Sustainable Energy Future: A System Dynamics approach to solving
the Electricity shortfall in Ghana
Benjamin Batinge
System Dynamics Group, University of Bergen
Fantoft Studentboliger, Postboks 298
5075 Bergen, Norway
+96749372
benjaminb50@gmail.com
Abstract
Ghana has been experiencing electricity supply deficit over the past decade. The annual gap
between the electricity demand and supply has been a major concern in the country. Even though
this challenge often seems temporary, it has never been fully resolved. The electricity gap in Ghana
is attributed to underutilization of existing capacity, significant loss of power generated through
transmission and distribution, low investment in the electricity sector, and low electricity tariffs.
A System Dynamics model is developed to create a vivid understanding of the complex feedback
loops within the electricity sector. The results present an outlook of the electricity situation in
Ghana. Policies discussed include the ideal investment pathways for sustainable electricity supply
in the future.
Declining cost of solar coupled with the constant gas shortages for thermal plants makes solar
ideal power source Sor, future energy needs in Ghana. The government of Ghana should review the
existing I I ry fr k to encourage private sector participation. A pricing system
determined by free market activities will reduce government's debt on electricity subsidy and also
offer an incentive for private investors.
Keywords: Electricity, Energy, Ghana, System Dynamics, Investment, and simulation.
Introduction
Energy is an essential sector of every economy. Different economic sectors; education, health,
manufacturing, construction, among others are heavily reliant on energy to function (Ackah et al.,
2014). It is a paramount objective of government to institute measures that ensure sufficient
provision of electricity for economic and social development (Winkler et al., 2011). Studies
(Ferguson et al., 2000; Apergis and Payne, 2011) have established a positive correlation between
electricity consumption and economic growth rates and development.
Ghana has witnessed considerable economic growth in recent times. In 2011, Ghana became one of
the fastest growing economies in the world (approximately 14% growth rate). This has resulted in
an increase in commercial demand and household consumption for electricity due to growth in
industry and extension of Rural Electrification Project respectively. Since 2007, Akosombo
hydropower, which supplies nearly 50% of the total electricity consumed in the country experienced
significant decline in water level as a result of inconsistent rainfall patterns. Consequently, two of
the four turbines in the dam have been shut down in 2014. The Thermal power sector has also failed
to produce at maximum capacity due to frequent breakdown of plants. The sector’s capacity was
heavily constrained in 2013 when a ship anchor severed the gas pipeline which transport gas from
Nigeria to Ghana to power the Thermal plants. As regards, accessing constant and reliable
electricity supply in Ghana for domestic and industrial activities has become a growing challenge.
The country has experienced rampant load-shedding and erratic blackout. Shortage of electricity
access is a leading cause of low levels of economic and social development (Medlock, 2011).
Various projects have recently been instituted to deal with the electricity supply shortage in Ghana
(Tema thermal power project 1 & 2, Takoradi Themal Plant Company, Takoradi International
Company, and the West African Gas Pipeline). In the midst of the severe energy crises in 2007, the
Ghana Energy Commission undertook an energy saving project that led to the distribution of free
compact fluorescent bulbs to replace the high energy consuming incandescent bulbs. All public
buildings were also fitted with capacitor to reduce public sector electricity consumption (Ghana
Energy Commission, 2013). These notwithstanding, the gap seem to be widening. Ghana’s
electricity market demand is forecasted to grow annually between 10%-15% (Acheampong et. al,
2014). The Energy-led-Growth-led-Energy hypothesis (Masih & Masih, 1997; Fatai et al, 2004;
Ghali and El-Sakka, 2004; Akinlo, 2008), which asserts that there is bidirectional causality between
energy consumption and economic growth could be attributed as the reason for the energy crisis
especially given that Ghana recorded significant economic growth of approximately 14%) in 2011.
Data on Electricity Demand, Supply, and Gap
Demand in MW
Effective supply in MW
a y i "Gap
SNtLVHOSNKOHDOSNXO BOS
SsSsSsesSeHHHSANAANA SD
sssseseseoeoqgessosssos
ANgNgannannanrnnaranaan
Time
Figure 1: Annual electricity demand and effective supply between the period; 2000-2014
The figure above shows the total electricity demand and the effective electricity supply in Ghana.
The effective electricity supply refers to the total capacity installed less the transmission and
distribution losses as well as intermittent turbine shut downs as a result of low water level (in the
case of hydro) and inadequate gas supply (for the thermal plants).
There are different studies that have been conducted (Gyamfi, 2007; Ackah ef al., 2014;
Acheampong et. al, 2014) in Ghana concerning electricity issues. Most of these studies (Ackah et
al., 2014; Acheampong et. al, 2014) adopt an econometric approach. A deeper analysis of the
structure and systemic layout of the Ghanaian electricity sector as well as the major parameters
responsible for the electricity demand and supply gap has not been examined. An adoption of the
system dynamic methodology in the analysis of this dynamic phenomenon remains nonexistent.
The objective of this study is to assess the central dynamics that characterize the electricity sector in
Ghana. It also sought to identify the ideal energy investment portfolios (and distribution policy) for
addressing current electricity needs and ensuring a sustainable electricity provision for future
demand in Ghana. The study discusses the implications of the current regulatory framework, market
mechanism, and electricity pricing system on electricity demand and supply. The study also
evaluates policy pertaining to post-generation/transmission losses on the demand and supply gap.
The study raises issues such as:
1. How is the electricity gap in Ghana developing?
2. What is the best energy source to solving Ghana’s electricity challenges in the future?
3. What power investment choices should the government of Ghana adopt?
This presents a dynamic decision point for various stakeholders; the government of Ghana, and
private energy companies that have earmarked Ghana as an investment destination. It also identifies
leverage points for mitigating the persistent power challenges the country encounters annually.
A management simulation model that represents the structure of the electricity sector in Ghana is
developed to provide insights on the sector dynamics and also inform stakeholders on possible
trends of electricity demand and supply. The design of this simulation model is based on past and
alternative future investment patterns of the three major electricity sources (hydro, thermal, solar) in
the country.
The study explores and tests some policy options relating to the regulatory framework of the energy
sector operations in Ghana. A policy of market regulated by demand and supply force referred to as
the Automatic Tariffs-Adjustment Formula introduced by the Public Utilities Regulatory
Commission (PURC) in 2011 is discussed as an alternative to the regulated market flooded with
subsidies and full control. Different policies relating to investment portfolio (hydro, thermal, and
solar) scenarios in electricity, power sources, market share, and price effect are also explored to
identify the ideal policy options for current and future electricity demand/needs in Ghana.
The paper presents a quantitative and qualitative assessment of the dynamics of the electricity sector
in Ghana. The first part (Introduction) identifies the research gap by stating the challenges that face
the electricity sector in the future. It proceeds to state the objective and goals as well as research
questions that help define the scope of the study. The second part is the theoretical background,
which highlights previous studies conducted in the energy sector. It examines the regulatory
framework based on the concepts of system dynamics energy models to produce a brief and in-
depth description of the energy sector. The third part describes the structure of the model developed
to help understand the internal dynamics in Ghana’s electricity sector. It explains the main
equations in the model. The forth section of the study presents the results based on the simulation.
The outcome of the proposed policy options captured in the model is declared. The fifth section
discusses the results from the simulation, the implications of the results, and how it relates to
similar studies in the past. In the last part, conclusions are then drawn based on the discussion and
recommendations made for both policy makers and future researchers on Ghana’s electricity sector.
Background/theory
Among the major challenges in the twenty-first century are increase in climate change and a gradual
depletion in fossil fuel. The global oil consumption is expected to peak and start a gradual decline in
the next twenty years (Randers, 2010). As more international treaties including the United Nations
2
Framework Convention on Climate Change, the Kyoto Protocol, the Copenhagen Accord, and the
Cancun Agreements respond to climate change issues (United Nations, 1998), the decline in the
demand for fossil fuel could be even more rapid. The integral nature of energy in today’s highly
industrialized world has made it impossible to trivialize the repercussions of misplaced energy
investment portfolio in the future. The need for consideration and diversification of investment to
alternative energy sources beyond fossil is inevitable. The global economy is laying the foundation
for the transition to a sustainable energy future. Issues of global warming, the quest for clean energy
(Erdogdu, 2007), the desire for a stable macroeconomic environment and the need to reduce
operational cost in the energy market is driving the energy investment decisions in this modern era.
The global society is currently predominantly dependent on resource-limited fossil fuel. A
sustainable energy transition would be a defining moment for society’s sustainability in the future,
ushering in an economy based on renewable energy flows from an economy based on fossil energy
stocks (Sgouridis & Csala 2014). Energy transitions in the past have often been partial. Biomass is
still a significant energy source (especially in developing countries) and exceeds nuclear energy
notwithstanding the general belief that, the fossil fuel dominance has replaced the use of biomass
(International Energy Agency, 2013). This is similar to the case of the transition from coal to
petroleum and natural gas. These transitions took over a century of innovation and diffusion for
scale sufficiency (Fourquet, 2010). Energy transitions follow the s-curve technology diffusion
pattern that consist of an experimentation phase followed by the dominance stage as a result of
universal adoption, steady stage through standardization, then the emergence of network
externalities, saturation, and possible phase-out (Christensen, 1997; Wilson & Grubler, 2011).
Grubler (2012) also observed that the downstream demand of energy is higher than the upstream
supply. This implies that, the upstream supply services can spearhead energy transition and the
downstream energy demand stock may exert a lock-in effect on the energy supply.
The Regulatory Framework of Ghana’s Electricity Sector
The Public Utilities and Regulatory Commission Act, 1997, established the Public Utilities
Regulatory Commission (PURC) (Act 538). The Public Utility Regulatory Commission (PURC)
and the Energy Commission are the regulatory bodies in the energy sector in Ghana. The PURC is
responsible for setting electricity tariffs. This is often done in consultation with key stakeholders
made up of the electricity generators, distributors and the representatives of major consumers. The
Energy Commission is responsible for technical regulation. In 2006, it established a licensing
framework for licensing electricity service providers. The Licensing Manual for service providers in
the electricity supply industry sets the requirements and guidelines for entities desiring to acquire
licenses to operate in the electricity supply industry. Provisional and full licenses have been issued
to entities engaged in the various segments of electricity supply. Besides adding generating capacity
to existing capacity and enhancing service delivery to customers, the licensing regime enhances the
Commission’s authority to hold the licensees to terms defined in the license.
There are four main state organisations involved in the electricity supply chain in Ghana. The Volta
River Authority (VRA) is responsible for generating electricity. After power is generated, the
Ghana Grid Company (GRIDCo) takes charge of transmitting the power generated through the
grids. Two organisations; the Electricity Company of Ghana (ECG) and the Northern Electricity
Department (NED) distribute the transmitted power to the final consumer. Ghana’s electricity
production is based on three main sources: Hydropower (water from dams), Thermal (gas, Light
Crude Oil, Distillate Fuel Oil), and Solar. According to the Volta River Authority (VRA), the total
installed capacity is 2,814 Megawatts (MW) whiles the effective capacity stood at 2,492 as of
December 2013. About 57% of this supply is from hydropower and approximately 42% resulting
from thermal plants powered by oil and natural gas. The remaining 1% is made up of solar
photovoltaic (PV) and other renewables such as biomass. Figure two below presents an overview of
the major parties involved in the electricity sector in Ghana, from generation to consumption.
Ghana’s Energy Policy from 2010 sets out to increase the share of renewable energy in the national
energy mix, through focusing on improved efficiency of fuel wood use, as well as shifting from use
4
of biomass to use of other alternative renewable energy sources, such as wind and solar. Currently,
the share of modern renewables in the energy mix is insignificant (Ministry of Energy, 2011). The
energy sector goals include increasing installed capacity from about 2,000 MW to 5,000 MW by
2015, and establishing universal energy access by 2020 (Ministry of Energy, 2010). The current
strategy aims to increase the renewable energy share in the country’s electricity mix to 10% by
2020 (Ministry of Energy, 2010). As at 2010, the installed electricity production in the country had
reached 2,185.5 MW, with 1,865 MW available (Energy Commission of Ghana, 2011). Most of the
generated power in 2010 came from hydroelectric sources and accounted for nearly 70%, with 30%
generated from thermal power. Electricity demand in the country is estimated to be growing at a
rate of 10% per year in 2012 (Ministry of Energy 2012). The government estimates that at this
growth rate, an additional 200 MW capacity every year is required in order to meet the demand.
PLANT FUEL TYPE Tam INSTALLED CAPACITY (AW)
ree eres
Adiiens Woter 1,020 900
But water ‘200 342
mich water 60 120,
Santora 2580 382
ahaanNaanarniew
; 378 300
252 200
su 220 iso
tee uo
Tema Thermal 2 Power Plant (TT2PP) a5 as
Takoraai T3 tae 20
Mines Reserve Plant (MAP) es 20
Effasu Power Barge 125 100
SubTotal E49 aaa
cniaa Saran
Genser Power - IPP LPG 5 aa
Sub-Total 5 Za
conaeblas
RA solar Siacshin 25 19
Sab Total 33 ee
ssi PxaeE
1ot the utilities (VRA & ce) provide.
Table 1 - Source: Ministry of Energy 2014
The problem of energy investment is complex and important. Even though different studies have
been conducted in the energy sector highlighting the looming challenges of over reliance on fossil
energy (Randers, 2010; Humphreys, 2014), and also addressing global energy issues, very little
studies address the different investment scenarios to bridge power supply gap in Ghana. No study in
the extant literature adopts the system dynamics methodology to create a fundamental structure of
the Ghanaian energy sector for detailed analysis. Based on the trend of global energy demand, the
rate of shifting dominance between fossil and renewable energy, and emergence of renewable
energy amongst competitive energy technologies, different investment policy options are evaluated
in this study. The study also proceeds to evaluate policies related to price, and post-production
losses (transmission and distribution), which are very significant in the case of Ghana.
Model
The application System Dynamics modelling methodology in energy research is not novel. Besides
its application in energy market dynamics and economic indicators (Naill, 1977), System Dynamics
is used by different studies (Chi, et al., 2009; Connolly et al., 2010) to conduct simulations for
energy development and energy structure testing. It is also applied in studies such as Anand et al.,
(2005) and Feng et al., (2013) who studied the environmental aspect of energy and CO2 emissions.
Issues of energy security resulting from supply and demand in country specific cases have also been
examined by Wu et al., (2011) and Shin et a., (2013).
None of the studies on energy in Ghana to the best of my knowledge adopts the System Dynamics
Approach, which first underscores the fundamental complexities of a system and evaluates possible
scenarios through simulations to prescribe potent blueprint for sustainably secure energy future in
the country. A system Dynamics model is therefore ideal in many relative terms for understanding
such endogenous dynamics.
The simulation software, iThink, version 10.0.6 was used to constructed the model and conduct all
simulations. To improve readability of the figures, the results of the simulation were exported to
excel and the graphs constructed and transferred to the main thesis report. The simulation results
presented in graphs and tables are detailed and easy to read and interpret.
The central focus of the model is to create a structure that represents the electricity sector in Ghana.
This provides insight on the internal dynamics creating the persistent power crises and also makes it
easier to identify police leverage point to rectify the problem. The model also sets to determine the
investment for future capacity demands. The simulation period is 31 years. This is decided based of
the historical period under consideration. It is often advised that, one should look as far back as one
looks forward. The simulation period starts from 2000 to 2030.
The major stakeholder that this model most appropriately serves is the Government of Ghana and to
an extent, independent power producers (private investors) who identify Ghana as a prospective
energy investment destination. Needless to say, an improvement in the power situation in Ghana
positively affects the citizens and other entities that are not directly connected to the power sector.
From the government’s perspective, the model provides a better understanding of the underlying
causes of the electricity crises but most importantly the major contributing factors which when
addressed/leveraged can result in a much larger improvement. Investors can review the simulations
and analysis from the model to understand the trend of demand for the different power sources. This
offers a sense of direction for their future investments.
The model also includes the electricity-pricing sector. In Ghana, this falls under the jurisdiction of
the Public Utilities Regulatory Commission. Currently, this regulatory body determines tariffs from
time to time. The study assesses the effect a liberalized market would have compared to this ‘fixed-
term’ pricing system. A liberalized electricity market will not only reduce this burden, issues of
subsidy could also be addressed through that.
Model structure
The model structure is informed by empirical studies that have examined the major factors
accounting for electricity demand and supply gap and the dynamics of energy investment and
portfolio diversification such as Humphreys (2014). It takes into account the major power source
for electricity production, the investment made in these sectors, the price of electricity, and the
demand over the period under consideration.
The Electricity Supply Sector
The Electricity supply sector in Ghana connected to grid consists of three main power sources;
hydroelectric power, thermal power, and solar PV. In terms of off-grid power consumption,
biomass is the leading energy source. Off-grid supply is however not the focus of this study.
Hydroelectric Power: The model contains the hydropower sector which is made up of three main
sources: Akosombo hydropower, Kpong hydropower, and Bui hydropower. These together
constitute about 50% of electricity produced in the country. The first hydro plant was built in 1965.
The facility has since received improvements in capacity as the electricity demand increased over
time. The current capacity stands at 1,020 megawatts. Since 2000, other hydro plants have been
built to cater for the growing power needs. These were the Kpong hydropower, which has a total
installed capacity of 160 MW, and the Bui Hydro project, which has a total capacity of 400 MW.
Collectively, these hydropower sources make up 53.8% of the total installed power capacity as of
2013 (Ghana Energy Commission, 2013). The Energy Commission also projects a potential
undeveloped hydro capacity of 195 MW from three water sources. Even though there are other
potential sites for developing hydro plants, they are mini capacity projects, the sum of which is less
than half the current operating capacity. The total hydro capacity potential in Ghana is therefore
limited. The productivity of the hydro source is mainly resource (water) constraint.
The hydro capacity is shown the stock and flow diagram in figure 3 below. The structure starts with
the stocks of total hydro capacity installed and capacity under construction. When construction is
completed after a construction time of 5 years, the plant becomes ready for use and the capacity is
therefore added to the installed capacity. The capacity installed is depleted by the rate of
depreciation over time. The hydro constructed increases by the hydro project initiation rate. The
hydro initiation rate is the total annual amount of investment (in Ghana cedis) allocated to hydro
projects. This amount is then divided by the unit cost of installing a MW of hydropower. Initially,
the unit cost was set constant but increases slightly with challenges that affect the hydro production.
The hydro plants depend on rainfall for them to function to capacity.
As a result of seasonal rainfall inconsistencies, water level in the dam is often below capacity. Some
hydro turbines are therefore shutdown especially during certain seasons of the year, reducing the
utilization factor, which represents the percentage of installed capacity actually generating power.
Akosombo dam, the largest source of electricity supply suffers this setback of low utilization as a
result of low water level the most and produces below capacity. The average production
potential/utilization factor of hydro, which depends on rainfall, is approximately ninety percent.
The rainfall pattern determines the utilization factor for hydro since the number of turbines
operating at a time depends on the water level. Most of the hydro shutdowns are often water related
rather than damage and maintenance. The utilization factor in the model is therefore equivalent to
the rainfall pattern. The product of the utilization factor and the total hydro capacity installed is the
Effective hydro capacity (power generated) hence,
Ehc = (ufh* Ihc,
Where hic is the Effective hydro capacity, ufh is the utilization factor for hydro, and Jhc is the
hydroelectric power capacity installed. The hydro capacity installed changes with the hydro
depreciation rate and hydro project completion rate. The hydro depreciation rate is, Hdr =
(Ihc/hit), where Har is the hydro depreciation rate, and hit is the hydro lifetime which is hundred
years. The hydro project completion rate (Her) is given by: Her = (HC/ht), where HC is the
Hydro capacity under construction, and At is the hydro construction time which is three years.
The hydro capacity under construction depends on the hydro project initiation rate, which is a
function of the average hydro cost per MW and the total amount of investment in Ghana cedis
budgeted for hydro projects. The equation hydro project initiation rate is therefore: (Hp = Hi/Ch),
Where Hp is the hydro project initiation rate; Hi is the Hydro investment in cedis; and Ch is the
Actual Cost per MW Hydro. The actual cost per MW hydro is a function of the initial cost per MW
hydro (Init Ch) and the effect of rainfall relative to dam capacity on MW cost hydro. The actual cost
per MW hydro is given by:
Ch = Init Ch* (1+ eRf)
Where Ct is the Actual Cost per MW hydro (Jnit Ct) is the Initial cost per MW hydro, and eR/is the
Effect of rainfall/dam water level on cost MW of hydro. As the dam water level increases, the
operational cost decreases with the effect of rainfall. The effect of rainfall effect is given as:
eRf = (1— Rf) where Rf is the average annual rainfall/dame water level.
Thermal Power: The thermal sector is the second major source of electricity. It has a total installed
capacity of over 40% of total electricity consumed. Similar to the hydropower, this also has its own
limitations. All of Ghana’s thermal plants depend on gas. Most of the gas is supplied by Nigeria
through the West African Gas Pipeline. The effective thermal capacity therefore depends on the
availability of gas. When there is limited gas supply, the thermal facility utilization factor is below
1. Another thing that affects the thermal utilization is the investment in solar. As more solar PV
(with increasing returns) is constructed, its substitute (thermal) becomes less attractive over time.
Thermal construction in Ghana began in the late nineties as electricity demand increases and the
hydro capacity became overburdened. Most of the thermal capacity on Ghana was constructed after
2000. The initial thermal capacity as of 2000 is 100 MW. The stock of installed capacity of thermal
depends on the thermal depreciation rate and the thermal project completion rate. Thermal
depreciation rate (Tdr) is given as Tdr = (Itc/tlt) where Itc is the Thermal capacity installed, and
tit is the thermal lifetime. The thermal project completion rate is a function of the thermal capacity
under construction and the construction time, which is two years. The equation of the thermal
completion rate is given by: Tcr = (TC/tt), where Tcr is the Thermal project completion rate, TC
is the Thermal capacity under construction, and ¢t is the Thermal construction time which is two
years.
The stock of thermal capacity under construction depends on the thermal project initiation rate,
which is a function of the average thermal cost per MW and the total amount of investment in
Ghana cedis allotted to thermal. The equation of the thermal project initiation rate is given by:
(Tp =Ti/Ct), Where Tp is the Thermal project initiation rate; Ti is the Thermal investment in
cedis; and Ct is the Actual Cost per MW Thermal. The actual cost per MW thermal is a function of
the initial cost per MW thermal and the effect of gas availability on MW cost. The actual cost per
MW thermal is given by:
Ct = Init Ct * (1+ eAg)
Where Ct is the Actual Cost per MW thermal; Jnit Ct is the Initial cost per MW thermal, and eAg is
the Effect of gas availability on cost MW thermal. As the gas availability increases, the operational
cost decreases with that effect. The effect of rainfall effect is given as: eAg = (1 — Ag) where Ag
is the availability of gas. This is equivalent to the utilization factor. It is the same as the utilization
factor is plant redundancy is not as a result of damage. The total effective capacity is the product of
the utilization factor and the total installed capacity.
Etc = (uft * Itc)
Where Etc is the Effective thermal capacity, and ufi is the utilization factor for thermal. In recent
times, Ghana’s thermal faces gas shortages. There is inconsistency in supply forcing some of the
plants to be shut down. Increasing the number of thermal plants/capacity seemingly compounds the
problem.
Solar PV: Solar PV is another growing energy source in Ghana. It is one of the little renewable
energy that is reliable. Government has undertaken some pilot projects in the field of solar to
supplement the energy shortfall. In the long run, it could become the leading energy source. It is
however at the moment limited by cost and development in solar technology compared to other
major energy sources. The solar power could tend to have a negative correlation with the hydro. As
the dam water levels decline during the dry season and turbines are shut down leading to lower
utilization factor, Solar could emerge as an ideal substitute because it would have a higher
utilization factor with warmer and sunny weather that characterizes the dry season. The total
effective capacity is the product of the utilization factor and the total installed capacity. The
effective solar capacity is given by:
Esc = (ufs * Isc)
Where Esc is the Effective solar capacity, ufs is the utilization factor for solar, and Jsc is the solar
capacity installed. Solar utilization factor is relatively equivalent to hundred percent unless damages
occur on capacity installed, since the conditions in Ghana are favorable for all year solar
production. The installed solar capacity is dependent on the solar project completion rate (Scr),
which is a function of the solar capacity under construction (SC) and the solar construction time
(st), which is three years: Scr = (SC/st).
The solar capacity under construction is a function of the solar project initiation rate given by the
equation: (Sp = Si/Cs), where Sp is the solar projects under construction, Si is the annual
investment in Ghana cedis allocated for solar production, and Cs is the cost per MW of solar unit.
Contrary to thermal, the cost per MW solar is expected to decline over time with the effect of
learning improving technology and efficiency. The cost per MW solar is:
Cs = Init Cs * (1 * le)
Where Cs is the Actual Cost per MW solar; Jnit Cs is the Initial cost per MW solar, and /e is the
Effect of learning curve on cost per MW solar. The effect of learning (/e) is given as (/-/c), where Ic
is the learning curve. As the learning curve grows, the learning curve effect on cost becomes lower
and the multiplier effect on the cost unit of MW solar becomes smaller. Learning curve is an
essential part of the model as it determines the solar adoption rate.
The Learning curve
Different studies (Moxnes, 1992; Wang et al., 2012; IRENA, 2012) on output effect on learning
curve and Solar PV technology valuation assume price reduction consistent with cumulative
production. Solar cost is expected to decline over time as a result of the learning curve effect. This
will result in increasing returns on solar investment. Some of the features of technology, which
provides increasing returns, are the large set-up (initial) cost, learning effects, co-ordination effects,
and self-reinforcing expectations (Arthur, 1988). In the light of this, the real reason for capacity
changes might not be related to learning.
Evaluating PV technologies with single-factor learning curve would likely result in overestimation
of the effects of learning-by-doing (Chanwoong and Junesuek, 2014)). A solar PV valuation that
adopts a two-factor learning curve; cumulative production and technological innovation driven from
Research and Development (R&D) reduces the estimation deficiency of the single-factor learning
curve applicability especially in technologies where R&D leads to rapid technological change
(Kouvaritakis et al., 2000). Subsequent studies on renewable energy cost/price estimates
(McDonald and Schrattenholzer, 2001; Kobos et al., 2006) supported the two-factor learning curve
framework by incorporating it in their evaluations of learning curve. This study considered a
learning curve driven from the accumulated production of solar and the R & D. The solar capacity
is therefore not directly linked to the learning curve. Instead, a learning curve based on similar
studies that accounted for detail variables such as knowledge stock, depreciation, and R & D time
lag (Kobos et al., 2006) is adopted. The learning curve in this study is illustrated in the graph below:
Learning curve
0.35
°
we
0.25 Y= 0:0085x= 17.076
R2=0.8708!
S
i
= Learning curve
Percentage
S
a
—— Linear (Learning curve)
°
5°
aR
Figure 2: Learning curve
The learning curve pattern in this study is derived from the findings of previous studies, which
found relatively similar pattern of learning curve using the two-factor analysis. Kouvaritakis et al.,
(2000), deduced a cumulative production effect of 16% and R & D effect of 7%. Criqui et al.,
(2000) arrived at similar result. They found 16.4% and 4.4% respectively. In a subsequent study by
Miketa and Schrattenholzer (2004), the learning by doing rate of 9.7% and learning by searching
rate of 10% was illustrated.
The Electricity Demand Sector
The electricity demand in Ghana has witnessed considerable growth. Whiles this is attributed
mainly to economic growth and expansion of industrial (mining, construction, etc) activities, there
has also been a considerable extension in the grid connection. The execution of the rural
electrification project contained in the Strategic National Energy Plan resulted in a direct increase in
the demand for electricity. The fact that there was a gap prior to the enforcement of this policy in
the last decade aggravated the demand supply gap, as more strain is place on supply.
The indicated demand
Different energy researchers have estimated electricity consumption using different methodologies.
Ranging from the widely used reduced-form model of Engle and Granger (1987), to the structural
form model of Kokkelenberg and Mount, (1993), and the Genetic Algorttm of Ceylan and Ozturk
(2004), the electricity demand in this study is determined by the indicated demand. The word
“demand” is used here cautiously because, the demand does not necessary refer to what is
consumed but rather the electricity required. In the event of electricity shortage, which is the case in
recent times, the electricity consumed is equivalent to the total effective electricity
supplied/distributed. The indicated demand is a function of demand, price, annual growth, and the
price elasticity. The equation for indicated demand in the model is given by:
Init D * (1 + g) * gt * ((P/P0)””)
Where: Init D: is the Initial Demand in the year 2000. The amount of electricity demanded in 2000
according to historical data is 1161 MW. This represents the initial demand at the beginning of the
simulation. The demand growth rate is denoted by g, which is 6%. The historical growth of demand,
gt, is averaged at 6%. FP is the current price of electricity. P0, is the reference electricity price
which is adjusted to the electricity price over time. PE: represents the price electricity of electricity.
In Ghana, the price electricity of demand is -0.38. This means that a percentage change in price will
lead to a more than proportionate change in electricity demand. The simulation results of the
indicated demand can be seen in figure 17.
The demand for electricity is given by:
Total_electricity__demand(t) = Total_electricity__demand(t - dt) +
(Change_in_ electricity demand) * dt
Where, demand increases or decreases every time step according to the development of the
indicated demand. As seen in the structure above, the demand is adjusted to the indicated demand.
The Electricity Price Sector
The electricity price is another aspect that is very essential in the model. Currently, the Public
Utilities Regulatory Commission fixes the electricity price in Ghana over a period of time. Prices
are barely reviewed unless there is a huge global price change. The government, as the regulatory
framework limits private sector participation, mainly supplies the electricity in Ghana. This has
resulted in price stagnation most of the time, which is not reflective of the indicated market
situation. The electricity price is given by:
P = PO « (ds)?s
Where: P is the Electricity price and PO is the reference electricity price as represented in the
indicated demand formulation. Then, ds represent the demand/supply ratio and Ps denotes the price
sensitivity of the demand and supply ratio. The ds and the Ps are accountable for the price
dynamics. Unlike the indicated demand, the price sensitivity has no direct effect on the electricity
supply because the market is not liberalized. When ds > 1, it means the demand exceeds the supply
and the price sensitivity determines the electricity price. In Ghana; Ps is 0.8 (Ps < 1). This implies
that, the electricity price is lower than the indicated price in a free market. A Ps = / means the price
does not respond to the interaction of demand and supply. On the other hand, Ps > / means the
in
price is significantly high and demand exceeds supply in the free market. The effect of different Ps
values between 0 and 2 can be seen in figures under the sensitivity analysis section.
Electricity price in Ghana does not directly affect the supply. It must be noted however that, price
does have an effect on supply through the demand. As Ps increases, demand decreases and that
results in a decline in ds and subsequently and decrease in investment.
The Electricity Investment Sector
Investment over time in the electricity sector accounts greatly for the current available capacity. The
annual investment made in the electricity sector before 2014 is not readily available. Calculations of
the annual investment were conducted based on the total amount of megawatts of power installed
over the simulation period and the average unit cost of a megawatt. The results arrived at was taken
as the average investment in cedis made in the electricity sector from 2000 — 2014. Thereafter, the
investment is calculated based on the demand and supply gap and the fraction of GDP that
represents the electricity sector investment. The amount of cedis investment in the electricity sector
is distributed to the various power sectors. The ratio of distribution was based on data from the
Energy Commission.
The amount of funds made available for investment in the electricity sector depends on the demand
and supply gap. The gap is the difference between the demand and the total electricity distributed to
consumers and accounted for. The amount of power distributed is the difference between the total
effective capacity and the transmission and distribution losses. The demand supply gap in
megawatts is multiplied by the average cost per MW of power installation to arrive at the budgeted
investment in cedis:
Budgeted_Investment_in_cedis = Demand_Supply_gap*Average_cost_per_MW_unit
The indicated investment on the other hand does not depend directly on the demand and supply gap
but rather the trend of GDP. A constant fraction of 1.5% is estimated to be the annual investment
needed in the electricity sector. The indicated investment therefore increases according to GDP:
Indicated_investment_in_cedis = Fraction_of GDP__to_energy*GDP
The annual investment in cedis is therefore a function of the indicated investment and the budgeted
investment:
MIN(Budgeted_Investment_in_cedis,Indicated_investment_in_cedis)
This annual investment in cedis is apportioned between the three power sources: hydro, thermal,
and solar. Initially, a constant proportion is assumed for both over the historical period based on the
capacity installed within such period. Hydro is the ideal source of electricity production in Ghana.
Unfortunately, the potential hydro sites are limited. The divestment of investment became
inevitable. There are two main investment pathways for hydro; low investment scenario where no
more hydro capacities are developed because they are small sites used as tourist venues, and the
high investment scenario where the highest undeveloped remaining hydro potential of 400 MW is
developed. The remaining investment after hydro is then shared between thermal and solar. Based
on the installation over the period, the fraction of solar was about 1% of remaining investment after
hydro and the rest was invested in thermal. This fraction however changed as the challenges in
thermal become more apparent and the cost of solar declines. The new fraction of investment in
solar was therefore model as a logistic function.
The investment in Ghana cedis of hydro is given by:
Historical_Hydro_investment_in_cedis*(1-
Hydro_investment_switch)+Current_hydro__investment_in_cedis*Hydro_investment_switch,
Where the historical investment is the annual hydro investment from 2000 — 2013, which is the
annual investment multiplied by the fraction allocated to hydro
(Fractional_hydro__investment*Average_Annual__Investment). The fraction hydro investment is
40%.
The current hydro investment is the investment in Ghana cedis after 2014. This is given by :
0+ STEP(Desired_Hydro__ installation. MW*Actual_Cost_per_MW__Hydro,2015)
Where the desired hydro installation MW is the annual amount of new hydro capacity required
depending on the maximum hydro target. It is given by: (Hydro_Capacity_to_be_Installed-
Hydro_capacity _installed)/Capacity_adjustment_time,
Where the Hydro Capacity to be installed is the hydro target.
The Solar and Thermal Investment Distribution
The equation for the fractional investment in solar PV after 2014 follows the logistic function given
as:
. L
Si=z + ex(es—Ct)
Where: L is equivalent to 1, and denotes the upper limit of Solar PV investment, which is assumed
to be Annual electricity investment, less investment made in hydropower. Cs: is the cost of a
megawatt unit of solar in Ghana cedis over time, Ct: is the cost of megawatt unit of a Thermal in
Ghana cedis, o<: is the unit multiplier and, e: is the exponential growth of the cost fraction over
time.
The investment in thermal is therefore given by:
Ti=1-([a00)
The equation above is consistent with the reasoning of Christensen, (1997) and Wilson & Grubler,
(2011) that energy transitions follow the s-curve technology diffusion pattern. It is also consistent
with the market share distribution between two competing technologies by Moxnes (1992). The
investment in energy between solar and thermal is consistent with the theories of competing
technologies. The Success-to-the-Successful (Braun, 2002) is operating between solar and thermal
with the financial sector been the independent variable.
The investment in Ghana cedis of thermal is given by:
Investment_for__Thermal_and_solar-Solar_investment_in_cedis,
where investment for thermal and solar is the difference between total annual investment in cedis
and the hydro investment in cedis.
The investment in Ghana cedis of solar is given by:
IF Solar_Thermal_Policy_swicth =1 THEN
(Investment_for__Thermal_and_solar*Fractional_Solar__Investment*Fractional_change)+(Invest
ment_for__Thermal_and_solar*New_solar__investment_fraction) ELS
(Investment_for__Thermal_and_solar*Fractional_Solar__ Investment)
The ‘IF THEN ELSE’ function is necessary to formulate the change in solar investment, and also
account for scenarios where solar policy is active and dormant.
The Causal Loops
The gap in demand and supply is an opportunity for potential investments in renewable energy.
These investments could be focused on efficient and emerging technologies in renewable given
shortcomings of fossil energy such as CO2 emission. The feedback loops in the model are mainly
balancing loops. The more the investment in electricity, the smaller the demand and supply gap
becomes and lesser investment are consequently required. The investment decisions concerning the
different power sources are a major part of this paper.
There are about seven balancing loops and four reinforcing loops within the model. The major
dynamics are a result of interactions between sectors. The causal loop diagram below presents a full
overview of the dynamics in in the model. Parameters/variables highlighted indicate policy point for
addressing the electricity concerns. The next chapter discusses in detail the result from the
simulation and policy outcomes.
Te
to oS — gene
y Atak
een
@.
Calica
@Y |
a
Hydro investment —- MW Hydro
cod . depreciat
Sofa
investment . four deprecation
‘Thema capacity
Availity of gas
Soke capacity Weaming curve
oO 3
SxL_“Sobar actin of
investment
Average cost per
“Thermal
BS
‘Thermal Fraction of
“Thermal investment
Renual investment in
cebeetricity
GDP faction
electricity i
Figure 3: A Complete Causal loop diagram
The full model layout
Average cost per
nW Solr
Effective installed
capacity
lee Toate Doms
Distribution losses a af
es
‘Average cost per
MW
‘Total Demand.ag
Demand supply
sp
Budgeted yf
fnwestment
ma) GPP gown rate
apr
Reference price
ait price
Inte price
wm) +
Tndicated demgnd-<¢— Initial demand
Prive ekesticiy
Anmal demand °
‘growth
Figure 4: Stock and Flow diagram of the Model
The structure of the model consists mainly of four parts as in the description above. One thing that
is essential in the model structure is that, due to various limiting factors, there is a distinction
between installed capacity and effective capacity. The later refers to the fraction of the former that
is fully operational. There is also a gap between the total effective capacity and the capacity that is
consumed. The difference between these two is the power lost through transmission. There is a
significant loss of power through distribution.
Data and model Validation
Model validation is an essential part of system dynamics application. In order to ensure certainty
that a given model developed represents a given underlying system, it needs to be rigorously
examined before policies based on the model can be tested. There are difference forms of model
validation: boundary-adequacy test, structure-verification test, parameter verification test, extreme-
condition test, behaviour replication test, and dimensional-consistency test (Forrester and Senge,
1980). It is not necessary for all these tests to be conducted before a model can be deemed valid.
According to Barlas (1996), a behavioral validity for a system dynamics model can be a sufficient
to ensure that the model is valid.
The relevance of model validity is emphasized by subsequent studies in system dynamics. In order
to strengthen the model validity, other calibrations and tests such as extreme condition test, unit
consistency check (already conducted in the model), base run and reference behaviour comparison
(demonstrated under figure 13 & 14), parameter sensitivity analysis, and structure-behaviour test
(figure 15, 16 & 17) should be conducted (Sterman, 2001; Wheat & Saldarriaga, 2011).
The table shows the different power sources in the country and the total capacity of different plants
under the power categories. The annual capacities from 2000 — 2013 are then compared to the
simulation results.
Table 2 below shows the Plant Capacities of the diferent energy sources
Plant 2000 | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013
ration r i r i
Akosombo | 5,557 | 5,524 | 4,178 3211 | a,40a| 4,718 | 4,690 | 3,104 | 5,254 | 5,842 | 5,961 | 6,295 | 6,950 | 6,727
Kpong | 1,052] 1,085] 85a | 675 | a77 | 911 | 929 | 62a | 941 | 1,035 | 2,035 | 1,066 | 1121 | aaa
gui} va} na | wa | na] na | na | na | wa | sa | na | wa | na | na | 302
Tub Feral | 6,609 | 6609 | 5,036 | 3886 | 5281 | S620 | 5619 | 3,727 | 6195 | 6a77 | 6906 | 7561 | HOME | 8239
Thermal Generation | Te f— i i —a oF i i aii a ey 1 = i
Takoradi Power Company (Tapco) | 346 | 740 | aza | 1328] 36 | asx | 1416 | 1521 | ava | as3 | 1.234 | 1,137 | 1,001 | 1,783
‘Takoraul International Company (Tico) | 268 | sto | 1,363] 668 | 222 | 2a | 1,395 | 2,417 | 3,063| 1,040 | 1160 | 657 | 1,168 | 1,032
Toma Thermal 1 Power Ptane(rTieP) | wa | na | va | na | wa | na | wa | na | na | szo | sor | sso | 22 | a75
Tema Reserve Power Ptant(tReP) | wa | wa | na | wa | na | wa | wa | a2 | ss | wa | na | na | na | na
Emergency Reserve PowerPlant (erPp)| NA | na | wa | wa | na | na | wa | go | 45 | wa | wa | na | na | na
Kumasi Reserve Power Plant xApP)| NA | NA | NA | NA | NA | NA | NA | 33 | a6 | Na | NA | NA | NA | NA
‘Mines Reserve Plant (map)| NA | NA | NA | NA | Na | na | na | 32 | 46 | a8 | 20 | a2 | 20 | wa
Tema Thermal2 Power Plant(TT2pP) | wa | wa | na | wa | na | na | wa | na | na | wa | 28 | so | aa | 94
Sunon Asogh Power (Ghana) ted (sarp) | wa | na | na | wa | wa | wa | na | wa | wa | wa | 138 | 1224] saa | 694
cenit Energytrd(cet)| wa | wa | na | na | na | na | na | wa | na | na | na | na | 94 | asa
takoradit3| wa | ma [wa | na | na | na | na | wa | na | na | wa | na | na | 202
‘Sub-Torat | 614 | 1250) 2237 | 1996 | 758 | L1s9 | Zeit | 3251 | 2129 | Bost | Zit | sew | aasa | ais
Renewables
vrasolar| wa | na | Na} na | Na | na | wa | NA | na | na | wa | NA | NA 3
Tora 7.223 | 7.859 | 7,273 | 5RRD| 6.039] 67am | 8 a30 | GAR | 8394 | BSR | 10,167 | 11,200] 12,08 | 12,870
Tnstalled Capacity (MW) Tas [iss | tera] 1382] i730] 4730 | 4730 | 193s | geal | 4970 | ates | 2170 | 2280 | 2aa7
Source: Ministry of Energy
In the figure below, the trend of the different power sources is show for the simulation period. In
order to validate the model, the simulated results in compared to the historical behaviour in the table
above. It is clear that, the simulate results in consistent with the reference data.
Base run of Hydro, Thermal, and Solar capacity
4000
3000
2 2000 Hydro capacity installed
1000 ee Thermal capacity installed
0 Solar Capacity Installed
2000
2002
2004
2006
2008
2024
2026
2028
2030
2010
2012
2014
2016
2018
2020
2022
Figure 5: Simulated capacities of the different power sectors
Results of total capacity is not enough to affirm the validity of the model hence, other simulation
results are evaluated. The demand and supply gap, the foundation of this study, is also juxtaposed
with the reference data. As indicated in figure (6) below, the simulation results are consistent with
the reference behaviour. The electricity demand and supply gap are similar to the historical data.
Little disparity is observable. This is attributed to parameter assumptions such as average cost of
MW per unit, which is taken as a constant figure in the model for lack of data. Other parameters
such as the utilization factors were average values and not exact data over time. This does not
adequately represent the real cost per MW. Other assumptions include the effect of learning curve,
the cost of the different energy sources, the actual annual investment, among others.
Demand and Supply gap
B 200 SS —2: Base run (BAU)
== Gap Data
DPNYTOYCHDON*HTOCHASONYXtOAS
SSSSCSHSHHHAHANAN ANAT
{RS § § & RRS F F&A RSA
-300
Time
Figure 6: Demand and Supply Gap
Scenario and policy description
As part of validating the model, a parameter sensitivity test is also conducted to determine whether
the mode responds as expected to parameter variations. The efficiency of sensitivity analysis
depends on the extent to which variations in the model behaviour as a result of parameter changes
can be deemed to have occurred due to such parameter changes. To ensure this, the model is
initialized in equilibrium. Price is a central parameter any time demand and supply are involved. All
things being equal, supply is expected to increase when price is high and decreases when price is
low. Demand on the other hand is expected to increase when price is low and decreases when price
is high. This can be seen in figure (15) below. As prices sensitivity is increased from zero (0) to two
(2), demand falls from 7,000 MW to about 3,500 MW. This is expected based on the model
construction/structure.
Effect of price sensitivity on total Electricity demand
8000 1: Total electricity demand
> 6000 se ——2?: Total electricity demand
= 4000 .
"3: Total electricity demand
2000 ~ =
——4: Total electricity demand
0 SL S20 Se ee ee
ony ont ont S ames; ici
Sees sesSaenseeseseanganssGs 5: Total electricity demand
SERS SSRS ESHER SS
as “ss aN "6: Total electricity demand
Time
Figure 7: Effect of price sensitivity on Total Electricity Demand with varying price sensitivities (1:
0; 2: 0.4; 3: 0.8; 4: 1.2; 5: 1.6; & 6: 2.0)
The figure (16) below is the price sensitivity effect on supply. As price increases from zero (0) to
two (2), the supply of electricity falls from about 5,900 MW to 3,300 MW. This seems illogical,
however, it is the correct behaviour based on model structure. In the model, price has no direct
influence on supply. This makes sense because, in Ghana, electricity sector is owned and regulated
by the government. The absence of private sector participation eliminates possible competition and
grants autonomy to the state. Price changes in electricity are therefore not market determined but
rather fixed and reviewed by a state agency periodically. As price only affects the demand, an
increase in price will lead to a fall in demand, which in turn leads to a decline in supply. The price
increment does not set the incentive for an increase in supply as it would in the case of a free market
operation.
Effect of price sensitivity on Total Effective Electricity Supplied
7000
6000 A 1: Total capacity supplied MW
5000
= 4000 A ———2: Total capacity supplied MW
= 3000 —3: Total capacity supplied MW
2000 > r :
——=4: Total capacity supplied MW
100 —==— pachy:SupP!
0 — : : —— "5: Total capacity supplied MW
gePesovsseanexyes ity suppli
SSS S25 55 5 2 88 8 8 § 3 —6: Total capacity supplied MW
ANNAANANANNANAANNAAAN
Time
Figure 8: Effect of price sensitivity on Total Effective Electricity Demand with varying price
sensitivities (1: 0; 2: 0.4; 3: 0.8; 4: 1.2; 5: 1.6; & 6: 2.0).
Indeed, investment in energy is affected by the price. Even though the electricity market is
regulated, there is still a price effect on the investment decision. The investment needs created by
price sensitivity in this model is counterintuitive. Usually, an increase in price would be an
incentive to invest because there would be high returns, and a price reduction would prompt decline
in investment. However, because the electricity sector is state-owned and the price is below
indicated market price, an increase in price would result in a decrease in price.
Effect of price sensitivity on Electricity Required Investment
2500 Z 1: Current investment in cedis
2000 —— ——2: Current investment in cedis
_— 3: Current investment in cedis
1000
4: Current investment in cedis
500
04 - 5: Current investment in cedis
>
)
$9 =--6: Current investment in cedis
a
GHS (Millions)
B
a
ra)
S
Figure 9: Effect of price sensitivity on Electricity Investment with varying price sensitivities (1: 0;
2: 0.4; 3: 0.8; 4: 1.2; 5: 1.6; & 6: 2.0).
The price increment would prompt demand to fall leading to a lower demand supply gap and
indicating decline in the need for investment. This is depicted in figure 9 above. As price sensitivity
is set at zero (0), investment increases at its highest because consumers can use electricity without
paying. On the other hand, when price sensitivity increased to two (2), the need for investment is
low because very few people are attracted to use electricity.
Results
The results from the simulation of the model developed specifically to analyze the energy sector in
Ghana do not promise automatic solution. That is to say that, if this problem were not tackled now,
it would not in any way become better by itself. The Business As Usual (BAU) scenario indicates
that, the electricity problem in Ghana is only going to increase in the future. The results reveal five
major issues in the electricity sector: There is the problem of low investment in electricity
generation, significant amount of power lost, very low tariffs on electricity, underutilization of
capacity, and fossil intensive investment.
Base run
The base run presents the results from the simulation without any policy in place. It is supposed to
replicate the business as usual scenario. As seen in the results presented, it is clear that, the
simulated results do not exactly reproduce the reference mode. This is due to a number of
disparities between actual values and estimated values. Figure 18 below shows the electricity gap
between 2000 and 2013. It compares the demand and supply data to the simulated results and it is
evident that, with the current trend of events, the problem would persist.
rae Base run: Electricity Demand, Supply, and Gap
3000 —
Demand Data in MW
2 2000 ; ,
1000 Effective supply Data in MW
0 ———== Gap Data
PNYTOHRONtTOHASONHO DS =
-1000 Seo 9 3 3 3 oo 8 NA SA Total electricity demand
ANNANAANKNKNANNKNAAN
. Total capacity supplied MW
Time
Figure 10: Electricity Demand, Supply, and Gap (Data and Simulation results)
The demand is below the effective supply, which is the effective supply capacity, less transmission
and distribution losses. Around 2008 and 2009 when some capacity of the Bui hydro project was
commissioned, the electricity crises subverted momentarily.
Demand and Supply gap
The figure below shows the electricity gap since 2000. Demand exceeding supply presents power
deficit situation in Ghana. As demonstrated in figure 11 below, the demand of electricity is above
that of the effective supply creating a gap over time.
Demand and Supply gap
300 &S
——2: Base run (BAU)
100
=" Gap Data
MW
-100 _—_ _— _—
PpoocoOoOe ae aN NNN ANA
>sssosescsesooguggeeseegocsgasas
Nnannnnnnnnnnananan
-300 -
Time
Figure 11: Demand and Supply Gap
The addition of some megawatt units from some thermal plants reduced the crises in the middle of
the last decade. In 2007, the Akosombo hydroelectric power station suffered its first severe water
crises resulting in nationwide load shedding. This prompted the construction of more thermal
plants. Since the first crises, the dam has never been fully functioning as the thermal plants that
were supposed to share the burden are operating below capacity. In 2011, when a ship anchor
destroyed the gas pipeline that provide gas to the thermal plants all the way from Nigeria, the power
crises became intense. The simulation results in figure 19 depict the sudden rise of electricity gap in
2012. Intermediary solutions have merely scratched the surface of the problem.
Total Installed, Effective, and Supplied Capacities in MW
One of the main issues of concern that the model has revealed is the fact that, there is a significant
difference between the total capacity of power installed, what is effectively generated, and the
actual megawatts of power distributed to consumers (figure 20). Out of the total capacity installed,
only a fraction of this is effective. This is due to different reasons that affect the functionality of the
different power plants/sources. In the case of hydro, the total capacity is that is effective is only
90% of the capacity installed because the lack of rainfall during the dry season worsens the already
under-producing hydroelectric turbines. With thermal, the effective capacity is merely about 70% of
total installed capacity, on average. This disparity is mainly due to the unavailability of gas to
operate the plants. Relying on gas from Nigeria is a rather worrying dependency relationship. In the
case of solar, issues of underutilizations are often related to the darkness at night. The sum of these
effective production capacities is not fully accounted for at the downstream supply chain as
depicted in figure 12 below. The supplied capacity is less than the effective capacity.
Total Installed, Effective, and Supplied Capacities in MW
6000
5000
4000
z 3000 Total Instlled Capacity MW
2000 Total Effective capacity MW
1000
0 Total capacity supplied MW
ent vVaoNntTomnSONXOHS
SSCSSSHHSHSHAHANAAANST
sssseseooogoggcgegseogcs
NAAAAANAAAA AN AAA
Figure 12: Total Installed, Effective, and Supplied Capacities in MW
This is because, in Ghana, there is significant amount of power lost from generation to distribution.
About 5% on average, of the power generated is lost through transmission and about 20% of the
power is lost through distribution. This is a major issue that needs keen attention.
Electricity Pricing
Electricity pricing is another issue that is very important in the overall discussion of the energy
situation in Ghana. In Ghana, the prices of electricity are fixed over time and reviewed by a
commission that also regulates water tariffs. The PURC is responsible for reviewing electricity
prices in Ghana. As regards, the tariffs on electricity are very low compared to market prices.
End-User Electricity Tarrifs
0.1
2 0.08 S
& —
2 0.06
+ ————————
3 0.04 Reference price
& 0.02 Electricity price
0
ent evneo nt voansonrt COC wMs
Sesoeogcouyy HS HSHAANAAA
SsesecesSs se SESeESGgESESGCSCSES SS
ANNAN AAAMANATHANANRNAAA
Time
Figure 13: Annual Electricity tariffs
Electricity in Ghana is also heavily subsidized by the state. The end-user price is therefore very low,
an incentive for high electricity consumption. Figure 13 above shows the electricity price in Ghana.
It is apparent in the simulation, that electricity price increases slowly. This price is even lower than
many developed/industrialized nations that have a lower per unit cost of production.
The figure below shows how a variation in the price sensitivity could affect the demand and supply
gap. When price sensitivity is very low, the gap would increase because electricity is cheaper since
it does not respond to market mechanisms.
Effect of price sensitivity on Electricity Demand Supply Gap
1000 ZL ———1: Demand Supply gap
800 fo Hippel Samsip aap
= 600 ZZ 3: Demand Supply gap
400 4: Demand Supply gap
- —— ibid
0+ a a a |
6: Demand Supply gap
Figure 14: Effect of price sensitivity on demand supply gap with varying price sensitivities (1: 0; 2:
0.4; 3: 0.8; 4: 1.2; 5: 1.6; & 6: 2.0)
On the other hand, when the price is very sensitive, the demand for electricity declines leading to a
lower gap as depicted in figure 14 above. This presents another policy option for the government to
alleviate this growing concern.
Electricity Demand and Indicated Demand
Demand is derived from the indicated demand. Indicated demand is higher than demand throughout
the simulation period. Figure 17 represents the development of the demand and indicated demand
over time. The price effect is shown in the indicated demand. A lower demand supply ratio below 1
would suggest that, electricity price is lower than the reference price. In other words, when the
reference price of electricity is higher than the actual price, it would prompt the indicated demand to
fall below the actual demand, contrary to the simulation in figure 23 below. The need constant
rising trend of demand is indicative that, price is indeed very low hence, encourages more demand.
Electricity Demand and Indicated Demand
4000
3000
z 2000
Indicated Demand
1000
0 Total electricity demand
SNHTODCHOSOCNKTOCHMONYXYOAS
SssesesSaHseHHannanaa
sessseseocogcgegegeggsessa
ANNNAANNANAANANAANAAAA
Time
Figure 15: Electricity Demand and Indicated Demand
Scenario and policy analysis
Policy formulation is an essential aspect of decision-making. A system Dynamics model that
contains within it a policy structure also embeds in itself some implementation assumptions (Wheat,
2010). Identifying the issues causing undesirable dynamics in a system is only one part of problem
solving. After ‘troubleshooting’ the next step is to identify ways of overcoming rectifying the
issues. This study proceeds to evaluate the policy options considered in the model. There are five
main policy related issues covered in the model. They include the investment alternatives, the
transmission and distribution losses, the capacity issue, and the pricing system in the electricity
sector.
The Investment policy
Governments’ attempt to bridge the power deficit in Ghana has been in a goal-seeking pattern.
Power decisions are made with the demand supply gap as the target. Unfortunately, the gap is a
moving goal that changes from year to year. Whilst plans are advanced to bridging the gap observed
in the previous year, very little consideration is given to the annual demand increment. The growth
in industry, extension of power girds to rural communities, and potential seasonal and unforeseen
hindrances are not given enough thought. The results from the simulation indicate that, an
increased investment beyond the current gap is required to overcome this incessant problem.
The simulation below shows two investment pathways. The budgeted investment, which is the
investment government made based on demand and supply gap is insufficient to solve the crises.
This low investment is one of the reasons why the gap persists.
Annual Investment in Ghana cedis in the electricity sector
2
3,000
= 2,500
E 2,000 —
& a d Investment in
@ 2500 cedis
z 1,000
A 500 — Indicated investment in cedis
s 0
3
2 SOSSSSSSSSLSSINALE
>ssceoseseoogqgasceseeoagse
< ANNAANNANAAKRANANAAA
Time
Figure 16: Annual Investment in Ghana cedis in the electricity sector
If the government keep benchmarking investment with demand and supply gap would be as
indicated in the base run in figure 17 below. On the other hand, the indicated investment in figure
24 seem appropriate scenario to overcoming the electricity crises. This investment is derived from
the GDP trend. Indeed, as the Energy-led-Growth-led-Energy scholars (Masih & Masih, 1997; Fatai
et al, 2004; Ghali and El-Sakka, 2004; Akinlo, 2008) hypothesized, there seem to be causality
between economic growth and energy consumption. It is estimated that, electricity investment in
Ghana needs to be equivalent to about 1.5% of GDP. Investing at this rare will result in an
elimination of the gap as indicated in figure 25 below.
Investment policy on demand and supply gap
2: Base run (BAU)
4: Investment Policy
Time
Figure 17: Investment policy on demand and supply gap
The investment policy is therefore an effective policy for solving the electricity crises. This should
be a wakeup call for the government to collaborate with the private sector to boost investment. A
good incentive for private sector participants will be a free and fair market system where the price
of electricity is market regulated and not solely determined by the Public Utilities Regulatory
Commission.
The Capacity Policy
Hydro limit: The capacities policy does not appear to be viable unless it is paired with some other
policy options. The policy however is worth considering because; the hydro potential in Ghana is
highly limited. Although hydro constitutes a greater amount of power sources for electricity, there
are not that many viable hydro sites that could be developed in the future. This policy sets the
maximum hydro capacity at 2000 MW. This implies that, only about 400 MW of hydro can be
developed in the future. The policy sets a scenario where by all these potential sites are developed
by the end of the simulation period.
Solar/Thermal capacity proportions: This capacity policy evaluates the extent a deliberate focus
on varying the different power capacities over time with the available investment might resolve the
gap. The Solar/Thermal policy is a priority due to varying returns in the future. Solar is expected to
record increasing returns over time as the unit cost decline due to the learning effect. The Thermal
unit cost on the other hand is expected to increase as gas supply is exposed to unstable political
environment. Again, solar technologies are expected to advance in the future and reduce the unit
cost making competitively favorable. The policy on solar and Thermal is therefore activated with
the hydro capacity policy, which simulates the maximum potential hydro capacity over time.
As the learning curve effect reduces the cost per unit, solar becomes more favorable and attract
investment. Thermal on the other hand gradually loses market share and become less attractive. The
transition from solar to thermal follows the s-curve technology diffusion pattern (Christensen, 1997;
Wilson & Grubler, 2011). At the end of the simulation period however, thermal still has the highest
market share. Energy transition takes time. Technology adaptation that involves high setup capital
takes off slowly. In the long run however, solar will overtake thermal as a cheaper electricity
source.
Figure 17 below illustrates the gap if only the capacity policy is activated. This means that, hydro
investment reaches its maximum of 2000 MW in 2030. Additional capacity from now to the end of
the simulation period is 400 MW. A fraction of this is invested annually. Remaining funds after
hydro is allotted to solar and thermal based on the fraction of market share. At the end of the
simulation period, the policy outcome is equivalent to the base run but with a rather larger gap
between policy start to end time. The capacity policy though necessary needs to be combined with
other policy options to realized positive outcome and secure a sustainable energy future in Ghana.
Capacity policiy effect on denamd and supply gap
72: Base run (BAU)
3: Capacities policy
3
S
2000
2002
2004
2006
2008
2010
2012
2 2014 =
2018
2020
2022
2024
2026
2028
2030
Figure 18: Capacity policy effect on demand and supply gap
Transmission and Distribution losses
The transmission and distribution losses are a major contributor to the low supply of electricity in
Ghana. The model proposes a policy that reduces the transmission and distribution losses by a
fraction. This policy aspect uses Sabatier and Mazmanian’s, (1980) idea of parameter testing using
an estimate of the parameter value. The transmission policy assumes a reduction of the losses by
50% to see how that affects the demand and supply gap.
Figure 18 below shows how a 50% reduction in transmission and distribution losses could reduce
the demand supply gap significantly.
Transmission loss policy on electricity gap
400
300
200
z= 100 = 2: Base run (BAU)
0 > 5: Transmission policy
Figure 18: Transmission and Distribution losses
Government should therefore redirect some of the investment made in construction of new power
plants to improving the distribution system and reduce the power losses. Without proper attention to
the power, the need for construction of new power stations will continually increase. Proper
maintenance on existing grid and improvement in the metering system to eliminate power theft
would hugely improve the power supply in Ghana.
Combined policies effect
The results in figure 19 below show the gap development over time if all the policies are activated.
This among other reasons makes solar the competitively viable alternative for electricity needs. At
the end of the simulation period, it is estimated that the gap will begin to emerge again with all
policies activated. This is because; the weight of the ineffective policies dominates in the long-run.
The best policy option or combination of options should therefore be settled on. This is discussed
further under policy comparisons below.
Combined policies effect on demand/supply gap
we
— ——2: Base run (BAU)
=
= 100
SL 1} —1—_1—1— 6: Combined policies
>PNYTOHON oo S ow ae
-100 See Sf 3 5 aoa A Nig 9
NARRARARARA AA ARARAAR
-200
Time
Figure 19: Combined policies effect on demand/supply gap
Policy combinations and Comparisons
The policy combination presents the scenario where all policies in the model are activated. The
results are then juxtaposed with other policy options to determine whether a single policy option
presents better outcome.
Demand/Supply gap under policy scenarios
1: Equilibruim
——2: Base run (BAU)
——3: Capacities policy
——4: Investment Policy
5: Transmission policy
6: Combined policies
7: Investment & Capacity policies
8: Investment &Transmission policies
9: Capacity & Transmission policies
Time
Figure 20: Different policy combinations
The investment policy seems to be the best approach to the current electricity shortage in Ghana. As
depicted in figure 20, the investment policy is the one that has the highest impact on the demand
and supply gap. This policy is therefore recommended to policy-makers. On the other hand, the
capacities policy does not in itself solve the crises in the long-term. Policies that do not contain
investment are not effective in the log-run.
The capacities policy is ineffective because, the hydro capacity is limited and thermal and solar are
both still more expensive compared to hydro at the end of the simulation. All the policy options
that include investment present better results than all other policies. This reiterates call for more
investment to be made in the electricity sector. Whether this investment is in the form of
constructing new power stations or improving existing one, investment is certainly imperative.
Total installed capacities with all policies activated
The figure below shows the total capacities recorded at the end of the simulation period if all
policies contained in the model are activated. The high capacity reaches its peak of 2000 MW and
the effect of depreciation is setting in. The limited nature of hydro resources makes it least ideal a
major policy option notwithstanding the relative cost advantage of hydro to other energy sources.
Total installed capacities with all policies activated
= 4000 ™ Solar Capacity Installed
™ Thermal capacity installed
™Hydro capacity installed
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
2024
2026
2028
Figure 21: Total Hydro, Thermal, and Solar PV Capacities MW
The figure (21) presented above indicates a growing trend of the thermal capacity in the country.
This is partly due to the gradual liberalization in the electricity sector to encourage private sector
participation. Even though thermal seem affordable now, there are some pertinent challenges
associated with it. For instance, in 2012, a ship anchor cut the West African Gas Pipeline, which
transports gas from Nigeria to Ghana. This result in near zero production from the thermal plants
creating a severe electricity crises, the remnants of which are still experienced today though repair
work on the pipeline is almost complete.
Solar is increasing more rapidly because the unit cost MW of solar is declining. This is as a result of
the learning curve effect. Even though the cost of solar at the end of the simulation period is still
slightly higher than that of thermal, there is high potential for further decline as it gains more
market share. The simulation period is not long enough to accrue a quantitative net benefit.
Qualitatively, investment in solar is future-oriented, reliable, and sustainable compared to thermal.
Discussion
The government provides nearly all the electricity consumed in Ghana. In recent times, there have
been discussions of market liberalization to encourage private sector participation in the energy
sector. Active participation of the private sector will lessen the burden or responsibility on
government to independently provide all the energy needs in the country. Policies should be
instituted and strategies designed to relax the entry barriers that detract potential independent
investors. The regulations should be relaxed to accord private entities a convenient platform to
operate in the sector.
Simulation suggests that, even though there are some major issues with the power generation,
policy makers give little attention to the contributions of demand side to the overall problem. There
is little consideration on the magnitude of the effect of changes electricity demand poses to the
electricity problem in the country as postulated by Gyamfi (2007) and Adom et al. (2012). This
reemphasizes the findings of Ackah and Adu (2014) that determining the factors driving the
electricity demand can inform policy makers to institute the appropriate policies manage and bridge
the power deficit in Ghana.
Electricity supply should be a key priority to stakeholders. It is central to economic development
and individual well-being. Ghana’s electricity data indicate a rather narrow energy mix at the
national level. Little attention has been given to energy sources that are abundant and not heavily
constraint by seasonal/perennial factors (such as rainfall), which are major causes of the current
energy predicament. Emerging technologies far more efficient, effective, and environmentally
friendly are far less explored in Ghana. Renewable energy sources such as solar should be
considered. The current rampant plant shutdowns as a result of limited gas supply as made thermal
a less ideal choice for sustainable energy future in Ghana. The result from the simulation of the
solar and thermal market share in future based on current cost values indicates that, solar will
eventually emerge as a cheaper source of electricity to thermal.
The results from the investment analysis suggest that, the target annual investment should not be the
demand and supply gap as this will always result in the government responding late to investment
needs and also consistently recording a gap between demand and supply of electricity. The
indicated investment computed based on the GDP growth over the simulation period is the best
investment alternative for future electricity sufficiency.
It is estimated that, approximately 22% of the electricity generated in the country cannot be
accounted for. This is lost through transmission and distribution. There are numerous grid cables
that are absolute and sub-standard. As a result, they are inefficient in transmitting and distributing
the total generated power to final consumers. Illegal connections, billing and revenue collection,
and poor metering system account for commercial losses in downstream process. The results from
the study suggest that, a critical policy focus on reducing the transmission and distribution losses
could improve the electricity supply enormously.
25
The current regulatory mechanism in the electricity market is ineffective and partly responsible for
the high-energy consumption in Ghana. The electricity price is not determined by the demand and
supply pressures. Electricity in Ghana is also highly subsidized compared to many developed
economies. As regards, there is very little effect on price even when demand far exceeds supply.
Conclusions
The Public Utilities Regulatory Commission, which is responsible for setting electricity tariffs
adjusts price in significantly large time step (one year). This creates a slow reaction of price to
demand and supply. It also serves as a disincentive to private entities that have the desire to invest
in the energy sector. Preliminary results on pricing policy option suggest that, a market determined
price is much more beneficial than the ‘fixed’ pricing system currently in use. In this model, the
policy returns are a bit complicated to compute. This is especially the case since most of the policies
are qualitative in nature. For instance, improving management efficiency to reduce distribution loss
would require some of incentives or repression upon failure. These incentives or deterrents are not
easily measurable in monetary terms.
The electricity problem in Ghana seems to be emanating from low investment, especially in
renewable energy, excessive transmission and distribution losses, and underutilization of capacity
due to seasonality factors and gas supply inconsistencies. This places emphasis on the need to
investment in renewable energy. Indeed, the set-up capital is high, the cost ratio to fossil is currently
high, but the long-term benefit must not be overlooked for short-term political income that leaves
the problem far worse every year. It is high time the government liberalizes the electricity market,
encourage private sector participation, promote a free market system, and clampdown on illegal
grid connections that accounts for nearly a quarter of total electricity generated.
A market-based price reduces energy ‘wastage’, reduces supply gap, and also reduces the subsidy
required on electricity. Reduction in subsidy could lead to more funds for expansion capacity. The
government should therefore adopt a free/market-based mechanism where prices instantly respond
to demand and supply capacities. One advantage of this is that, there is an incentive to conserve
energy as it is rewarding to reduce consumption.
The regulatory agencies need to be more proactive in discharging their duties. The Energy
Commission needs to set out clear directives to Independent Power Producers who procure license
to produce electricity. Licenses should only be issued to reliable private sector entities capable of
producing the amount of electricity for which such license has been issued. A clear timeline should
also be set for such entities to attain certain production level. This reduces the repercussions
government endure in response to address perennial electricity supply gap.
In existing electricity grids, measures can be taken to reduce technical energy losses. The losses are
relatively higher when there is low voltage. Options include upgrading the voltage of a transmission
and distribution system, and replacing existing transformers with more efficient ones. Electricity
conservation and efficiency is very low in Ghana. Though this can largely be attributed to the nature
of household appliances (old and sub-standard) and “stolen power”, the high subsidy on electricity
resulting in prices far lower than some highly developed countries is a huge incentive for excessive
usage.
There is the need for further studies in the Ghanaian energy sector that takes into consideration
complete historical data of utilization over time and not average values as contained in this study, to
ensure a much concrete prediction of the electricity scenario in the future. Future research should
also find actual data on investment in the different power sectors. This could yield more resilient
results than the calculations in this study, which computed the annual investment by calculating
backwards with the overall capacity installed over the period and evenly spreading that over the
historical period. The actual cost of a MW unit of solar and thermal should be carefully computed to
ensure a better judgment of market share distribution/development over time.
References
Ackah, I., & Adu, F. (2014). Modelling Gasoline Demand in Ghana: A Structural Time Series
Approach. International Journal of Energy Economics and Policy, Vol. 4, No. 1, pp. 76-82.
Ackah, I., Adu, F., & Takyi, R. O. (2014). “On The Demand of Electricity in Ghana:
Do Exogenous Non-Economic Variables Count?” International Journal of Energy Economics and
Policy. Vol. 4, No. 2, pp. 149-153.
Adom P.K., Bekoe, W. Akoena, S.K.K. (2012) Modeling aggregate domestic electricity demand in Ghana:
An autoregressive distributed lag bounds cointegration approach, Energy Policy, 42, 530-537.
Akinlo, A.E., (2008), Energy consumption and economic growth: evidence from 11 African countries:
Energy Economics 30, 2391-2400.
Anand, S., Vrat, P., & Dahiya, R. P. (2005). “Application of System Dynamics Approach on assessment
and mitigation of CO2 emissions from the cement industry”. Journal of Environmental
Management. Vol. 79, pp. 383-398.
Apergis, N., & Payne, J. (2011). “A dynamic panel study of economic development and electricity
consumption-growth nexus”. Energy Economics. Vol. 33, pp. 770-781.
Arthur, W. B. (1988). “Competing Technologies, Increasing Returns, and Lock-in By Historical Events”
Economics Journal Vol. 99, pp. 116-131.
Barlas, Y. (1996). “Formal aspects of model validity and validation in system dynamics.” System
Dynamics Review Vol. 12, pp. 183-210.
Braun, W. (2002). The system archetypes. The Systems Modeling Workbook.
Ceylan, H., & Ozturk, H. K. (2004). “Estimating energy demand of Turkey based on economic indicators
using genetic algorithm approach”. Energy Conversion and Management Vol. 45, pp. 2525-2537.
Chanwoong, J. & Juneseuk, S. (2014). “Long-term renewable energy technology valuation using system
dynamics and Monte Carlo simulation: Photovoltaic technology case”. Energy Vol. 66, pp. 447-
457.
Chi, C. H., Nuttall, W. J., & Reiner, D. M. (2009). “Dynamics of UK natural gas industry: system
dynamics modelling and long-term policy analisys”. Technological Forecasting and Social Change.
Vol. 76, pp. 339-357.
Christensen, C. (1997). “The Innovator’s Dilema: When New Technologies Cause Great Firms to Fail.”
Harvard Business Press: Watertown, MA, USA.
Connolly, D., Lund, H., Mathiesen, B. V. & Leahy, M. (2010). “A review of computer tools for analyzing
the integration of renewable energy into various energy systems”. Application Energy. Vol. 87, No.
4, pp. 1059-1082.
Criqui, P., Klaasen, G., & Schrattenholzer, L. (2000). “The Efficiency of Energy R & D Expenditures.
Economic Modelling of Environmental Policy and Endogenous Technological Change Workshop”.
Institute for Environmental Studies, Amsterdam. November 16-17.
Engle, R.F. and Granger, C.W.J. (1987) “Co-integration and error correction: representation, estimation,
and testing”, Econometrica, Vol. 55, No. 2, pp.251—276.
Erdogdu E. (2007) “Electricity demand analysis using cointegration and ARIMA modelling: a case study
of Turkey”. Energy Policy, Vol.35, pp. 1129-46.
Energy Commission of Ghana (2013). Energy statistics. Energy Commission, Accra, Ghana (2013).
Energy Commission of Ghana, (2011), Energy (Supply and Demand) Outlook for Ghana, April 2011.
Fatai,-K, Oxley,-L and Scrimgeour,-F.G (2004), Modelling the causal relationship between energy
consumption and GDP in New Zealand, Australia, India, Indonesia, the Philippines and Thailand:
Mathematics and Computers in Simulation, 64, 431— 45.
Feng, Y. Y., Chen, S. Q. & Zhang, L. X. (2013). “System Dynamics modelling for urban energy
consumption and CO2 emissions: A case study of Beijing-China’”. Ecological Modelling. Vol. 252,
pp.44-52.
Ferguson, R., Wilkinson, W., Hill, R. (2000). “Electricity use and economic development”. Energy policy.
Vol. 28, pp. 923-934.
Forrester, J. W. & Senge, P. (1980). Test for building confidence in system dynamics model, TIMS Student
Management Science, Vol. 14, pp. 209-228.
Fourquet, R. (2010). “The slow search for solutions: Lessons from historical energy transitions by sector
and service”. Energy Policy 2010, 38, 6586-6596.
Ghali, K.H., El-Sakka, M.I.T., (2004), Energy use and output growth in Canada: a multivariate
cointegration analysis. Energy Economics 26, 225-238.
Grubler, A. (2012). “Energy transition Research: Insights and cautionary tales. Energy Policy 50, 8-16.
Gyamfi, S. (2007) “The Role of Demand Response in Managing Ghana's Electricity Supply”
http://www.ghanaweb.com/GhanaHomePage/features/artikel.php?ID=125746 Retrieved 18 May
2015.
Humphreys, J. (2013). Institutional Pathways to fossil-free investing”. Endowment Management in a
warming World. May 3013.
International Energy Agency (2013). “World Energy Outlook; (EA: Paris France 2013.
Masih, A.M.M., Masih, R., (1997), “On temporal causal relationship between energy consumption, real
income and prices; some new evidence from Asian energy dependent NICs based on a multivariate
cointegration/vector error correction approach”. Journal of Policy Modeling Vol. 19, No. 4, pp.
417-440.
Kaouvaritakis, N., Soria, A., & Isoard, S. (2000). “Modelling energy technology dynamics: methodology
for adaptive expectations models with learning by doing and learning by searching”. /nternational
Journal of Global Energy Issues. Vol. 14, No. 1-4, pp. 104-115.
Kobos, P. H., Erickson, J. E. & Drennen, T. E. (2006) “Technological learning and renewable Energy
Costs: Implications for the US Renewable Energy Policy”. Energy Policy. Vol. 34, No. 13, pp.
1645-1658.
McDonald, A., & Schrattenholzer, L. (2001). “Learning rates for energy technologies”. Energy Policy. Vol.
29, No. 4, PP. 255-261.
McGrath, M. (2012). “Ghana solar energy plant set to be Africa’s largest”. BBC News. Retrieved 18 May
Medlock, K. B. (2011), Empirical evidence of the Efficiency of National Oil Companies. Empirical
Economics (May 2011)
Miketa, A., & Schrattenholzer, L. (2004). “ Experiments with a methodology to model the role of R & D
expenditures in energy technology learning processes; first results”. Energy Policy. Vol. 32, No. 15,
pp. 1679-1692.
Ministry of Energy, (2011), Ghana Country Report, June 2011,
http://nrec.mn/data/uploads/Nom%20setguul%20xicheel/Water/badrakh%20china/Ghana.pdf
Ministry of Energy, (2010), Energy Sector Strategy and Development Plan, February 2010.
Moxnes, E. (1992). “Positive feedback economics and the competition between “hard” and “soft” energy
supplies”, Journal of Scientific and Industrial Research. Vol. 51, pp. 257-265.
Naill, R. F. (1977) “Managing the energy transition: a system dynamics search for alternatives to oil and
gas”. Cambridge, MA: Ballinger Publishing Company.
Randers, J. (2012). “2052: A Global Forecast for the Next Forty Yea’r. White River Junction,
Vermont.Chelsea Green Publishing.
Sabatier P, Mazmanian D. (1980). The implementation of public policy: a framework for analysis. Policy
Studies Journal Vol.8, No. 4, pp. 538-560.
Sgouridis, S., & Csala, D. (2014). “A framework for Sustainable Energy Transitions: Principles, Dynamics,
and Implications.” Sustainability 6, 2601-2622.
Shin, J., Shin, W. S., & Lee, C. (2013). “An energy security management modl using quality function
deployment and system dynamics”. Energy Policy. Vol. 54, pp. 72-86.
Sterman, JD. (2000). Business Dynamics: Systems Thinking and Modeling for a Complex World.
Irwin/McGraw-Hill: Boston, MA.
United Nation (1998). Kyoto Protocol on Climate Change. Kyoto, Japan
Vaughan, A. (2012). “Africa largest solar power plant to be built in Ghana”.7he Guardian. Retrieved 18
May 2015.
Wheat, D. & Saldarriaga, M. (2011). Using the Epidemic Game Model to Illustrate Guidelines for a
Modelling Report.
Wheat, D. (2010). “What Can System Dynamics Learn From the Public Policy Implementation
Literature?” Systems Research and Behavioral Science Systems Research Vol. 27, 425-442
Wilson, C., & Grubler, A. (2011). “Lessons from the history of technological change for clean energy
scenarios and policies.” Natural Resources Forum; Wiley: New York,, NY, USA. 35, 165-184.
Winkler, H., Simeos, A. F., La Rovere, E. L., Alam, M., & Rahman, A. (2011). “Affordability of electricity
in Developing Countries”. World Development, Vol. 39, No. 6, pp. 1037-1050.
Wu, J. H., Huang, Y. L., & Liu, C. C (2011). “Effect of floating pricing policy: an application of system
dynamics on oil market after liberalization”. Energy Policy. Vol. 39, pp. 4235-4252.
Appendix: Policy structure
ao eT ry
