Energy for Sustainable Development 68 (2022) 350–363

Contents lists available at ScienceDirect

Energy for Sustainable Development

Coal vs. renewables: Least-cost optimization of the Indonesian

power sector
Jose Antonio Ordonez a,b,⁎, Marek Fritz a, Johannes Eckstein a
a
Fraunhofer Institute for Systems and Innovation Research (ISI), Breslauer Str. 48, 76139 Karlsruhe, Germany
b
Mercator Research Institute on Global Commons and Climate Change, EUREF-Campus 19, 10829 Berlin, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Indonesia, Southeast Asia's largest economy, ranks globally among the top countries in terms of the envisioned
Received 3 February 2022 expansion of coal-fired power plants. Despite strongly falling costs for key renewable energies in recent decades,
Revised 19 April 2022 most outstandingly solar PV and wind power, their role in Indonesia's power sector planning is negligible. To as-
Accepted 30 April 2022
sess the potential role of renewable energies in Indonesia's power sector, we develop a cost optimization power
Available online xxxx
sector model. Based on four different scenarios assuming a comparatively slow and a comparatively rapid de-
Keywords:
crease in the costs for renewable energies, each with and without carbon pricing, we assess capacity expansion,
Coal: Energy system modelling electricity generation, resulting CO2 emissions, and total system costs until 2040. We compare our results to
Variable renewable energy Indonesia's utility latest power expansion business plans RUPTL, as well as to the country's energy master plan
Cost optimization RUEN. We find that official power sector development plans would lead to comparably higher electricity gener-
Carbon pricing ation costs, due to overcapacity and negligence of future least-cost technologies, especially solar PV. Carbon pric-
Renewable energies ing as low as 5 USD per ton of CO2 would make coal an economically unviable alternative and foster the
Indonesia integration of biomass and geothermal power plants. In the context of rapidly decreasing costs for renewables,
Climate change mitigation
solar PV would be cost competitive with coal even in the absence of carbon pricing by the middle of the
2020–2030 decade. Wind power remains largely uncompetitive across the whole time horizon, due to the lack
of substantial wind resources. Our results highlight that Indonesia's power sector planning could be substantially
improved, from both a cost and a climate protection perspective.
© 2022 The Authors. Published by Elsevier Inc. on behalf of International Energy Initiative. This is an open access
article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Introduction power capacities in Indonesia (Enerdata, 2021), while the remainder is

provided by hydro and geothermal power plants. Coal plays a dominant
Indonesia is Southeast Asia's largest economy and the 4th most pop- role, accounting for more than half of installed power capacities,
ulous country in the world. As a signatory to the Paris Agreement, followed by natural gas, with around one third. With 33 GW of coal ca-
Indonesia has committed to the goal of limiting the global temperature pacities under active development, the country ranks 3rd globally after
increase to well below 2 °C above pre-industrial levels and to pursue ef- China and India in terms of the future expansion of coal-fired power
forts to limit the increase to 1.5 °C. Achieving this goal will require a plants (Global Energy Monitor, 2021). The large expansion of coal-
carbon-neutral global economy by the second half of the century, for fired power plants in recent years has led to substantially increasing
which the global phase-out of coal use without carbon capture and stor- CO2 emissions from electricity generation, which shows the steepest
age in the power sector is a prerequisite (Luderer et al., 2018; Tong et al., rise in emissions across all sectors (WRI, 2021).
2019). Recently, global electricity markets have seen a surge in renewable
Strong economic and demographic growth in Indonesia has led to energies (RE), most outstandingly solar PV and wind power, as econom-
rising electricity demand over the last decade, which has increased at ically viable and carbon-free electricity sources. Globally, solar PV capac-
an annual rate of 6% (ESDM, 2021). To date, this demand has been ity increased 17-fold from 2010 to 2020, and onshore wind capacity
largely covered by fossil fuels, which represent over 90% of the installed increased fourfold (IRENA, 2021a). The global weighted-average
levelized cost of electricity (LCOE) for utility-scale solar PV electricity
generation fell by 85%, and by 56% for onshore wind, bringing these
RE into a competitive range with fossil fuels (IRENA, 2021b). Despite
⁎ Corresponding author at: Fraunhofer Institute for Systems and Innovation Research
this global development on the electricity markets, Indonesia has not
ISI, Breslauer Str. 48, 76139 Karlsruhe, Germany. yet integrated substantial shares of solar PV and wind power.
E-mail address: [email protected] (J.A. Ordonez). Indonesia's first wind power plant was inaugurated in the Sidenreng

https://doi.org/10.1016/j.esd.2022.04.017
0973-0826/© 2022 The Authors. Published by Elsevier Inc. on behalf of International Energy Initiative. This is an open access article under the CC BY license (http://creativecommons.org/
licenses/by/4.0/).
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

Rappang Regency in 2018, and its total installed wind capacity is cur- drivers of its continued use (Ordonez et al., 2021; Ordonez & Eckstein,
rently around 150 MW. Similarly, its total installed solar PV capacity 2020). Handayani et al. (2017) modelled different scenarios based on
is around 160 MW, with the first larger utility-scale projects coming both simulation and optimization in LEAP, without specifically focusing
online in recent years (ESDM, 2021). Indonesia's latest electricity on renewable cost reductions. More recently, some studies have
utility business plan RUPTL 2021–2030, which is Indonesia's most in- analysed how to reach carbon neutrality using modelling tools (IESR,
dicative 10-year power sector development plan, foresees the addi- 2019, 2021a, 2021b, 2021c). Our study aims to help fill this gap by
tional development of 14 GW1 coal power plants in the next analysing the cost-effectiveness of Indonesia's power sector develop-
decade, while the share of variable renewable energies, solar PV ment under different future assumptions.
and wind, remains small (PLN, 2021). Indonesian stakeholders re- This paper is structured as follows. Section 2 describes the ratio-
port that solar PV and wind projects are non-competitive with coal, nale behind the considered scenarios. Section 3 presents the model-
in particular due to the prevalence of double-digit interest rates for ling methodology. Section 4 shows the results of the future capacity
capital-intensive renewable projects (Ordonez et al., 2021; composition in different scenarios, how this compares to policy
Ordonez & Eckstein, 2020). plans, the role of key technologies, and associated CO 2 emissions
In contrast to the trend in Indonesia, many other countries previ- and total system costs under each scenario. Section 5 discusses key
ously relying strongly on coal, such as China, India or Vietnam, have findings in an integrated perspective with an outlook to policy
decidedly integrated variable renewable energies to their power sec- implications. Section 6 concludes.
tors in recent years, thereby reducing the share of coal-fired power
plants in their power development plans in favour of renewables Scenario assumptions under falling LCOE
(Enerdata, 2021; Global Energy Monitor, 2021). While Indonesia's
state-owned utility PLN has recently pledged to strive for carbon We developed four different scenarios assuming (a) moderate
neutrality by mid-century, no official targets or plans to achieve and (b) comparably stronger cost reductions for renewable energies,
this exist (Rahman, 2021). Indonesia's climate target is to reduce each (1) with and (2) without carbon pricing (see Table 1). Scenarios
greenhouse gases by 29%–41% by 2030 compared to a Business as (a) and (b) assumed different reductions for both total upfront cap-
Usual scenario (see Fig. A1 in the Annex A), while it aims to reach a ital costs and financing costs of electricity generation technologies.
share of 23% of RE in the power sector by 2025 (Bridle et al., 2018). RE projects are capital intensive and associated with high upfront
Most recently, the role of carbon taxes to reach Indonesia's energy expenditures, such as costs of generation equipment, balance of
and climate targets has been discussed, with the Government of plant components and installation (Steffen, 2020). Thus, total up-
Indonesia reportedly being in the process of drafting an amendment front costs, while driven to a large extent by the globally falling
to Indonesia's tax law to incorporate carbon taxes (Indonesian Law costs of key generation components, such as solar PV cells, are also
Blog, 2021; Rahman, 2021). a function of local conditions (IRENA, 2020). Indonesia's ‘local con-
In light of the contradiction of relying on coal in the context of tent requirements’, which mandate that 44% of the total value of
plummeting costs of renewables, our study aims at analysing the inte- components in solar PV projects are manufactured locally, further
gration of renewable energy technologies into Indonesia's power sector. disconnects the country from globally reported cost reductions
This paper investigates cost-optimized pathways to integrating higher (GBG Indonesia, 2021). To a large extent, financing costs, which are
shares of RE into the Indonesian power system. It aims to answer the determined by the interest rates of commercial debt and equity to fi-
following research questions: 1) Are Indonesia's official power sector nance upfront expenditures, reflect local risks associated with the
development plans cost-efficient? 2) What role could solar PV and project as a function of regulatory, technical, political, and other
wind power play in Indonesia's power sector given their strong cost risk factors (Feyen & Huertas, 2020). The high upfront expenditures,
reductions? 3) What effect would different levels of carbon pricing i.e. the capital intensity of renewable projects, make them particu-
have on Indonesia's power sector composition? 4) What impacts larly sensitive to interest rates (Egli et al., 2019; Egli, 2020). While
would the falling costs of RE and the introduction of carbon Indonesian stakeholders report double-digit interest rates for the
pricing have on CO2 emissions and electricity generation costs? By development of RE projects, typical commercial debt interests as re-
employing a power sector simulation and cost optimization energy ported by the Internationally Monetary Fund are around 10% per
model, we systematically explore the different future power sector year (IMF, 2022).
compositions, its associated CO2 emissions and electricity generation Table 1 presents an overview of the assumptions underlying the
costs. Technically, we use the Low Emissions Analysis Platform (LEAP, scenario rationale. The baseline scenario assumes moderate cost re-
Heaps (2021)), which has also been applied to Indonesia in numerous ductions for solar PV and wind energy costs. These cost reductions
other studies (DEN, 2019). We examine four different power develop- could be driven by the globally falling manufacturing costs for
ment scenarios from a cost optimization perspective, as well as two of- main technology components (IRENA, 2020, 2021b), yet hampered
ficial Indonesian power development plans. We systematically assess by the local content requirements in place. With regard to financing
capacity expansion, electricity generation, the resulting CO2 emissions, costs, the baseline scenario assumes that these will remain in a
and total system costs until 2040, and draw conclusions from the double-digit range, reflecting the general country risk and other fac-
comparison. tors, e.g. constantly changing regulations in the power sector or the
There are very few scientific studies available that analyse the eco- reportedly politically-driven aversion of the monopolistic utility to-
nomic efficiency of Indonesia's power sector. Previous literature exam- wards renewables (Ordonez et al., 2021; Ordonez & Eckstein,
ined the drivers for the development of the power sector from a political 2020). The assumption of prevalent high financing costs is relaxed
economy perspective, concluding that coal's cost competitiveness is in the RE cost reduction scenario, which features decreasing financing
better able to ensure sustained low electricity prices while supporting cost rates until 2025 (reaching parity with those of fossil fuels),
PLN's financial health. Coal-related royalties as a source of public fi- while capital costs are also assumed to experience a stronger decline,
nance, the role of coal in regional economic development and the vested reaching comparable levels to India by 2025 (with figures for India
interests of key political functionaries owning coal assets are the major derived from IEA (2020c)). India is taken as the reference country,
as the socioeconomics and power sector structure of many of
India's utilities are arguably comparable to those in Indonesia (IEA,
1
The total coal-fired capacities under active development are larger than this figure
2020c). We also examined the impact of CO 2 prices on the
(approx. 33 GW), as they comprise newly announced capacities, as well as capacities in electricity sector development for the two previously defined
the pre-construction and construction phases (Global Energy Monitor, 2021). scenarios, that is, in the context of moderate cost reductions for

351
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

Table 1
Overview of main assumptions for the cost-optimized scenarios (own elaboration).

Scenario Rationale Capital costs for solar PV and wind Financing cost Carbon pricing
name turbines

Baseline Moderate cost reduction for Moderate reduction of 3% per year Prevalent high financing cost rates (10% variable RE, 8% other RE, None
renewables 7% for fossil technologies)
CO2-price CO2 price and moderate cost Moderate reduction of 3% per year Prevalent high financing cost rates (10% variable RE, 8% other RE, USD 30 per
reduction for renewables 7% for fossil technologies) ton of CO2
RE cost Strong cost reduction for renewables Decreasing, reaching parity with Decreasing financing cost rates for all renewables, reaching None
reduction Indian levels by 2025 parity with fossil fuels (7%) in 2025
Combined CO2 price and strong cost reduction Decreasing, reaching parity with Decreasing financing cost rates for all renewables, reaching USD 30 per
policies for renewables Indian levels by 2025 parity with fossil fuels (7%) in 2025 ton of CO2

renewables (baseline scenario) and in the context of strong cost (RUPTL until 2028 and 2030, RUEN until 2050), while electricity gen-
reductions (RE cost reduction scenario). In the CO2 price scenario, we eration, costs and emissions are calculated endogenously based on
assumed carbon pricing is introduced in the form a CO2 tax of USD the lowest operational costs of the installed capacity.
30 per ton of CO 2 under the baseline scenario conditions. In the In the cost-optimized modelling environment, the LCOE of a given
combined policies scenario, we likewise assumed the introduction technology are, under the specified technical constraints, the most rele-
of a CO2 price of USD 30 per ton of CO2 but under the RE cost reduc- vant factor in determining the optimization objective function (see sub-
tion scenario conditions. Accordingly, cost reductions for variable re- sequent methodology section). The LCOE are thus illustrative to
newable energies are equivalent in the baseline and the CO2 price understand the decision logic of the model for key technologies and dif-
scenario. Likewise, the stronger cost reductions for renewables are ferent years and scenarios. Fig. 1 shows the LCOE composition of solar
the same in the RE cost reduction scenario and in the combined pol- PV and coal for different years and scenarios. Because of high financing
icies scenario. cost rates in Indonesia, the share of financing costs is almost 55% for the
In addition to the four scenarios, our results are compared to the LCOE of solar PV in 2018. Based on the assumptions presented in Table 1
two most recent utilities' power sector capacity development plans and Table 2 the LCOE of solar PV would decrease by 18% in 2025 from
by the utility, RUPTL 2019–2028 and RUPTL 2021–2030, (PLN, 2018 levels in the baseline scenario, as well as in the CO2 price
2019, 2021), as well as to the capacity expansion envisioned in the scenario, and by 61%in the RE cost reduction scenario and combined
long-term energy master plan RUEN by Indonesia's national energy policies scenario for the same year. For 2040, the LCOE of solar PV
council (DEN, 2017a). RUPTL is the utility business plan and indica- would experience a 46% reduction in the baseline and CO2 price
tive of future installations, while RUEN is a long-term strategy docu- scenario, and 73% reduction in the other scenarios. While coal is also
ment, which serves as the overarching reference for energy sector assumed to experience a cost reduction over the examined horizon,
planning in Indonesia (Ordonez & Eckstein, 2020). For these power this technology is mature and cost reductions negligible (approx. 2%
development plans, the future capacity expansion is exogenous to reduction between 2018 and 2040). Fig. 1 thus represents coal with
the model and corresponds to the specifications of the plans uniform LCOE.

120

100 6
Total costs: -18%
8 Capital costs:-49% Coal: 32% lower
Financing costs:-49% (CO2 price)
80 6 Total costs: -46%
6 Coal: 56% lower
USD / MWh

(no CO2 price)

60 56 Total costs: -61% Capital costs:-71%
Financing costs:-81% 24
6 Total costs :-73%
45 4
40
6
4 29 23
6
16 2
20 6
32 10
26
9
14 16
9 6
0
Solar PV Solar PV Solar PV Solar PV Solar PV Coal
2018 2025 2025 2040 2040 2018-2040

Historical Baseline scenario RE cost reduction Baseline scenario RE cost reduction All scenarios
scenario scenario

Capital costs Financing costs Annual fixed costs Operation costs Carbon price USD 30 per ton

Fig. 1. LCOE of solar PV for the historical year 2018, as well as under the baseline scenario (and the CO2 price scenario) and RE cost reduction scenario (and the combined policies scenario)
for 2025 and 2040, as well as for coal for all scenarios and years. Numbers in the bars indicate costs for the respective component in USD / MWh, while bubbles indicate relative cost
reductions or comparison to 2018 solar PV LCOE.

352
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

Table 2
Economic and technical parameters of different power generation categories in 2018 and 2040; BAS=Baseline Scenario, RE = RE cost reduction scenario, (source: own compilation based
on (A) DEN (2017b), (B) Kalirajan and Syed (2019) (C) IRENA (2020) (D) Cole and Frazier (2019), (E) PricewaterhouseCoopers (2018), (F) ACE (2019), (G) IEA (2020a), (H) BPS (2020),
(I) Mone and Hand (2017), (J) DTU (2021), (K) ESMAP (2020).

Power plant type Capital costs (USD/kW) Fixed O&M costs Variable O&M costs Efficiency Lifetime Fuel Maximum
in 2018/2040 (USD/kW/year) in (USD/MWh) in (years) costs capacity
2018/2040 2018/2040 (USD/GJ) factor

Coal 1400/1340 A,B 41.2/39.3 A,B 0.12/0.12A,B 37% A,B 30 A 2.4 G 80% A, B
Open cycle gas turbine (OCGT) 770/700 A,B 23/22.5 A,B 0 A,B 33% A,B 25 A 6.2E 95% A, B
Closed cycle gas turbine 750/680 A,B 23/22.5 A,B 0.13/0.13 A,B 56% A,B 25 A 6.2 E 95% A, B
(CCGT)
Large hydro 2000/2000 A,B,C,D 38/34.7 A,B 0.65/0.6 A,B – 50 A – 36% H
Small hydro 2600/2600 A 53/48.8 A 0.5/0.46 A – 50 A – 50% C
Geothermal 3700/3050 A,B,C,D 18/15.9 A,B 0.25/0.22 A,B – 30 A – 80% A, B
Biomass 1700/1500 A,B,C,D 48/40.8 A,B 3/2.59A,B 31%A 25 A 1.3 F 80% A, B
Solar PV 1190/BAS: 610 RE:340C,D 11.9/BAS: 6.1 0 A,B – 25 A – 17% K
RE: 3.4C,D
Wind 1500/BAS: 1210 RE: 1020 30/26.5 A,B 0 A,B – 27 A – 25% I, J
A,B,C,D

Diesel 800/800 A,B 8/8 A,B 6.4/6.4 A,B 45% A 25 A 15.9 G 80% A, B
Pumped hydro storage 860/860 A 8/8 A 1.3/1.3 A 80% A 35 A – 100%
Battery storage 1570/730 D 39.3/18.2 D 0D 85% D 15 D – 100%

Methodology

This section presents a general overview of the modelling methodology. The model was implemented in the Low Emissions Analysis Platform
(LEAP), while the least-cost optimization was computed using the Next Energy Modelling system for Optimization (NEMO), which was directly
linked to LEAP (Heaps, 2021; Veysey et al., 2021)2.
Starting in 2019, the model determines the least-cost capacity expansion and electricity generation for each year until 2040 for the different elec-
tricity generation technologies (Table 2). We limit our analysis until 2040 given that the energy system development in next two decades are crucial
for the achievement of the goals of the Paris Agreement (Cf. (Luderer et al., 2018)), and the fact that break-even points in the cost-competitiveness of
renewables against coal are expected to occur in this timeframe (Fig. 1). Electricity generation is determined for 24 hourly time steps in two seasons
(wet and dry) each year. This results in 48 time steps per year, corresponding to 1056 time steps throughout the modelling time horizon.
The modelling is split into five regions (Java-Bali, Sumatra, Sulawesi, Kalimantan, and Papua-Maluku) to account for differences in the load curves,
regional renewable capacity limits for each technology, as well as annual and daily RE variations. A reference load curve is taken from Huda et al.
(2018) for Java-Bali, and from Winarko et al. (2019) for all other regions. For each region, intraday and seasonal patterns for solar PV and wind
power output data are compiled from the Global Solar Atlas (Solargis, 2020) and Renewables Ninja (Staffell & Pfenninger, 2016). Given the lack of
interconnection lines between the regions, electricity exchange between regions is not considered.
The growth in annual electricity demand for all scenarios is taken from RUPTL 2019–2028 (PLN, 2019). The growth predicted in RUPTL 2019
(close to 6%) is regarded as a realistic compromise, because it is significantly lower than Indonesia's energy masterplan RUEN (8%), yet higher
than the low growth demand projection in the latest business plan of the utility, RUPTL 2021–2030 (4.9%). (KESDM, 2019). Fig. A2 in Annex A pre-
sents a visualization and comparison of different demand projections. Additional consumption caused by transmission and distribution losses is im-
plemented as a percentage of electricity demand, and future transmission and distribution losses are based on the projection in the national
electricity plan RUKN (KESDM, 2019).
Integrating variable RE into the power system causes additional costs elsewhere in the system (Hirth & Steckel, 2016). These integration costs can
be categorized as balancing costs, grid expansion costs, backup costs, full-load hour reduction costs, and overproduction costs (Ueckerdt et al., 2013).
Backup costs, full-load hour reduction costs, and overproduction costs are considered endogenously by the model setup. Transmission and balancing
costs are specified as additional operational costs. Based on IEA (2018) and Heptonstall and Gross (2020), we integrated these costs with reference
values of 6 USD/MWh for solar PV and 4 USD/MWh for wind.
The model also considers battery and pumped hydro storages, which are associated with cost parameters corresponding to the generation tech-
nologies. Pumped hydro storage is considered to be off-river storage, for which ample sites exist in all regions of Indonesia (Stocks et al., 2019). For
pumped hydro and large-scale battery storage3, the associated costs are subject to relatively high uncertainty. Large hydro is considered to be hydro
power produced using dams, which is constrained in the model to supply electricity in the evening only, while small hydropower plants are assumed
to be run-of-the-river plants, implemented without a daily cycle.
In order to calculate the least-cost capacity expansion, the minimum total net present value of all annualized costs of the electricity system over
the entire simulated time period is determined by the objective function (Eq. 1).
( " #)
1
obj ¼ min ∑ y  yb ∑½annualized capital costsþ fixed costs∗capacityPGT,y þ operational costs∗∑generationPGT,y,t  ð1Þ
y ð1 þ dÞ PGT t

2
While the aforementioned sources Heaps, 2021; Veysey et al., 2021 provide an in-depth overview of the modelling methodology and this section elaborates the key modelling assump-
tions and parameters, additional information can be supplied on request.
3
Of note, the costs for batteries and pumped hydro storage are likewise dependant on the specific capacity costs of each technology (see IRENA, 2017 for a comprehensive overview on
costs of electricity storage technologies. Moreover, the values presented in Table 1 are reference values that might vary from source to source.

353
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

In which:

y ¼ year;yb ¼ base year

PGT ¼ power generation technology

t ¼ timestep

d ¼ discount rate

The net present value of total costs for a given technology was computed based on exogenously determined cost components, classified as
upfront capital costs, fixed yearly costs, and operational costs (see Table 2 for an overview of relevant cost assumptions). Conventional oper-
ational cost parameters are fuel costs and maintenance costs resulting directly from operation. Carbon pricing costs and variable RE integration
costs were also modelled as operational costs. No retirement costs or compensation payments for early retirements are considered. The endog-
enous model variables are the capacity expansion in each year and electricity generation in each time step, subject to technical constraints on
demand and supply balance, availability of power generators, reliable capacity reserve margins, regional renewable capacity limits, and stor-
age constraints.
In the model, the output of a power generation category is constrained only to its available capacity in the respective time step and
ramping, i.e. differences between time steps are not limited. This is, however, a technical constraint of real-world applications which also
leads to additional costs. These costs are implicitly covered by the model by allocating them to variable RE as integration costs. Incorporating
these ramping costs and constraints explicitly into optimization modelling could improve the model in the future once these are available in
LEAP.
The model does not represent individual power plants but average patterns of seasonal variations, demand and RE potentials. Therefore, the
results should be interpreted as an indication of the cost-optimal dispatch of an average unit rather than a specific single unit. Given the
ongoing research and respective uncertainties in the values related to storage facilities, it will be essential to reassess this in the future
using updates.

Results Capacity

This chapter presents the results of the cost-optimized scenarios. We Cost-optimal capacity development up to 2040
first discuss capacity expansion capacities in the cost-optimized scenar- Fig. 2 shows the capacity development for all official power develop-
ios up to 2040, the end of the modelling period, and in comparison to ment plans considered and the cost-optimized scenarios for selected
Indonesia's latest power sector expansion plans RUPTL (PLN, 2019, years. The corresponding values are provided in Table A3 in Annex A.
2021). In view of its prominent role in several scenarios, solar PV capac- In 2020, Indonesia had a total installed power capacity of 70 GW,
ity development is examined in more detail. A sensitivity analysis ex- strongly determined by coal-fired power plants (51%, 37 GW), followed
plored the influence of different CO2 price levels on power capacity by natural gas (21%, 19 GW), and 9% diesel generators (6 GW). Together,
composition in 2040. We then present hourly power generation fossil fuels accounted for over 90% of installed capacities, with the re-
profiles by scenario in 2040, and compare system costs, resulting CO2 mainder being mainly hydro (5 GW) and geothermal power plants (2
emissions and carbon revenues. GW).

300

250
Installed capacity in GW

200

150

100

50

0
2020 2028* 2030 2030 2040 2030 2040 2030 2040 2030 2040 2030 2040
Historical RUPTL RUPTL RUEN Baseline RE cost reduction CO2 Price Combined policies
2019-28 2021-30

Coal CCGT OCGT Diesel Geothermal Biomass Wind Large Hydro Solar PV Pumped Hydro Storage

Fig. 2. Installed capacities in 2020, 2030 and 2040 for official power sector plans and cost-optimized scenarios. (*) RUPTL 2019–2028 is shown until 2028, as this is the last year of the plan.

354
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

RUPTL 2019 and RUPTL 2021 assume that the strong prevalence of Total capacity in RUPTL (121 GW in 2028 for RUPTL 2019 and 119
fossil fuels continues up to the end of the decade. In RUPTL 2019, coal GW in 2030 for RUPTL 2021 and the RUEN (estimated 159 GW in
is planned to retain its dominant role in the mix, growing to 60 GW 2030)) is significantly above the baseline scenario, with 93 GW by
by 2028. Natural gas is planned to grow to 30 GW, and diesel generators 2030, implying a non-optimal use of installed power capacity. As
to remain at 6 GW. Expansion of carbon-free electricity is planned for noted in the methodology section, the cost-optimized scenarios are
hydro (10 GW) and geothermal power plants (6.6 GW), while solar based on demand growth aligned with RUPTL 2019–2028 (PLN,
PV (1 GW), biomass (0.9 GW) and wind power plants (1.2 GW) are 2019), which ranges between that of RUEN and of RUPTL 2021. This ex-
envisioned to play a rather limited role. RUPTL 2021 paints a similar pic- cludes the possibility that overcapacities in RUPTL compared to our
ture to RUPTL 2019 for 2030 values, but reduces coal capacity expan- baseline scenario are attributable to substantially different demand
sions by 5.7 GW and natural gas by 2.7 GW. Solar PV increases to 4.8 projections. As discussed in Section 4.4, installing more capacity than
GW and biomass to 1.9 GW, while geothermal capacity declines by 1 required results in higher system costs, as the capital expenditures
GW compared to RUPTL 2019. There is an increase in pumped hydro incurred generate less electricity than cost-optimized capacity
storage from 3.5 GW in RUPTL 2019 to 4.2 GW in RUPTL 2021. utilization.
The energy master plan RUEN envisions a strong expansion of fossil As noted above, the coal capacity expansion in the cost-optimized
capacities, particularly coal, to an estimated 123 GW by 2040, more than baseline scenario is similar to that in RUPTL 2019. The cost-optimized
40 GW beyond the baseline scenario, the most favourable scenario for baseline scenario also foresees an expansion of natural gas, diesel, geo-
coal development. We regard this master plan as only broadly indica- thermal and hydro, but requires 30 GW less capacity at the end of the
tive, as it is based on a very high electricity demand growth (8% per decade, which is broken down into 11 GW less gas, 5 GW less geother-
year compared to around 6% historically) and thus an unrealistically mal, 4.5 GW less hydro, 3.5 GW less solar PV, but 6.6 GW of additional
high power capacity expansion. pumped hydro storage capacities.
In the cost-optimized baseline scenario, i.e. assuming a moderate de- In the context of stronger cost reductions for renewables (RE cost re-
crease in costs for variable renewable energies and no CO2 price, coal duction scenario), 15 GW of additional solar PV and 5 GW of geothermal
also experiences a large expansion, reaching 58.6 GW by 2028, which are installed until 2028 compared to RUPTL 2019. The RE cost reduction
is very similar to RUPTL 2019 (59 GW for RUPTL 2019). Towards 2040, scenario also requires 14 GW less coal in the same year than RUPTL 2019
there is another major expansion of coal to 85 GW in the baseline sce- or the baseline scenario. Likewise, it requires 12 GW less natural gas
nario, or an additional 50 GW coal compared to 2020. Geothermal en- (CCGT) and 5 GW of pumped hydro storage. Differences to the 2030
ergy plays an equally important role in all cost-optimized scenarios, values of RUPTL 2021 are smaller, although coal and natural gas are
including the baseline scenario, reaching 16 GW by 2040 and still higher in RUPTL 2021 (by 5 GW and 9 GW, respectively), while
constrained in its upper bound by regional maximum capacity limits. the difference to geothermal capacity is higher in compared to this edi-
Flexibility is provided by pumped hydro storage (20 GW), while other tion of RUPTL. The difference in pumped storage is reduced to 2.2 GW.
renewables such as solar PV, wind or biomass remain strictly absent. With a CO2 price of 30 USD per ton in the context of moderate cost
The share of coal in 2040 increases to 60%, natural gas only comprises reductions for renewable energies (CO2-price scenario), coal becomes
6%, while geothermal increases to 12% of total installed capacities. economically less viable, and 30 GW less coal capacity is built by 2028
In the RE cost reduction scenario, i.e. without a carbon price, but compared to RUPTL 2019 (or compared to the baseline scenario, and
with stronger cost reductions for renewable energies, solar PV reaches 24 GW less when compared to the 2030 values of RUPTL 2021). Like-
20 GW by 2028, and 111 GW by 2040. Indeed, given the lack of substan- wise, 6 GW less gas are required compared to RUPTL 2019, while the dif-
tial wind potentials, any increase in variable renewable capacities is ference is smaller compared to RUPTL 2021 (0.2 GW). Fossil capacities
only due to solar PV in all cost-optimal capacity mixes except the base- in RUPTL are replaced by 17 GW of biomass power plants and 5 GW of
line scenario. Coal still plays an important role in the RE cost reduction geothermal in the CO2 price scenario. Given that there are no
scenario, reaching around 50 GW by 2030 and remaining at that level substantial reductions in the costs of solar PV, this scenario sees no
until 2040. solar PV capacities being built between 2020 and 2030.
In the CO2 price scenario, i.e. assuming moderate cost reductions for Assuming a CO2price of 30 USD per ton and stronger cost reductions
RE but adding a CO2 price of 30 USD per ton, coal capacities are reduced for renewable energies (combined policies scenario) leads to a rapid ex-
to 30 GW by 2030 (30 GW less than RUPTL and the baseline scenario) pansion of RE, with 52 GW of additional solar PV power installed by
and 20 GW by 2040. Due to its lower carbon content, natural gas is 2028, as well as 14 GW of biomass and 9.5 GW of geothermal power
cost-competitive and reaches 14 GW by 2030. In this scenario, no plants. Coal and natural gas are reduced by 30 GW and 12 GW, respec-
solar PV capacity is built in the 2020–2030 decade, but reaches 60 GW tively, compared to RUPTL 2019. In comparison with the 2030 values of
by 2040, suggesting solar PV becomes competitive during the RUPTL 2021, the difference in coal decreases slightly to 24 GW, with ad-
2030–2040 decade. Biomass would become competitive in the decade ditional increases in solar PV (69 GW) and geothermal (11 GW).
2020–2030, expanding to 26 GW by 2040, but its further development
is capped by its regional capacity limits4. Solar PV capacity development
Combining a carbon price with strongly falling prices for RE technol- Solar PV could experience the strongest cost reduction across all re-
ogy leads to a rapid expansion of solar PV, which reaches 75 GW by newables in the years to come, and consequently plays a key role in all
2030 and 140 GW by 2040. Coal use becomes uncompetitive and is sub- the cost-optimized scenarios. Fig. 4 shows the solar PV capacity addi-
stantially replaced by natural gas and the rapid expansion of solar PV. tions in 5-year intervals throughout the modelling horizon. In the base-
Biomass and geothermal expansion occur in the 2020–2030 decade, re- line scenario, no solar PV capacity is deployed. Pricing CO2 emissions
stricted in their development by regional capacity limits. under moderate cost reductions for renewables (CO2 price scenario)
sees no solar PV capacities installed in a cost-optimized power sector
until the end of the examined period, i.e. in the timeframe 2036–2040.
Comparing the cost-optimized scenarios to RUPTL
In this scenario, solar PV costs break even with fossil technologies by
Fig. 3 shows the differences in capacities by technology in the cost-
2037, leading to rapid capacity expansion, reaching 60 GW in total in
optimized scenarios compared to RUPTL 2019 and RUPTL 2021.
2040. In the context of rapid cost reductions for renewables but no car-
4
bon pricing (RE cost reduction scenario), solar PV already increases until
Of note, in our modelling framework, biomass is considered to be managed sustain-
ably thus as of being carbon-neutral. Such a large expansion of biomass power plants
2035, with a particularly large increase in the last five years of the
would bring along challenges related to land use, land use change and forestry. Annex B modelled period. Pricing CO2 emissions under rapid cost reductions
provides an overview on limitations within the considered modelling framework. for renewables (combined policies scenario) already leads to rapid

355
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

120

100

80
Capacity difference in GW

60

40

20

-20

-40

-60
Baseline RE cost CO2-price Combined Baseline RE cost CO2-price Combined
reduction policies reduction policies
compared to RUPTL 2019-2028 compared to RUPTL 2021-2030

Geothermal Biomass Small hydro

Wind Coal CCGT
Diesel Large Hydro Solar PV
Pumped Hydro Storage

Fig. 3. Capacity differences to RUPTL 2019–2028 for the year 2028 and for RUPTL 2021–2030 for the year 2030 for cost-optimized scenarios.

expansion of solar PV capacities up to 2030, which accelerates even an increase in pumped hydro storage compared to RUPTL 2021
more as costs drop further after 2035. (3.5 GW in 2028 of RUPTL 2019 compared to 4.2 GW in 2030 of
RUPTL 2021).
The role of energy storage
Indonesian stakeholders often address the availability of energy The effect of CO2 emission pricing
storages as crucial to the development of variable RE (Ordonez &
Eckstein, 2020). If no storage capacities are considered, solar PV capac- Charging USD 30 per ton of CO2 emissions would trigger a
ities by 2040 (approx. 110 GW in the RE cost reduction scenario) are re- substantial reduction in the use of coal, as presented in the previous
duced by around 20 GW but 90 GW of solar PV would still be installed. section, regardless of the assumptions made concerning RE cost reduc-
Thus, solar PV shows a high potential for economic and technical inte- tions. In order to analyse the effect of carbon prices in more depth, a sen-
gration, even without considering energy storages. However, energy sitivity analysis was performed, in the context of moderate cost
storage increases the economic potential of solar PV as it enables the reductions for renewable energies (by applying a carbon price to the
provision of peak power. Based on the underlying cost assumptions baseline scenario) and in the context of more rapidly decreasing costs
(Table 2) in this model, pumped hydro storage is cost-competitive for renewable energies (by applying a carbon price to the RE cost reduc-
with battery storage throughout the model. Of note, RUPTL 2021 plans tion scenario). We varied the carbon prices for USD 5, 15, 30, 50 per ton

80

70
Capacity additions in GW

60

50

40

30

20

10

0
2020-2025 2026-2030 2031-2035 2036-2040

Solar PV - RE cost reduction

Solar PV- Combined policies
Solar PV - CO2 price

Fig. 4. Development of solar PV capacity in different timeframes for cost-optimized scenarios.

356
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

Carbon pricing in the context of moderate cost Carbon pricing in the context of strong cost
200 reductions for renewables reductions for renewables

150
2040 capacity difference in GW

100

50

-50

-100
5 USD/t 15 USD/t 30 USD/t 50 USD/t 5 USD/t 15 USD/t 30 USD/t 50 USD/t
CO2-price effect on Baseline scenario CO2-price effect on RE cost reduction scenario

Geothermal Biomass Small hydro

Wind Coal CCGT
Diesel Large Hydro Solar PV
Pumped Hydro Storage

Fig. 5. Capacity differences in 2040 for the baseline scenario (left-hand panel) and the RE cost reduction scenario (right-hand panel) when varying the CO2 prices in each scenario.

of CO2 and examined the resulting capacity mix at the end of the Daily generation profiles
modelling horizon.5 The resulting capacity mix is presented in Fig. 5,
which shows the capacity difference in 2040 compared to the 2040 In view of the large share of solar PV in all cost-optimal scenarios
capacity mix of each scenario. apart from the baseline scenario, we also investigated the potential flex-
With moderate cost reductions for RE (i.e. basing our sensitivity ibility challenges of different solar capacity mixes by analysing hourly
analysis of carbon prices on the baseline scenario), a carbon price as power generation profiles of a typical day in 2040. We based our analy-
low as USD 5 per ton of CO2 reduces the total installed coal capacities sis on a dry season day in Java-Bali for 2040, because this is projected to
by 25 GW by 2040 and adds approximately the same installed capacity have the highest share of solar PV generation of all seasons and regions,
of biomass power plants (23 GW) by the same year. Increasing the car- and thus the most challenging conditions for balancing supply and de-
bon tax to USD 15 per ton of CO2 reduces coal by 35 GW, increases bio- mand. Fig. 6 shows the corresponding daily generation profiles.
mass by 25 GW, and causes small additions of natural gas and energy There is no solar PV in the baseline scenario, and generation and
storage facilities by 2040. A price of more than USD 30 per ton of CO2 flexibility are provided by pumped-hydro storage or natural gas
reduces coal by 63 GW compared to the baseline, and increases solar (Fig. 5, A). In the RE cost reduction and the CO2-price scenarios (Fig. 5,
PV capacities to 60 GW, as well as adding 35 GW of natural gas. Increas- B and C), maximum solar PV generation occurs between 11:00 and
ing the price to USD 50 per ton of CO2 additionally expands the share of 12:00 if skies are clear, reaching 59 GW in the RE cost reduction scenario
solar PV to 100 GW. and 39 GW in the CO2-price scenario. Solar PV generation is highest in
With stronger cost reductions of renewable energies (i.e. basing our the combined policies scenario, with a constant supply of 61 GW be-
sensitivity analysis of carbon prices on the RE cost reduction scenario), a tween 09:00 and 14:00. High amounts of variable solar PV generation
carbon price of USD 5 per ton of CO2 decreases coal capacity by 5 GW, result in the need to ramp down other power plants. In the RE cost re-
although the total installed coal capacities in this scenario are approx. duction scenario, as in the combined policies scenario, coal-fired
50 GW by 2040 compared to 85 GW in the baseline scenario. Thus, power plants need to be ramped down from 34 GW to close to zero be-
the resulting coal capacity is around 45 GW, compared to 60 GW in tween 06:00 and 09:00, as sunrise initiates solar PV generation, and
the baseline scenario. This carbon price also triggers the additional in- then ramped up again between 12:00 and 15:00, as solar generation de-
stalment of 15 GW of solar PV, from a total of 111 GW to 125 GW in creases, requiring a very high degree of flexibility. This mode of opera-
2040. Increasing the tax to between 15 and 30 USD per ton of CO2 tion would certainly require refurbishment of coal-fired power plants
reduces coal use by around 30 GW to a total installed capacity of 20 to provide increased flexibility, as conventional power plants have flex-
GW by 2040. This is equivalent to the capacity mix of the combined pol- ibility limits depending on their age, design and technology6 (IRENA,
icies scenario. Natural gas use then becomes competitive, with an addi- 2019). In the CO2 price scenario, the lower share of solar PV and the
tional 20–25 GW by 2040. A carbon tax of USD 50 per ton of CO2 triggers comparatively high share of gas compared to the scenarios assuming a
the instalment of 25 GW of wind power, highlighting that wind only be- stronger decline in cost reductions for RE result in natural gas balancing
comes cost-competitive in the context of strong cost reductions for solar PV generation, without the need to ramp down coal or other
wind LCOE combined with high carbon taxes. technologies.

5 6
Under these assumptions, the sensitivity of USD 30 per ton of CO2 corresponds to the These refurbishments might make conventional power plants more expensive, but
CO2 price and combined policy scenario, respectively. any such additional costs are not considered in the model.

357
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

(A) Baseline (B) RE cost reduction

70 70
60 60
50 50

Power in GW
Power in GW 40 40
30 30
20 20
10 10
0 0
-10 -10
-20 -20

(C) CO2-price (D) Combined policies

70 70
60 60
50 50
Power in GW

Power in GW
40 40
30 30
20 20
10 10
0 0
-10 -10
-20 -20

Pumped Hydro Storage Solar PV Large Hydro CCGT

Coal Biomass Small hydro Geothermal
All others Total

Fig. 6. Daily power generation profiles for different scenarios in Java-Bali in 2040 (dry season).

Electricity-related CO2 emissions and total system costs CO2 emissions

In the context of constantly increasing energy demand, the develop-
The following paragraphs discuss CO2 emissions, total electricity ment of CO2 emissions reflects the carbon intensity of the resulting
production costs as well as the total system costs for the different electricity mix in a given power sector development scenario. In the
scenarios. Fig. 7 shows the CO 2 emissions and total electricity baseline scenario, the reliance on carbon-intensive coal causes CO2
production costs in conjunction. emissions to increase from 180 million tons (MT) in 2019 to 470 MT

A) B)
500
75
450
Electricity production costs in USD/MWh

70
CO2-emissions in million tonnes

400

350
65
300

250 60
200
55
150

100
50
50

0 45

Fig. 7. A) Development of annual CO2 emissions and B) development of total electricity production costs (carbon pricing costs not included) by scenario and power sector development
plans, in the timeframe 2019–2040.

358
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

Table 3 base year and putting these costs in relation to total electricity genera-
Cumulated emissions until 2040 for varying CO2 prices (under RE cost reduction tion. While the costs associated with carbon pricing represent a cost
assumptions).
block that affects the cost-optimal future capacity mix, carbon pricing
CO2-price 0 USD/t (RE 5 30 USD/t 50 70 also represents a revenue stream for the government that can be
cost USD/t (Combined USD/t USD/t recycled in different schemes. Hence, adopting a systemic perspective,
reduction) policies)
we excluded these costs under the assumption that carbon pricing is
Cumulated 5.5B T 4.3B 2.9B T 1.5B 1.2B revenue-neutral (c.f. Indonesian Law Blog (2021)).
emissions T T T
Official power expansion plans, above all Indonesia's master plan
RUEN, but also the utilities' business plans RUPTL 2019–2028 and
RUPTL 2021–2030, lead to the highest system costs and electricity
by 2040 (Fig. 7). The installation of geothermal power plants between prices of all the scenarios, due to the instalment of substantial overca-
2029 and 2032 reduces the CO2 intensity of electricity generation and pacities (Fig. 7). RUEN reaches 70 USD per MWh by 2040. Costs associ-
emissions remain constant in this period. ated with a cost-optimal operation of the capacity mix of RUPTL
In the context of strong reductions in the costs for renewables (RE 2019–2028 are higher than in the optimized scenarios, but still signifi-
cost reduction scenario), Indonesia's power sector still has coal capacity cantly lower than in RUEN, reaching costs of 56 USD per MWh by
additions, especially in the first half of the 2020–2030 decade, before 2025 and 53 USD per MWh by 2028. After a significant increase in the
geothermal and solar PV power plants become competitive and capacity first half of the 2020–2030 decade, RUPTL 2021 reaches 51 USD per
additions of these technologies reduce the CO2 emission intensity MWh in 2030, close to the CO2 price scenario. The baseline and the RE
between 2025 and 2030. However, in the 2030–2040 period, the cost reduction scenarios, reaching values of USD 45–47 per MWh,
power sector still sees a combination of renewables and coal capacity have the lowest costs over the time horizon, as their optimization is a
additions, leading to more or less constantCO2 emissions. By 2040, the function of least-cost technologies, disregarding their carbon context.
RE cost reduction scenario leads to annual emissions of approx. 240 Accordingly, the CO2 price and the combined policies scenarios yield
MT CO2, which are half the baseline scenario emissions. higher costs, of USD 52–48 USD per MWh, as cost optimization is a func-
Putting a price on CO2 emissions (CO2 price scenario) would render tion of the least costs considering emissions. In addition, the costs do not
coal uncompetitive, and thus decrease emissions from the start of the fall continuously because the system costs minimized by the objective
modelling horizon. Capacity additions in the 2020–2025 timeframe function include carbon pricing costs, which are excluded in this figure
are based on biomass and geothermal power plants in this scenario, as discussed above.
which decrease the CO2 emission factor before natural gas plays a To analyse differences in costs over the entire modelling time hori-
dominant role between 2025 and 2035. Towards the end of the zon for the various scenarios and power sector plans, we used cumu-
2030–2040 decade, solar PV breaks even and capacity additions of lated electricity production costs until the end of the time horizon in
solar PV rapidly reduce the CO2 emission intensity. Putting a price on 2040 (see Table 4). Electricity production costs are presented including
CO2 leads to emissions of 167 MT of CO2 by 2040, or 35% of the and excluding carbon pricing costs. RUEN has the highest cumulated
baseline scenario emissions. costs at 709 billion USD. The lowest costs are achieved with falling
Finally, putting a price on CO2 in the context of strongly falling costs costs for renewables (509 billion USD in the RE cost reduction scenario),
of renewables (combined policy scenario) leads to the strongest while the CO2 price scenario (550 billion USD) and the combined
decrease in CO2 emissions of all the scenarios considered. The carbon policies scenario (536 billion USD) are in-between these two. For the
tax makes coal uncompetitive, so that solar PV takes a big share of the scenarios with CO2 pricing, the difference between the total costs
power sector's composition. This leads to CO2 emissions of around 120 including and excluding CO2 emission costs represent the carbon
MT of CO2 by 2040, or 60% of the 2019 emissions and only 25% of the revenues. Pricing emissions under moderately falling costs for
baseline scenario emissions. RUPTL's last planning year is 2028, but its renewables (CO2 price scenario) yields 110 billion USD, and 86 billion
strong reliance on coal means the CO2 emission development is USD with strongly falling costs (combined policies scenario), ranging
aligned with the baseline scenario, reaching 280 MT of CO2 by 2028, between 4 and 6 billion USD annually.
or an increase of 2019 emissions by 50%.
Finally, we computed cumulated CO2 emissions by the end of the Discussion and policy implications
analysed time horizon (2040) resulting from different levels of carbon
price building on the RE cost reduction scenario (Table 3). Cumulated The results of the cost-optimized power sector modelling show that
emissions decline continuously with increasing CO2 prices (and are Indonesia's power sector plans offer scope for both substantial CO2
highest in the baseline with 7.1 billion tons (BT)). Notably, emissions emission savings and reducing the overall costs of electricity. While
are significantly lower for a CO2 price of 5 USD/t (−1.2 BT) compared the current official power expansion plans are strongly based on the
to no CO2price, even though the capacity mix in this scenario is use of carbon-intensive coal, renewables could play a pivotal role in
comparable (see Fig. 5). Indonesia's future power mix, particularly solar PV, but also geothermal
and biomass. Reconciling climate protection with sustained low elec-
Electricity production and system costs tricity prices is based on a combination of three elements: i) avoiding
Electricity production costs in 2018 USD per MWh of electricity gen- the installation of unnecessary power plants ii) introducing carbon pric-
erated a determined by discounting all cost components to the same ing, and iii) politically fostering cost reductions for RE, in particular for

Table 4
Cumulated system costs throughout the modelling time horizon (2019–2040) excluding and including CO2 emission costs. The difference represents carbon revenues, the average yearly
revenue is computed dividing carbon revenues by 21 years.

Baseline RE cost reduction CO2-price Combined policies RUEN

System costs (excluding CO2 emission costs) 517B USD 509B USD 550B USD 536B USD 709B USD
System costs (including CO2 emission costs) 517B USD 509B USD 660B USD 622B USD 709B USD
Difference 0 (No carbon price) 0 (No carbon price) 110B USD 86B USD 0 (No carbon price)
Average yearly carbon revenue 0 (No carbon price) 0 (No carbon price) 5.5B USD 4.3B USD 0 (No carbon price)

359
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

solar PV. This section discusses the effects of these elements on the contexts, but it remains unclear to what extent such subsidies reduce
power system's development, as well as policies to achieve cost reduc- the LCOE of RE technologies (Ragwitz & Steinhilber, 2014).
tions while reducing CO2 emissions. Decreasing the LCOE of RE technologies can only be achieved by re-
ducing its cost components: either lowering technology and installation
Optimal capacity development costs, or financing costs. Technology costs for solar PV have shown a
sharp decrease in recent decades, a decrease that is explained by the
Indonesia's official power expansion plans lead to the highest sys- well-documented relationship between the cumulative installed vol-
tem costs of all considered scenarios, largely due to the installation of umes of a given technology and its overall manufacturing costs
unnecessary power plants. This becomes obvious when comparing the (IRENA, 2020). The global market of solar PV manufacturing is domi-
two RUPTLs with the baseline scenario, which has a similar technologi- nated by China, followed by Malaysia, Vietnam, Taiwan and Korea,
cal composition, but lower capacity requirements and overall electricity while Indonesia plays a negligible role in the production of solar PV
costs. Avoiding the installation of unnecessary power plants prevents equipment (IEA, 2020b). Arguably, prevalent local content require-
unnecessary capital expenditures, thus lowering electricity costs. ments in Indonesia, which require solar PV projects to have a 44%
Using realistic electricity demand projections7 on the one hand, and op- share of locally manufactured components, cut the country off from
timally planning and operating the power plants on the other, is crucial the continuous reduction in technology costs elsewhere. Even more, in
to avoid unnecessarily high electricity costs in the decades to come. absence of an Indonesian local manufacturing solar PV industry, such
local content requirements might represent a prohibitive barrier to de-
velop solar PV projects. On the other hand, the establishment of a local
Carbon pricing solar PV manufacturing industry requires both a policy mix with
technology-push and demand-pull policies, as well as the existence of
Carbon pricing would make the integration of new coal power plants related industrial capabilities, a favourable investment context, among
an economically unviable alternative, even at prices as low as USD 5 per other (Hayashi, 2020). Thus, either abolishing local content require-
ton of CO2 while fostering low-carbon alternatives. Carbon pricing at ments to enable reductions in solar PV equipment costs by imports, or
USD 30 per ton of CO2 (carbon price scenario) would also create developing a comprehensive strategy to foster the establishment of a
revenue stream ranging between 4.3 and 5.5 billion USD annually. To cost-competitive local manufacturing capabilities, will be key to lower-
put this in perspective, the yearly public revenue collection target ing solar PV technology costs in the Indonesian context.
from coal and minerals mining imposed on Indonesia's Ministry of En- Financing costs are the other large LCOE cost component of capital-
ergy and Minerals by the Ministry of Finance is around 40 trillion intensive renewable energies. Lowering these costs, which are reflected
Indonesian Rupiah, or approx. 3 billion USD (Harsono, 2020). Three in the weighted average cost of equity and debt, requires a reduction in
quarters of Indonesian coal production and therefore most of this reve- regulatory uncertainty and thus the risk premiums on capital to finance
nue stream is related to coal production for exports. Thus, carbon reve- such projects. Specific project risks are often constituted in the numer-
nues from domestic carbon taxation in the power sector at USD 30 per ous administrative, regulatory, economic, political and other barriers
ton of CO2 have the potential to overcompensate the fiscal losses of that project developers face when developing a project (Woodman
reduced coal mining for domestic use, while leading to lower et al., 2017). De-risking RE projects thus requires guaranteeing a stable
emissions and lower electricity production costs than 2019 levels. regulatory environment and reducing administrative and regulative
These revenues could be used for different purposes, including public barriers is key to lower financing costs (Schmidt, 2014; Waissbein
infrastructure development (Franks et al., 2018; Jakob et al., 2015), et al., 2013). Reforming the monopolistic structure of the power market,
social protection (Budolfson et al., 2021), regional economic which in Indonesia has a strong bias towards favouring coal-fired power
development of coal-dependent regions (Reitzenstein et al., 2021) or capacities over RE, might also facilitate RE integration. Indeed, experi-
to foster the clean energy transformation (e.g. retrofitting the conven- ences from multiple countries show that competitive markets have re-
tional power fleet for greater flexibility, demand-side management, placed coal with cheaper alternatives, i.e. renewable energies and
etc.). In Indonesia's decentralized fiscal system, a part of these carbon natural gas in recent years (Jakob & Steckel, 2022). However, such far-
revenues could compensate public income losses of coal-dependent re- reaching reforms represent a major institutional change that is hard to
gions, such as East- and South Kalimantan, or South Sumatra. Also, in- achieve in developing countries and emerging economies, given their
creasing electricity prices for end-consumers comes at the possible weaker legal, economic and political institutions (Victor & Heller, 2007).
cost of popularity losses of elected officials, and reducing fuel and elec- In Indonesia's non-competitive, monopolistic power market, both
tricity subsidies has proven to be represent a major challenge in the past falling costs of renewables and a carbon tax might not automatically
(Inchauste & Victor, 2017; Ordonez et al., 2021). Carbon revenues could translate into a high share of low-carbon technologies (Bauer et al.,
thus also be directed back to citizens, in particular towards low-income 2021; Eckstein, Nascimiento, et al., 2020). Electricity procurement by
segments of society, thereby avoiding a trade-off with poverty allevia- the state-owned utility PLN by developing domestic power plants or
tion and increasing the acceptability of important political constituen- by entering into long-term power purchase agreements with indepen-
cies (Klenert et al., 2018; Maestre-Andrés et al., 2019). dent power producers implies that careful consideration of least-cost
technologies is pivotal in order to avoid a long-term lock-in into high
Fostering cost reductions for renewable energies electricity costs. A sub-optimal choice of technologies ultimately implies
a lock-in to either higher electricity prices for end consumers, or higher
Solar PV could become economically viable in Indonesia by the mid- electricity subsidies.
dle of the 2020–2030 decade, if costs fell quickly (RE cost reduction sce-
nario). Support schemes for RE, such as feed-in tariffs8 or quotas, have Conclusions
proven to foster the integration of RE technologies in a variety of
In this paper, we developed a cost-optimized power sector model for
7
Electricity and energy modelling reports in Indonesia and ASEAN neighbors often rely Indonesia and examined the cost-optimal power sector composition
on overall optimistic economic growth and corresponding electricity demand growth pro- until 2040 in the context of moderate and strongly falling costs for RE
jections (Cf. ACE, 2015, 2017, 2020; DEN, 2016, 2019). technologies, with and without carbon pricing. We thereby analysed
8
Of note, the resulting capacity expansion of the RE cost reduction scenario could also be
attained by integrating renewable support schemes and thus subsidizing the additional
the role of different technologies in the electricity supply, their associ-
costs of renewables. Yet, this would not yield the system cost reductions as presented in ated costs and CO2 emissions, and compared our results to those
the previous chapter. obtained by using the power development plans RUPTL 2019 and

360
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

2021 of Indonesia's utility and the energy masterplan RUEN by the Vietnam, demonstrates how countries with coal endowments, a
National Energy Council. comparable socio-economic development level, and similar
Our findings show that Indonesia's power sector plans offer scope power sector structures can overcome prevalent barriers and sub-
for improvement. Indonesia's power sector development could recon- stantially exploit their vast solar PV potentials. Against this back-
cile climate mitigation with reducing the overall costs of electricity. In ground, our results highlight how climate protection in Indonesia
particular, we found that avoiding overcapacities, implementing carbon might not represent a trade-off with economic development or
pricing, and fostering cost reductions for solar PV, are of major impor- poverty alleviation, but could provide a clear opportunity to re-
tance in tapping this potential. Combining these elements is crucial to duce electricity costs for industry, commerce and households
achieve a rapid reduction in CO2 emissions while sustaining low alike.
electricity generation costs, thereby aligning Indonesia's power sector
development with the goals of the Paris Agreement. Funding
Our study highlights that some main objectives behind the contin-
ued use of coal-fired power plants, i.e. keeping electricity prices low No external funds were received for this paper.
and collecting public revenue from coal royalties, could still be achieved
by using low-carbon technologies and introducing carbon pricing. Declaration of competing interest
The experience of other countries which have installed sub-
stantial shares of solar PV in the last decade, such as India and The authors declare no conflict of interest.

Appendix A
GHG emissions in MtCO2e

3500
3000
2500 Forestry**
2000 Agriculture
1500 IPPU
1000
Waste
500
Energy*
0
2010 2030 2030 2030
Historical BAU CM1 CM2

Fig. A1. Projected emissions in Indonesia's NDC in the Business as Usual Scenario (BAU), and the two conditional scenarios (CM1, CM2) in 2030 against 2010 historical emissions. *Includ-
ing electricity, transport, heating, fugitive emissions, **including peat fires (Eckstein, Ordonez, & Wachsmuth, 2020).
Final electricity demand in TW h

1600
1400
1200
Historical
1000
RUEN 2015-2050
800 RUTPL 2019-2028
600 RUPTL 2021-2030
400 RUPTL 2021-2030
200
0
2034
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
2024
2026
2028
2030
2032

2036
2038
2040

Fig. A2. Historical and projected total final electricity demand by using electricity demand growth rates as presented in RUPTL 2019–2028, RUPTL 2021–2030, and the energy masterplan
RUEN. Own elaboration based on PLN, 2019; PLN, 2021; DEN, 2017b; ENERDATA, 2021).

361
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

Table A3
Power capacity in GW by technologies for official policy plans RUPTL 2019–2028, RUPTL 2021–2030 and RUEN, as well as for cost optimized scenarios for the years 2020, 2028, 2030, and
2040. Own elaboration based on scenario results and DEN, 2017a; PLN, 2019; PLN, 2021.

Geo-thermal Biomass Solar PV Wind Large Hydro Small hydro Pumped Hydro Storage Coal CCGT OCGT Diesel Total

RUPTL19 2020 2,3 0,1 0 0,1 5,9 0,5 0 39,8 14,3 8 6,5 77,5
2028 6,6 0,9 1 0,9 10 1,7 3,5 59,3 20,4 9,6 6,6 120,5
RUPTL21 2020 2,3 0,1 0,2 0,1 5,9 0,5 39,8 14,2 8 6,1 77,2
2028 4,6 1,3 4,5 0,7 9,5 1,6 2 53,6 18 9,9 6,2 111,9
2030 5,7 1,9 4,8 0,8 10,9 1,6 4,2 53,6 18,1 10 6,2 117,8
RUEN 2020 3,1 1,1 0,9 0,6 5,6 1 0 37,6 15,1 6,6 5,4 77
2028 8,4 5,1 10,4 4,1 20,3 3,5 0 66 22,9 10,9 7,6 159,2
2030 9,3 7,8 14,2 7 22 3,8 0 74,6 25,6 12 8,3 184,6
2040 13,4 17,6 29,6 17,5 30 5,4 0 122,8 42,2 19,8 13,6 311,9
Baseline 2020 2 0 0 0 5,4 0 0,1 32,5 11,3 6,3 6,4 64
2028 1,8 0 0 0 5,4 0 6,6 58,6 8,9 5,7 5,9 92,9
2030 5,8 0 0 0 5,4 0 7,8 60,4 8,9 5,1 5,6 99
2040 16,5 0 0 0 5,4 0 21,2 85,7 5,3 2,4 3 139,5
RE cost reduction 2020 2 0 0 0 5,4 0 0,1 32,5 11,3 6,3 6,4 64
2028 12 0,2 18,6 0 7,7 0 5 45,2 8,9 5,7 5,9 109,2
2030 14,3 0,3 20,4 0 7,7 0 6,4 48,3 9 5,1 5,6 117,1
2040 16,6 16,6 111,2 0 7,9 0 24 50,7 7,6 2,4 2,9 239,9
Combined policies 2020 2 1,9 0 0 5,4 0 0 32,2 11,2 6,3 6,4 65,4
2028 15,1 15,3 55,6 0 7,7 2,9 1,6 29,5 8,8 5,7 5,9 148,1
2030 16,3 15,3 73,9 0,9 7,7 2,9 4,9 29,2 8,8 5,1 5,6 170,6
2040 16,6 23,8 139,1 0,9 8,4 2,9 12,3 22,9 32,2 2,4 3,4 264,9
CO2 Price 2020 2,1 10,7 0 0 5,4 0 0 32,2 11,2 6,3 6,4 74,3
2028 11,5 18,2 0 0 7,7 2,9 0,2 29,5 14,9 5,7 6,1 96,7
2030 14,5 18,5 0 0 7,7 2,9 0,2 29,2 18,3 5,1 7,1 103,5
2040 16,6 25,9 60 0 8,1 4,3 2,8 22,9 39,4 2,4 9,4 191,8

Appendix B. Modelling limitations

Of note, our model has several limitations. Cost reductions for RE in the absence of carbon pricing would require higher shares of solar PV in a power
system with high shares of coal, and would demand much greater flexibility from the conventional power plant fleet. The applied modelling frame-
work might lack the complexity needed to address the ramping up or down of conventional power plants, which may affect the modelling results,
especially for scenarios with high RE shares towards the end of the modelling framework. The daily operation patterns should therefore only be seen
as indicative. In any case, both demand side management measures and refurbishments of coal-power plants would be required to enable their
ramping-up and down more flexibly (IRENA, 2021a). Such refurbishments, while theoretically possible, are subject to uncertainty as they might
be costly or technically challenging. By focusing on the power sector, our model omits grid expansion constraints and transmission and distribution
challenges in this regard. Another limitation is in assessing the sustainability of biomass use, as well as the technical and financial viability of pumped
hydro storages. Adding substantial capacities of biomass comes along with the challenge of guaranteeing its sustainable management, in particular in
a country where the bulk of emissions stem from the land use and forestry sector. Likewise, while large potential of pumped hydro storage is assumed
to exist for Indonesia, its feasibility remains to be proved.

References Cole, Wesley, & Frazier, A. Will (2019). Cost Projections for Utility-scale Battery Storage.
Available online atNational Renewable Energy Laboratory. https://www.nrel.gov/
docs/fy19osti/73222.pdf.
ACE (2015). The 4th ASEAN Energy Outlook (AEO4) 2013-2035. Available online atJakarta,
Indonesia: ASEAN Centre for Energy (ACE). https://aseanenergy.org/. DEN (2016). Indonesia energy outlook 2016. Jakarta, Indonesia: Dewan Energi Nasional
(DEN) - National Energy Council of Indonesia.
ACE (2017). The 5th ASEAN Energy Outlook (AEO5) 2015-2040. Available online atJakarta,
DEN (2017a). Rencana Umum Energi Nasional. Jakarta, Indonesia: Dewan Energi Nasional
Indonesia: ASEAN Centre for Energy (ACE). https://aseanenergy.org.
(DEN) - National Energy Council of Indonesia.
ACE (2019). Levelised Costs of Electricity (LCOE) for Selected Renewable Energy Technologies
DEN (2017b). Technology data for the Indonesian power sector. Catalogue for generation
in the ASEAN Member States II. Available online atJakarta, Indonesia: ASEAN Centre for
and storage of electricity. Jakarta, Indonesia: Dewan Energi Nasional (DEN) - National
Energy (ACE). https://aseanenergy.org/.
Energy Council of Indonesia.
ACE (2020). The 6th ASEAN Energy Outlook (AEO6) 2017-2040. Available online atJakarta,
DEN (2019). Indonesia energy outlook 2019. Jakarta, Indonesia: Dewan Energi Nasional
Indonesia: ASEAN Centre for Energy (ACE). https://aseanenergy.org/.
(DEN) - National Energy Council of Indonesia.
Bauer, Nico, Stephen, Bi, Diluiso, Francesca, Manych, Niccolò, Ohlendorf, Nils, Ordonez, DTU (2021). Global wind atlas 3.0. Available online atTechnical University of Denmark
Jose Antonio, & Steckel, Jan Christoph (2021). Phasing-out coal in Indonesia. Available (DTU). https://globalwindatlas.info.
online atPolicy brief. Berlin: MCC. https://www.mcc-berlin.net/fileadmin/data/C18_
Eckstein, Johannes, Nascimiento, Leonardo, Kurdziel, Marie, & Ordonez, Jose Antonio
MCC_Publications/INDONESIA_Policy_Brief_Coal_Phase-out_20211101.pdf.
(2020a). Decreasing costs of renewables – Insights on energy sector planning and climate
Blog, Indonesian Law (2021). Available online atIn Fransiscus Rodyanto, & Aldilla S. policy from three country case studies. Available online atWorking paper sustainability
Suwana (Eds.), Putting a price on emissions: Indonesian government plans to impose and innovation no. S 17/2020Karlsruhe: Fraunhofer ISI updated on 2020https://
carbon tax. Indonesian Law Blog. https://www.ssek.com/blog/putting-a-price-on- www.isi.fraunhofer.de/content/dam/isi/dokumente/sustainability-innovation/2020/
emissions-indonesian-government-plans-to-impose-carbon-tax. WP-17-2020-Decreasing_costs_of_renewables_Overview.pdf.
BPS (2020). Electricity Statistics. Available online atBadan Pusat Statistik- Central Sta- Eckstein, Johannes, Ordonez, Jose Antonio, & Wachsmuth, Jakob (2020b). Decreasing
tistics Agency of Indonesia. https://www.bps.go.id/menu/1/informasi-umum. costs of renewables - Implications for Indonesia’s climate targets. Working paper
html#masterMenuTab1. sustainability and innovation no. S 13/2020. Fraunhofer Institute for Systems and
Bridle, Richard, Gass, Philip, Halimajaya, Aidy, Lontoh, Lucky, McCulloch, Neil, Innovation Research ISI. https://publica.fraunhofer.de/eprints/urn_nbn_de_
Petrofsky, Erica, & Sanchez, Lourdes (2018). Missing the 23 per cent target: 0011-n-6156249.pdf.
roadblocks to the development of renewable energy in Indonesia. Available Egli, Florian (2020). Renewable energy investment risk: an investigation of changes over
online atGSI and IISD Report. https://www.iisd.org/system/files/publications/ time and the underlying drivers. Energy Policy, 140, Article 111428. https://doi.org/10.
roadblocks-indonesia-renewable-energy.pdf. 1016/j.enpol.2020.111428.
Budolfson, Mark, Dennig, Francis, Errickson, Frank, Feindt, Simon, Ferranna, Maddalena, Egli, Florian, Steffen, Bjarne, & Schmidt, Tobias S. (2019). Bias in energy system models
Fleurbaey, Marc, et al. (2021). Climate action with revenue recycling has benefits with uniform cost of capital assumption. Nature Communications, 10(1), 4588.
for poverty, inequality and well-being. Nature Climate Change, 11(12), 1111–1116. https://doi.org/10.1038/s41467-019-12468-z.
https://doi.org/10.1038/s41558-021-01217-0. Enerdata (2021). Global statistical databaseAvailable online at. https://www.enerdata.net.

362
J.A. Ordonez, M. Fritz and J. Eckstein Energy for Sustainable Development 68 (2022) 350–363

ESDM (2021). Handbook of Energy & Economic Statistics of Indonesia 2020. Ministry of En- access. Climate Change and Sustainble Development Series 230602. Fondazione Eni
ergy and Mineral Resources. Enrico Mattei (FEEM).
ESMAP (2020). Global Photovoltaic Power Potential by Country. Washington, D.C: World Kalirajan, Kaliappa, & Syed, Arif (2019). The Indonesia energy technology assessment (IETA)
Bank. 2017.
Feyen, Erik, & Huertas, Igor Zuccardi (2020). Bank lending rates and spreads in EMDEs KESDM (2019). Rencana umum ketenagalistrikan nasional (RUKN) 2019-2038. Jakarta:
evolution, drivers, and policies. Policy Research Working Paper 9392. World Bank Kementarian energi dan sumber daya mineral.
Group. Klenert, David, Mattauch, Linus, Combet, Emmanuel, Edenhofer, Ottmar, Hepburn,
Franks, Max, Lessmann, Kai, Jakob, Michael, Steckel, Jan Christoph, & Edenhofer, Ottmar Cameron, Rafaty, Ryan, & Stern, Nicholas (2018). Making carbon pricing work for cit-
(2018). Mobilizing domestic resources for the Agenda 2030 via carbon pricing. Nat izens. Nature Climate Change, 8(8), 669–677. https://doi.org/10.1038/s41558-018-
Sustain, 1(7), 350–357. https://doi.org/10.1038/s41893-018-0083-3. 0201-2.
Global Energy Monitor. (2021). Boom and Bust 2021. Tracking the Global Coal Plant Luderer, Gunnar, Vrontisi, Zoi, Bertram, Christoph, Edelenbosch, Oreane Y., Pietzcker,
PipelineAvailable online at. https://endcoal.org/wp-content/uploads/2021/04/ Robert C., Rogelj, Joeri, et al. (2018). Residual fossil CO2 emissions in 1.5–2 °C path-
BoomAndBust_2021_English.pdf. ways. NatureClimatic Change, 8(7), 626–633. https://doi.org/10.1038/s41558-018-
Handayani, Kamia, Krozer, Yoram, & Filatova, Tatiana (2017). Trade-offs between electri- 0198-6.
fication and climate change mitigation: An analysis of the Java-Bali power system in Maestre-Andrés, Sara, Drews, Stefan, & van den Bergh, Jeroen (2019). Perceived fairness
Indonesia. Applied Energy, 208, 1020–1037. https://doi.org/10.1016/j.apenergy.2017. and public acceptability of carbon pricing: A review of the literature. Climate Policy,
09.048. 19(9), 1186–1204. https://doi.org/10.1080/14693062.2019.1639490.
Harsono, Norman (2020). Energy ministry rejects miners' request for relaxation on ocal Mone, Christopher, & Hand, Maureen (2017). 2015 cost of wind energy review. National
royalties. Available online atThe Jakarta post, 2020. https://www.thejakartapost. Renewable Energy Laboratory.
com/news/2020/06/08/energy-ministry-rejects-miners-request-for-relaxation-on- Ordonez, Jose Antonio, & Eckstein, Johannes (2020). Decreasing costs of renewables –
coal-royalties.html. Analysis of energy sector planning and climate policy in Indonesia. Working paper
Hayashi, Daisuke (2020). Harnessing innovation policy for industrial decarbonization: Ca- sustainability and innovation No. S 14/2020. Fraunhofer Institute for Systems and Inno-
pabilities and manufacturing in the wind and solar power sectors of China and India. vation Research ISI.
Energy Research & Social Science, 70, Article 101644. https://doi.org/10.1016/j.erss. Ordonez, Jose Antonio, Jakob, Michael, Steckel, Jan Christoph, & Fünfgeld, Anna (2021).
2020.101644. Coal, power and coal-powered politics in Indonesia. Environmental Science & Policy,
Heaps, C. G. (2021). LEAP: The low emissions analysis platform. Version 2020.1.30. Available 123, 44–57. https://doi.org/10.1016/j.envsci.2021.05.007.
online atSomerville, MA, USA: Stockholm Environment Institute. https://leap.sei.org. PLN (2019). Rencana Usaha Penyediaan Tenaga Listrik (RUTPL) 2019-2028. Available online
Heptonstall, Philip J., & Gross, Robert J. K. (2020). A systematic review of the costs and im- atPLN. https://web.pln.co.id/stakeholder/ruptl.
pacts of integrating variable renewables into power grids. Nature Energy, 6(1), 72–83. PLN (2021). Rencana Usaha Penyediaan Tenaga Listrik (RUTPL) 2021-2030. Available online
https://doi.org/10.1038/s41560-020-00695-4. atPLN. https://web.pln.co.id/stakeholder/ruptl.
Hirth, Lion, & Steckel, Jan Christoph (2016). The role of capital costs in decarbonizing the PricewaterhouseCoopers (2018). Power in Indonesia. Investment and taxation guide - 6th
electricity sector. Environmental Research Letters, 11(11), Article 114010. https://doi. EditionAvailable online at. https://www.pwc.com/id/en/pwc-publications/industries-
org/10.1088/1748-9326/11/11/114010. publications/energy–utilities—mining-publications/power-guide-2018.html.
Huda, Muhammad, Aziz, Muhammad, & Tokimatsu, Koji (2018). Potential ancillary ser- Ragwitz, Mario, & Steinhilber, Simone (2014). Effectiveness and efficiency of support
vices of electric vehicles (vehicle-to-grid) in Indonesia. Energy Procedia, 152, schemes for electricity from renewable energy sources. WENE, 3(2), 213–229.
1218–1223. https://doi.org/10.1016/j.egypro.2018.09.172. https://doi.org/10.1002/wene.85.
IEA (2018). Partner country series - Thailand renewable grid integration assessment. Avail- Rahman, Dzulfiqar Fathur (2021). PLN pledges carbon neutrality by 2050. Available on-
able online atParis: IEA. https://www.iea.org/reports/partner-country-series- line atThe Jakarta post, 2021. https://www.thejakartapost.com/news/2021/05/07/
thailand-renewable-grid-integration-assessment. pln-pledges-carbon-neutrality-by-2050.html.
IEA (2020). Projected costs of generating electricity (2020 ed.). International Energy Reitzenstein, Alexander, Popp, Rebekka, Oei, Pao-Yu, Brauers, Hanna, Stognief, Nora, &
Agency (IEA). Kemfert, Claudia (2021). Structural change in coal regions as a process of economic
IEA (2020b). Available online atSolar PV module shipments by country of origin, 2012-2019. and social-ecological transition – Lessons learnt from structural change processes in
https://www.iea.org/data-and-statistics/charts/solar-pv-module-shipments-by- Germany. Available online atGerman Environment Agency.
country-of-origin-2012-2019. Schmidt, Tobias S. (2014). Low-carbon investment risks and de-risking. Nature Climate
IEA (2020c). World Energy Outlook 2020. Paris: International Energy Agency (IEA). https:/ Change, 4(4), 237–239. https://doi.org/10.1038/nclimate2112.
https://www.iea.org/reports/world-energy-outlook-2020. Solargis (2020). Global solar atlas 2.0Available online at. https://globalsolaratlas.info.
IESR (2019). A roadmap for Indonesia’s power sector: How renewable energy can power Staffell, Iain, & Pfenninger, Stefan (2016). Using bias-corrected reanalysis to simulate cur-
Java-Bali and Sumatra. Jakarta: Institute for Essential Services Reform (IESR). rent and future wind power output. Energy, 114, 1224–1239. https://doi.org/10.1016/
IESR (2021a). Beyond 207 gigawatts: Unleashing Indonesia's solar potential. Institute for Es- j.energy.2016.08.068.
sential Services Reform (IESR). Steffen, Bjarne (2020). Estimating the cost of capital for renewable energy projects.
IESR (2021b). Deep decarbonization of Indonesia’s energy system: A pathway to zero emis- Energy Economics, 88, Article 104783. https://doi.org/10.1016/j.eneco.2020.104783.
sions by 2050. Available online atInstitute for Essential Services Reform (IESR). Stocks, Matthew, Blakers, Andrew, Cheng, Cheng, & Lu, Bin (2019). Towards 100% renew-
https://iesr.or.id/en/pustaka/deep-decarbonization-of-indonesias-energy-system-a- able electricity for Indonesia: The role for solar and pumped hydro storage. 2019 in-
pathway-to-zero-emissions-by-2050. ternational conference on technologies and policies in electric power & energy. 2019
IESR (2021c). Indonesia Energy Transition Outlook 2021. Jakarta: Tracking Progress of En- International Conference on Technologies and Policies in Electric Power & Energy.
ergy Transition in Indonesia. Institute for Essential Services Reform (IESR). Yogyakarta, Indonesia, 10/21/2019 - 10/22/2019. (pp. 1–4). IEEE.
IMF (2022). International Financial Statistic. Available at2021.IMF Data. Interna- Tong, Dan, Zhang, Qiang, Zheng, Yixuan, Caldeira, Ken, Shearer, Christine, & Hong,
tional financial statistics. https://data.imf.org/?sk=4c514d48-b6ba-49ed- Chaopeng (2019). Committed emissions from existing energy infrastructure jeopar-
8ab9-52b0c1a0179b. dize 1.5 °C climate target. Nature, 572(7769), 373–377. https://doi.org/10.1038/
Inchauste, Gabriela, & Victor, David G. (2017). The political economy of energy subsidy re- s41586-019-1364-3.
form. . https://doi.org/10.1596/978-1-4648-1007-7 Available online at. Ueckerdt, Falko, Hirth, Lion, Luderer, Gunnar, & Edenhofer, Ottmar (2013). System LCOE:
Indonesia, G. B. G. (2021). Global business guide Indonesia - Going local: Understanding What are the costs of variable renewables? Energy, 63, 61–75. https://doi.org/10.
Indonesia’s local content requirementsAvailable online at. http://www.gbgindonesia. 1016/j.energy.2013.10.072.
com/en/main/business_updates/2014/upd_going_local_understanding_indonesia_s_ Veysey, Jason, Kamp-Benedict, Eric, Binnington, Taylor, & Heaps, Charlie (2021). Next en-
local_content_requirements.php. ergy modelling system for optimization (NEMO). Version 1.4. Stockholm Environment
IRENA (2017). Electricity storage and renewables: Costs and markets to 2030. International Institute.
Renewable Energy Agency (IRENA). https://www.irena.org/publications/2017/Oct/ Victor, David, & Heller, Thomas C. (2007). The political economy of power sector reform.
Electricity-storage-and-renewables-costs-and-markets. In Adilson Oliveira, Chi Zhang, Rahul Tongia, & Armando Jiménez Carreón-Rodríguez
IRENA (2019). Innovation landscape brief: Flexibility in conventional power plants. Available (Eds.), The Experiences of Five Major Developing Countries. Cambridge University Press.
online atAbu Dhabi: International Renewable Energy Agency checked on 10/22/ Waissbein, Oliver, Glemarec, Yannick, Bayraktar, Hande, & Schmidt, Tobias S. (2013).
2021https://www.irena.org/-/media/Files/IRENA/Agency/Publication/2019/Sep/ Derisking renewable energy investment. A framework to support policymakers in
IRENA_Flexibility_in_CPPs_2019.pdf?la=en&hash= selecting public instruments to promote renewable energy investment in developing
AF60106EA083E492638D8FA9ADF7FD099259F5A1. countriesAvailable online at.
IRENA (2020). Renewable power generation costs in 2019. Abu Dhabi: International Re- Winarko, Tatok, Hariyanto, Nanang, Rahman, Fathin Saifur, Watanabe, Masayuki, &
newable Energy Agency. Mitani, Yasunori (2019). Cost-benefit analysis of PV penetration and its impact on
IRENA (2021a). Capacity statistics time seriesAvailable at. https://www.irena.org/Statistics/ the frequency stability: Case study of the south-central Kalimantan system. 2019
View-Data-by-Topic/Capacity-and-Generation/Statistics-Time-Series. IEEE PES International Conference on Innovative Smart Grid Technologies Asia (ISGT
IRENA (2021). Renewable power generation costs in 2020. Abu Dhabi: International Re- 2019). May 21-24, 2019, Chengdu, China. 2019 IEEE Innovative Smart Grid Technologies
newable Energy Agency. - Asia (ISGT Asia). Chengdu, China, 5/21/2019 - 5/24/2019. (pp. 1700–1705).
The political economy of coal. Available online atJakob, Michael, & Steckel, Jan C. (Eds.). Piscataway, NJ: IEEE.
(2022). Obstacles to clean energy transitions. Taylor and Francis Group. https:// Woodman, Bridget, Mitchell, Catherine, Ragwitz, Mario, & Ordonez, Jose Antonio (2017).
www.taylorfrancis.com/books/oa-edit/10.4324/9781003044543/political-economy- Economic and non-economic barriers and drivers for the uptake of renewablesAvailable
coal-michael-jakob-jan-steckel. online at. https://escholarship.org/uc/item/1nt311xc.
Jakob, Michael, Chen, Claudine, Fuss, Sabine, Marxen, Annika, Rao, Narasimha, & WRI (2021). Climate watch historical GHG emissions. Washington, D.C. Available online at
Edenhofer, Ottmar (2015). Using carbon pricing revenues to finance infrastucture checked on 9/1/2021https://www.climatewatchdata.org/ghg-emissions.

363