M. Kemal DEMİRKOL2,
Mücahit SAV1, Erdem ERGİN3, Edward Byers4,
Zeren ERİK5, Arif Cem GÜNDOĞAN2
1 EÜAŞ (Turkish State Electricity Generation
Company Inc.), Ankara, Turkey
2 GTE Carbon, Ankara, Turkey
3 Disaster
Risk Management Specialist, Ankara, Turkey
4
School of Civil Engineering & Geosciences, Newcastle University, Newcastle,
UK
5 Climate Policy Specialist,
Ankara, Turkey
Abstract
The aim of this study is to assess climate change related risks on
selected EÜAŞ (Turkish State Electricity Generation Company Inc.) coal based
thermal power plant assets from the water-energy nexus perspective. We discuss plant
specific climate risk assessments, water dependency and temperature related
risks on EÜAŞ assets by conducting two case studies which focus on Soma B and
Afşin-Elbistan B (AETS B) plants. Our findings reveal that the most important
risk for Soma B Plant in the medium term is identified as floods induced by
increasing precipitation levels and efficiency loss in the facility due to
expected increases in air and water temperatures. For AETS B, both decreasing
precipitation as well as increasing air and water temperatures in the medium
term are expected to impact water efficiency and the operating efficiency of
the facility. The study concludes that both Soma B and AETS B Plants, the
operational design parameters have already been strained on a number of
occasions and this is expected to increase with climate change. Therefore, we
recommend the power generation industry should include climate change
adaptation measures in their strategy plans to mitigate adverse impacts of
climate change on generation capacity.
Keywords:
climate change, thermal power
generation, adaptation, water-energy nexus, water intensity
1.
INTRODUCTION
It is known that thermoelectric power
generation technologies heavily depend on water availability hence climate induced
changes in water resources significantly affects generation (Van Vliet, 2016). In
this study, we scrutinize selected EÜAŞ thermal power plant assets and examines their associated climate change related risks. Possible performance changes of selected EÜAŞ
assets are studied by utilizing historical data and future forecasts. In this
introductory part, we provide a quick overview of climate, energy and water
interactions. We present two case studies which focus on Soma B and
Afşin-Elbistan B (AETS B) plants and illustrate plant specific information and
risks.
1.1. Climate Change & Electricity Generation at
Thermal Power Plants
Electricity generation is dependent on
numerous climate variables such as temperature, precipitation as well as
climate induced parameters such as surface water availability, surface or
ground water temperatures. It is already evident that climate change has
significant impacts particularly on power generation from thermal sources
(Mideksa & Kallbekken, 2010; Koch et al, 2015). Anthropogenic climate
change is likely to have adverse impacts on electricity generation through
affecting efficiency and cooling water requirements of thermal power plants
(Schaeffer et al, 2012). Changes in climate, such as higher temperatures and
sea level rise, may result in adverse consequences, and subsequently, both
governments and the private sector are increasingly undertaking assessments
across sectors and considering impacts across asset portfolios (Byers &
Amezaga, 2015). To give an example, “an
increase in ambient temperature results in a decrease in the difference between
ambient and combustion temperature, reducing the efficiency of boilers &
turbines” (Contreras-Lisperguer & de Cuba 2008, Wilbanks et al. 2008 in
ADB, 2012). Table 1 provides a
summary of potential climate change impacts on thermal power generation.
Table 1. Key Climate
Change Impacts on Thermal Power Plants (ADB, 2012)
Climate Variable
|
Physical Components
|
Key Impacts
|
Precipitation
increase or decrease
|
•
Fuel (coal) storage
•
Boiler/furnace
•
Turbine/
generator
•
Cooling system
|
•
Increase could cause reduced coal quality (and combustion efficiency) due to
higher moisture content of coal
•
Decrease could affect availability of freshwater for cooling (all thermal
systems)
|
Higher
air
temperature
|
•
Boiler/furnace
•
Turbine/
generator
|
•
Lowered generation efficiency
•
Decreased IGCC system efficiency (converting coal to gas)
•
Lowered CCGT efficiency (gas)
|
Higher
wind
speed
|
•
Buildings, storage,
generating
plant
•
Air pollution
control
|
•
Damage to infrastructure
•
Wider pollutant dispersion
|
Sea
level rise
|
•
Buildings,
storage,
generating plant
|
•
Increased sea levels and storm surges could damage coastal infrastructure
|
Extreme
events
(including
flooding)
|
•
Buildings,
storage,
generating plant
|
•
Hurricanes, tornadoes, ice storms, severe lighting, etc. can destroy infrastructure
and disrupt supplies and
offshore
activities
•
Possible soil erosion and damage to facilities
|
1.2. Adaptation
in a Changing Climate: Thermal Electricity Generation at the Water-Energy Nexus
Climate Change Adaptation refers to “adjustments
in ecological, social, or economic systems in response to actual or expected
climatic stimuli and their effects or impacts” (Smit & Pilifosova, 2003).
It also refers to the changes in processes, practices, and structures to
moderate potential damages or to benefit from opportunities associated with
climate change (IPCC TAR WG2, 2001). The primary impacts of climate change will
be on the ecosystems and the secondary impacts will be on socio-economic
systems such as health, food, water and energy systems. The socio-economic
systems on which the societies depend are highly interconnected to provide
greater efficiency. As seen in Figure 1
failure on one component can have cascading impacts on services and assets.
Figure 1.An
illustration of interdependencies (Wilbanks et al, 2014)
Some failures may affect the system at
local scale; some others can have devastating effect at regional, national or
even transboundary scale. Therefore, systems approach should be at the heart of
climate risk assessments. The water-energy
nexus is the term used to describe the interface between water and energy
systems (Schnoor 2011; Scott et al 2011). It includes all interactions where
energy is used for water systems and water is used in energy systems (Figure 2). Provision of water requires
energy particularly for treatment and pumping (Siddiqi & Anadon, 2011).
Figure 2. Water,
Energy and Climate Change (WCSBD, 2009)
It is evident that climate change will
exert increasing pressure on existing water resources (Vörösmarty et al, 2000)
and this will affect water dependent electricity generation capacity. The
dependency of the electricity sector on water is viewed as a growing risk to
both energy and water security (Gadonneix et al, 2010; IEA, 2012; WWAP, 2014). Study
shows that the average reduction in capacity availability across Europe will be
6.3-19% in Europe with an increased probability of extreme events by a factor
of three (Van Vliet et al, 2012). The large majority of the impacts are
expected to be felt by thermal power plants operating with once through cooling
systems. Climate change is expected to bring impacts in different ways. Rising
mean temperatures may have small but continuous effects on the efficiency of
facilities. For example, rising air temperatures reduces the cooling efficiency
of cooling towers thus requiring a greater throughput of pumped water and/or
mechanical fans. Efficiency reductions are likely to occur if the design limit
of the cooling system is exceeded. It is expected that the mean and extreme
values of air temperatures to change and this will impact on the cooling
operations of power plants. Changing means may impact the performance of system
over a long period. For example, gas combustion turbines (Brayton cycle) have
reduced efficiency at higher temperatures because the air has less density and
as a result, efficiency reduces in the order of 0.3-0.5% per °C increase
(Sathaye, et al., 2013; Maulbetsch & DiFilippo, 2006) (see Figure 3). Therefore, a mean air
temperature change over the life of the plant will result in higher fuel inputs,
higher emissions and reduced profits.
Figure 3.Change
in turbine capacity as a function of ambient temperature (adapted from
Kehlhofer et al., 2009 and Maulbetsch and DiFilippo, 2006 in Sathaye et al,
2013).
2. PLANT
SPECIFIC CLIMATE RISK ASSESSMENTS
Hereby we present plant-specific case
studies which offer in-depth look at selected EÜAŞ assets. Two plants (Soma B
and Afşin-Elbistan B) were selected based on the strategic interest of EÜAŞ and
included consideration of factors including: age; generation type; cooling
system type; and historical performance. One of the selected plants Soma B is an old
coal powered thermal power plant. Initially commissioned in 1953, 2 units of 22
MW each started operating in 1954. In 1967, it was decided to build new units
of 165 MW each and 6 units started operating in 1983, 1984, 1985, 1986, 1991,
and 1992 respectively. Therefore, it is achieved to Soma B Power Plant’s
current total capacity of 1,034 MW (5.4 % of national capacity). The second
plant subject to in-depth analysis is Afşin Elbistan B Thermal Power Plant
(AETS B), a new coal powered thermal power plant. Initially commissioned in
1996 as a 4x360 MW plant, 2 units started operating in 2004 and the other 2 in
2005. AETS B has a total capacity of 1,440 MW, consisting 7.5 % of the national
capacity. This plant is next to AETS A, commissioned in 1973 and in operation
since 1984 with a total capacity of 1,376 MW (4x344 MW). Used coal exploited
from the Kışlaköy open-cast (A sector) up until 2009 and then started to use
the Çöllolar coal mine of 544 million tons for 1.5 years. EÜAŞ is now in
planning stage to build 3 more power plants to exploit sectors C, D and E. The
choice of 2 facilities with similar production mode (steam turbine) and fuel
(lignite) but of different age, cooling technology and in different regions
offers the ideal conditions to elaborate unique yet comparable solutions. Considering
the context of Soma B and AETS B plants, this study aims to provide a future
oriented perspective on how climate change can impact existing facilities’
performance.
2.1. Soma B
Plant Specific Climate Projections & Water Resources
The region has a semi-arid climate. It is
characterized with dry hot summers and mild winters with limited precipitation.
In terms of projections, Turkish State Meteorological Service (MGM) provides
broad patterns for both RCP 4.5 and RCP 8.5 scenarios over 3 period intervals
(2011-2040, 2041-2070 and 2071-2100); some on the watershed scale, some for the
province of Manisa (see Table 2).
Table 2. Climate change projections for Manisa
|
|
2011-2040
|
2041-2070
|
2071-2100
|
Daily Mean Air Temperature
(Changes in °C)
|
RCP 4.5
|
1.5 – 2
|
2 – 2.5
|
2.5 – 3
|
RCP 8.5
|
1.5 – 2
|
2.5 – 3.5
|
4 – 5
|
Annual Precipitation
(Changes in %)
|
RCP 4.5
|
10 – 15
|
0 – 5
|
5 – 10
|
RCP 8.5
|
5 – 10
|
0 – 5
|
-5 – 0
|
For daily mean air temperature and annual precipitation,
latest MGM projections (Demir et al, 2013;
Demircan et al, 2014) are used. Accordingly, air temperatures show a
continuous increasing trend but precipitation decreases in the first period and
either increases or decreases depending on scenario in the following period. Concerning
water resources, there are two creeks in the rea. These are Bakır Çay and
Yağcılar Çayı. Bakır Çay starts at the mountains located northeast of Soma and runs
for 104 km through the agricultural plains to join the Yağcı Cay and ultimately
the Aegean sea near Çandarlı. It is known for its irregular flow and often creates
flash flooding. Bakır Çay passes just near the Soma Plant but is not directly
used by Soma B Plant. Yağcılar Çayı starts on the mountains located north of
Soma, Yağcılar Çayı feeds the Sevişler Dam, which provides irrigation in the
plains located east of Soma, controls flooding and feeds Soma B with cooling
water. Yağcılar is also known to cause flash floods occasionally. Soma B Plant
is connected with a 13 km pipeline to Sevişler Dam located 9 km to the north
west of the power plant for its water intake. This reservoir is a substantial
resource for the surrounding region. Water levels in the reservoir have been
recorded on a monthly basis since 1987. Filling of the reservoir tends to occur
from December through to May-June, whilst water levels fall from July through
to November. Some calendar years have experienced as many as eight months of
recession. In most years the winter and spring rains raise water levels
substantially, in the order of 5 to 10 m above the low water levels experienced
in November and December. On a few occasions, in 2001, 2007 and 2008 (red
circles in Figure 3), the rains have
failed to increase water levels substantially leading to further reductions in
water levels in the following years.
Figure 3.
Hydrograph of water levels in Sevişler Dam between 1987-2011
It is seen in the Figure 4 that the initial 13 years experienced regular filling and
withdrawal cycles. But from 2000 onwards, the inter-annual variability is
higher and less regular. More recent reservoir levels have been on average 8m
lower in the 2000-2011 period, compared to the 1987-1999 period.
Figure 4.
Annual Profiles (More recent years -shown in dark grey- have been falling to
lower levels, hence average from 2000-2011 is lower)
2.2. AETS B Plant
Specific Climate Projections & Water Resources
The region has a semi-arid climate with
similar properties to the one in Soma. In terms of projections, MGM provides
broad patterns for both RCP 4.5 and RCP 8.5 scenarios over 3 period intervals
(2011-2040, 2041-2070 and 2071-2100); some on the watershed scale some for the
province of Kahramanmaraş (see Table 3).
Table 3. Climate Change Projections for Kahramanmaraş (where AETS
B is located)
|
|
2011-2040
|
2041-2070
|
2071-2100
|
Air Temperature
(Changes in °C)
|
RCP 4.5
|
1.5 – 2
|
2.5 – 3
|
2.5 – 3
|
RCP 8.5
|
2 – 2.5
|
3 – 4
|
4 – 5
|
Annual Precipitation
(Changes in %)
|
RCP 4.5
|
-5 – -10
|
-10 – -15
|
-10 – -15
|
RCP 8.5
|
0 – -5
|
-10 – -15
|
-10 – -15
|
MGM projections point to a sustained,
strong increase in temperatures and sustained, strong decrease in
precipitation. Regarding water resources, there are one major river and several
creeks in the area. These are Ceyhan River, Hurman Creek, Söğütlü Creek. Ceyhan
River is one of the most important river of Turkey runs for 509 km from
Elbistan where it starts to the Mediterranean Sea. It is the water resource of
the city, its agricultural plain, and also is one of main irrigators of
Çukurova. This is a strong river with continuous stream although the drought of
2014 has seriously reduced its output. Elbistan has experienced several floods
in the past but most of the river banks are now strengthened for flood
protection. Hurman Creek starts on the north of Afşin and running 64 km, it
joins into the Ceyhan River on west of Elbistan. It feeds Afşin and its
surrounding agricultural plains and runs along the eastern side of AETS B.
Although not currently used for thermal plants, there are plans to build a dam
on it for serving the new power plants. It is irregular in flow volume but
there is no information regarding flood history. Söğütlü Creek starts on north
east of Elbistan in borders of Malatya, Söğütlü feeds into the Keban dam before
feeding the agricultural plains located on the east of Elbistan and running
through the city. It also has an irregular flow and has a known history of
flooding. In addition to these water sources, Afşin Mağara Gözü and other
smaller creeks are present and mainly used for irrigation purpose. Another
significant source of water are the open-cast coal mine operations, which
result in large volumes of groundwater ingress. Extracted water from Afşin
Elbistan Lignite Management is pumped into nearby streams which join the Ceyhan
River. But they are not used for water needs of AETS A or B. Water is extracted
at approximately 2 m3/ s on a continuous basis. This could be
considered as a viable alternative water source if the quality can be managed
and treated. The water could be stored in a reservoir in order to reduce
fluctuations in water availability, not only for the power plant but also
agricultural producers in the region. AETS B takes its water from Ceyhan River,
at the exit of the city of Elbistan. The original plan was to get water from
the same pumping station feeding AETS A, but as a result of local authorities’
objection, water intake of AETS B displaced to where Söğütlü creek joins Ceyhan
River, just the exit of the city. This change has significant impact on the
plant: (i) water quality is poorer and can be disrupted after heavy rain fall
when the water is muddier, (ii) the site can suffer occasional flooding from
either streams, although protective measures were recently put in place, and
(iii) there may be limited water left in times of drought. Both AETS A & B
get their cooling water from Ceyhan, a conflicting issue with other users
(sugar beet, city, etc.). The river has an average output of 5 m3/s
and AETS A (1.5 m3/s) and AETS B (1 m3/s) get
approximately half of it.
3. WATER DEPENDENCY & AIR-TEMPERATURE IMPACTS ON
SELECTED EÜAŞ PLANTS
3.1. Water Use
& Dependency at Selected EÜAŞ Power Plants
As investigated in the previous section,
water use at the plants had significant inter-annual variability. Water use at
AETS B has been almost constant, except a slightly higher value in 2013, likely
due to an extended outage at one of the units (see Figure 5).
Figure 5.
Water use at AETS B and Soma B Power Plants (2004-2013)
The results show some similarity to the
rate of fuel use (see Figure 6).
Fuel use varies primarily due to the quality of the coal and consistency in
quality.
Figure 6.
Fuel use rate per unit of electricity generated (2004-2013)
Fluctuations in Net Calorific Value would
explain some of the variability in Soma’s fuel use rate, but other factors,
such as inter-annual climate variability and regional demands, are also likely
contributors. In both cases, fuel use rate is only a mediocre parameter for
predictor of water use with correlation R2 values of 0.56 and 0.52
for AETS B and Soma B Plants, respectively (see Figure 7).
Figure 7.
Correlation between fuel use rate and water use of Afşin B and Soma B Plants
3.2. Impacts
of Change in Air Temperature on Selected EÜAŞ Plants
We determined the magnitude of expected
temperature changes for selected power plant sites and estimate the impacts of
those temperature changes, both mean and extreme, on the power plants. Climate
projections obtained from the HadGEM2-ES RegCM4.3.4 regional climate model were
used with Representative Concentration Pathways (RCP) 4.5 and 8.5.
3.2.1. Impacts
on Soma B Power Plant
Observing historical climate record from
1960-2013, it can be seen that the higher 44°C threshold has never been
breached, whilst temperatures in excess of 42°C, the highest temperature, were
recorded in the months of June and July 2007 with temperatures of 42.7°C and
43.7°C, respectively. These temperatures are 10-12°C higher than the average
maximum temperatures for those months (see Figure
8).
Figure 8.
Historical temperature profiles for Soma 1960-2013
With climate change, impacts on air
temperatures indicate a rising trends, with fewer cold days, higher average
temperatures, and more frequent and hotter hot days. Table 4 indicates the expected extent of these temperature impacts,
both as absolute temperatures and as changes. In Table 4, the historical records show that the lowest 10% of monthly
mean temperatures were lower than 6.2 °C. However, in the 2050s, the lowest 10%
of monthly mean temperatures will be at 7.4-7.6 °C or below, depending on the
emissions pathway (RCP4.5 or RCP8.5). Similarly, the 50% threshold values
represent the median and the 90% values represent the highest 10% of air
temperatures.
Table 4. Monthly mean and change in monthly mean temperatures at different
threshold levels for Manisa region
Temperature
|
°C
|
Monthly mean
|
Change in
monthly mean
|
Threshold
|
Historical
|
2020s
|
2050s
|
2080s
|
2020s
|
2050s
|
2080s
|
10%
|
6.2
|
7.2-7.0
|
7.4-7.6
|
7.6-8.3
|
1.0-0.8
|
1.2-1.4
|
1.4-2.1
|
50%
|
15.1
|
16.7-17.0
|
17.6-18.4
|
18.5-20.7
|
1.6-1.9
|
2.5-3.3
|
3.4-5.6
|
90%
|
25.0
|
27.8-28.6
|
29.6-30.9
|
30.7-33.8
|
2.8-3.6
|
4.6-5.9
|
5.7-8.8
|
The 50% monthly mean air temperature
indicated in Table 4 is 15.1 °C, but
in future climate scenarios, this temperature will be exceeded more frequently.
For example, this temperature threshold is now approximately 40% for RCP 8.5
model in 2050s, thus exceeded 60% of the time period. This data can be used to
interpret critical temperatures and understand how the frequency of exceedance
increases with time. Critically, this indicates how temperature-sensitive
design and performance parameters must be adjusted to take into account climate
change. In reference to the design threshold temperatures of 42 °C and 44°C for
Soma B Plant, projected monthly maximum air temperatures at the 90% threshold
approach these levels in the 2020s, suggesting that in extreme circumstances
(i.e. 95th or 99th percentiles) these temperatures may be exceeded.
3.2.2. Impacts
on AETS B Power Plant
Observing historical climate records from
1960-2013, it can be seen that the higher 40°C threshold has never been
breached. Temperatures in excess of 38°C, the highest temperature, were
recorded in the months of June 2006, July 2011 and August 2006 (see Figure 9). These temperatures are
8-12°C higher than the average maximum temperatures for those months.
Figure 9.
Historical temperature profiles for Afşin 1960-2013
Table
5
similarly indicates the combined threshold temperatures for the
Adana/Diyarbakir region. The impacts are more severe than for the Manisa
region, both proportionally and in absolute terms. In the near term, monthly
mean temperatures are expected to rise by around 2.3-2.7 °C whilst in some
months this will be 4.2-4.8 °C higher.
Table 5. Monthly mean and change in monthly mean temperatures at
different threshold levels for the Adana/Diyarbakır region
Temperature
|
°C
|
Monthly mean
|
Change in monthly mean
|
Threshold
|
Historical
|
2020s
|
2050s
|
2080s
|
2020s
|
2050s
|
2080s
|
10%
|
-1.7
|
-0.4--0.2
|
0.1-0.3
|
0.3-1.1
|
1.3-1.5
|
1.8-2.0
|
2.0-2.8
|
50%
|
10.8
|
13.1-13.5
|
14.2-15.2
|
15.5-17.7
|
2.3-2.7
|
3.4-4.4
|
4.7-6.9
|
90%
|
22.5
|
26.7-27.3
|
29.2-30.8
|
30.0-33.0
|
4.2-4.8
|
6.7-8.3
|
7.5-10.5
|
Changes in monthly maximum temperatures
are also significant and approach the design threshold temperatures indicated
by EÜAŞ, of 38 °C and 40 °C in the top 10% of monthly maximum temperatures. As
with the Manisa region, the extreme cases will be in excess of the 90%
threshold values indicated in the table. The analysis of the climate projections indicated that the impacts are
expected to be more severe in the Adana/Diyarbakir region compared with Manisa
region.
3.2.3. Discussion
Redevelopment, retrofit,
maintenance and expansion of plants should take into account air temperature
impacts on a variety of temperature-sensitive components, such as cooling system capacity
(including pump and fan power), susceptibility of steam turbines to unit tripping due to low backpressure, oil coolers, alternators and compressors. The expected working life of
components that are replaced or maintained should be taken into account when
determining which level of climate impacts to be used for design parameters. It’s better to be aware that the following conditions also depend on
temperature impacts: Reduction in performance (such as cable de-rating); outage of operation (such as unit-tripping). The susceptibility to high temperatures
for some components increases with age and use, and also the timing of
maintenance. Air
temperature changes are also expected to impact on performance and safety of
transmission and substation infrastructure, to which generation assets are
connected. This depends on the design parameters for
which this infrastructure was specified. For the Soma B power plant, extended operation of the plant
should not be too challenged in the near term future, given that the plant is
approaching the end of its economic life. The
relevant recommendations depend largely on the planned strategy. Prospective
predictions related to climate change should be taken into account for revisions and new
investments that will be made. Complete redevelopment of the site however should take into account at least the 2050s’ climate
impacts. In particular, strengthening and expansion of the existing cooling
system is probably required in order to increase capacity. For AETS B, the
expected economic lifetime of the plant takes its operation
into the 2030s and 2040s. Thus we should take into account both 2020s and 2050s
climate projections. Moreover, the impacts on the Adana/Diyarbakir region are
also expected to be more severe than on Manisa. This is especially important
given the plans to build of new 3 (C, D, E)
coal-fired thermal plants in the local vicinity. The control room monitoring
and operation of the cooling systems is sufficiently advanced such that
different climate profiles can be managed and optimized. Nonetheless, operation in extreme conditions may be limited by performance
reductions and increased wear and tear on components
that must be considered in initial economic studies.
4. RESULTS & RECOMMENDATIONS
According to climate change projections,
the most important risk for Soma B Plant in the
medium term is identified as possible floods induced by increasing
precipitation, and
efficiency loss in the facility due to expected increases in air/water temperature. For AETS B, it is
projected that both decreasing in precipitation as
well as increasing in air/water temperature in the medium term and the
efficiency of the facility will be affected. When other power plants planned to
be built nearby, urban development, the agricultural and industrial use are
taken into consideration, for both power plants, it
is possible to say that their activities will likely be affected. For both Soma
B Plant as well as AETS B Plant, the operational design parameters (air/water temperature, water
availability etc.) have already been strained on a number of occasions and this is expected to increase
with climate change. To develop an efficient climate
resilience strategy, a wide range of measures are
required. Foremost, the energy industry should include climate change adaptation measures in their strategy plans.
5.
ACKNOWLEDGEMENTS
Hereby we acknowledge that this research
was supported by the UK Foreign and Commonwealth Office (FCO) through the
Prosperity Fund as a part of the project titled “Climate Resilient Thermal
Energy Production and Improving Energy Supply for Turkey”. We would like to
thank our colleagues from EÜAŞ who provided data, insight and expertise that greatly
assisted the research.
6.
REFERENCES
§ Asian
Development Bank – ADB (2012). Climate Risk and Adaptation in the Electric Power
Sector. Asian Development Bank.
