Natural Resources, 2010, 1, 69-79
doi:10.4236/nr.2010.12007 Published Online December 2010 (
Copyright © 2010 SciRes. NR
The Economic and Environmental Impacts of
Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Levent Aydın*
Ministry of Energy and Natural Resources, Moussa Bouh Odowa, Turkey.
Received October 31st, 2010; revised November 24th, 2010; accepted November 25th, 2010.
Since Turkeys economy and population is rapidly growing, Turkey mostly meets its energy demand from imported fos-
sil sources due to the very limited indigenous oil and natural gas resources. However, Turkey has abundant renewable
resources especially, hydro power potential to be used for generation of electricity. But only one-third of this significant
economical potential could be used. This usage seems insufficient when compared with that of European countries. In
order to analyze the potential long term impacts of the hydro power expanding shock on some macroeconomic vari-
ables of interest such as GDP, real consumption, real investment, exports, imports, trade balance, and carbon emis-
sions, we developed TurGEM-D, a dynamic multisectoral general equilibrium model of the Turkish economy. Using
TurGEM-D, we analyzed the impact of hydro power shock under policy scenario doubling hydro power generation. The
simulation results show that doubling hydro power have slightly positive effects on macro indicators and carbon emis-
sions for Turkish economy.
Keywords: Hydro Power Generation, Dynamic CGE, Turkey, Carbon Emission
1. Introduction
Turkey is seeking to cover an imminent shortfall in elec-
tricity as well as cut its dependence on foreign energy
resources, mainly natural gas. One of the Turkish energy
policies is to designate hydro and nuclear power as an
essential source of energy, meeting at least one-fifth of
Turkeys power needs within the next decades.
On the other hand, Turkey is not a rich country in
terms of the hydrocarbon (oil and natural gas, etc.) po-
tential to be used for generation of electricity. Therefore
Turkey has a strategy for developing the hydropower
potential and expects a few hundred small hydro power
plants to be constructed in the long run. Moreover Tur-
key must discover new and renewable energy resources.
However, new and renewable resources other than hydro
will not be sufficient to produce large amounts of elec-
tricity in the coming decades even if major efforts were
made to develop them.
Eventually, Turkey must base its power generation
strategy on constructing nuclear and hydro power plants
for coming decades in order to minimize foreign de-
pendency of natural gas and carbon dioxide (CO2) emis-
The other main characteristic of strategy for power
generation is the fact that the strategy is highly supply-
oriented. Emphasis has been placed on ensuring addi-
tional power supply to meet the growing demand while
energy efficiency has been lower priority. In spite of new
regulations and more activities have recently been
launched to enhance energy efficiency, there is obviously
a long way to go. As such, studies conducted by the Tur-
kish government officials have demonstrated that Turkey
has 25-30% energy conservation potential.
The aim of this paper is to evaluate the economic and
environmental impacts of constructing new hydro power
plants as alternative fossil-fired power plants. In order to
analyze this policy option we specifically developed
TURGEM-D (Turkey General Equilibrium Model-Dy-
namic) which is dynamic, multisectoral and applied gen-
eral equilibrium model of the Turkish economy. The
model structure of TURGEM-D mainly was based on
ORANI-INT model except production structure. The
production structure is divided into two types: fossil fired
power generation and hydro power generation.
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
TURGEM-D database was compiled from the I/O ta-
bles of Turkish economy with reference year of 2004.
Both sectors and commodities of this data base were
firstly aggregated into 8 sectors and commodities as fol-
lows: agriculture, coal, oil, gas, oil products, energy in-
tensive industries, electricity, other industries and ser-
vices. The electricity sectors further disaggregated 2 sec-
tors and its commodities with additional data of Turkish
power sector.
We expect hydro power generation of Turkish econ-
omy is to be about twofold by increasing average annual
production from 62 billion kWh to 118 billion kWh in
the next decade (2010-2020). The policy scenarios are to
diversify fuel sources as well as supply routines and ori-
gin and they also aims to reduce import dependence of
natural gas and coal in power generation while increase-
ing the share of renewable hydro and nuclear power in
Turkey. In the long term options we expect that hydro
power make significant contribution to power generation.
We can use real GDP and CO2 emission as a macroeco-
nomic variable to evaluate impacts of hydro power plants
on Turkish economy.
2. Hydro Power Potential and Policy in
Turkey has theoretical hydropower potential of 433 bil-
lion kWh1, technically feasible potential of 216 billion
kWh, and technically economical potential of 140 billion
kWh. In 2009, there were 213 hydroelectric power plants
in operation in Turkey. Total hydropower plants capacity
was 14,300 MW with average annual total production of
50,000 GWh which corresponds to 36% of the total eco-
nomical potential. In addition 145 hydro plants with total
capacity of 7,286 MW are under construction, corre-
sponding to about 23,770 GWh of additional annual power
generation. The remaining potential of 66,230 GWh
of total 140 billion kWh will use to construct the 200
hydropower plants in the coming decades. This would
bring the number of hydropower plants to 558 and the
total installed capacity to 44,200 MW [1,2].
Although Turkey has a big potential for hydro power,
the utilization of this potential is a question of determin-
ing and implementing sound, long-term energy planning
and politics that should prepare the best and reliable en-
vironment for the national and international investors
[3,4]. Place and being a source of re- newable energy in
power, the lack of negative influence of the environment,
low operating and maintenance costs make it necessary
to develop as possible. Giving priority to the use of water
resources in Turkey has been adopted as a national policy.
Installed hydroelectric power, even though significant
progress achieved to date, is not enough. Hydroelectric
energy has to be addressed with the development of a
new strategy.
Table 1 indicates that total the installed capacity will
increase to 48,817 MW in 2010 and to 96,349 MW in
2020. The installed hydropower capacity is anticipated to
increase to 18,234 MW in 2010 and to 34,076 MW in
2020. Thus, an additional 1,584 MW of hydro capacity
should be added to the system annually over the next 10
years [5].
The production of hydroelectric power plants, is de-
pendent on rainfall conditions change in the share of total
production each year, however, the 20-and 30% of elec-
tric energy in Turkey is produced from water. As shown
Table 1 the renewable power generation with rainfall
conditions will increase to 62 Billion kWh in 2010 to 118
Billion kWh in 2020.
In the recent years, more emphasis has been put on the
environmental integration of small hydro plants into river
systems in order to minimize environmental impacts,
incorporating new technology and operating methods in
Table 1. Turkeys long-term electricity supply projections.
Year 2010 2020
Rain Rainless Rain Rainless
Power Plant Type Installed capacity (MW) Billion kWh Installed capacity (MW) Billion kWh
Fuel based 30,583 211 211 62,273 425 426
Renewable based 18,234 62 46 34,076 118 77
Total supply 48,817 273 257 96,349 543 503
Sources: TEIAS, Turkey Electricity Generation Planning Study (2005-2020/October 2004)
It is almost 1% of world total potential and 16% of the total hydropower capacity in Europe (Balat, 2007).
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
By the end of 2008 Turkey was producing 17% of its
electricity from renewable energy sources. The revised
strategy paper for the electricity sector set a target of
producing 25% of the countrys electricity from renew-
able sources by the end of 2020. Considering that elec-
tricity consumption is expected to double by the same
date, this objective will require significant efforts. Tur-
keys energy policy has been revised in line with EU
policy in the context of enlargement process.
In addition to the publication of Electricity Market
Law (Law No. 4628) led to the establishment of Electric-
ity Market Regulatory Authority and the publication of
Renewable Energy Law (LawNo.5346) guarantees to buy
electricity from legal entities with a price of €5.5cent/
kWh by government for duration of 10 years, the Law
No. 5784, was published stating that the legal entities are
not required to apply for a license to generate electricity
from renewable energy up to a capacity of 500kW and
government guarantees the buy the excess electricity.
These laws on renewable energy utilization for electricity
generation in Turkey have brought some improvements
to the market. However, they must be revised or redes-
igned to fulfill the requirements of EU policy as Turkey
is an EU- candidate country [6].
3. Modelling Framework and Simulation
Before proceeding with the model simulation, we must
first briefly discuss the TurGEM-D model and its data-
base, originally developed by Malakellis [7]. We take the
standard ORANI-INT model and introduce some
changes to make it consistent with power generation. We
first show that the structure of TurGEM-D model allows
for analysis of increasing supply of renewable power
generation. The most significant features that distinguish
TurGEM from ORANI-INT are the inclusion of inter-
fuel substitution, power generation, capital energy sub-
stitution/complementarity, and dynamic mechanism ca-
pable of projecting the development of the economy
through time. With TurGEM-D we have produced an-
nual projections of the Turkeys CO2 emission, GDP
growth rate, and other economic variables.
3.1. Model Structure and Data
TurGEM-D is dynamic computable general equilibrium
model of Turkish economy. The dynamic Turkish model
described below was developed from the Australian
model originally presented in ORANI [8] and has its core
the dynamic CGE model described in ORANI-INT mod-
el. Model structure is modified from ORANI-INT model
to make suitable for analyzing energy, natural resources
and climate change issues.
The model database was compiled from the 2004 Tur-
kish Input-Output table [9] and Energy Statics [10]. The
64 sectors in Turkish economy are first aggregated to 8
sectors, which are thought to be critical for this analysis.
Electricity sector further disaggregated into fossil fired
electricity generation (oil-fired, coal-fired, gas-fired) and
hydro power generations. The electricity industry is able
substitute alternative power generation technologies in
response to changes in relative costs. The output of the
power sector is an aggregate of the power generated from
each of these technologies. The production structure of
the power generation sectors in TurGEM-D model is
illustrated by the nested structure and all electricity gen-
erated from these technologies as shown in Figure 1. In
addition, power generated from renewable energy sour-
ces is designed as a separate sector so as to analyze Tur-
kish energy policies in line with the EU policy.
The production structure of non-electricity sectors in
TurGEM-D model is represented by the nested structure
of energy composite (coal, oil products, gas) and primary
factor composite (land, labor, capital).
The dynamic structure of TurGEM-D is illustrated in
Figure 2. The model is replicated T times by indexing
all variables in the model with respect to time, where T
is the length of time horizon (in years). Sectoral invest-
ments and aggregate household expenditures are exoge-
nous in the model. The model equations are dynamic:
they express relationships among variables at different
points in time.
As shown in Figure 1, on the production side of the
model, energy is taken out of the intermediate input
nest to be incorporated into the value-added nest. In-
corporation of energy into the value-added nest is han-
dled in two steps. First, following Burniaux and Truong
(2002), energy commodities are first divided into elec-
tricity and non-electricity. Certain degree of substitu-
tion is allowed among the non-electricity inputs (σ
as well as between the electricity and the non-electricity
inputs (σ
ENER). The energy composite is then combined
with capital to produce an energy-capital composite. This
is in turn combined with other primary factors in a val-
ue-added-energy (VAE) nest through a CES structure.
The substitution elasticity between capital and energy
composite (σKE) is still assumed to be positive (indicating
energy and capital are substitutes in the inner nest).
However, provided the value of σKE is set at a level lower
than σVAE , the overall substitution elasticity (as viewed
from the outer nest) between capital and energy may
still be negative [11]. More precisely, we make use of the
formula derived by Keller [12] which specifies the rela-
tionship between the inner and outer elasticity of sub-
stitution between K and E as follows:
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
σ=(σσ )/S+σ/S
where SKE is the cost share of the KE-composite in the
outer (value-added) nest, and σKE-inner and σKE-outer indicate
the inner and outer substitution elasticities between K
and E, respectively.
In TurGEM-D, the (inner) value of σKE is assumed to
be 0.5 for most industries (including electricity), and is
set equal to 0.0 for coal, oil, gas, petroleum and coal
products, and agriculture. We followed Burniaux and
Truong [13] in adopting the parameter values. The value
of σVAE ranges from 0.2 to 1.45 and this seems to be
slightly larger than the values adopted by other models.2
As indicated by the directions of arrow in Figure 2
some variables are linked through time via forward and
backward linkages. For investment (It), the backward
linkages are provided by the capital accumulation equa-
tions (Kt) while the forward linkages are provided by the
specification of forward looking rates of return. Similarly
consumption (Ct) in each period is linked (forward) to
future prices and income while current debt (Dt) is linked
(backwards) to past savings. Only one representative
household is modeled in the model and capital is as-
sumed to be sector-specific. This means that the model
incorporates 9 independent types of forward-looking
behavior where 8 relate to the planning of sectoral in-
vestments while the other relates to the planning of ag-
gregate consumption. There are also 9 accumulation
Figure 1. Production structure of the model.
See Tables 9 and 10 in Burniaux and Truong (2002: 32) for more details on the values of
used by other models.
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
identities modeled: again 8 of these relate to the accu-
mulation of sectoral capital stocks and the final one re-
lates to the accumulation of foreign debt [7].
Investors can allocate their funds across 9 types of
perfectly substitutable assets. These include shares in the
8 industries and foreign and domestic bonds. Zero pure
profits are assumed in all activities with commodity and
carbon taxes on commodity flows putting wages between
basic prices and purchasers prices.
The demand and supply of commodities is determined
by the optimizing behavior of producers, investors and
consumers in the context of competitive markets. Gov-
ernment demands for commodity are assumed to be fixed
or exogenous while foreign demand is specified in an ad
hoc manner. The demand for factors and supply of capi-
tal are also derived from optimizing behavior and com-
petitive market assumptions. The supply of labor and
agricultural land can be specified exogenously or they
can be determined by demand [7].
We account for carbon dioxide emissions arising from
the combustion of fossil fuels such as coal, natural gas,
and petroleum products. We assume that carbon dioxide
emissions are closely related to energy consumption. We
assign user, fuel, and source specific emissions coeffi-
cients (CO2 per dollar, at 2004 value) and prorate the fuel
specific 2004 national CO2 inventories among users. This
produces the CO2 emissions matrix by fuel commodities,
commodity sources and users. Table 2 shows CO2 emis-
sions from 3 fuels (domestic plus imported): coal, natural
gas, oil.
3.2. Simulation Design
For policy simulation, model is solved over 16-year time
horizon and results are reported as percent deviations
from the baseline scenario. In setting up the simulation,
we need to specify the closure for the model and the set
of relevant shocks for the exogenous variables.
Figure 2. Dynamic mechanism of the model.
Table 2. Turkeys CO2 emissions by user (2004, million tons).
Domestic Imported
Gas Oil Products
Gas Oil Products.
Agriculture 0 0 10 0 0 3
Energy intensive industries 0 0 4 1 1 1
Coal fired power plants 48 0 0 11 0 0
Oil fired power plants 0 0 10 0 0 3
Gas fired power plants 0 0 0 0 29 0
Other industry services 14 0 42 8 5 11
Households 4 3 20 5 4 5
Total 65 3 86 26 39 23
Source: GTAP 6 database
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
In the balanced growth baseline scenario used as the
control scenario, the economy converges to a balanced 5
percent3 average annual growth asymptotically, all real
variables grow at 5 percent per annum and all prices are
The past behavior of agents is taken as given in the
model. This implies that any variable defined for year 0
is exogenous. Because we have specified a 1-year gesta-
tions lag in the capital creation process. The amount of
capital that sectors have at disposal in year 1 is charac-
terized by a short-run equilibrium in which the supply of
sectoral capital stock cannot be altered. Beyond year one,
however, the supply of sectoral capital stocks is allowed
to change so as to equalize the rate of return on capital.
Arbitrage conditions that relate risk adjusted sectoral
rates of return to the interest rate are enforced by making
the capital stock shifter variable exogenous [7].
For many variables TurGEM-D has no formal theory
and, typically, the values of these variables are specified
exogenously. These variables are technical change and
consumer tastes, indirect and carbon tax tools, risk fac-
tors, foreign prices, foreign interest rates, transfer over-
seas, population, and aggregate real government expen-
ditures. TurGEM-D model is implemented and solved by
GEMPACK4 software [14-16]
To analyze the results of simulation it is convenient to
divide the 16 years time horizon of the experiment into
three sub-intervals. The period of 2004-2009 is the pre-
shock years. The second sub-interval analyzed includes
year of 2010, the year in which the hydro power plant
shock is initially implemented. The supply of hydro
power plant was annually increased by 7%. Next, the
transition from year to the long run represented by the
period of 2011-2020 is discussed.
The initial database of model is the 2004 input-output
tables. Two situations are specified as follows.
1) Historical or pre-shock closure in the period of
2004-2009: Since official data on private consumption,
investment, government consumption, exports, exchange
rate and labor employed are available from the Turkish
Statistics [9], we set growth rates of these variables as
2) Forecast closure in the period 2010-2020: Most ex-
ogenous variables in the historical closure for the period
2004-2009 are set endogenous in the forecast closure. In
the baseline forecast, private consumption, investment
expenditure, government consumption, exports and im-
ports are determined in the model.
4. Simulation Results
This section discusses the simulation results. We first
discuss the impact of power generation shock to certain
macroeconomic indicators and sectoral output as well,
followed by a discussion on the impact of the shock on
carbon emissions.
4.1. Impact of Expanding Hydro Power
Generation on Macro Indicators
The simulation results are presented in a series of real
GDP growth and real consumption, real investment, ex-
ports, imports, and sectoral outputs as the deviation from
baseline scenario as depicted in Figure 3.
The scenario looks into the economic and environ-
mental effects of increasing supply of hydro power gen-
eration on Turkish economy. The supply of hydro power
generation target for 2020 is doubling the supply of 2010.
In order to achieve this energy policy target, Turkey
needs to provide an incentive. Production and investment
subsidy in this sector is used as the needed incentives in
this simulation. Therefore, Figure 3 shows that real GDP,
real consumption, real investment annually increase by
0.14, 0.13, and 0.07 percent respectively in the period of
2010-2020. While exports increase by 0.31 percent, im-
ports also increase by 0.19 percent in the same period.
The terms of trade and allocative efficiency induced
by expanding hydro power generation allow expanding
GDP by 0.144 percent per annum under fixed exchange
rates5. Given that the trade account must be balanced in
year 16, the deterioration in terms of trade means that
any increase in the volume of imports must be more than
offset by an increase in export volumes. In year of 2020
the share of exports and imports are 0.34. As shown Fig-
ure 3, in year of 2020 since both export volumes in-
crease by 0.31 percent and import volumes increase by
0.19 percent, terms of trade deteriorates. To see show
how the increase in real GDP is divided among foreign-
ers and among the various domestic agents we use the
definition of percentage change in real GDP from the
expenditure side (gdp),
where the variables,
are the per-
centage changes in real; consumption, investment, gov-
ernment spending, exports and imports. The coefficients
OECD statics indicate a 5% average annual GDP growth for the last
decade. (OECD country statistical profile, 2010) downloadab
le at queryid =2357#.
GEMPACK is developed by the Centre of Policy Studies, Monash
University, Australia.
Exchange rates are assumed to be fixed in the model. Even though
Turkey has switched to flexible exchange rate sys
tem since 2001, there
have not been big fluctuations in recent years. .
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
Figure 3. Deviation of macro indicators from baseline (%, Turkey).
are shares in GDP of consumption, investment, govern-
ment spending, exports and imports.
Evaluating Equation (2) using values of shares and the
simulation results for year of 2020 we obtain;
gdp16 = 0.66*(0.13) + 0.2*(0.07) + 0.14*(0.0)
+ 0.34*(0.31) 0.34*(0.19)
= 0.144
we can deduce from the evaluation of Equation (2) that
about 0.04 percentage {0.34*(0.31) 0.34*(0.19)} of the
0.14 percent increase in real GDP in 2020 is not avail-
able for domestic absorption (0.1 percent). Since gov-
ernment expenditure is assumed to remain at its control
scenario level that is constant, the deterioration in the
terms of trade is absorbed mostly by the Turkish house-
holds {0.66*0.13 = 0.09 percent}.
We can reports the macroeconomic costs of imple-
menting the hydro power generation scenario in terms of
the percentage change in per capita utility of the repre-
sentative household and the associated terms-of-trade
changes. While the terms of trade deteriorate, welfare of
Turkish household (measured in terms of utility of the
representative household) increases due to the increased
in real consumption.
As can be seen in Figure 4, imposing doubling hydro
power generation considerably reduces crude oil produc-
tion. But, Turkey has no big oil and natural gas reserves.
The most promising and significant domestic energy re-
sources in Turkey are coal (mainly in the form of lig-
nite), hydro and geothermal. The share of domestic en-
ergy resources in terms of world reserves is coal, 0.6%;
geothermal energy, 0.8%; and hydroelectric energy, 1%.
Proven recoverable oil reserves in Turkey are 38.7 mil-
lion tons by the end of 2007. In 2007, total oil consump-
tion was 27.69 million tons, of which 25.56 million tons
were imported and 2.13 million tons produced domestic-
cally. Turkeys oil production in 2007 met only 8.0% of
demand and the rest (92%) was imported, mainly from
Russia, Iran, Saudi Arabia, Libya, Iraq, Syria and Algeria.
Therefore expanding hydro power generation contributes
the lessening the dependency to the imported oil.
The bigger winner from expanding hydro power is en-
ergy intensive industries which records 0.36 percent in-
crease in output. The performance of energy intensive
industries is dependent on competitiveness of its highly
capital intensive commodities and gets additional cost
advantage from the fall in the price of electricity.
Simulation results show that expanding hydro power
plants have a negative effect on agriculture (0.04%).
Electricity are also intermediate inputs used by other
industries, causing considerable indirect real output rise
in the fossil fired electricity (0.15%), oil product indus-
tries (0.06%). Other industries getting particularly af-
fected are gas (0.3%) and coal (0.18%) and other indus-
tries and services (0.03%).
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
Figure 4. Percent deviation of sectoral output from baseline (%, 2020).
It should be noted that these are long run effects, since
substitution normally occurs in the long run, hence re-
sulting in changes in energy structure. In the end, re-
sources will shift from oil extraction industries to hydro
power generation and energy intensive industries.
4.2. Impact of Expanding Hydro Power
Generation on Carbon Emissions
Expanding of hydro power shock has two interrelated
impacts: GDP growth rate impact and energy impact that
induce a change in carbon emission level. This section
seeks to explore whether the possible power generation
shock of the next decade will have any significant effect
on carbon emission growth rate in the Turkish economy.
The impact of expanding hydro power on the CO2
emissions is positive but it is not so significant. The car-
bon emissions is on average 0.012% lower than the base
case over the ten-year simulation period while expanding
hydro power generation are assumed to be doubled.
As is known, carbon emissions are closely related to
energy consumption. Therefore, CO2 emissions are asso-
ciated with all emitting activities, including current pro-
duction, capital formation, and household and govern-
ment consumption [17].
The growth rate of carbon emission is defined as the
weighted average of a firms usage and private and gov-
ernment consumption. Commodities emit carbon into the
atmosphere when they are burned. We calculate the rate
of carbon emission for each region and fossil fuel com-
modities, gco2(i), as the sum of the carbon emissions of
all sources (domestic and imported) and all users (see
Equation (3));
Where i = {coal, oil, oil products and gas} and j = {all
tradable and capital goods}, the coefficients CO2DF is
the emissions from firms demand for domestic product,
CO2IF is the emissions from firms demand for imports,
CO2DP is the emissions from the private consumption of
domestic product, and so on. The corresponding vari-
ables, gco2fd is emissions from firms demand for do-
mestic product, gco2fm is emissions from firms demand
for imports, gco2pd is emissions from the private con-
sumption of domestic product, and so on. We assume
that emissions are proportional to demand: for instance,
emissions from firms demand for domestic product can
be written as follows:
gco2fd (i,j) = qfd(i,j).
[ ]
COIF(i,j,t) gco2fm(i,j,t)CODF(i,j,t) gco
co2(i,t)COIG(i,t) gco2gm(i,t)CODG(i,t) gco2gd(i,t)
CO(i,t) [COIP(i,t) gco2pm(i,t)CODP(i,t) gco2pd(i,t)]
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
Others follow the same pattern. Similarly, we calculate
the rate of the economy-wide CO2 emissions for each
region, gco2t, as the sum of emissions from the com-
modities subject to the carbon tax.
gco2(t)CO(i,t) gco2(i,t)
= (5)
where i = coal, gas, oil products and
By using Table 3 and simulation results we can evalu-
ate growth of carbon emission in line with Equation (4)
for year of 2020. Table 3 shows the amount of carbon
emission released by fuels and percentage changes due to
the expanding hydro power generation.
As shown in Figure 3 doubling hydro power genera-
tion supply leads to a fall in output of the crude oil pro-
duction derived from fossil fuels, thereby contributing to
environmental protection through the reduction of carbon
dioxide emissions associated with the fossil fuels. The
carbon dioxide emissions by the Turkish economy de-
creases by about 0.012 percent per annum as shown in
Figure 5.
In general, the results suggest that the long run effects
of expanding hydro power generation in Turkish econ-
omy with respect to CO2 emissions and GDP growth
move in the reverse direction due to the renewable en-
ergy usage instead of fossil fuels emitting carbon into
atmosphere. Cumulative output gains over the ten-year
projection period resulting from doubled hydro power
generation can be as large as 1.5 percent, while cumuli-
tive CO2 emissions loss can be nearly 0.1 percent devia-
tion from baseline.
Table 3. Total carbon emission and its percentage change by fuel type.
Fuel type Emission
(Million Ton) Share of CO2
Emission Growth of CO2 Emission
(Average of 2010-2020) Percentage change of
CO2 Emission
Coal 91 0.38 0.046 0.017
Gas 42 0.17 0.065 0.011
Oil Products 109 0.45 0.089 0.040
Total 242 1.00 0.012
Source: GTAP database and own calculations.
Figure 5. Percent deviation of CO2 emissions from baseline (%, Turkey).
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
The model presented in this article shares many of
features incorporated in the models in the intertemporal
CGE literature. In order to place TurGEM-D in this lite-
rature we outline key features of four models chosen for
comparison in table A1 in Appendix. The four models
are described by Mercenier and Sampio de Souza [18],
Bovering and Goulder [19], Jorgensen and Wilcoxen
[20], and McKibbin and Wilcoxen [21]. These models
are chosen due to the fact that they represent the latest
developments in the class of models to which Tur-
GEM-D model belongs.
5. Conclusions
One of the most important conclusions of this study for
Turkey as a developing country as implied by the re-
sults of the simulation - is that increasing renewable en-
ergy source that is hydro power decrease carbon emis-
sions without reducing economic growth dramatically.
The net effects of this scenario would reduce even fur-
ther the cost of adopting environment friendly energy
While evaluating the results of energy and environ-
mental policies, one should keep in mind that this model
measures only deviation from the baseline as to the costs
and benefits of these policies. Further potential advan-
tages or disadvantages of hydropower in the context of
power generation have not been captured by the model.
As to the policy recommendation for policy makers
who have to consider carbon abatement policy without
giving up economic development as an ultimate target,
we can say the following. Given the fact that some sort
of a carbon tax reform is a must in the context of adjust-
ment to the EU energy policies, one option is that carbon
tax revenues can be used to finance the adoption of
technological change in the form of shifting more to-
wards renewable energy sources.
As another option, they could be used to minimize the
burden sharing of energy tax in favor of the producers.
Due to the long term positive implications in creating a
less carbon-emitting, more energy efficient economy, we
suggest that energy tax revenues should be used to fi-
nance shifting towards renewable energy-based technol-
ogy and environment-friendly production structure. Be-
cause this is the best policy option in achieving cleaner
environment without harming the capital stock, invest-
ment possibilities and indirect tax revenues. Utilization
of renewable energy sources at a higher degree would
further contribute to reducing dependency of Turkey to
imported energy sources, hence reinforcing energy sup-
ply security.
The introduction of the energy-environmental dimen-
sion in TurGEM-D is only one step towards the elabora-
tion of a ORANI framework that is suitable to analyze
GHG issues. It is hoped that the current version of Tur-
GEM-D could be further extended in order to analyze
some other renewable energy issues, such as new con-
structing or expanding capacity in nuclear power in elec-
tricity generation and using bio-fuels as a fuel in trans-
port sector.
[1] B. Dursun and C. Gokcol, The Role of Hydroelectric
Power and Contribution of Small Hydropower Plants for
Sustainable Development in Turkey, Renewable Energy,
[2] DSI, State Hydraulic Works, Statistics on Hydropower,
Ankara, Turkey.
[3] M. Ozturk, N. C. Bezir and N. Ozek, Hydropower Wa-
ter and Renewable Energy in Turkey: Sources and Pol-
icy, Renewable and Sustainable Energy Reviews, Vol.
13, No. 3, 2009, pp. 605-615.
[4] H. Balat, A Renewable Perspective for Sustainable En-
ergy Development in Turkey: The Case of Small Hydro-
power Plants, Renewable and Sustainable Energy Re-
views, Vol. 11, No. 9, 2007, pp. 2152-2165.
[5] TEIAS, Directorate-General of Turkish Electricity Trans-
[6] S. Kucukali and K. Baris, Assessment of Small Hydro-
power (SHP) Development in Turkey: Laws, Regulations
and EU Policy Perspective, Energy Policy, Vol. 37, No.
10, 2009, pp. 3872-3879.
[7] M. Malakellis, Integrated Macro-Micro-Modeling under
Rational Expectations with an Application to Tariff Re-
form in Australia, Physica-Verlag, 2000.
[8] P. B. Dixon, B. R. Parmenter, J. Sutton and D. P. Vincent,
ORANI: A Multisectoral Model of the Australian Econ-
omy,Amsterdam, North-Holland, 1982.
[9] TURKSAT, The Supply-Use and Input-Output Tables,
Ankara, 2008.
[10] World Energy Council, Energy Statistics, Turkish Na-
tional Committee, Istanbul, 2004.
[11] A. M. Borges and L. H. Goulder, Decomposing the Im-
pact of Higher Energy Prices on Long Term Growth, In:
H. E. Scarf and J. B. Shoven, Eds., Applied General Equi-
librium Analysis, Cambridge University Press, Cambridge,
[12] W. J. Keller, Tax Incidence, A General Equilibrium
Approach, North Holland, 1980.
[13] J. Burniaux and T. Truong, GTAP-E: An Energy-Envi-
ronmental Version of the GTAP Model, GTAP Techni-
cal Paper No. 16, Center for Global Trade Analysis, Pur-
due University, West Lafayette, 2002.
[14] B. V. Dimaranan and R. A. McDougall, Global Trade,
Assistance, and Production: The GTAP 6 Data Base,
Center for Global Trade Analysis, Purdue University,
The Economic and Environmental Impacts of Constructing Hydro Power Plants in Turkey:
A Dynamic CGE Analysis (2004-2020)
Copyright © 2010 SciRes. NR
[15] W. J. Harrison and K. R. Pearson, An Introduction to
GEMPACK, Document GPD-1, Centre of Policy Stud-
ies, Monash University, Clayton, Melbourne, 2000.
[16] W. J. Harrison and K. R Pearson, Computing Solutions
for Large General Equilibrium Models Using GEM-
PACK, Computational Economics, Vol. 9, No. 2, 1996,
pp. 83-127.
[17] L. Aydin and M. Acar, Economic and Environmental
Implications of Turkish Accession to the European Union:
A CGE Analysis, Energy Policy, Vol. 38, No. 11, No-
vember 2010, pp. 7031-7040.
[18] J. Mercenier and M. Sampaio de Souza, Structural Ad-
justment and Growth in a Highly Indebted Market Econ-
omy: Brazil, In: J. Mercenier and T. N. Srinivasan, Eds.,
Applied General Equilibrium and Economic Development,
University of Michigan Press, Ann arbor, 1994, pp. 281-
[19] A. L. Bovenberg and L. H. Goulder, Introducing and
Open Economy Features in Applied General Equilibrium
Models, In: H. Don van de Klundert and J. Sinderen,
Eds., Applied General Equilibrium Analysis, Kluwer
Academic Publishers, Dodrectht, 1991, pp. 47-64.
[20] D. W. Jorgensen and P. J. Wolcoxen, Intertemporal Gen-
eral Equilibrium Modeling of U.S. Environmental Regu-
lation, Journal of Policy Modelling, Vol. 12, No. 4, 1990,
pp. 715-744.
[21] W. J. McKibbin and P. J. Wilcoxen, The Theoretical and
Empirical Structure of the G-Cubed Model, Economic
Modelling, Vol. 16, No. 1, January 1999, pp. 123-148.
Table A1. Key features of models.
Single country with open economy features
Sector produces its own capital using sector specific technology
Energy substitution in its structure
Investment decision are driven by forwards-looking rates of return
Time-to-build investment specification that does not incorporate convex
adjustment costs
Infinitely lived representative household
Labor supply decision are not determined by the solution of utility
maximization problem
Government decision is endogenous
Budget deficit is determined residually.
Sources: Malakallis, 2000
Notes: M-S denotes Merceiner and Sampaio de Souze, B-G denotes Bovenberg and Goulder, J-W denotes Jorgensen and Wilcoxen, M-W denotes McKibbin
and Wilcoxen model.