Available via license: CC BY 4.0
Content may be subject to copyright.
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2017.Doi Number
ACT on RE+FLEX: Accelerating Coal Transition
through Repurposing Coal Plants into
Renewable and Flexibility Centers
Z. Huang, I. Smolenova, D. Chattopadhyay, C. Govindarajalu, J. de Wit, T. Remy, I. Curiel
The World Bank, Washington DC 20433, USA
1
Corresponding author: Debabrata Chattopadhyay (dchattopadhyay@worldbank.org)
ABSTRACT Decarbonizing the power sector forms a critical part of the global combat against climate
change. This requires inter alia retirement of the global coal power plant fleet of 2,100 GW. Although a
significant part of this capacity is aging, there are complex issues that need to be addressed including the
economic viability of existing coal plants in some countries relative to renewable projects and barriers to exit
of coal. We have used detailed power plant level operational cost data for ten developing countries with
significant share of coal and compared these with levelized cost of renewables, to demonstrate that
competitiveness of coal varies significantly across different geographies. Countries like India where
renewable projects have been highly competitive and there is an aging fleet of coal plants many of which are
far away from mines, are already highly uncompetitive. On the other hand, countries like South Africa that
have relatively inexpensive coal plants, but the average cost of renewable projects have not yet dropped
sufficiently (as of 2020), will require special efforts to phase out coal completely beyond plants that have
reached, or gone well past their technical life. Accelerated retirement of coal would require a new business
model that allows repurposing some of these sites for alternative usage including generation from renewables,
conversion of the incumbent generator into a synchronous condenser coupled with a fly wheel to provide
reactive power and inertia; and installation of energy storage systems. As a repurposed coal plant for energy
related activities can retain part of the workforce, it can also address some of the complex social issues. In
order to develop a comprehensive repurposing program at a national level, the process needs to follow a least-
cost planning methodology to identify prospective coal plant candidates for repurposing and then undertake
a cost-benefit analysis of individual projects. We have demonstrated this methodology using a case study for
Morocco.
INDEX TERMS Climate Change, Renewable Power, Coal Plant Repurposing, Least Cost Planning.
I. INTRODUCTION
A. CONTEXT
Global CO2 emissions reached 33.4 billion/giga tonnes (bt
or gt) in 2019 with significant contribution from the
developing nations over the last two decades, reducing the
contribution from US and Europe to approximately a third of
total emissions. China had emissions in excess of 10 bt
followed by India (2.6 bt). USA, in comparison, had 5.3 bt
1
The opinions and views presented in this paper are his own and do not necessarily represent the views of the International Bank for Reconstruction and
Development/World Bank or its affiliated organizations.
emissions in 2019 or half of that in China, albeit its cumulative
emissions till 2019 of 410 bt is close to double that of China.
Coal-fired power generation globally accounts for more than
10 bt of emissions and is the single largest source of emissions.
The growth of coal-based generation capacity in the
developing world has been very significant including China
(1,000+ GW), India (200+ GW), South Africa (40+ GW).
There are as many as six developing countries in the top ten
coal-based power systems. Reducing emissions from power
generation, especially coal-fired power generation, has been
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
extensively discussed over at least past three decades. Yet, the
global installed coal-fired capacity has doubled from 1,066
GW in 2000 [1] to 2,125 GW in 2020 despite the falling cost
of wind, solar and battery storage, particularly over the past
decade. As of May 2021, there is additional 180 GW coal
capacity being constructed and another 320 GW in planning
stage [2]. A large part of the new capacity is in the developing
nations. The number of countries with coal capacity stands at
80 and there are still as many as 13 new countries that may be
added to this list [1]. On the bright side, coal-based generation
(as opposed to capacity) plateaued over the last five years and
even decreased for the first time in 2020, a lbeit a serious
contraction in demand due to COVID-19 contributed to it.
More importantly, retirement of the older fleet has
accelerated and nearly 300 GW of coal capacity has shut down
over the past five years, mostly in the United States, China,
and Europe. There are also voluntary initiatives promoting
coal phase out such as the Powering Past Coal initiative led
by the UK and Canada [3], that promises to phase out coal
from 19 countries. This however accounts for less than 5% of
the installed 2.1 TW capacity. As the developed nations that
account for more than 60% of the cumulative CO2 emissions,
have already started taking actions including phasing out coal,
our endeavor here is to focus on some of the coal-heavy
developing nations to explore the economics of coal and ways
to accelerate the transition in these countries. There are
analyses that suggest a large share – somewhere in the range
of 65%-80+% will need to be retired by 2030 to contain the
global temperature increase to two degree Centigrade. There
has been significant optimism that coal capacity will decrease
as cleaner generation has reached grid parity and this is going
to happen rapidly with 42% of the global capacity already
making a loss back in 2018 [4]. Others have a more cautious
view on grid parity and noted that parity will depend on system
and country characteristics [5]. Given that the optimism on
coal retirement has been out there for several years but at the
same time growth in coal capacity has largely continued
unabated over the last decade, there is clearly something amiss
about the predictions that heralded a rapid end of coal.
It is also worth noting that there have been multiple policy
initiatives not the least of which is a Nationally Determined
Contributions (NDC) formulated under the Paris Agreement
in 2015 that were submitted by 192 countries worldwide
accounting for 96% of the global greenhouse gas emissions.
However, a lack of transparency of the NDC targets limited
their effectiveness including any discernible trend in reduction
of coal capacity. There have been other policy drives like
renewable energy targets that have been more effective in
pushing fossil fuel out of the generation mix. Policies around
energy efficiency initiatives, specific clean technology
promotion including battery storage in recent years have also
played a role, albeit to a much lesser extent compared to RE
targets.
One of the factors contributing to coal plants not shutting
down anywhere near as fast as was predicted is that existing
coal plants that are 20 years or older have their capital
investment largely depreciated and the cost of operating these
plants is relatively low in many cases. In several countries that
have substantial domestic coal resources, the cost of keeping
coal generators online is particularly low (e.g., $10-30/MWh).
At the same time, notwithstanding the low winning auction
prices for solar/wind celebrated in the media, average
solar/wind power purchase agreements (PPA) in many
developing countries tend to be significantly higher than those
record low prices [6]. There are in fact several factors that
might explain the low auction prices [7] and there is routinely
a difference found between auction and PPA prices (as well as
levelized cost of electricity or LCOE) [6]. If the LCOE of
solar/wind is not below the short run marginal cost (SRMC)
of operating an existing plant, the coal plant could
economically operate, as long as the gap between LCOE and
SRMC is high enough to cover the fixed operation and
maintenance (FOM) costs. Our first focus in this paper is to
undertake a rigorous comparison of LCOE of renewable
projects based on IRENA’s 2019 data and realistic cost data
for coal plants in a number of developing countries deployed
as part of power system planning studies in these countries. As
a simple first step, it can provide useful insights into the share
of coal fleet and specific power plants that have already
become (as of 2020) uncompetitive to renewables. It is
important to develop a clear perspective on where coal sits
relative to renewable energy so that policymakers can make
informed decisions with respect to decarbonization.
Repurposing old coal plants/mines [8] can form a part of
such policy decisions because they can support
renewable/storage development at the existing coal plant site
(e.g., solar PV on ash ponds, battery storage or BESS near the
substation, etc) as well as re-use some of the assets (e.g.,
substation and connection assets, generator, etc). A
repurposed coal plant site can accommodate some renewable
(RE) generation, albeit it may replenish only a small share of
the incumbent coal plant generation. More importantly, the
site ca n become a “flexibility” (FLEX) center to replenish
bulk of the reactive power by converting an existing generator
into a synchronous condenser (SYNCON) as well as
frequency control ancillary service (e.g., through flywheel
attached to SYNCON and from a BESS) and inertia (through
flywheel). As many power systems make a transition to a high
level of RE penetration, such RE+FLEX centers on
repurposed coal plant sites could be an attractive proposition.
This can be a critical part of programs like Accelerating Coal
Transition (ACT) [8] because repurposing can make a good
business case for plants to shut down well before they
complete their technical life [9]. The new project can also re-
employ part of the workforce and continue an energy business
on the site that can partially address some of the difficult social
issues that often prolong the life of coal plants [10].
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
Identification of coal projects that have become
uneconomic is a key first step that can start with a simple
LCOE analysis. However, this will need a more extensive
planning study to assess the requirements for generation (and
ancillary services) from a system perspective, namely: can
these coal plants be safely retired and repurposed for the
system to meet demand and continue to operate in a secure
way? Such a planning analysis may identify economic
retirement opportunities ahead of reaching the technical plant
life, especially as RE costs drop and the O&M costs of an old
plant typically rise rapidly towards the end of its life. Once a
set of plants is identified to be potential candidates, a cost-
benefit analysis for each individual plant will need to be
conducted to ascertain if repurposing as a RE+FLEX facility
can bring sufficient benefits to warrant the investment and
mitigate the coal decommissioning impacts.
B. SCOPE OF THIS PAPER
In this paper, we present a set of 11 developing country case
studies where coal plays a significant role (Table 1). First, we
present a set of statistics to compare the relative merit or coal
vs RE across 10 of these countries with a large fleet of older
coal plants; Second, we provide a more detailed examination
of a few key countries where coal retirement is a genuine
prospect; Third, we discuss the case for Morocco which is a
smaller system with a relatively newer coal power plant fleet.
Since Morocco has taken a strategic position to integrate large-
scale RE, we show how planning studies should be conducted
to identify coal plants that can be retired ahead of their
economic life for which repurposing benefits outweigh costs.
Countries covered in this study are India, Turkey, South
Africa, Ukraine, Vietnam, Pakistan, Kazakhstan,
Uzbekistan, Bulgaria, Bosnia-Herzegovina, and Morocco.
All of these countries rely on coal for power generation for a
significant share of its total generation, but also possess
significant renewable resource potential. These 11 countries
account for 11% of the global generating capacity but 17% of
the global coal power capacity. These countries are home to
26% of the global population but represent only 6% of the
global GDP (Table 1) and electricity consumption growth will
continue to be higher than their developed counterparts.
Economics of coal and renewable assumes special
significance when we consider the trade-off between the need
for growth to be fueled by low-cost electricity and a
disproportionate level of new investments needed in clean
technologies. This work is important to establish the relative
economics of coal and renewable which vary significantly
across geographies – an issue that is surprisingly arcane, but
quite fundamental to formulate carbon reduction strategies.
We also introduce a methodology to evaluate coal plant
repurposing embedded into sophisticated least-cost planning
models. Repurposing as we discuss can also be an important
part of a strategy to accelerate retirement of coal units and
support development of renewables in a financially
sustainable way.
II. KEY LITERATURE ON COAL vs RENEWABLES
Coal based generation has been under scrutiny since the
nineties, but serious intent and action to retire coal generating
assets been visible only in a select number of countries over
the last decade. Inadequate systematic and detailed evidence
of competitiveness of existing coal plants vis-à-vis
renewable energy has contributed to significant confusion
over the share of coal generation that should be closed down
purely on economic grounds. Some predicted that up to 42%
of global coal power plants were running at a loss in 2018,
which will grow to 96% by 2030 [4]. More than 50% of
renewable capacity added in 2019 is considered to cost less
than the cheapest new coal plants [6], and 54% of coal-fired
power generation in the EU is cash-flow negative after debt
servicing, with the US trailing slightly behind at 48% [4].
Other more conservative sources estimate a minimum of
TABLE I: COAL-DOMINATED DEVELOPING COUNTRIES COVERED IN THE ANALYSIS
Country
System Size
(MW)
Coal (MW) and
%*
Electricity
consumption
(TWh)**
Coal
consumption
(mt)***
GDP (B$)****
Population
(M)****
India
350,149
226,500 (65%)
1,309
986
2,869
1,366
Turkey
80,353
16,979 (22%)
272
134
761
83
South Africa
56,121
41,674 (75%)
229
196
351
59
Ukraine
52,047
24,663 (48%)
130
51
154
44
Vietnam
41,862
14,491 (35%)
227
74
262
96
Pakistan
34,178
3,045 (9%)
126
21
278
217
Kazakhstan
20,179
12,216 (61%)
101
97
182
19
Uzbekistan
14,455
2,650 (19%)
58
6
58
34
Bulgaria
11,672
4,614 (40%)
35
34
69
7
Bosnia-Herzegovina
4,211
2,050 (49%)
12
9
20
3
Morocco
8,278
2,663 (32%)
35
8
120
36
TOTAL
673,505
351,545 (52%)
2,534
1,617
5,124
1,965
World
6,303,594
2,007,214 (32%)
2,4739
8,622
87,799
7,674
Global share of selected
countries (%)
11%
18%
10%
19%
6%
26%
Source: World Electric Power Plants Database, Platts, 2018 ; International Energy Agency (IEA); Energy Information Administration (EIA); World Bank
* Share of coal-based capacity ** 2018/19 data from IEA ***2018 data from EIA ****2019 data from World Bank
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
20% of coal power generation (or 400 GW) was
uncompetitive in 2018 relative to solar and wind, which will
increase to 50 % (or 1000 GW) by 2030 [8]. In India, roughly
71 GW of privately-owned coal-fired capacity is expected to
soon face potential financial distress [11]. The phaseout of
coal power has been claimed for several years in these
studies [4,6,8,11], yet a lack of consistency in data and
objective analysis makes it difficult to assess when this is
likely to occur. Although the genesis of this finding has been
in low winning auction prices for wind followed by solar
over the past decade, we discuss below several important
nuances that would render a direct comparison between
auction prices and cost of coal generation difficult.
The decline in solar and wind technology costs and their
‘grid parity’ is the key rea son why m a ny new coal plants
around the world are struggling and face delay or are being
abandoned. The LCOE of newly constructed coal plants is
undercut by the LCOE of new onshore wind, utility-scale
solar, and combined-cycle gas turbine generation [11]. Low
auction prices and power purchase agreements for solar in
UAE (Abu Dhabi and Dubai), Chile, Ethiopia, Mexico, Peru
and Saudi Arabia, dropping as low as USD 0.03/kWh [6],
reinforced the view that unsubsidized solar will soon
overtake fossil fuel-based generation. The average price of
auctions and power purchase agreements of solar PV
projects that will be commissioned in 2021 in India is
currently at USD 0.039/kWh according to IRENA [6].
Winning auction prices, however, represent only a fraction
of the total renewable capacity to be commissioned and are
heavily influenced by country's resource potential, financing
condition, and auction design [12]. The lowest winning
auction price of solar/wind in one country/region may not be
representative of conditions elsewhere, or representative of
the full spectrum of projects in the same country/region.
Replicating such promising results in other countries and
regions can be challenging, particularly in some developing
countries where RE projects continue to be expensive. A
close examination of the rapid price decline in UAE and
Saudi Arabia points to several factors depressing the price,
such as forward-bidding on expected future decline in
hardware prices, low operation and maintenance labor costs
in the Gulf region, higher utilization via scaling production,
extension of PPAs to 25 years as opposed to typical 20 years,
favorable financial terms and low or no land costs [7].
LCOE, on the other hand, represents the minimum price
that offsets all direct costs associated with generation,
including capital expenditures, operational expenditures, and
debt service costs, and has generally been higher than
winning auction prices. Distinct from winning auction price,
LCOE is considered to be a more appropriate metric for
economic analysis, as it fully and transparently captures
economic costs over the lifetime of generating technology.
In the past decade, LCOEs for RE have indeed dropped
massively and are poised for further reduction. The global
weighted-average LCOE of utility-scale solar PV declined
by 82% between 2010 and 2019, followed by concentrated
solar power (47%), onshore wind (39%) and offshore wind
(29%) [6]. In general, LCOE reflects the substantial lifetime
cost of renewable generation better than auction prices.
A comparison between investing in renewable energy and
operating coal power based on LCOE of renewable energy is
also problematic. This is because the latter rarely, if ever,
incorporates indirect costs associated with grid integration.
The costs associated with renewable supply intermittency,
the higher transmission and transportation costs from
geographic locations favorable for renewables to urban
areas, and expenses associated with prematurely closed
generations [13] increase the societal costs of solar and wind
technologies that ultimately need to be borne by customers.
Due to such indirect costs, US states that adopted Renewable
Portfolio Standards (RPS) incurred 11% higher electricity
prices seven years after passage of the policy, compared to
states that had not adopted RPS [13].
Retirement of a coal plant, even those that are well below
LCOE of solar/wind, has its own challenges. Coal plant
decommissioning is expensive and is further constrained by
hidden social and political costs. The investment
requirement for technical decommissioning of a coal-fired
facility can be $100-200/kW or up to $200 million for a
1,000 MW coal plant [8]. A basic social program that ensures
‘Just Transition’ for workers is estima ted to add 20% to the
decommissioning costs of a project [10].
Once we consider all of the hidden costs that need to be
added to the observed low auction prices for solar and the
full decommissioning costs of a coal plants – it is not hard to
see why coal plants are not shutting down as rapidly as has
been promulgated in some of the forums. It often requires a
political commitment by countries who can afford to make
the transition going beyond economics. While coal-fired
generation has decreased in most high-income countries,
most notably in the EU and the USA (-19% and -14% in 2019
respectively [14], coal still reigns in many parts of Asia.
Across ASEAN countries, coal-fired generation grew by
14% in 2019 and the region is on its path to become the third-
largest coal consuming region by 2025, followed by the US
and the European Union [14]. According to a study published
in 2020, 90% of the new coal-fired capacity that was built in
the first half of 2020 was commissioned in China as China
quickly recovered from COVID-19 pandemic with new coal
proposals and permits [15]. Coal consumption is expected to
increase in the world’s most populous country by 2.6% in
2021 [14].
As the discussion so far alludes to, one has to take a
sobering view on the speed at which coal plants are likely to
be decommissioned. This is not to suggest that the Great
Transition [16] will not happen, nor to deny a substantial cost
advantages of renewables over coal that will emerge [17].
However, we will need to understand far better the
economics of coal relative to renewables in different parts of
the world and also develop transition trajectories and
business models for shutting down coal plants. There is a
growing volume on this topic [18-42] that we have discussed
over the remainder of this paper including the case studies
that are presented in subsequent sections.
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
Coal plant repurposing [8,9,18] is a new business model
tha t requires more a ttention. It is not a “new” technology per
se as it is more about packaging technologies like converting
existing generators into synchronous condensers [19] and
using the site for solar PV, BESS, thermal storage, biomass
etc. As a recent analysis [20] shows a repurposing project (in
India) can pay as much as five times of the cost of
decommissioning. This is particularly good news as the
significant cost of decommissioning in the US [21] has
demonstrated, such costs can be a significant barrier to exit.
That said, some countries have also managed the retirement
of coal generators well. Canada exploited the existing
infrastructure and turned the decommissioned coal plant site
into a 44-MW solar power station in Nanticoke, Ontario [22].
Through building solar park projects in coal mining regions,
Germany created employment for 80,000 people in
Brandenburg, Saxony, and North Rhine-Westphalia,
outperforming the employment from coal power industry in
these regions [23, 24]. Such models can build optimism
towards a faster decline of coal in places like Chile [25] and
South Africa [26] where the concept of repurposing is
gaining momentum. Wider technology choices are also
available for coal repurposing projects such as concentrated
solar power (CSP) and heat exchangers, conversion of coal
plants to run on biomass, conversion of adjoining coal mines
to pumped storage hydro, usage of the site to develop wind
projects, etc.
III. ANALYSIS OF COST OF COAL VS RENEWABLES
The first part of our analysis covers an objective
comparison of LCOE of renewables with the short run
marginal cost of older coal plants in ten developing nations.
This is a stringent but practical test of competitiveness of
renewables over coal.
In most power systems around the globe, fossil fuels such
as coal have taken a dominant position in electricity
generation for decades, and the commercial deployment of
large-scale renewable energy projects was not rolled out
until the 2010s. As of 2018, 2007 GW of coal plants were
operating in the world, 17% of which built in the
1950s~1970s and have essentially reached, if not gone well
past, the end of the design life, given an average age of these
plants of 50~60 years [27]. China, United States, and India
are the top three countries in terms of capacity in operating
coal plants, contributing to over 70% of the global coal
capacity (Figure 1).
FIGURE 1. Operating coal capacity by commissioning year
Source: World Electric Power Plants Database, Platts, 2018
A. METHODOLOGY
To explore the economic merits of RE displacing coal “at
the margin”, this study compares the generation cost of coal
plants with that of renewable energy technologies at the
country level. With the capital investment of extant coal
plants considered as sunk costs, the relevant generation cost
for a coal plant is limited to its short-run marginal cost
(SRMC) including levelized fixed operation and
maintenance costs. In contrast, replacement with renewable
energy will require the construction of additional capacity.
Therefore, the cost of renewable electricity is assessed using
long-run marginal cost (LRMC). We have used the LCOE as
a proxy for LRMC. LCOE is a project-centric metric as
opposed to LRMC of supply which is based on the premise
of an optimized system to meet incremental demand [28].
We are effectively making an implicit assumption that solar
(or wind whichever is cheaper in a system, or more likely a
mix of the two) will be the dominant form of supply in the
future system.
In other words, the analysis is based on the comparison
between SRMC of operating coal capacity and
LRMC/LCOE of renewable energy, to see if building new
RE capacity to produce a marginal kWh would displace a
marginal kWh of coal. This is a more stringent test for
renewables as the fixed capital cost of existing coal assets are
effectively treated as sunk. It differs from most of the
comparisons in the literature that compare LCOE of coal
against renewables. We argue that this test, notwithstanding
its stringent nature, is a better reflection of reality. A coal
plant will continue to be in service as base load as long as it
can recover its SRMC given the significant cost of exit
(decommissioning costs) and the fact that many of these
older coal plants have depreciated fully. A comparison of
LCOE of renewables with SRMC of coal can be helpful to
identify countries with uneconomic coal capacity that can be
shut down to accelerate energy transition, as well as to
identify the old coal plants suitable for repurposing.
102
417 371
137
110 75
6857
100 123 93
72
106
0
100
200
300
400
500
600
700
Ealier
than
1950
1950s 1960s 1970s 1980s 1990s 2000s 2010s
Capacity (GW)
Other countries
United States
India
China
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
1) Definition of costs
SRMC of coal generation consists of the fixed operation
and maintenance cost (FOM), variable operation and
maintenance cost (VOM) proportional to the generation
activities, as well as fuel cost. Although the conventional
definition of SRMC does not include FOM, a significant part
of these ‘fixed’ costs relate to maintaining the dispatch status
of coal plants, especially the older coal fleet and as such we
have included a levelized component of FOM (for instance,
regular maintenance and inspection costs) in our definition
that holds significant implications for its relative economics
as these plants lose dispatch over the years. SRMC (in
$/MWh) of coal electricity was calculated at the plant level
in the study, the formula of which is written below.
𝑆𝑅𝑀𝐶 = 𝐹𝑂𝑀
𝐶𝐹 ∗8760 ℎ𝑟𝑠 + 𝐹𝑢𝑒𝑙 𝐶𝑜𝑠𝑡 × 𝐻𝑅 + 𝑉𝑂𝑀
The units of FOM, Fuel Cost, HR and VOM are
respectively $/MW/year, $/MMBtu, MMBtu/MWh, and
$/MWh. CF stands for the capacity factor of the plants, a
ratio of actual annual generation over maximum generation
to evaluate the plant utilization. HR is the Heat Rate.
LCOE of renewable energy covers capital costs, FOM,
VOM, and fuel cost:
𝐿𝐶𝑂𝐸 = 𝐶𝐶 × 𝐶𝑅𝐹 + 𝐹𝑂𝑀
𝐶𝐹 ×8760 ℎ𝑟𝑠 + 𝐹𝑢𝑒𝑙 𝐶𝑜𝑠𝑡 × 𝐻𝑅 + 𝑉𝑂𝑀
CC is the capital cost to build the solar/wind farm
overnight ($/MW). CRF, Capital Recovery Factor, is a rate
of repayment that translates overnight capital costs ($/MW)
into annualized costs ($/MW/year), determined by discount
rate and lifetime of the investment. We have used the LCOE
of wind and solar PV at the country level from IRENA
estimates [6] as a proxy for LRMC.
2) Assumptions
Several assumptions were necessary due to data
limitations and also to make results comparable across the
country case studies: (a) baseline capacity factor is assumed
to be 60% for all existing coal plants to maintain
comparability; (b) the commissioning year of the plant is
determined by the oldest operating unit; (c) additional
transmission costs to connect renewable energy and grid
balancing costs are ignored that may in fact be quite
significant at higher level of renewable penetration[29] – this
assumption in a way offsets for the fact that capital costs for
existing coal is treated as sunk; (d) no carbon tax is imposed
on the generation; (e) all costs are in USD (2020 value).
3) Sensitivity Analysis
Sensitivity analysis was also conducted to explore how the
relative economics of the older fleet of coal plants would
change as maintenance costs increase or as plant availability
and efficiency of these plants and hence utilization, drop over
the years. On the other hand, renewable energy LCOE is
projected to drop sharply. Therefore, relative economics of
coal would deteriorate considerably. Important factors to
consider in the sensitivity analysis include decreasing
renewable energy cost, declining utilization of coal plants,
and increasing FOM of coal plants. These factors have been
chosen to reflect the policy initiatives that most countries
have adopted. As we have alluded to in the introductory
section, NDCs per se have been ineffectual due to a lack of
transparency, coherence, practical steps to implement and
also a lack of sufficient degree of ambition [30]. That said,
promotion of renewable energy either as a plank in the NDC
in some cases, or as a standalone renewable energy target in
significantly more cases, has been a major driver. This has
directly and through indirect paths too, led to a reduction in
RE technology costs, especially for solar and wind, rapid
adoption of these technologies, and these in turn have
impacted on utilization of coal in some countries. A drop in
utilization of coal also translates into a higher level of
cycling, hence more wear and tear, and overall a significant
rise in the levelized fixed O&M costs for coal plants.
B. LIMITATIONS OF THE ANALYSIS
The cost comparison analysis makes simplifying
assumptions to match available data and keep it transparent,
namely:
a. It ignored the utilization difference of coal plants across
age groups for all countries, conservatively estimating
the SRMC of some old plants with low utilization, while
overestimating the SRMC of highly utilized plants. We
overcome this limitation to some extent using the
sensitivities conducted varying the utilization rate of
coal plants. We also recognize that a proper assessment
at a system level would ultimately require a planning
study to identify the prioritized plants for retirement as
we discuss in the next section;
b. We have relied on an average estimate of LCOE of
renewables at the country/region level instead of the
LCOE of individual projects; and
c. It did not cover the extra costs to integrate renewable
electricity from the remote wind/solar farms, the carbon
costs associated with coal generation, or systems
balancing costs.
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
C. DATA SOURCES AND ISSUES
Data for coal plants, such as capacity, status, date of
commissioning, and fuel type, was extracted from the World
Electric Power Plants Database of Platts [27]. Cost data of
coal plants, such as FOM, VOM, heat rate, and fuel costs,
was extracted from the country-level or regional World Bank
- Electricity Planning Model (EPM) database, a least-cost
power modeling platform developed by the World Bank to
simulate economic dispatch at the plant level [31]. This
database is collected from the power system planning
reports, or directly from the utilities in the selected countries.
LCOE projections of wind and solar were collected from
IRENA [6].
As power plant data needed to be merged from Platts and
EPM, one challenge we encountered is that individual plant
names or even capacity did not always match across these
databases. We had to therefore cross-check details through
Platts, EPM, as well as WRI
2
datasets for plant name,
capacity size, and commissioning year. We also needed to
update the status of power plants as plants under construction
may have been commissioned and others that have been
retired since the databases were last updated. Fuel costs used
in EPM represent economic costs that are varied from
countries, zones, and coal types – they do not represent the
plant specific costs or contract prices.
Our economic analysis using Platts and EPM data focuses
on the first ten countries that have significant older coal
capacity noted in Table 1. These countries represented ~349
GW or 17% of the global operating coal capacity in 2018.
India is the largest coal country, accounting for 61% of the
operating coal capacity. South Africa, Turkey, and Ukraine
also occupy a significant share of capacity.
FIGURE 2. Operating coal capacity in selected countries (2020)
Note: IND-India, ZAF-South Africa, TU R-Turkey, VNM-Vietnam, UKR-Ukraine,
KAZ-Kazakhstan, BGR-Bulgaria, PAK-Pakis tan , BIH-Bos nia and Herzegovina,
UZB-Uzbekistan
Morocco, with its relatively new fleet of coal plants,
represents a special case where there is only one 280 MW
unit (or 10% of coal capacity) currently in operation that is
relatively old (34 years) compared to the rest of the fleet (less
than 20-years old units). Morocco is, however, included in
our analysis to demonstrate how a country with significant
2
Global Power Plant Database, WRI.
potential for renewables can accelerate the retirement of its
coal capacity and in particular aim to economically shut
down the 280 MW unit and repurpose the site. We present
this analysis in section V.
IV. RESULTS: COAL vs RENEWABLES ANALYSIS
A. CROSS-COUNTRY ANALYSIS
1) SRMC of coal generation by age group
For the purpose of the analysis, three age groups were
considered: less than 10 years (new); 10~30 years and over
30 years (old). Of the examined countries, 42% of coal plants
are old, especially those in India, South Africa, Ukraine and
Kazakhstan. 36% are new with India accounting for 75% of
the new units, though Vietnam and Pakistan also have
relatively young but small fleets. Countries that have a higher
proportion of old coal capacity are South Africa, Ukraine,
Kazakhstan, Bulgaria, Bosnia and Herzegovina, and
Uzbekistan.
With the default assumptions, the average SRMC of each
age group mostly falls in between $30~$40/MWh, with very
few extremes. Older plants (>30 years) generally have higher
SRMC ($41~$59/MWh) due to higher heat rates, which can
be observed for Turkey and Vietnam. However, this cannot
be generalized as fuel costs account for vast majority of
SRMC and fuel prices vary significantly across geographies.
Older coal plants in India tell quite a different story owing to
their low fuel costs. In India, the older plants are more likely
to use cheaper local coal than some new plants which are
fueled by imported coals or are located far away from the
mines. The abundance of coal reserves in Kazakhstan and
South Africa can also bring down the SRMC to $25/MWh
and $34/MWh, respectively (Figure 3).
2) Competitiveness of coal generation
Coal generation can gradually lose cost advantages as
renewable electricity costs fall. Figure 4 plots the SRMC
range of coal generation of each country (in boxes) to
compare with the LCOE of renewable generation (in dots).
In most countries, the LCOE of solar PV, and to a lesser
extent wind, is higher than the SRMC of coal. There are few
exceptions, such as India and Turkey, showing overlaps
between the LCOE of renewable energy and the SRMC
range of coal plants (Figure 4). This implies installing new
renewable energy capacity can be more economic than
operating an existing coal unit.
Other countries…IND 10%
ZAF 2%
TUR 1%
UKR 1%
VNM 1%
KAZ 0%
BGR 0%
PAK 0%
UZB 0%
BIH 0%
MOR 0%
Selected
countries
17%
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
Among all selected countries, India has the lowest average
LCOE to produce renewable power, namely, $49/MWh for
wind and $45/MWh for solar PV [6]. With SRMC of coal
power ranging from $17/MWh to $51/MWh, India has
already showed some uneconomic coal capacity in the
country, as operating costs are becoming higher than the
average LCOE of renewable electricity.
In Turkey and Ukraine, though the LCOE of renewable
energy was relatively high in 2019/20 ($63~$84/MWh), but
the SRMC of coal generation is even higher in many cases
because of the aging nature of these fleets. The SRMC of
coal ranges from $40/MWh to $72/MWh in Turkey, and
from $61/MWh to $74/MWh in Ukraine (Figure 4). Using
2019/20 data even with relatively high cost of renewable
projects, it was economically feasible to repurpose some coal
plants that are more costly than wind or solar in both
countries. As the cost of renewable projects in these two
countries poised to drop sharply, retirement of coal is a
distinct possibility in the near future.
The remaining countries have an average LCOE of
renewable energy ranging from $68/MWh to $86/MWh,
much higher than the coal generation costs, especially for the
countries with cheap coal resources. South Africa is one of
the most representative countries, where coal generation
leads in the power dispatch with an estimated SRMC of
$33~$42/MWh (Figure 4). Other countries, such as Vietnam,
Kazakhstan, and Pakistan, also follow very similar patterns
as South Africa. Although South Africa has already seen 5.4
GW of its retired coal capacity that are under active
consideration for being repurposed with solar PV [32] and
has extensive plans to retire up to 10.5 GW of capacity over
this decade [33], early retirement and repurposing of the rest
of the fleet will need a close economic scrutiny in the short
term given relatively low SRMC of coal in the country. As
solar/wind and battery storage cost drop over the years and
the remaining fleet loses efficiency, the relative
competitiveness of coal will erode in the long term, as
discussed further below.
3) Sensitivity analysis
The scale and scope of uneconomic coal capacity can be
affected by multiple factors, such as learning curve of
renewable energy cost and utilization of coal plants.
Sensitivity analysis is helpful to quantify the impacts from
some input variables to simulate their trends in the future,
including aggressively downward trend of renewable energy
cost, decreasing utilization of coal capacity, and rising FOM
costs as more plants age. Policy initiatives inter alia RE
FIGURE 3. Operating coal capacity and short-run marginal cost (SRMC) by age group
Note: All LCOE data was from the Renewable Power Generation Costs of IRENA [6]. When country-level data was not available, the LCOE of the region or a nearby
country was used for the cou ntries, such as Bulgaria, Pakistan, BIH, and Uzbekistan.
FIGURE 4. Cost competitiveness of existing coal capacity versus local renewable resources
Note (Figure 3 & 4): IND-India, ZAF-South Africa, TUR-Turkey, VNM-Vietnam, UKR-Ukraine, KAZ-Kazakhstan, BGR-Bulgaria, PAK-Pakistan, BIH-Bosnia and
Herzegovina, UZB-Uzbekistan
92.3
5.6 4.5
15.2
0.7 4.9 0.3
58.9
4.1 8.5 2.2 0.1 0.1 0.1
55.7
33.7
6.0
23.7
1.4
12.1 3.9 1.7 2.7
33 33
46
36
45 40
34 34
45
36
54
32 34
59 68
41
25
39 39
30
-10
10
30
50
70
0
20
40
60
80
100
IND ZAF TUR UKR VNM KAZ BGR PAK BIH UZB
SRMC ($/MWh)
Total Capacity (GW)
Capacity: < 10 years Capacity: 10~30 years Capacity: > 30 years
Average SRMC: < 10 years Average SRMC: 10~30 years Average SRMC: > 30 years
32.9 33.9
49.8
67.6
36.5
25.1
39.9 40.9 39.4 30.0
45
81 78 84 82 86 78 68 78 86
49
68 63
75 80 64 67
85
67 64
0
20
40
60
80
100
IND ZAF TUR UKR VNM KAZ BGR PAK BIH UZB
Unit: $/MWh
SRMC (Range) SRMC (Min) SRMC (Max) SRMC (Average) PV LCOE (2019) Wind LCOE (2019)
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
targets that provided the scale up, coupled with technology
breakthrough have already lowered solar and wind capital
costs significantly. Projections by IRENA among others
indicate average annual capital cost reduction from 1%-3%
pa or approximately 10%-30% over this decade. Coal plant
fleetwide average utilization in countries like India has
already dropped significantly from 75% in 2011 down to 50%
in 2020 with the advent of cheaper renewables being one of
the key drivers. Major RE policy initiatives such as the 450
GW target by 2030 in India would push this utilization
further down to 40% [34]. This drop in turn would increase
the levelized FOM charges faced by these plants. We have
adopted these RE cost and coal plant utilization reduction
parameters uniformly across all countries for comparability.
The following sections discuss how these results are
sensitive to changes in key assumptions driven by the RE
policy initiatives:
a. Decreasing renewable energy cost: By assuming a
10%, 20%, and 30% drop in the capital cost of local PV and
wind power from present level, the analysis measured the
influence on coal economics from a steeper learning curve of
renewable electricity cost in the near term (Figure 5,
A1&A2). Given current cost level (0% decrease), India and
Turkey have at most 14 GW and 1 GW coal less economic
than solar PV and wind, respectively.
If costs of solar generation fall by 30% from the level in
2019, the amount of uneconomic coal capacity, with SRMC
higher than solar costs, will reach as high as 135 GW in
selected countries, nearly ten-fold of the amount in the base
case (Figure 5, A1). If LCOE of wind energy reduces by 30%,
the scale of uneconomic coal capacity, with SRMC above
wind costs, will increase massively to 113 GW from 4 GW
in the base case (Figure 5, A2).
Under most circumstances, uneconomic coal capacity will
be concentrated in India, Turkey, and Ukraine, with India
contributing to at least half of this capacity. Specifically, a
drop by 30% in renewable electricity cost from that in 2019
will mean the average LCOE of solar PV or wind falls in
$31~$34/MWh range in India and $40~$66/MWh in other
countries. As renewable cost decreases in the future, both
solar and wind can be more competitive than a large
proportion of coal in India, while wind has a greater
advantage in Turkey and Ukraine. Hence, coal transition can
be significantly accelerated in India, Turkey, and Ukraine,
especially when costs of renewable generation lower by 30%
from the current level.
1. Uneconomic coal capacity with SRMC > LRMC of solar PV
2. Uneconomic coal capacity with SRMC > LRMC of wind
A1) Decreasing renewable energy price (solar)
A2) Decreasing renewable energy price (wind)
B1) Decreasing utilization of coal
B2) Decreasing utilization of coal
C1) Increasing fixed O&M for coal plants
C2) Increasing fixed O&M for coal plant s
Note: IND-India, ZAF-South Africa, TU R-Turkey, VNM-Vietnam, UKR-Ukraine, KAZ-Kazakhstan, BGR-Bulgaria, PAK-Pakistan, BIH-Bosnia and Herzegovina, UZB-
Uzbekistan
FIGURE 5. Sensitivity analysis of unprofitable coal capacity across countries
14 51 69 110
11
4
8
20
0
50
100
150
0% 10% 20% 30%
Uneconomic coal (GW)
Increase of solar LCOE (%)
314
57 81
13
9
11
8
20
20
0
50
100
150
0% 10% 20% 30%
Uneconomic coal (GW)
Increase of wind LCOE (%)
14 14
24
1
0
5
10
15
20
25
30
0.6 0.5 0.4
Unecono mic coal (GW)
Capacity factor of coal plants
3 3 8
1 1
8
2
0
5
10
15
20
25
30
0.6 0.5 0.4
Unecono mic coal (GW)
Capacity factor of coal plants
14 14 14 16
1
0
5
10
15
20
0% 10% 20% 30%
Uneconomic coal (GW)
Fixed O&M of coal plants increases by %
3335
1113
0
5
10
15
20
0% 10% 20% 30%
Uneconomic coal (GW)
Fixed O&M of coal plants increases by %
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
b. Decreasing utilization of coal capacity: When the
capacity factor of all coal plants is reduced to 50% from the
base level (60%), uneconomic coal capacity will increase
from 14 GW to 25GW across the selected countries, mostly
in India (Figure 5, B1). A further decrease of the capacity
factor to 40% pushes the SRMC of an additional 10 GW coal
capacity in Turkey and Ukraine beyond the LCOE for wind
(Figure 5, B2).
c. Increasing FOM of coal plants: Aging coal plants may
require higher fixed operation and maintenance expenditures
to keep them in working order, which pushes up their SRMC.
However, given the relatively small share of FOM in the
SRMC calculations, we find that even a 30% increase in the
FOM only increases uneconomic coal capacity slightly by
2~3 GW from the base case, primarily in India and Turkey
(Figure 5, C1&C2).
In general, our analysis suggests that a reduction in
renewable energy cost may be a bigger driver of global coal
transition than costs related to the age and inefficiency of the
coal power plants. On average, the difference between the
LCOE for RE and the SRMC for coal is the smallest in India,
Turkey, and Ukraine making the economic viability of coal
plant shutdown in these countries worthy of a deeper
investigation. In other countries, such as South Africa, where
LCOE of renewable electricity in 2019/20 [6] is almost twice
the SRMC of existing coal, accelerating the coal transition,
beyond the planned retirement, will rest more heavily on
steeply declining costs of renewable energy. It is possible for
these countries to achieve such a decline, especially for
South Africa, which is naturally endowed with high-quality
solar and wind resources. It should be noted that in 2019/20,
solar costs in South Africa are 50% higher than in Argentina,
70% more than in Brazil, and about twice the costs in India
and China [6], suggesting ample room for improvement.
B. COUNTRY SPECIFIC RESULTS
We have next focused on the four larger systems of India,
South Africa, Turkey, and Ukraine and understand the
economic retirement opportunity of coal in these systems.
1) Comparative costs of coal and RE
To identify the uneconomic coal capacity, we ranked coal
plants in a reverse order of merit and considered the LCOE
of solar/wind as the cost ceiling for economic plants in each
country. Specifically, in Figure 6, SRMC in the order from
the highest to the lowest can be regarded as the supply curve
of uneconomic coal, while the LCOE of renewable
electricity is represented by a horizontal line to determine the
highest possible SRMC for economic coal plants.
a. India: Though a coal-dominated country, India is
leaping ahead in renewable energy development with the
third fastest capacity volume additions around the globe in
recent years [35]. Coal remains of course by far the largest
generation resource in India, with an installed capacity of
about 207 GW and contributing to over 60% in the domestic
capacity mix. Meanwhile, economies of scale in the
manufacturing and installation of utility-scale solar and wind
projects have pushed down the cost of renewable energy.
With the average LCOE of wind ($49/MWh) and PV
($45/MWh) in 2019 as benchmarks, the country has around
14 GW coal plants already unprofitable, that is, 6% of the
operating coal capacity with an SRMC higher than LCOE of
PV (Figure 6). At a weighted average SRMC of $46.4/MWh
for this 14 GW coal, it is around $4/MWh higher than the
LCOE of PV in 2020. This translates into a negative cost of
carbon reduction of (-)$4.46/t. The cost of CO2 reduction of
course turns positive when we look at the cheaper end of coal
plants up to $27.8/t for the cheapest coal plant but on average
below $10/tCO2e across the entire fleet. Furthermore, if
compared with the solar power purchase agreement (PPA)
pricing estimates, $42/MWh in 2020 and $33/MWh in 2021,
more coal plants will be rendered uncompetitive increasing
the share to 16% and 43%, respectively, of the existing coal
capacity (Figure 6).
Around 30% of the coal capacity being operated in India
has an SRMC falling in the range of $30~$40/MWh. If
LCOE of renewable generation drops to about $30/MWh,
over half of the existing coal plants, or more than 110 GW,
will be no longer economic in power generation, especially
the ones fueled by expensive imported coals. Two thirds of
these plants were built after 2005 and are in fact far away
from the planned retirement year, such as Bellary Thermal
Power Station with an estimated SRMC of $49/MWh, due to
cost of coal transportation. Therefore, as costs of renewable
energy are further reduced in the next few years, there is
plenty of potentials for India to repurpose a greater number
of existing coal plants. This conclusion is also supported by
the recent IEEFA analysis [20].
It should be noted that the choice between coal and
renewables in any location is far more complex than a simple
comparison of SRMC of coal and LCOE. The situation in
India underlines these complexities. The cost of integrating
renewables that we have alluded to before, requires attention
by system planners. Even at a relatively low 9.13%
penetration in terms of generation mix in 2019/20, it was
estimated (in July,2020) that the cost of integrating
renewable is around $15/MWh (Rs., 1,110/MWh) [36]
which considers the additional transmission and balancing
costs, but not storage costs. This cost is expected to go up as
the penetration of variable renewable increases over the
years. On the other hand, due to the nature of the legacy
generation contracts, there is a “fixed cost” component for
coal that is akin to capital cost and can be in the range of $10-
26/MWh. From the perspective of a buyer (distribution
com pany in India), the “capital cost” is therefore not sunk. It
has been argued that shutting down 54 coal plants that are
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
older than 20 years would save the distribution companies
a massive $7.2 billion (Rs 530 billion) [37]. However, there
are other views e.g., Tongia [38] that aptly note that a more
holistic comparison of the options needs to be made to
account for integration including storage costs that might
make a case for a combination of renewable and coal to be
economic. The debate around coal in India [36] also noted a
need for substantial reform to tariff paid to coal generators
tha t often include a significant “fixed cost” payment that
would allow coal generators to continue to be in existence
even if its utilization drops.
b. South Africa: Despite high-quality renewable
resources, the renewable generation for South Africa
according to IRENA [6] data is still costly, with an average
LCOE of wind and solar between $68/MWh and $81/MWh.
The low price of local coal resources makes thermal coal
electricity more affordable for the country, with a relatively
low SRMC falling somewhere between $33/MWh and
$43/MWh, almost half of the current wind or solar costs.
Hence, in terms of the existing price gap, coal plants will still
stay prioritized in the power dispatch because of lower costs
(Figure 6).
As of 2020, South Africa has 44 GW coal capacity in
operation, accounting for three quarters of the installed
generation capacity in the country. About 18 GW of existing
coal capacity was built before 1980 and is gradually
approaching the end of technical life (50~60 years) in this
decade. Hence, South Africa also envisioned the renewable
energy development in the Integrated Resource Plan (2019)
[33] to replace 10.5 GW of coal by 2030 and 35 GW by 2050.
The economic and business case of coal retirement would
hinge critically on how fast cost of solar and wind projects
drop over the coming years. When LCOE of renewable
generation drops by at least 40% in the country, some coal
plants will become uneconomic to operate. Such a degree of
cost reduction within the next 10 years is equivalent to a
minimum cost reduction rate of 3.8% per annum which is in
the projected range of cost decline. Given that global
weighted average LCOE of newly commissioned solar PV
and onshore wind has dropped by about 13% and 9% year-
on-year in 2019 [6], it is foreseeable that coal plants will
potentially turn uneconomic in the next 10 years.
c. Turkey: The capacity mix of Turkey is diverse, with
hydro, gas, and coal as three main generation sources in
terms of capacity size, accounting for 80% of the installed
capacity. Coal accounts for more than 20% of the generation
capacity in Turkey (Table 1) and 37% of generation in 2019.
As of 2020, the operating coal capacity is around 19 GW.
The coal produced in Turkey is predominantly lignite, which
is powering over half of the coal plants in capacity terms.
India
(Installed coal capacity = 207 GW)
South Africa
(Installed coal capacity = 44 GW)
Turkey
(Installed coal capacity = 19 GW)
Ukraine
(Installed coal capacity = 24 GW)
Note: Orange dots of SRMC are displayed when cost information of coal plants is available.
FIGURE 6. Inverse supply curve of coal capacity (India, South Africa, Turkey, and Ukraine)
0
10
20
30
40
50
60
0 50,000 100,000 150,000 200,000
SRMC ($ /MWH)
CUMUL ATIVE CAPACITY (MW)
Wind LCOE (IRENA, 2019) = 48.9 $/MWh
PV LCOE (IRENA, 2019) = 44.7 $/MWh
PV PPA Estimate (IRENA, 202 0) = 42.4 $/MWh
PV PPA Estimate (IRENA, 202 1) = 32.8 $/MWh
SRMC of coal plants
0
10
20
30
40
50
60
70
80
90
0 10,000 20,000 30,000 40,000 50,000
SRMC ($ /MWH)
CUMUL ATIVE CAPACITY (MW)
SRMC of coal plants
PV LCOE (IRENA 201 9) = 80.5 $/MWh
Wind LCOE (IRENA 2019) = 67.7 $/MWh
0
10
20
30
40
50
60
70
80
0 5,000 10,000 15,000 20,000
SRMC ($/MWH)
CUMULATIVE CAPACITY ( MW)
PV LCOE (IRENA, 201 9) = 77.6 $/MWh
Wind LCOE (IRENA, 2019) = 63.3 $/MWh
SRMC of coal plants
0
10
20
30
40
50
60
70
80
90
0 5,000 10,000 15,000 20,000 25,000
SRMC ($ /MWH)
CUMULATIVE CAPACITY (MW)
Wind LCOE (IRENA, 2019) = 75.3 $/MWh
SRMC of coal plants
PV LCOE (IRENA, 2019) = 83.8 $/MWh
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
Lignite is less efficient in generation, so the rest of the coal
power stations are also fueled by imported bituminous coal.
Thus, the SRMC of coal electricity is generally high in
Turkey, ranging from $40/MWh to $72/MWh (Figure 6).
Generation from Afşin-Elbistan - a power station, which was
commissioned in 1984, is estimated to be the most
expensive, while the other coal plants have SRMC ranging
from $40/MWh to $60/MWh.
Given present cost level of renewable electricity in
Turkey, the average LCOE of wind ($63/MWh) has already
dropped below the upper bound of the SRMC of coal
generation, indicating that wind resources already have some
cost advantage. Generation from solar PV also has a
competitive LCOE ($78/MWh), close to the SRMC of some
coal capacity. To sum up, considering the generally high cost
of coal generation in Turkey, as renewable energy industry
further develops, it is economically feasible to expand
renewable capacity to substitute the uneconomic coal.
d. Ukraine: There is currently 24 GW of coal power
plants in the Ukraine that account for around 40% of installed
power generating capacity (Table 1) and 30% of the annual
generation in 2018 [39].
In Ukraine, the cost of coal generation is also relatively
high. This is due to the high cost of coal itself, being sourced
from inefficient mining facilities, as well as old coal plants
with high operations and maintenance costs. The SRMC of
coal plants ranges between $61/MWh and $74/MWh (Figure
6).
Currently, the cost of renewable energy is also high,
although the LCOE of wind resources ($75/MWh) comes
close to the upper bound of the SRMC for coal plants. It is
only a matter of time before wind and solar will provide more
affordable power than coal plants in Ukraine. For example,
if the LCOE of these renewable technologies were to drop
by more than 20% (from current levels of $84/MWh for local
solar PV, and $75/MWh for wind), it would fall below the
average SRMC of all coal power plants ($68/MWh).
2) Sensitivity analysis: combined effect of RE cost
reduction and O&M cost increase of coal
In fact, the scale of uneconomic coal capacity will be more
likely to simultaneously influenced by the decreasing cost
trend of renewable energy and hence lower utilization of
aging coal plants. Figure 7 displays the sensitivity analysis
conducted for India, Turkey, and Ukraine, which clearly
demonstrates that even absent any carbon price, relative
economics of renewable coupled with rising costs of
inefficient older coal fleet, would make a strong case for
natural transition away from coal. The combined sensitivity
analysis evaluated the amount of uneconomic coal capacity
under decreasing LCOE of RE (by 0%, 10%, 20%, and 30%)
as well as decreasing capacity factor of coal power (60%,
50%, 40%).
IND
TUR
UKR
Note: *CF stands for “capacity factor. ** IND-India, TUR-Turkey, UKR-Ukraine.
FIGURE 7. Sensitivity analysis of unprofitable coal capacity in India, Turkey, and Ukraine (over LCOE of VRE and CF of coal)
14 14 24
51 55 60
69 72 87
110 118 124
338
14 21 34
57 59 63
81 94 100
0
100
200
CF=0.6 CF=0.5 CF=0.4
Uneconomic coal (GW)
0% decrease PV LCOE 10% decrease PV LCOE 20% decrease PV LCOE 30% decrease PV LCOE
001112
13
9
4
911
1 1
8
3
8
11
911
19
11
19 19
0
5
10
15
20
25
CF=0.6 CF=0.5 CF=0.4
Uneconomic coal (GW)
000002
8911
20 20 20
0 0 2
8 8 9
20 20 2020 20 20
0
5
10
15
20
25
CF=0.6 CF=0.5 CF=0.4
Unecono mci coal (GW)
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
a. India: Approximately 60% of the existing coal
capacity, or 124 GW will be rendered uneconomic when
LCOE of PV reduces by 30% and capacity factor of coal
drops to 40%. This is a 13% increase over 107 GW, relative
to the scenario where we considered the baseline capacity
factor of 60% (Figure 7). Capacity factor of Indian coal
plants has been dropping steadily over the years from ~78%
in 2009/10 to 60% in 2019 and an estimated 51% in 2020/21
[40]. It is therefore envisaged that a drop in solar/wind costs
coupled with increasing cost of coal plants could potentially
make a stronger case for repurposing coal plants in India.
b. Turkey: With costs of wind generation dropping by
30%, another 8 GW Turkish coal capacity will become
uncompetitive if capacity factor of coal generation is reduced
from 60% to 40% (Figure 7). In other words, when
renewable energy cost and coal utilization decreases at the
same time, all existing coal capacity (about 19 GW) will be
uneconomic making a strong case for several of these
projects to be repurposed to provide flexibility services.
c. Ukraine: As renewables and coal generation costs are
relatively close at present, we find a very striking change in
balance as we make coal more expensive while dropping cost
of renewable projects. If renewable generation costs drop by
30% together with a drop in utilization of coal plants, almost
all 20 GW operating coal capacity in Ukraine is rendered to
be more expensive than renewables (Figure 7). Even when
LCOE of solar/wind drops by only 10%, a capacity factor of
40% of coal plants will increase the level of SRMC and make
some coal capacity uneconomic. Given the aging coal fleet
in Ukraine and relatively high average cost of renewable
projects observed to date, this sensitivity is deemed to be
realistic and once again should make a resounding case for
coal plant repurposing.
V. COAL PLANT REPURPOSING: PLANNING AND
COST BENEFIT ANALYSIS
While the cost analysis of coal plants provides insights
into the competitiveness of coal at a high level, identification
of coal plants that can be retired and repurposed requires:
a. A least-cost planning analysis that decides both new
entry and retirement of plants in the system to meet
projected demand over next several years; and
b. A cost-benefit analysis of retired coal plants that may
potentially be repurposed.
In this section, we present such an analysis for Morocco,
which currently has more than 50% of its generation (Figure
8) coming from four coal power plants (Table 2). It is a
relatively new coal fleet with only 3 units that are older than
25 years including Mohammedia (280 MW unit, 33 years old
and has 34% thermal efficiency). Morocco also has nearly 2
GW of heavy fuel oil (HFO) capacity that is very expensive
to run with SRMC ranging from $140-177/MWh. Coal
plants in comparison are far cheaper with SRMC between
$38-53/MWh. LCOE of wind and solar for Morocco are
estimated at $47/MWh and $43/MWh, respectively using
data collected for planning analysis.
TABLE II: MAJOR COAL PLANT DERATED CAPACITY IN
MOROCCO (2019)
FIGURE 8. Current generation mix in Morocco
Morocco has significant RE resources and has the largest
concentrated solar plant (510 MW) in the world at
Ouarzazate. There is significant policy development in the
country to increase the share of RE going forward to more
than 50% by 2030. Demand is projected to grow rapidly from
~40 TWh pa to somewhere between 65-100 TWh by 2035
according to the official “Low” a nd “High” demand
projections. Morocco is also currently a net exporter of
electricity to Spain and has aspiration to increase its exports.
The high demand growth presents a challenge to any coal
retirement as the system needs to increase its capacity from
10 GW at present to 25-35 GW by 2035 depending on the
demand growth.
A. LEAST-COST PLANNING ANALYSIS
The least-cost planning analysis is conducted to check if
one or more of the older coal (and HFO) generating units are
likely to be economically retired and yet for the system to
build sufficient cleaner/cheaper capacity that meets the
operating reserve and reliability standards. We used the
Electricity Planning Model [31] developed at the World
Bank to assess the optimal capacity expansion and retirement
strategy for High and Low demand scenarios for 2019-2035.
The model is a state-of the-art planning tool that has been
deployed for more than 80 countries.
As the results of the planning analysis in Figure 9 show,
the coal generation share for both low and high demand
Comm.
Year
Capacity
(MW)
Fixed O&M
($/MW-year)
Var. O&M
($/MWh)
Heat Rate
(MMBTU/MWh)
2016 320 32,490 2.90 8.76
1987 280 32,490 2.90 9.91
Units 1&2 1994 630 25,850 0.99 8.27
Units 3&4 2000 624 25,850 0.99 8.24
Units 5&6 2013 626 35,200 1.30 8.76
2018 1,283 55,000 0.83 8.25
Jorf Lasfar
Jerada
Mohammedia
Safi
Plant Name
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
scenarios, drops significantly to 20% or below by 2035 even
without considering a carbon limit/price. In absolute terms,
both scenarios decrease coal-based generation from 27 TWh
at present to 16-17 TWh by 2035. Solar and wind generation,
on the other hand, increase to 52%-61% by 2035 suggesting
Morocco’s 52% renewable target by 2030 may be eminently
achievable. A significant share of the existing coal
generation will become uncompetitive compared to RE over
the next 15 years relative to solar and wind.
FIGURE 9. Generation mix and export (TWh – secondary axis)
Mohammedia (280 MW) is found to have a low capacity
factor in 2019 and is deemed to be retired economically from
2022 as its fixed O&M is not justified from a further drop in
dispatch. A significant part of the HFO capacity (1,590 MW)
is also retired economically over the planning period. In total
1,870 MW of thermal capacity is retired in line with a drop
in thermal share that saves $221 million.
B. COST-BENEFIT ANALYSIS OF REPURPOSING
MOHAMEDDIA (280 MW) COAL PLANT
Cost-benefit analysis (CBA) compares the incremental
benefits and costs of one or more alternative scenarios versus
a status quo (business as usual or BAU). Mohammedia (280
MW) is retired in 2022 in our alternative scenario (repurpose
scenario) as per the least-cost analysis. The BAU keeps
Mohammedia running until 2031 or 9 more years after the
economic retirement year found in the least-cost planning
analysis. The time horizon of the CBA is aligned with the
power system study and considers the next 15 years (up to
2035). Benefits and costs are discounted at a 6% discount
rate. Further details of the BAU and our alternative repurpose
scenario are given below.
(1) BAU (Mohammedia continues to be in operation):
a. Coal plant is used at 50% capacity factor (CF) until
2031;
b. Plant is decommissioned as from 2032; and
c. 150 workers are laid off in 2032.
(2) Repurpose scenario (Solar PV, BESS and SYNCON):
a. Coal plant is retired in 2022;
b. New RE plant reuses the site (50 MW Solar PV +
50 MW 3h battery storage (150 MWh)) leading to a
15% reduction in CAPEX relative to using a
greenfield site for the RE;
c. New plant converts coal generator into a SYNCON
(synchronous condenser);
d. New plants provide ancillary services through
SYNCON (Reactive power), BESS (frequency
control) and new flywheel (inertia);
e. 50 workers retain employment at the new plant; 100
people are laid off in 2022;
f. Avoided CO2 benefits from reduced coal power
generation from the site is attributed to the
repurposed project if generation from the
incumbent coal plant can be displaced by cleaner
forms of generation including renewable power
generated on the repurposed site and also elsewhere
in the system. This is estimated as part of the
planning optimization; and
g. Since the Mohammedia coal power plant is only a
small part of the Moroccan power system we
assume no change in generation costs vs. BAU.
There are incremental benefits and costs of the
repurposing option relative to the BAU scenario, namely: (a)
Benefits that include avoided carbon emissions over a 9-year
period, some of the jobs are retained on the site, the
RE+FLEX site can provide substantial ancillary services,
reduced cost of site and connection costs for the repurposed
project, and the salvage value of the assets that are
decommissioned is collected earlier; vs (b) Costs that include
new investments in RE+FLEX, decommissioning costs need
to be incurred early and a larger share of the employees will
need to be retrained/relocated early. As the World Bank [9]
study discusses in greater detail, there are reasons to believe
that benefits can outweigh cost by a significant margin,
especially if there are avoided carbon emissions due to the
fact that economic retirement not only reduces system costs
from uneconomic fixed O&M costs, but also avoids CO2
emissions that have a significant cost as reflected in available
estimates of the shadow price of carbon used in project
analysis [41]. Ancillary services from the RE+FLEX center
even after the coal plant is retired may also have significant
benefits and enhances the ability of the system to integrate
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
higher volume of variable RE generation. There are other
important issues such as retention of part of the workforce.
Table 3 lists the assumptions that are used in this analysis.
TABLE III: KEY ASSUMPTIONS ON REPURPOSING
IMPACT
Parameters
GHG
emissions
1) CO2 cost: $41/ton (2020) - $56/ton (2035)
[41]
Decommission
ing costs
1) Land surface: 2 km2
2) No ash disposal area (recycled by cement
companies)
3) Coal storage area: 0.032 km2
4) Coal cleanup cost: ~ 0.1 M$ (estimate based
on [9])
5) Other decommissioning costs*: ~ 11 M$
(estimate based on [9])
Social costs
1) 150 workers employed at coal plant
2) 50 workers re-employed at PV plant
3) Average cost of support program for laid-off
workers: ~$12k/worker (one-off) (estimate
based on [10])
Ancillary
services
provision
1) Reactive power value: 5$/Mvarh (used
8760h/year) [42]
2) Frequency control services: 3$/MWh (4h of
storage)
Salvage value
of coal plant
1) CAPEX of old coal plant: 2.5 M$/MW
2) Scrap value: 10% of initial CAPEX
3) (estimate based on [9])
4) 5% of initial CAPEX of old plant is reused
for repurposing (estimate based on [9])
Change in
CAPEX with
new plant
1) ~15% savings in CAPEX of PV plant
(estimated CAPEX of PV plant is 1
m$/MW)
2) CAPEX of SYNCON + Flywheel: 0.1
M$/Mvar
3) CAPEX of BESS: 200$/kWh with 3h of
storage (150 MWh)
* Other decommissioning costs include (i) employee, station overheads and
O&M expenses incurred from pre-decommissioning to the completion of
the decommissioning phase; (ii) Pre-demolition environmental regulation
costs (i.e. asbestos removal); (iii) demolition costs. For more details see [9].
Figure 10 shows how the benefits and costs stack up for
the Mohammedia (280 MW) early retirement in 2022. Total
discounted benefits of $518m over 2022-2031 far outweigh
the costs of $74.5m, i.e., a benefit-cost ratio of nearly 7.
Avoided CO2 emissions valued at the World Bank social
costs render the highest component of benefit ($385m or
74% of total benefits). Bulk of the costs of the project are due
to the investments in solar, BESS and SYNCON of $67m. It
should be noted that even if the avoided CO2 benefits are
completely ignored, benefits remain at $133m which still
yields a healthy benefit-cost ratio of 1.8. A potentially
significant source of benefits for the repurposed plant can be
the provision of frequency control and dynamic reactive
power services. The incremental benefit for such services is
estimated at $98m using international market price and can
by itself justify the cost of the project.
FIGURE 10. Benefits vs costs of repurposing (inc. CO2 benefits)
VI. CONCLUDING REMARKS
There have been countless discussions on reducing the role
of coal in power generation since the nineties, but these have
become increasingly intense only in the last decade as cost
of renewables followed by that of storage, plummeted. These
discussions have started being translated into policy actions
in some countries – most notably in the UK in recent years.
However, the lack of actions outside a limited number of
developed nations through focused policy actions is also
conspicuous. A casual comparison of winning auction prices
of solar and wind in some part of the world, with the
levelized cost of a new coal plant, would suggest that coal is
vastly unprofitable and should retire. Indeed, a number of
studies in the recent years have been making this claim
heralding the death of coal. This is, however, a complex
subject that requires detailed data at country/system level to
see the competitiveness of existing coal based on its short run
costs with the levelized cost of renewables which is also very
specific to each geography depending on project costs and
resource quality.
The answer on economic attractiveness of renewables in
the short term can be surprisingly different across
geographies as a comparative analysis we have constructed
for ten developing countries demonstrates. Countries like
India where renewable projects have been highly
competitive and there is an aging fleet of coal plants many of
which are far away from mines, are already highly
uncompetitive. On the other hand, countries like South
Africa that have relatively inexpensive coal plants, but the
average cost of renewable projects have not yet dropped
sufficiently (as of 2020), will require special efforts to
dislodge coal completely beyond plants that have reached or
gone well past their technical life. There are still other
countries like Ukraine and Turkey where both renewable
projects as well as coal tend to be expensive, making them
“com petitive” to ea ch other. As renewable costs are
projected to go down over this decade and older coal fleets
0
100
200
300
400
500
600
Benefits Costs
$m
Social costs (earlier termination) Incr. decomm. costs
CAPEX for new plant Re-employment
CAPEX redn new plant Incr. Salvage value
Ancillary services CO2 reduction
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
continue to incur higher maintenance costs and lose
efficiency – the situation will change rapidly in all three
categories as we have tested using a set of sensitivity cases.
We find that countries like India will see solar and wind costs
that can be built at prices that challenge the economics of 124
GW of operating coal plants, or 60% of its coal fleet. Ukraine
and Turkey where coal is only marginally competitive will
see a more drastic switch for coal to become uncompetitive
in the coming years. The transition process will require
policies like carbon pricing to ensure coal is phased out.
However, as the preceding discussions suggest, a relatively
low cost of carbon will be adequate in most cases as coal in
several systems are only marginally more expensive than the
LCOE of renewables. The process can in fact start in the
developing world immediately and gain momentum as costs
of renewables and storage continue to fall. As we have
alluded to in the context of India’s transition to clean energy,
the first 14 GW of old and expensive coal plant has a
negative cost of carbon reduction and on average the entire
fleet has an average cost of CO2 reduction below $10/ tCO2e
. As costs of solar, wind and storage fall, the cost of CO2
reduction in the long term will be negative making it a “win
win” situa tion to get cheaper and cleaner electricity. It is well
worth noting that the cost of carbon reduction in India from
coal based generation that we have presented is competitive
relative to those that prevail internationally, e.g., $16/tCO2e
in Spain carbon tax, $12/tCO2e in Beijing pilot ETS,
$10~$30/tCO2e in EU ETS, and $15/tCO2e in California
Cap and Trade [43].
As the reality suggests, in the near term, coal will not pave
the way for renewables without a policy intervention
notwithstanding the potential low cost of carbon. More
importantly, such policies also need to address some of the
complex issues around the integration of renewables
including very real cost of such integration and removing the
barriers to exit for coal. The latter includes significant local
development, social, and environmental issues as well as
significant cost of decommissioning. There is an acute need
for development of innovative business models to remove
the exit barrier. Repurposing old coal plant sites and
equipment can be a formidable option that not only enhances
the business case for exit including recovery of
decommissioning costs, but also partially addresses social
and environmental ones. Development of RE and flexibility
(RE+FLEX) centers on old coal plant sites can also be part
of the solution to integrate large-scale renewables at a
modest cost. It is an opportunity to reduce the barriers to coal
retirement and to propel the low-carbon development. There
are wider benefits of coal plant repurposing to further boost
utilization of renewable energy, stimulate new employment,
and strengthen the grid. Such efforts are only at a nascent
stage and will need to be scaled up quickly and efficiently to
accelerate coal transition.
VI. ACKNOWLEDGEMENT
The authors would like to take this opportunity to thank the IEEE
Access Editor and two reviewers for the comments provided on the
original submission. This research was funded by ESMAP and
benefitted from the country level data that was collated from
various other ESMAP projects (with the exception of Bulgaria).
The team would also like to thank our two internal reviewers from
the World Bank, Rahul Kitchlu (Senior Energy Specialist) and
Phillip Hannam (Energy Economist), for their comments.
VII. REFERENCES
[1] Carbon Brief, “Mapped: The world’s coal power plants”, March,
2020. [Online]. Available: https://www.carbonbrief.org/mapped-
worlds-coal-power-
plants#:~:text=Global%20coal%20capacity%20grew%20in,1%2C06
6GW%20to%202%2C045GW
[2] Global Energy Monitor, “Global coal plant tracker”. [Online].
Available: https://globalenergymonitor.org/projects/global-coal-
plant-tracker/, Accessed on: May, 2021
[3] Government of the United Kingdom, “Powering past coal alliance:
Declaration”. [Online]. Available:
https://assets.publishing.service.gov.uk/government/uploads/system/
uploads/attachment_data/file/660041/powering-past-coal-alliance.pdf
[4] Carbon Tracker, “42% of global coal power plants run at a loss, finds
world’s world-first study”, November 30, 2018. [Online]. Available:
https://carbontracker.org/42-of-global-coal-power-plants-run-at-a-
loss-finds-world-first-study/
[5] A. Rosenbaum and D.W. Gao, “Understanding grid parity”,
IEEE/PES T&D Conference, Dallas, TX, USA, May, 2016.
[6] IRENA, “Renewable power generation costs in 2019”, Abu Dhabi,
2020. [Online]. Available: https://www.irena.org/-
/media/Files/IRENA/Agency/Publication/2020/Jun/IRENA_Power_
Generation_Costs_2019.pdf
[7] H. Apostoleris, S. Sgouridis, M. Stefancich, M. Chiessa, “Evaluating
the factors that led to low-priced solar electricity projects in the
Middle East”, Nature Energy, vol. 3, pp. 1109-1114, 2018.
[8] D. Chattopadhyay, M. Bazilian, B. Handler and C. Govindarajalu,
“Accelerating the coal transition”, The Electricity Journal, vol.34,
no.2, January, 2021.
[9] G. Shrimali and A.Jindal, “Coal plant repurposing in developing
countries: Concepts and illustrative case study for India”, World
Bank, February, 2021.
[10] World Bank, “Review of social transition programs associated with
closure of coal mines and power plants”. Energy and Extractives
Practice, 2020.
[11] A. Benn, P. Bodnar, and J. Mitchell, “Managing the coal transition”,
Rocky Mountain Institute, 2018. Available online:
https://rmi.org/insight/managing-coal-capital-transition/
[12] International Energy Agency, “Have the prices from competitive
auctions become the new normal prices for renewables?”, February
4, 2019. [Online]. Available: https://www.iea.org/articles/have-the-
prices-from-competitive-auctions-become-the-new-normal-prices-
for-renewables
[13] M. Greenstone, I. Nath, “Do renewable portfolio standards deliver
cost-effective carbon abatement?”, Energy Policy Institute at the
University of Chicago, Working Paper No. 2019-62, November
2020. [Online]. Available: https://bfi.uchicago.edu/wp-
content/uploads/2020/11/BFI_WP_201962.pdf
[14] IEA, “Coal 2020: Analysis and forecast to 2025”, Demand,
December, 2020. [Online]. Available:
https://www.iea.org/reports/coal-2020/demand#abstract
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3087081, IEEE Access
VOLUME XX, 2021 1
[15] C. Shearer, “Analysis: the global coal fleet shrank for first time on
record in 2020”, August 3, 2020. [Online]. Available:
https://www.carbonbrief.org/analysis-the-global-coal-fleet-shrank-
for-first-time-on-record-in-2020
[16] L.R. Brown. “The great transition: S hifting from fossil fuels to solar
and wind energy.” WW Norton & Company. 2015.
[17] M. Ram, M. Child, A. Aghahosseini, D. Bogdanov, A. Lohrmann, C.
Breyer et al., “A comparative analysis of electricity generation costs
from renewable, fossil fuel and nuclear sources in G20 countries for
the period 2015-2030.” Journal of Cleaner Production, vol. 199, pp.
687-704, October 20, 2018. DOI: 10.1016/j.jclepro.2018.07.159.
[18] IEEFA, “Out-to-pasture coal plants are being repurposed into new
economic endeavors.” Institute for Energy Economics and Financial
Analysis. 2019. [Online]. Available: https://ieefa.org/ieefa-update-
out-to-pasture-coal-plants-are-being-repurposed-into-new-economic-
endeavors/
[19] A. Deecke and R. Kawecki, “Usage of existing power plants as
synchronous condenser.” Przeglad Elektrotechniczny, vol. 91, pp.
64–66, 2015.
[20] IEEFA, “Repurposing coal plants into solar and battery can pay up to
5 times more than decommissioning”, December 2020. [Online].
Available: https://ieefa.org/ieefa-repurposing-coal-plants-into-solar-
and-battery-can-pay-up-to-5-times-more-than-decommissioning/
[21] D. Raimi, “Decommissioning US power plants: Decisions, costs, and
key issues.” Resources for the Future, 2017. [Online]. Available:
https://www.rff.org/publications/reports/decommissioning-us-power-
plants-decisions-costs-and-key-issues/.
[22] Ontario Power Generation, “Nanticoke solar now generating
renewable power for Ontario”, March, 2019. [Online]. Available:
https://www.opg.com/story/nanticoke-solar-now-generating-
renewable-power-for-ontario/
[23] Energy Post Europe, “Coal regions are ideally suited for utility-scale
wind, solar and jobs”, December, 2020. [Online]. Available:
https://energypost.eu/coal-regions-are-ideally-suited-for-utility-scale-
wind-solar-and-jobs/
[24] German Renewable Energy Federation, “Von der kohle zum
vorreiter bei erneuerbaren energien: Fünf praxisbeispiele für einen
erfolgreichen strukturwandel in Deutschland (A pioneer from coal to
renewable energy: Five practical examples of successful structural
change in Germany)”, November, 2018. [Online]. Available:
https://www.bee-
ev.de/fileadmin/Publikationen/Studien/BEE_Strukturwandel_Best-
Practice.pdf
[25] ReNow, “Chile reaches 4. 9 GW of renewables on its way towards
coal-free future” Renewables Now. [Online]. Available:
https://renewablesnow.com/news/chile-reaches-49-gw-of-
renewables-on-its-way-towards-coal-free-future-657570/. Accessed
in: November 2020
[26] REW. “Renewables in, coal out: South Africa’s energy forecast.”
Renewable Energy World, October, 2019. [Online]. Available:
https://www.renewableenergyworld.com/2019/10/18/renewables-in-
coal-out-south-africas-energy-forecast/
[27] PLATTS, “World electric power plant database”, S&P Global,
March, 2018. [Online]. Available:
https://www.spglobal.com/platts/ko/products-services/electric-
power/world-electric-power-plants-database
[28] Australian Energy Regulator, “Approach to wholesale electricity
market performance monitoring”, March, 2017.
[29] X. Yao, B. Yi, Y.Yu, Y. Fan et al., “Economic analysis of grid
integration of variable solar and wind power with conventional
power system”, Applied Energy, vol. 264, 2020.
https://doi.org/10.1016/j.apenergy.2020.114706
[30] W. P. Pauw and R.J.T Klein, “Beyond ambition: increasing the
transparency, coherence and implementability of Nationally
Determined Contributions”, Climate Policy, vol. 20, 2020, pp. 405-
414. https://doi.org/10.1080/14693062.2020.1722607
[31] D. Chattopadhyay, P. Chitkara, I. Curiel and G. Draugelis, “Cross-
border interconnectors in South Asia: Market-oriented dispatch and
planning”, IEEE Access, July, 2020. [Online]. Available:
https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9127966
[32] E. Bellini, “Eskom launches tender to ‘repurpose’ coal plants with
low-carbon growth tech”, PV Magazine, April, 2020. [Online].
Available: https://www.pv-magazine.com/2020/04/23/eskom-
launches-tender-to-repurpose-coal-plants-with-low-carbon-growth-
tech/
[33] Department of Mineral Resources and Energy of South Africa,
“Integrated resource plan (IRP2019)”, October, 2019. [Online].
Available: http://www.energy.gov.za/IRP/2019/IRP-2019.pdf
[34] S. Mukhopadhyay, P. Gupta, B. Arya, A. Rajput et al., “Long term
planning for Indian power sector with integration of renewable
energy resources”, IEEE Ninth India Power Conference, February,
2020.
[35] International Energy Agency, “Solar PV -tracking report”, June,
2020. [Online]. Available: https://www.iea.org/reports/solar-pv
[36] R. P. Singh, “Revisiting the national renewable energy policy”, The
Economic Times, July, 2020. [Online]. Available:
https://energy.economictimes.indiatimes.com/energy-speak/opinion-
revisiting-the-national-renewable-energy-policy/4380
[37] B. Tripathi, “How retiring old coal plants could save Rs 53,000
crores for power discoms”, Business Standard, September, 2020.
[Online]. Available: https://www.business-
standard.com/article/economy-policy/how-retiring-old-coal-plants-
could-save-rs-53-000-cr-for-power-discoms-120091200179_1.html
[38] R. Tongia, “RE versus Coal: A false framing as both have a role to
play”, Policy Brief, October, 2018. Brookings India. [Online].
Available: https://www.brookings.edu/wp-
content/uploads/2018/10/Renewable-Energy-
%E2%80%9Cversus%E2%80%9D-coal-in-India.pdf
[39] International Energy Agency, “Ukraine energy profile”, Country
Report, April, 2020. [Online]. Available:
https://www.iea.org/reports/ukraine-energy-profile
[40] Ministry of Power of India, “Power sector at a glance”. [Online].
Available: https://powermin.nic.in/en/content/power-sector-glance-
all-india, Accessed in: February, 2021
[41] World Bank, “Guidance note on shadow price of carbon in economic
analysis”, November, 2017. [Online]. Available:
http://documents1.worldbank.org/curated/en/621721519940107694/p
df/2017-Shadow-Price-of-Carbon-Guidance-Note.pdf
[42] K.L. Anaya, M.G. Pollitt, “Reactive power procurement: lessons from
three leading countries”, Cambridge Working Papers in Economics:
1854, 2018, Available online: https://doi.org/10.17863/CAM.33807
[43] World Bank, “State and trends of carbon pricing 2020”, May, 2020.
[Online]. Available:
https://openknowledge.worldbank.org/bitstream/handle/10986/33809
/9781464815867.pdf?sequence=4&isAllowed=y
