothermal sector in Indonesia,
and to the issuance of a new tariff:

Calculations of project financial returns following conventional IFI practice are often
unreliable.
Actual financial returns to developers equity are critically dependent on the proposed
financial structure. Financial returns should be based on estimated cash flows and cash
revenue requirements, calculated at nominal prices. The resulting FIRR may be quite different
to project financial returns.
134APPENDIX 4

Simple calculations of WACC for geothermal projects are unreliable. Not only are they
independent of loan tenors, they also assume equity pari passu with debt, unrealistic for
geothermal projects.
Tariffs issued on the basis of production costs will be highly dependent on financial structure
and financing terms. FIRR may be much more sensitive to the financial structure than to
individual technical assumptions. We have not seen the financial model being used by
MEMR, but it seems that the tariffs are calculated on the basis of achieving a 16.5% project
IRR. The details of the MEMR production cost tariff model should be examined (just as the
details of the Sinclair Knight Merz Limited model require review).
Projects financed with concessional IFI funding, that meet equity investor FIRR targets, will
also have healthy DSCRs: for such projects, meeting DSCR standards are of little concern.
Because of long loan tenors, and low interest rates (particularly on CTF funds), revenue
requirements to meet target international rate of return are not front-loaded, so the current
PPA escalation and/or indexation approach is not problematic. Generous grace periods and
low first year debt service payments result in good cash flows even in the early years. Phrased
differently, in concessionary financed projects, equity IRR, not DSCRs (to the extent that
they apply at all) are likely to be the binding condition for bankability.
Front-Loaded Tariffs in Other Countries

The best example of front-loaded tariffs for capital intensive renewable energy projects is Sri Lanka.
While it has no geothermal resources, its FITs for wind and small hydro provide for two options:

a three-tier fixed payment for recovery of investment and equity return, plus an escalable
component for O&M; or,
a fixed levelized payment, non-escalable (which then declines in real terms).
The idea of a three-tier tariff for investment cost recovery is precisely because in Sri Lanka, most small
hydropower and wind power projects are financed locally, with loans of short tenor. The three-tier
tariff rates are as follows:
Table A4.7: Three-Tier Tariff Rates in Sri Lanka
Technology/Source
Non-escalable
Escalable
Escalable
Year 18
Year 915
Year 16+
Base O&M
(year 120)
Base Fuel
(year 120)
Mini-hydro
12.64
5.16
None
1.61
None
Mini-hydro-local
12.92
5.28
None
1.65
None
Wind
17.78
7.26
None
3.03
None
Wind-local
18.28
7.47
None
3.11
None
Biomass (115 years)
7.58
3.10
None
1.29
9.10
Biomass 16 year onwards
7.58
3.10
None
1.61
9.10
Agro & Indus (115 years)
7.58
3.10
None
1.29
4.55
Agro and Indus 16 year

onwards
7.58
3.10
None
1.61
4.55
15.16
6.19
None
4.51
1.75
Waste Heat
7.13
2.65
None
0.43
None
Escalation per year
None
None
None
7.64%
Municipal Waste
Agro = agriculture, hydro = hydropower, Indus = industry, O&M = operation and maintenance.
Source: Public Utilities Commission of Sri Lanka.
5.09%
So, for example, for investment recovery, a wind project gets 17.78 SLRs/kWh (13.5 US/kWh) in
years 18, 7.26 SLRs/kWh (5.5 US/kWh) in years 915, and nothing from year 16 onward. The PPA
stipulates that the energy deliverable in years 915 must be at least the same as that delivered in years
18. The resulting time patterns are illustrated in Figure A4.7.
Figure A4.7: Three-Tier Front-Loaded Tariffs in Sri Lanka

30.00
Mini hydro
Wind
Coal
Unit Price (SLRs/kWh)
25.00
20.00
15.00
10.00
5.00
-
2 3 4 5
6 7
8 9 10 11 12 13 14 15 16 17 18 19 20
hydro = hydropower plant, SLRs/kWh = Sri Lanka rupees per kilowatt-hour.

Source: Authors calculations based on data from the Public Utilities Commission of Sri Lanka.
Appendix 5
Possible Technology Paths to

Faster Development
Use of Ancillary Plant
By far the largest geothermal generation capacity in Indonesia uses condensing steam turbines with
evaporative (wet) cooling towers. This is similar to the situation in most of the rest of the geothermal
industry worldwide, but there are other mature, well proven technologies that offer some advantages
particularly in terms of mobilization time and that have not been taken up in Indonesia to any
significant degree as yet. They may well present an opportunity to accelerate the Indonesian program.
Stand-Alone Back Pressure Plants

A back pressure turbine exhausts usually to atmosphere (or in some cases to another second-stage
turbine), unlike a condensing turbine which exhausts to a subatmospheric condenser. These two
turbines are diagrammed in Figure A5.2 and Figure A5.4, respectively. In the geothermal energy
context, back pressure plants (BPPs) differ in the following ways from the more common condensing
plants:
Advantages:

Simple and robust.
Do not require a cooling circuit, so smaller footprint.
Available in small sizes and widely available.
Much shorter manufacture and commissioning time: typically 1 year or less compared to 2 to
3 for a 55 MW condensing plant.
Much cheaper: roughly half the capital expenditure/MW delivered.
If running on a liquid dominated resource (the most common), still require reinjection wells
for separated water, but do not require condensate injection wells.
Because of their small size and simplicity, can easily be remobilized from one project to
another.
Disadvantages:
136
Relatively inefficient, as they require about twice as much steam per MW as a condensing
plant. This impacts the number of production wells and can raise concerns about the
efficiency of resource use and sustainability if used long term.
Probably require more frequent maintenance than larger condensing units, in part because
they tend to be connected to relatively unsophisticated steam gathering systems (e.g., no
steam scrubbers), though because of their simplicity this should not be expensive.
Possible Technology Paths to Faster Development137
Potentially greater environmental impact (noise, exhaust condensate spray), but not great if
properly designed and operated as has for example been well demonstrated in Mexico.
If distributed around a field, add complexity to the electrical interconnection.
Mode of Deployment
There has been only one such small (1.5 MW) unit deployed in Indonesia in recent years, at Sibayak.
That was a relatively unsophisticated installation, on a less than ideal resource, and should not be
regarded as typical. PLN has also run a BPP at Ulumbu (2 x 2.5 MW). The typical set-up of a backpressure steam turbine is shown in Figure A5.1. Elsewhere in the world such units in the 1 MW5 MW
range have been used either on individual wells (well head generators) or with small numbers of wells
interconnected to multiple back pressure units. This has been particularly commonly used in Mexico,
but has also been successfully done in several other countries including El Salvador, Nicaragua
(FigureA5.2), Papua New Guinea, and the Philippines.
Figure A5.1: Back Pressure Steam TurbineStand Alone Operation

Steam
Two Phase
Non-Condensing
High Pressure Turbine
Atmospheric Vent
High Pressure
Separator
Brine
Production Wells
Injection Wells
Rock Muffler
Source: J. Lawless. 2008. Cost Reduction Through Improved Geothermal Well Targeting. Paper for the Western Pacific Regional Branch
of the International Geothermal Association/INAGA Joint Technical Seminar. Nusa Dua, Bali. 26-28 April.
Figure A5.2: San Jacinto Project, Nicaragua: 2 x 5 Megawatt Back Pressure Plant
Source: Authors photograph.
138APPENDIX 5
Figure A5.3: Schematic of Condensing Steam Turbine Plant
Off-Gas
CO2, H2S, N2, H2, CH4
Evaporation ~255 t/h
Hot Basin (44C)

Forced Draft Cooling Tower
Steam Flow
~350 t/h
Cold Basin (28C)

Cooling Tower
Outlet
9500 t/h
Intercooler
55 MWe
Turbine
NCG
0.1 bara
Direct Contact
Condenser
Riser
Steam
Ejector
Liquid Ring
Pump
44C
Hot Well
Pump
Sump
Caustic
Injection
(48% NaOH)
~95 t/h
Blow-Down
Injection
Bara = absolute pressure, C = Celsius, CH4 = methane, CO2 = carbon dioxide, H2 = dihydrogen, H2S = hydrogen sulfide, MWe = megawatt
electric, N2 = nitrogen, NaOH = sodium hydroxide, t/h = ton per hour.
of the International Geothermal Association/INAGA Joint Technical Seminar. Nusa Dua, Bali. 2628 April.
Figure A5.4: McLachlan Plant, Wairakei, New Zealand:

55-Megawatt Condensing Steam
Note: This plant uses a second-hand Fuji turbine. Note the evaporative cooling towers.
In very remote locations,

BPPs may be operated on a
semi-permanent basis, but more
commonly they are used for a
few years before being replaced
by a larger condensing and/or
binary plants. A schematic of a
condensing steam turbine plant
is shown in Figure A5.3.
The huge advantage of such a
scheme is the early availability
of generation and revenue. Even
the successful early exploration
wells can be put to use within 12
months of being drilled, which
could otherwise sit unused for
many years (Figure A5.5). More
significantly, there are projects in
Indonesia such as Karaha where
for various reasons as many as
half the production wells for a
large scheme have been drilled
and left closed for a decade
or more.
Figure A5.5: Time from First Exploration Well to Commercial Operation Date
Years
0
10
20
30
40
Bulalo
Tongonan
Wyang Windu
Tiwi
Kamojang
Palinpinon
Wairakei
Les Azufres
Momotombro
Ahuachapan
G. Salak
Cerro Prieto
Lahendong
Kizildere
Los Humeros
Mokai
Darajat
Dieng
Olkaria
Ohaaki
Rotokawa
There are also other less tangible advantages. Early operation of a BPP gives an early test of the reservoir
in a way that is far more comprehensive and reliable than individual short-term well tests. This adds
confidence to the estimation of resource capacity, provides information on likely well run-down rates,
fluid chemistry, and permits closer optimization of the design parameters such as separation pressure
and turbine inlet pressure for the larger ultimate scheme. Perhaps, just as significantly, it provides
investors with additional confidence in the feasibility of the project.
The question then has to be asked, why has there not been greater use of this development pathway
in Indonesia and what is hindering its use in the future? Several possibilities arise:

The power price is unlikely to be a disincentive, since the cost per megawatt of such schemes
must be less than for condensing plants, especially if the wells are regarded as free for this
purpose. Nevertheless, the financial viability needs to be demonstrated.
Current power purchase agreements (PPAs) do not appear to allow for early small-scale
generation. Some developers acknowledged that this could be an obstacle. However, one
pointed out that activating the PPA early had the advantage of initiating the escalation
clausesthough it would mean it also terminated earlier than otherwise.
The business model with the developer providing steam and a different party owning the
power plant does not allow the developer to also undertake early generation, whereas it is the
developer who has expended the early investment and has the incentive to do so. If PLN is to
own the eventual larger plant there, but has invested nothing in drilling, there is little incentive
for PLN to undertake early generation.
The developers are reluctant to either install temporary pipelines (and in most cases
separators) to connect the wells to the plant or to pre-invest in part of the permanent
steam above ground system (SAGS), especially at a time when not all of the wells have been
drilled so the permanent SAGS cannot be fully designed. The same applies to the electrical
switchyard. Neither of these should be insurmountable.
140APPENDIX 5

Depending on the location PLN may be reluctant to install a transmission line, including land
acquisition, for the sake of a small amount of generation. There are ways around that: for
example, at the San Jacinto project in Nicaragua a 2 x 5 MW BPP was installed 3 years before
the full 2 x 36 MW condensing plant. The transmission voltage and towers were sized for
the full-sized plant, which would run in a double circuit mode, but for the smaller BPP only a
single circuit was initially installed. Because of the small size of the units, there should not be
issues with grid control.
The environmental impact assessments and environmental permits may have to be modified.
The Government of Indonesia may be reluctant to permit relatively inefficient use of the
resource (though we have no evidence that this is the case). Where this issue has arisen in
other countries, one approach has been to limit the time that the BPP can operate for on any
particular project.
All of these issues may in the part simply have made the concept too hard for the developers and
PLN who were focused on the larger developments needed to recoup their significant investments in
exploration and development. However, in the current context, with the development program falling
well behind schedule, it may be worth re-examining what can be done to facilitate this approach in the
national good.
Bottoming Binary Plants

A binary cycle plant uses heat exchangers and a secondary working fluid of low boiling point rather
than passing geothermal steam directly through a turbine. In the geothermal context, they differ in the
following ways from condensing steam plants:
Advantages:

Usually (though not necessarily) are air-cooled rather than using evaporative cooling
towers, and so conserve water usage (whether of geothermal or other origin). This can be
an important issue for arid areas, or where minimizing pressure drawdown in the geothermal
reservoir is important, but so far neither of these has been perceived to be a major factor in
Indonesia.
Available in small sizes (<1 MW to 15 MW) and available from several manufacturers in
modular form. This also means that the economies of scale for binary plant are less significant
than for a condensing steam plant.
Much shorter manufacture and commissioning time: typically 1 to 1.5 years compared with
2 to 3 years for a 55 MW condensing plant.
Because of their small size and modular construction simplicity, can be remobilized from one
project to another.
Most significantly, can make use of lower temperature geothermal fluid than condensing
steam plants. They can, however, also make use of moderately high temperature resources.
Can cope with higher noncondensable gas contents than condensing steam plant.
Have somewhat reduced environmental impact as all liquid is reinjected, though there are a
few additional environmental issues that have to be considered, most notably the storage and
use of relatively large amounts of hydrocarbon working fluids. In practice, unless maintaining
reservoir pressures is of great concern because of effects such as subsidence, the differences
are small.
Disadvantages:

Higher cost per MW than a condensing turbine by a factor of about three.
If used in a bottoming mode (which is what is proposed here), the reservoir fluid is cooled
before reinjection to about 100C rather than the 160C, which would be more common
for condensing plants. This raises the potential for mineral scaling particularly of silica in the
reinjection system. Modern practice has however developed several means of dealing with
this, including recombining separated water (which can have chemical advantages as well as
providing dilution), and condensate and chemical dosing, which has proven to be effective
when properly designed and operated, at moderate cost.
The fact that the reinjection temperature is lower can potentially have adverse reservoir
impacts if not properly managed. However, if done properly it can also lead to greater overall
energy recovery.
In a high-enthalpy resource that is intensively exploited, particularly without full reinjection,

the steam fraction produced by the wells can rise with time. This is advantageous from the
point of view of a condensing plant as it increases the steam flow and reduces the reinjection
load, but it can lead to an associated binary plant being under supplied with liquid. In the
worst case, part of the binary plant may have to be decommissioned, but since it is modular
and can be redeployed elsewhere it does not have to be a fatal constraint, especially for an
operator with a large portfolio. Alternatively, the modular nature allows a more conservative
approach to be taken and the binary units installed progressively over time as the reservoir
response to exploitation becomes better known.
Mode of Deployment
One binary plant has been deployed in Indonesia at Lahendong, and that is perhaps better regarded
as a prototype rather than as a typical modern commercial unit. Elsewhere in the world binary plants
have been deployed either as single units or in combinations up to 155 MW. They are used in three
very different ways, only one of which is proposed in the present context (Figure A5.6):
1. Mode 1: as stand-alone units or combinations of units operating on the whole of the fluid
produced by the wells:
a. Mode 1a: to take advantage of their ability to generate from lower temperature resources, or,
b. Mode 1b: to minimize reservoir impact because of the greater reinjection fraction, or,
c. Mode 1c: to deal with difficult fluids such as those with a high gas content.
2. Mode 2: in effect, to replace the condensers on a condensing steam plant, utilizing a
combination of a back pressure steam turbine and a binary unit used as a steam condenser.
3. Mode 3: in combination with a steam turbine, to extract extra energy from the separated
water. Thus in a very real and defensible sense, they add to the extractable capacity of a given
resource.
It is only the third mode that is relevant in the present context. Use on low temperature resources
(mode 1a) is very important in certain countries but the probable power prices in Indonesia would
not support the use of such resources except in the special case of very remote locations with few
alternative generation options. Such projects are important in terms of social impact but are likely to
be small in terms of additional MW in the national portfolio. The need to reinject a greater fraction, as
used in mode 1b, is not currently perceived to be an issue in Indonesia. Mode 1c or mode 2 could be
used at certain locations, but that would be an engineering and cost design decision on a case by case
basis and in the present context would replace a condensing plant rather than adding to the national
generation total.
142APPENDIX 5
Figure A5.6: Bottom Binary Plant Operating on Separated Water
Forced Draft Cooling Tower
Condensing Turbine
Bottoming OEC Unit
Air-Cooled
Condenser
Organic
Fluid Circuit
Brine injection
Binary
Turbine
Production
Separator (~ 8 b.a.)
Separated Brine
Binary Fluid
Heat Exchanger
Production Wells
of the International Geothermal Association/INAGA Joint Technical Seminar. Nusa Dua, Bali. 26-28 April.
Figure A5.7: Bottoming 17-Megawatt Binary Plant (Foreground) at Wairakei, New Zealand
Note: Notice the bank of air-cooling fans. The original 50-year old condensing steam plant is in the background. It lacks cooling towers,
because it has an unusual cooling system using direct contact river water. That has no particular relevance in terms of the retrofitted
binary plant.
The significant advantage of the third mode of operation is that it can rapidly (within as little as 1 year)
produce 15%20% of additional electricity without requiring any additional wells, since the separated
fluid is already available on the surface. It is therefore practically free of resource risk, provided the
issues of scaling and reservoir management can be dealt with. Examples in many fields worldwide
demonstrate that such issues are manageable. Additional environmental issues are likewise small.
Significantly, so-called bottoming binary units can be retrofitted to existing condensing steam plants
with only very minor engineering modifications, which would not require the condensing plants to be
out of operation for a long period.
As with BPPs, the question is why has there not been greater use of this development pathway in
Indonesia, and what is hindering its use in the future?
Some possibilities are:

The higher capital cost per MW compared to a condensing plant, coupled with historically low
tariffs. However, tariffs have or will rise, and more significantly drilling costs (especially when
the averaged effect of unsuccessful wells is taken into account) have risen faster than power
plant costs.2 Therefore, the cost differential between a bottoming binary plant (excluding
wells) and that of a whole scheme (including wells) based on a condensing steam plant is
much less than it used to be. Nevertheless, the financial viability (taking proper account of
the shorter lead time) needs to be demonstrated.3
PPAs are not set up to allow for additional small-scale generation.
PLN may be reluctant to install additional transmission capacity. Because of the small size of
the units, there should not be issues with grid control.
The environmental impact assessments and environmental permits may have to be modified.
The developers may have historically been concerned about silica deposition because of the
greater fluid cooling before reinjection. This is a real issue, but in recent years ways of dealing
with it have been proven.
An unusually higher proportion of the projects developed so far in Indonesia are on dry or nearly
dry steam resources. Darajat and Kamojang are dry steam, Wayang Windu and part of Dieng are
almost dry, to the point where bottoming binary units would be of limited applicability. But the
proportion of liquid resources where binary units are applicable will rise with time.
Developers are concerned about adverse reservoir impacts,4 but as noted above there are
many examples worldwide where such have been successfully avoided.
As with the BPP concept, with the development program falling well behind schedule, it may be worth
re-examining what can be done to facilitate this approach especially with regard to retrofitting to
existing plants.
Potential Impact
An estimate is made of additional generation possible by retrofitting binary units to existing condensing
steam plants or including them in the design of future plants is provided in Appendix 2. The actual
2
Because they are very efficient at drilling, this advantage is less for Chevron than for the other operators. It said that it had
been, and was now, evaluating binary options but in its projects the best financial return was in adding more condensing
units. That does not mean that the financial return for binaries was nil, and in the future, as it has probably more or less
installed the full capacity of condensing units that the existing wet resource at Salak may support, it may be more willing to
consider this option.
When AECOM/Sinclair Knight Merz Limited prepared financial models in 2010 for Pertamina Geothermal Energy (PGE)
as part of the feasibility studies for Lumut Balai, Ulubelu, and Tompaso, they included binary options and concluded they
were financially inferior to a purely condensing plant. But that was within the scope of a fixed PPA capacity (i.e., the binary
plants were to replace part of the condensing plant, not to be incremental to it), and in the event the differences in net
present value (NPV) were small.
This is the case for PGE at Ulubelu where there has already been some cooling of certain production wells by reinjections.
The issue is about to be analyzed using reservoir modeling, and should be manageable provided PGE stops injecting close
to a major fault which conducts the reinjected fluid back to the production wells.
144APPENDIX 5
amount of additional generation possible will depend on the fluid enthalpies and separation pressures
in each case, but for the present purpose a default and conservative increment of 15% of the output
of the condensing plant operating on wet resources is assumed. That figure is conservative: in some
projects it is going to be as high as 20%. Kamojang, Patuha, and Darajat are excluded as they use
dry steam. Because of the small separated water fraction at Wayang Windu, a smaller estimate of
possible binary capacity is given than for the other fields. It is assumed that a retrofitted binary plant
could be installed within 12 months for existing plants, and simultaneously with commissioning of
new condensing plants. Some additional delays in the program are anticipated so the estimates are
conservative.
Looking only at Java and Sumatra, it can be seen from the estimates in Section 2, Table 2.3 that there is
potential for about 125 MW of binary to be added immediately (i.e., within 12 months), possibly plus
another 5 MW at Wayang Windu, though it is uncertain whether that small increment there would be
considered worth pursuing. The possible additional total by the end of 2020 is in excess of 400 MW.
Opportunities for Technical Improvements in Planned

Developments Using Condensing Steam Plant
Use of Larger Units
For condensing steam turbine plants, the type planned for all developments in the near future in
Indonesia, there are significant economies of scale both in terms of the size of the schemes overall and
the size of the individual units used. It can be assumed that the overall size of the planned developments
has been thought through in the context of prudent incremental development, resource capacity,5
grid and market constraints, and availability of finance6though those factors could well change in
the course of development.
However, most of the developments and in particular those planned by Pertamina Geothermal Energy
(PGE) are based around unit sizes of 55 MW. In some cases there are good reasons, as given in the
previous paragraph, why that should be so. But in other cases such as Ulubelu 3 and 4, where two
55MW units are to be installed on the same resource simultaneously, there appear to be no good
reasons to do so rather than installing a single 110 MW unit.
It is suspected that the reason for settling on multiples of 55 MW units is largely historical: at one
stage these were the largest units available and became a de facto industry standard. But larger, well
proven units are now available from the same manufacturers with the largest currently installed in
New Zealand, having a name plate rating of 132 MW and in fact producing up to 140 MW. The first 110
MW unit was installed in Wayang Windu in the late 1980s, so it is not as if the concept is unknown in
Indonesia.
This issue was discussed with PGE during the feasibility studies and financial negotiations for Ulubelu3
and 4, and it was pointed out by the consultants that a saving of about 20% would be possible by
installing a single 110 MW unit. Based on an analysis by AECOM, there were no grid constraints on the
use of a single larger unit. Despite this PGE has persisted with the concept of a 2 x 55 MW plant, and the
only explanation given has been that it was considered too difficult to change the PPA that had already
Noting that steam turbines do not run as efficiently at part load, so if there is uncertainty about the long term resource
capacity, it may be better to install smaller units.
It is worth noting that the amount of equity required to get to the resource proven stage for a 55 MW development will
not be much less than for a 110 MW development.
been signed with 2 x 55 MW units specified. This appears to be a very weak reason for accepting a more
expensive development.
It takes approximately the same length of time to manufacture and install a 110 MW unit as a 55 MW
unit, so in that sense there are no time savings to be made. But logically, if a particular plant can be
commissioned for a significantly lower cost, it ought to be easier to finance so in that regard could
accelerate the program. This is one area where the involvement of more experienced commercial
partners may help with the PGE development plan (Section 9).
Use of Other Than New Units

Elsewhere in the world, good use has been made of secondhand (some unused) and reconditioned7
geothermal steam turbines, which are available at lower prices than new units, but more significantly
may be available much faster, in less than 1 year compared to 23.
There is also a specific opportunity with the availability of unused but technically obsolete steam
turbines originally manufactured for naval marine use in the US, which are now being reconfigured for
geothermal use elsewhere. These are not only cheaper, but available much more quickly than newly
manufactured units. In both cases there may be issues of efficiency: the older plants may require more
steam per MW, and therefore more wells, but this is usually more than offset by the cost of the plant.
Reliability is less of an issue since manufacturers or reconditioners guarantees are often available.
The biggest obstacle to the use of such plant in Indonesia so far is believed to have been that
government procurement processes preclude the use of other than newly manufactured plant.
However it is a moot point whether such restrictions would necessarily apply to independent power
producers (IPPs) or to joint ventures with PGE.
Flexibility in Deployment
There are several instances in the world where plants have been ordered for a particular project and
then switched to a different project by the same owner when for some reason the original project
became stalled or delayed. This included some of the Philippine National Oil CompanyEnergy
Development Corporation plants in the Philippines and others in New Zealand and Nicaragua.
Such opportunities may arise, and can be a means to keep the overall program moving when there are
unexpected difficulties such as poor drilling results or land access issues with any particularly project.
To be able take advantage of these opportunities requires significant flexibility by the developer, the
lenders, and the off-taker, to an extent that has not been apparent so far in Indonesia. But with the
incoming of more flexible business arrangements, such as the commercial partnerships with PGE, it
may be worth discussing such concepts.
There may be a need to convert them from 60 Hz operation as in the US to 50 Hz operation as in Indonesia, but this has
been successfully done elsewhere, e.g., for the 55-MW Poihipi plant in New Zealand (shown in Figure 3.3).
Appendix 6
Cost of Exploration
This appendix presents an estimate of the minimum total cost of an exploration program in Indonesia
over the next few years.
It is assumed there has been significant basic geological reconnaissance already in Indonesia to
identify a sufficient portfolio of prospects. The following basic assumptions can be made as to average
costs per prospect:
Stage 1: Surface geoscientific exploration, baseline environmental studies, prefeasibility study, etc.
(i.e., everything prior to drilling): $1 million.
Stage 2: In the feasibility study, the cost of exploration drilling and well testing for two to four wells
is conservatively estimated at $25 million, while noting that depending on location the actual costs
could be significantly higher. That includes environmental permitting and infrastructure, but it does
not include corporate costs in tendering etc., which may be significant but would never be funded by
an external agency. With regard to infrastructure, it is worth noting that the easiest prospects to access
have probably already been developed, so infrastructure costs can be expected to rise as more and
more remote prospects are accessed. A figure as high as $50 million could be correct in some cases.
Stage 3: Costs at this stage comprise delineation drilling, preliminary engineering designs, and updated
feasibility studies i.e., everything additional of a technical nature that is needed prior to financial
closure. Once again, corporate costs are excluded. At this stage, costs will be proportional to the size
of the project. If it is assumed that:

one successful production well and one successful reinjection well are drilled in stage 2;
70% of the total wells are required to get to the end of stage 3;
an average 7 MW output;
a 2:1 production to reinjection ratio (but an integral number of wells must be drilled);
a 75% success rate in drilling;
a cost of $6.5 million per well including infrastructure and testing; and
preliminary engineering and feasibility costs $1 million per project.
Then stage 3 costs for various project sizes will be as follows:

55 MW:
$62 million
110 MW:
$136 million
220 MW: $278 million

These costs will be applied to the development schedule in Table A3.1, making the following
assumptions:
146
The number of wells is based on the project capacities ignoring binary plants.
Projects with a total capacity of 10 MW or less are omitted.
Cost of Exploration147
The total MW excludes binary plants that could add about 15% to the MW total.
Exploration costs are assumed to be spent an appropriate amount of time before power
plant commissioning, i.e., for example stage 3 costs will be complete at least 2 years before
commissioning.
Where it is known that certain milestones have already been achieved (e.g., stage 2 complete),
that is taken into account.
Where projects are undertaken in a number of small stages, it is assumed that only limited
further exploration is required for the later stages.
This excludes projects scheduled for development but considered unfeasible.
Therefore the true total cost could be higher, assuming someone does decide to explore
those.
No preinvestment for projects that will be commissioned after 2020 is included.
MW for projects that have already had their exploration completed but which are not yet
completed (e.g., Ulubelu 3 and 4) are included.
Table A6.1 reflects the additional expenditures required from now until financial closure (i.e., funding
which cannot be obtained from conventional commercial sources), together with the MWs achievable
by certain milestone dates.
Table A6.1: Expenditures Required through Financial Closure and Milestone MWs
Year
2014
2015
2016
2017
2018
2019
2020
Total
Expenditure ($ million)
434
704
799
609
322
2,869
40
40
273
993
1,633
2,408
3,008
3,008
Cumulative new MW
MW = megawatt.
Appendix 7
The Geothermal Tariff Issuance

of June 2014
The recommendations on the mechanism of ceiling tariff and tender process improvements were provided
to the Directorate General of New and Renewable Energy and Energy Conservation of the Ministry of
Energy and Mineral Resource (MEMR) as Tariff Methodology Report in April 2014. After several follow-up
meetings between the staff of MEMR, geothermal stakeholders, and the Asian Development Bank (ADB)
World Bank team, MEMR issued a new regulation (MEMR Regulation No. 17, 2014) in June 2014 to replace
the previous regime of FITs with a competitive tender scheme with benefit-based ceiling prices. This
change will likely be a major step forward in improving the Indonesian geothermal regulatory framework.
Key Provisions of Ministry of Energy and

Mineral Resources Regulation No. 17/2014
Ceiling Prices Based on Benefits
The new regulation sets the ceiling prices by region and target commercial operation date (COD)
year. Three regions are determined based on main generation sources. Region 1 consists of Sumatra,
Java, and Bali, in which geothermal would replace power from large coal-fired power plants. Region 2
includes other areas where small coal-fired power plants are planned to be the main source of power
such as Sulawesi, West Nusa Tenggara, East Nusa Tenggara, etc. Region 3 covers any areas where
isolated diesel generation is the primary source of power.
The values of the new ceiling prices are shown in Table A7.1. The prices are slightly lower than the
recommended figures in this report due to adjustments on some of the parameter values by MEMR.
However, these prices appear to be in line with the calculation of avoided costs in the tariff methodology
report by ADBWorld Bank.
Table A7.1: New Ceiling Prices
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
Region 1
11.8
12.2
12.6
13.0
13.4
13.8
14.2
14.6
15.0
15.5
15.9
Region 2
17.0
17.6
18.2
18.8
19.4
20.0
20.6
21.3
21.9
22.6
23.3
Source: Government of Indonesia, Ministry of Energy and Mineral Resources. Ministerial Regulation No. 17/2014.
148
Region 3
25.4
25.8
26.2
26.6
27.0
27.4
27.8
28.3
28.7
29.2
29.6
The Geothermal Tariff Issuance of June 2014 149
Provision of Model Power Purchase Agreements at Tender

The new regulation also requires PLN to provide a model power purchase agreement (PPA) at the
tender stage. In the previous regulations, the role of PLN in tender was not clear. In some cases, PLN
was expressly excluded from the process by an MEMR instruction.
Revocation of Izin Usaha Pertambangan (Mining Business Permit)

Without Progress
The regulation requires existing Izin Usaha Pertambangan (mining business permit) (IUP) holders
to complete their PPA negotiation processes with PLN and preparation for exploration between
August and December, depending on the progress of licensing processes in each field. For example,
for an existing IUP holder that has obtained the MEMR approval on the selling price but cannot
conclude a PPA with PLN by August 2014, its IUP will be revoked and the corresponding wilayah kerja
pertambangan (geothermal work area) (WKP) will be retendered.
Key Differences Between Ministry of Energy and Mineral

Resources Regulation No. 17/2014 and ADBWorld Bank
Recommendations
Updating Mechanism for Ceiling Prices
The ADBWorld Bank team recommended that the new scheme incorporate a mechanism to update
the values of ceiling prices annually. The need for updates arises as parameters from the international
energy markets such as coal prices may change significantly over time. MEMRs regulation makes no
provision for such updating.
Transparent Mechanism for Price Setting

While the regulation does not preclude seeking access to the methodology used to set the ceiling
prices, the ADBWorld Bank team recommended that the methodology should be transparent so
that the governments assumptions and ceiling prices are well justified. It would help if the government
posted the methodology report on MEMRs website.
Clarity on the Role of the Geothermal Fund in Tender Processes

There is no indication of the mechanism to incorporate the use of the Geothermal Fund in the
new regulation. If the recovery cost for the exploration by the Geothermal Fund is clearly stated as
bidders obligation in tender documents, the ceiling prices may not need adjustment, as developers
can incorporate the cost in their price calculations. However, such requirements for central and local
governments at the tender stage are not clarified in the regulation. In our previous discussions, MEMR
plans to clarify the role of Geothermal Fund in the revision of government and ministerial regulations
on the tender mechanism (PP59/2007 and MEMR 11/2008) as soon as the geothermal law revision
is approved.
150APPENDIX 7
Central Tender Authority

The MEMR Regulation No. 17/2014 is based on the tender mechanism by central or local governments.
However, the new bill on geothermal has shifted this authority to the central government. While this
change is in line with our recommendation, technical assistance will likely be needed to ensure the
quality of tender documents and evaluation processes.
Transparent Escalation Mechanism

The current escalation mechanism is determined at the PPA negotiation stage between the winning
bidder and PLN. This ad hoc process has led to time-consuming negotiations. Developers are also
left with uncertainties over the escalation mechanism until after they submit their bids. Therefore, a
transparent mechanism set by the government would facilitate this negotiation process and accelerate
the conclusion of PPAs for both PLN and developers.
Appendix 8
The New Geothermal Law 2014

On 26 August 2014, Indonesias House of Representatives passed the Bill on Geothermal Energy as a
revision to the current Geothermal Law No. 27 of 2003. The bill provides improvements to some of
the issues that have hindered geothermal projects in Indonesia, namely:

distribution of authority of government institutions over direct and indirect use of geothermal
resources;
licensing procedures; and
forestry issues in geothermal development.
The bill distinguishes between the direct use (for example, tourism, industry, and agribusiness) of
geothermal resources and their indirect use (electricity generation). While the 2003 Geothermal
Law assigns authority over geothermal resources for both direct and indirect use to central and/or
provincial and/or regency governments, the 2014 bill assigns authority for licensing indirect use of
geothermal resources to the central government only, represented by MEMR. However, the authority
for direct use of geothermal resource remains the same as the 2003 law, and is held by central and/or
provincial and/or regency government based on its location.
The new arrangement for government authority over the indirect use of geothermal resource will
allow MEMR to conduct all geothermal concession tenders. While a new government regulation is
required to set out the new requirements for the tender mechanism, centralized tender management
is in line with the recommendations in this report. The interest of local governments in geothermal
development is secured by a new production bonus, starting on the date of commercial operation,
levied in addition to any applicable local taxes.
The 2003 law stipulated only that the Minister of MEMR could conduct exploration, but did not
specify whether any outside entities could be assigned to conduct exploration. The new bill states
that the Minister of MEMR can appoint other entities to conduct exploration. Again, while the details
are left to the new implementation regulation, this change may provide resolution to the issues of the
exploration authority for the Geothermal Fund.
The licenses of geothermal development consist of a Geothermal License (for indirect use and
issued by central government) and a Direct Use License (issued by central or provincial or regency
government).

The Geothermal License for electricity generation may last up to a total of 37 years
(a maximum of 7 years for exploration, which includes feasibility study preparation, and
30 years for exploitation). Prior to the expiration of license, developers may apply for an
extension of a maximum of 20 years.
151
152APPENDIX 8

When the bill officially becomes law, the legacy geothermal concessions will be valid for
30years from the date of the enactment. The existing concessions (IUPs) under Geothermal
Law 2003 and joint operation contracts will also be valid until the expiration of the license or
the joint operation contract.
The declassification of geothermal as mining activity allows greater latitude for the geothermal
development in the protected and conservation forests. Developers are required to secure borrow-touse permits if the wilayah kerja pertambangan (geothermal work area) is located in a production forest
or protected forest, or a forestry utilization permit if the work area is located in a conservation forest.
Unlocking Indonesia's Geothermal Potential

This report was produced jointly by the Asian Development Bank and the World Bank and is based on a
series of technical assistance activities conducted during 2013-2014. The study documents key issues that
have constrained the development of Indonesia's geothermal power development sector, including tariffs,
tendering processes, financial considerations, permitting, and inter-agency coordination. The report then
makes a set of comprehensive recommendations to unlock the potential of the sector, including a new tariff
regime, improvements to the tendering process, re-negotiation of power purchase agreements, and innovative modes of financing and project de-risking.
About the Asian Development Bank

ADBs vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member
countries reduce poverty and improve the quality of life of their people. Despite the regions many successes,
it remains home to two-thirds of the worlds poor: 1.6 billion people who live on less than $2 a day, with
733 million struggling on less than $1.25 a day. ADB is committed to reducing poverty through inclusive
economic growth, environmentally sustainable growth, and regional integration.
Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for
helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants,
and technical assistance.
About the World Bank
The World Bank Group (also known as the Bank Group) is the largest anti-poverty institution in the world,
offering loans, advice, knowledge, and an array of customized resources to more than 100 developing
countries and countries in transition. Established in 1944 and headquartered in Washington DC, the Bank
Group is a specialized agency of the United Nations that is made up of 188 member countries. It works with
country governments, the private sector, civil society organizations (CSOs), regional development banks,
think tanks, and other international institutions on a range of issuesfrom climate change, conflict, and food
crises to education, agriculture, finance, and tradein its efforts to accomplish two goals: end extreme
poverty by 2030 and boost shared prosperity for the bottom 40 percent of the population in all developing
countries.
THE WORLD BANK

1818 H Street NW
Washington, DC 20433
United States
www.worldbank.org
ASIAN DEVELOPMENT BANK
6 ADB Avenue, Mandaluyong City
1550 Metro Manila, Philippines
www.adb.org

Unlocking Indonesias Geothermal Potential

Uploaded by

Copyright:

Available Formats

Unlocking Indonesias Geothermal Potential

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Unlocking Indonesias Geothermal Potential

Uploaded by

Copyright:

Available Formats

UNLOCKING

ASIAN DEVELOPMENT BANK

Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO)

Asian Development Bank

I. Asian Development Bank.

List of Stakeholder Comments

Tariff Ceilings and Escalation

4 Tariff Ceilings Based on Benefits

Premium for Price Volatility

Power Purchase Agreements

6 Recovery of Incremental Costs

Cost of Existing Projects

8 The Tariff Impacts of Front-End De-risking

Cost Recovery Options for the Fund

9 Institutional and Financing Issues

9.1 Structural Issues

Unlocking the Sector: Study Team Recommendations

10.1 General Tariff Approach 

Power Purchase Agreement Renegotiation Policy

2 Transmission Interconnection Costs

4 Tariff Structure and Project Finance

Possible Technology Paths to Faster Development

Use of Ancillary Plant 

7 The Geothermal Tariff Issuance of June 2014

List of Tables, Figures, and Boxes

List of Tables, Figures, and Boxesvii

A3.2 Revised Targets

viiilist of tables, figures, and boxes

List of Stakeholder Comments

Greenhouse Gas Valuation

Why Use Deterministic and Not Probabilistic Calculations? 

The Tender Process

Use of the Geothermal Fund

A1.1 Use of International Code

geology, geophysics, and geochemistry

Asian Development Bank

African Rift Geothermal Development Program

binary bottoming plant

back pressure plants

clean coal technologies

Certified Emission Reduction

commercial operation date

Clean Technology Fund

debt service coverage ratio

financial internal rate of return

Fast Track Program

Fast Track Program 1

Fast Track Program 2

Geo Dipa Energi

gross domestic product

Geothermal Energy Development Program

global social cost of carbon

heavy fuel oil

International Energy Agency

International Finance Corporation

international financial institution (Asian Development Bank, World Bank)

Indonesia Geothermal Association

independent power producer

Izin Usaha Pertambangan (Mining Business Permit)

Japan Bank for International Cooperation

List of Stakeholder Comments

Tariff Ceilings and Escalation

4 Tariff Ceilings Based on Benefits

Premium for Price Volatility

Power Purchase Agreements

6 Recovery of Incremental Costs

Cost of Existing Projects

8 The Tariff Impacts of Front-End De-risking

Cost Recovery Options for the Fund

9 Institutional and Financing Issues

9.1 Structural Issues

Unlocking the Sector: Study Team Recommendations

10.1 General Tariff Approach

Power Purchase Agreement Renegotiation Policy

2 Transmission Interconnection Costs

4 Tariff Structure and Project Finance

Possible Technology Paths to Faster Development

Use of Ancillary Plant

7 The Geothermal Tariff Issuance of June 2014

A3.2 Revised Targets

Greenhouse Gas Valuation

Why Use Deterministic and Not Probabilistic Calculations?

The Tender Process

Use of the Geothermal Fund

A1.1 Use of International Code

2.4 Production Costs or Benefits as a Basis for