Table of Contents
SYSTEM DYNAMIC MODEL TO ANALIZE INVESTMETS IN POWER
GENERATION IN COLOMBIA
Santiago Arango*, Ricardo Smith**, Isaac Dyner**, and Silvia Osorio**
*University of Bergen, Norway
**National University of Colombia, Medellin
ABSTRACT
The Colombian electricity sector was restructured in 1994. Under the new legal framework,
investment decisions for building new capacity need to incorporate elements of risk and uncertainty.
In these circumstances it results advantageous for agents to leam about market risks for assessing
the implications of their decisions. Microworlds or simulators, as the one presented in this paper,
intend to help investors for better understanding uncertainty and its implications over their decision-
making processes.
The authors developed a micro world, supported on a System Dynamics model, in which it is
possible for the decision maker to assess his/her investments in capacity under simulated conditions.
The developed microworld for the Colombian electricity market is described in some detail and
applications are presented.
Key words: Investment, Electricity, Simulation, System Dynamics, Microworlds
1 INTRODUCTION
The electricity sector in Colombia was restructured in 1994, where new markets elements were
incorporated in the system. During the last decade, the scheme has changed from a central planning
structure towards free markets, seeking efficiency and better use of resources.
In this context, the criteria for investment decisions of new power plant need to incorporate
risk and uncertainty analysis. In these circumstances, it results advantageous for agents to
leam about market risks in order to assess the implications of their decisions as, a
consequence of competition, we have observed trends towards improvements in efficiency
and reductions in electricity prices.
It is important to note that these newly engineered markets have exacerbated different
sources of uncertainty in variables such as: electricity price, regulation, demand growth,
and technology development, among others.
However, it is difficult to assess the evolution of such variables, which makes even more
complicated to evaluate investment decision on power plant. This is why, the authors
developed a tool, supported on a System Dynamics model, in which it is possible for the
decision maker to assess his/her investment in capacity under simulated conditions. The
methodology and the developed tool are described next.
2 METODOLOGY APPROACH
There is a tradition in the electricity sector to support investment decisions on simulation
models. It seems even more reasonable today to continue this trend of though because of
the ever increasing interactions among the multiple factors that are involved in the created
markets, which include the economy, energy technology, normative policies, conservation
and environmental legislation, company strategy and privatisation policies (Bunn y Larsen,
1997).
Econometric simulation and optimisation have been the most common tools under central
planning. The main criticism regarding the use of these tools, under market conditions,
have pointed out issues related to the modelling process, the way that uncertainty is
incorporated and model credibility, among others (Lee et al., 1990; Dyner and Larsen,
2001).
Complexity and market forces generate new methodological requirements. Such
requirements represent a challenge for the modelling approach with respect to the
incorporation of feedback thinking and uncertainties, among others (Dyner, 2000) - System
Dynamics seems to be appropriate for these purposes (Nail, 1992; Bunn and Larsen, 1992;
Ford, 1997; Dyner, 1995; Montoya, 1997; INTEGRAL - UN - COLCIENCIAS, 1999).
Under the current market conditions, the chosen tool to support decision-making processes
should take into account at least some of the uncertainties that characterize the sector.
Hydrology, demand, fuel supply, prices and bids incorporate significant uncertainties. This
market characteristic creates the need for focusing on the behaviour analysis of the system
rather than on the search for an “optimal solution” that does not exist.
In order to address this issue, a system dynamics microworld was developed. The purpose
is to create an environment that facilitates leaming about the problem of investment in new
generation capacity. In this paper, we exhibit a model that has been built with the purpose
of assessing the economic impact of the electricity market over investments in new power
plant. In the following sections we present the model that has been built for this purpose.
3 MODEL DESCRIPTIONS
The developed tool is supported by a system dynamic model, which allows analysing the
system evolution under any scenario created by the potential investor. The main purpose of
the model is the estimation of cash flows and other financial indicators.
The model was built using components of related models that have been developed under
the framework of the Energy Institute at the National University of Colombia (UN-
COLCIENCIAS-INTEGRAL, 2000; UN-COLCIENCIAS-ISA, 2000; Montoya, 1996).
We now tum to present the general model structure and its corresponding modules.
The developed model has been based on a general causal loop that explains the dynamic of
electricity markets (Bunn and Larsen, 1997), which is shown in Figure 1. In this figure, on
the one hand, one can appreciate that a high margin (difference between Capacity and
Demand) tends to reduce low electricity prices. Also, an increase in the electricity price
implies incentives to investment, as investors should obtain larger revenues. When there
are more investments, the capacity of the systems increases and hence the margin increases,
closing a balanced loop. On the other hand, a high price tends to reduce electricity demand,
because of price elasticity. The demand is affected also by extemal variables such as the
population and the gross domestic product. And finally, to close the other cycle, the higher
demand the lower price, due to the margin’s definition.
a
ae Incentives to
our tice Invest
Population t
Ns, Margin
ty. : +
+ Demand a a ‘
Figure 1. General scheme of the dynamics of the Colombian Electricity Market
There are other relevant elements, which are not accounted in the diagram, such as: the
availability of the power plant to generate electricity, which is influenced by the hydrology,
fuels availability, and installed capacity; and investments incentives, which depend on
financial indicators and regulatory incentives are important.
The model was created in a modular architecture as can be appreciated in Figure 2. The
main modules in the model are: market, expansion, demand, hydrology and finances.
Exogeneous Finances
Variables
Prices
Market
Expansion
rao Rea?
Opens
meses
Hydrology
Figure 2. Modular structure of the SD model
Demand
Following we present a brief description of each component of the model, and some
elements related to the validation of the model.
Market
The main purpose of this module is to establish the price setting mechanism, which
basically depends upon supply and demand. The market module incorporates some sort of
economic equilibrium criteria to adjust the supply and demand curves. In this way,
electricity prices and plant dispatch are determined taking into account hydrology
conditions and technology composition.
According to the electricity market in Colombia, price formation is represented as shown in
Figure 3. The supply curve is obtained by adding the supply curves of all generation
technologies involved. The pool price (PB) is found at the intersection of the supply and
demand curves.
Offer Bid
Price Short run
US$MWh A
Whether
PP
—_—_—> New projects
Demand
= >
Generation Capacity Quantity mw
Figure 3. Price process formation
Supply functions have been estimated by technology - hydroelectric with reservoir, river plant, gas
plant and coal technologies (UN-COLCIENCIAS-ISA, 2000). Different sets of supply functions
have been estimated according to the season (whether it is a raining season or not), and also
depending on the macro climatic condition (whether a Nifo occurs or not). A typical supply
function is shown in Figure 4.
Verano-Normal
300.00
250.00
200.00
150.00
Precio (USD/MWh)
100.00
50.00
0.00
0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00 7000.00 8000.00
Disponibilidad Comercial (MW)
Figure 4. A typical Supply functions per technology.
In addition, the supply functions are modified depending on the entrance of a new project. Those
projects extend the supply functions to the right, which assumes that the new projects are more
efficient than the old ones. The availability of the hydro plants is estimated according to water
inflows into the reservoir and the reservoir level. Price volatility is estimated as an indicator of the
price variability.
Expansion
The expansion module is an abstraction of what actually happens in reality, with the
purpose of providing insights into the capacity evolution of the system. The transitions
from highly regulated systems to competitive structures may create problems in terms of
the required capacity to attend demand (Hirst y Hanley, 1999) because of the inappropriate
financial instruments used by investors. The model uses two altemative mechanisms describe
below: an adapted “real options” approach and minimum average cost.
For the real options approach, the fundamental idea is to compare the “critical price” (estimated
theoretically), P*, for each project with the expected electricity price, Pe. A new project will enter
to the system if it satisfies:
+ Pe =P,*, i represents the projects
* t 2tn (to: minimal time to entry)
P* is estimated according to the optimal nile to entrance in a real option model (Dixit and Pindyck,
1994; Osorio, 2000). The estimated values for the model are taken from INTEGRAL-
COLCIENCIAS-UN (2000). Pe is calculated as follows:
pk =pM (tk )* 7] Fi
i
Where the P," is the expected price for the technology k, PM (t*) is the move average of the pool
price and F;; is a factor which involves others aspects such as technology, incentives, etc,
represented by i. For details of the model, see UN-COLCIENCIAS-ISA (2000).
Hydrology
The hydrology module contains two components: one is the occurrence of the ENSO (El
Nino South Oscillation) and the other is a stochastic model, RAR(1), for the representation
of inflows (Salazar, 1994). Due to the fact that there are two kinds of hydro plants, it is
necessary to calculate water inflows. For this target we selected a RAR (1) model (more
details see 1994), which include the influence of the ENSO phenomenon and the monthly
dependence of the inflows. To see more details in the parameters estimations and
validation of this model see Arango (2000).
Finally, the evolution and occurrence of the ENSO phenomenon, we used a model based on re-
sample techniques available in the project INTEGRAL-COLCIENCIAS-UN, 2000.). With this
model, we picked up 5 hydrologic scenarios of the ENSO phenomenon, in order to have a wide
range of possibilities, in addition to a random one.
Demand
Demand is modelled according to the forecasts of the Unit of Energy Planning in Colombia
(UPME, 1999), which includes variables such as population, and GDP, among others.
Those are time series, which can be modified if the user wants to have a personalized
demand scenario.
Finance
Finally, a finance module takes into account the new plant being evaluated according to
indicators such as net present value, profits, cash flow, etc. This module takes into account
the investment decisions and evaluates its implications on the system.
The investment is made in certain period of time. It is done in terms of capacity of the
project, technology, availability conditions (hydrology or gas availability), costs and debt
capacity, among others.
According to the investment decision, the microworld estimates cash flows. With these
cash flows, the model estimates some finance indicators such as Net Present Value, Internal
Retum Rate and Recuperation Period of Capital.
4 VALIDATION
The model validation was undertaken using the data available for the period January 1996
to December 1999. Using this data, we observe how in general, the model represents
system behaviour, despite not only the small amount of data available but also because of
the quality of the data. The data does not have significant measure error; the problem is
that this is a market in an infantile stage, where elements such as the leaming process
within the market, transitional problems and the high risk aversion that is taking place.
The behaviour of the systems, in terms of the gas and hydro generation, is well represented as
shown in Figure 5. In addition, in terms of price behaviour, the model closely follows historical
data as shown in Figure 6 (especially during the occurrence of the ENSO phenomena during 1997 -
1998). To see more details of the validation process, see Arango (2000).
Generacién Midrulia: Modelo Vs Real [GWh Generacién Térmica: Modelo Vs Real [GWh]
CROCE, PONS
56 wf nttiisig ai
7 ae ee
al soo} 2 a
w
Figure 5. Simulated and Real generation. Hydro to the rigth, and thermo to the left.
Precio de Bolsa: Modelo Vs Real
120. |
—- PB_mod
Brea
a | Pe
Precio [US$/MW1
20.
1.996 1.997 1.998 1.999 2.000
Tiempo
Figure 6. Simulated and Real pool price
Reservas Hidricas: Modelo Vs Real [GWh]
12.500.
10.000-
7.500. Reservas_mod
Ri 7
eservas_real
5,000-
2.500-
1.996 1997 1.598 1999 2.000
Tiempo
Figure 7. Water in reservoirs: modelled and real
5 DESCRIPTION OF THE TOOL
The microworld was developed as leaming and analysis tool applied to the Colombia
Electricity Sector. It allows the user to invest in a simulated project, under a risky and
uncertain market, where the user can assess the performance of his/her investment. The
tool allows the analysis of several expansion criteria, understanding the investment context,
valuing the consequences of the decisions, and defining investment strategies in power
generation in Colombia. Moreover, the users can also improve skills such as scenarios
analysis, work group, and mental model revaluation about the Colombian electricity
market, among others.
The tool has two general functions: as an analysis platform, where the user decides its
investment at the beginning of the simulation period; and as an analysis tool of the
evolution of the system and investment performance, where the user can make decisions
about electricity trading.
5.1 The tool as a Platform for Investment analysis
With the platform for Investment analysis, the user provides some initial condition of the system,
defines some features of the scenario and decides about his/her investment. The decisions are made
by pressing buttons of the main window and, at the end of simulation, it is possible to appreciate the
system evolution and the performance of investments. The main window is shown in Figure 7.
simulador, en donde
condiciones del sistema
toman dichas decisiones.
el tiempo se podran obser
econémicos del proyecto y el
sistema.
INICIAR CON "CONDICIONES DE MODELAMIENTO
Figure 7. Main window of the tool
The main components are the following:
* Condiciones de modelamiento (model setup): As is shown in Figure 8, this windows allows
setting-up some of the main variables of the system, such as hydrology conditions, demand,
discount rate, and expansion criterion, etc.
* Descripcion del modelo (model description): display a general view of the system dynamic
model.
* Decisiones de inversion (Investment decision): it is the input of all data needs for the investment,
such as investment and operation costs, capacity, location, entrance date, etc.
* Indicadores del proyecto (project’s indicators): this button show the financial indicator not only
of the project but also of the investor. See Figure 9.
+ Estado del sistema (State of the system): This button allows the view of the system. Variables
such as pool price, demand, margin, expansion, and hydrology are displayed.
Figure 8.
TASA DE DESCUENTO
Din tsa de desu
on que desea Vabjar %
CRITERIO DE EXPANSION
Defer de
een ie
© Precio Critico y Precio Esperado
© Minimo Costo Energia Media
‘cuadro respectivo
© Escenario_Alto
© Escenario_medio
© Escenario_bajo
© Personalizado
DEMANDA
Seleccione el escenario de demanda deseado. El
‘escenario personalizado, permite modificar las tasas
de crecimiento de la demanda haciedo click en el
Periodo
2000 - 2002
2003 - 2005
2006 - 2008
Tasa
200%
200%
200%
2009-2011 [200%
2012-2014
Cambiar datos “banco de proyectos" :
MARGEN MAXIMO PERMISIBLE
Margen en Potencia por a
duracién
‘encima del cual no
centraria ningun proyecto,
MARGEN DESEADO
Siselecciona el crterio de
HIDROLOGIA
Seleccione el escenario
hidrolégico de eventos ENSO.
Sisselecciona el personalizado,
puede modificarel inicio de
‘cada evento del ENSO y su
"Margen en Capacidad”
para expansién, debe
digitar el margen deseado
30 | %
PRECIO PROMEDIO DE CONTRATOS
[s/kivh constantes de enero de 2000)
Model conditions in the investment platform
INDICADORES FINANCIEROS
{ersten orenacon oe rnovecro ‘allo: 2009 MES! 1. )
VPN TR PRO FF
‘Del proyecto antes de impuestos (PAN) ~ao 650 nan [ru]
‘Del proyecto después de impuestos (PDI) 657 400 nan [ros]
De inversionista despues de impuestos (11) 3845.00 NAN
muss) tanec
Figure 9. Financial indicator of the investment’ s performance
In addition, the platform allows appreciating the evolution of the system and the cash flow of the
project continuously.
5.2 Micro world to invest in the Colombia electricity sector
The micro world is an interactive game, where the potential investor makes periodic decisions
under a defined scenario. From time to time, the user can observe the evolution of the system and
decide whether to carry out or defer the investment in power generation. Some of the main features
are:
* The occurrence of an ENSO event is a random variable. It is chosen from a database of events
ENSO (there is not possibility to define an hydrologic scenario).
* The step-time is a semester.
* Each semester the user decides about his/her investment, the selling profile (contracts and pool),
demand growth, availability of the project, among others.
The decision making windows is shown in Figure 10. This illustration presents the buttons
“Indicadores del proyecto” and “estado del sistema”, which have similar functionalities as in the
platform.
DECISIONES SEMESTRALES
eManon Co]
PORCENTA) EEN BOLSA | *
precioveconraaros [30] SW
isPoNeiLioAD ww
Penlovo ve constauccion [2] alos
Marques ca cuando
desee hacer su inversion
Figure 10. Decision making window of the microworld.
The user can observe the system evolution and decide whether to invest in the project at any time.
Once the decision to invest is made, the user can define his/her strategy to improve the performance
of the investment. Decisions about selling profile and contracting prices should be made after
investment.
The microworld shows the simulation system evolution. As an example, the following section
shows the behaviour of the system under de baseline scenario’, and the evaluation of a standard 150
MW CCGT (Combined Cycle Gas Turbine) under this scenario.
6 APLICATION
The tool has not been tested with real investors; nevertheless some experiments have been initially
conducted as indicated in this section. Next, the baseline scenario is presented and the performance
of a hypothetical investment in a small CCGT of 150 MW CCGT. A complete report of these cases
and additional cases are presented in Arango (2000).
6.1 Baseline Scenario
It could be considered as an intermediate scenario, called the baseline scenariol. This case has the
hydrologic conditions shown in Figure 11. In this figure, there are 4 events ENSO, with different
periods and durations, each one has associated a reduction in the aggregated inflows to the system.
This inflow allows calculating the availability of hydroelectricity, and the occurrence of the ENSO
determines mainly the bid curves. Due to the fact of the topology of the Colombian System (70%
Hydro and 30% Thermo), these variables give some of the main features of the scenario.
' Baseline scenario: low demand growth (UPME, 1999), expansion made using the criterion of critic and
expected price, defined hydrologic scenario, annual discount rate of 11%, and selling profile of 30% in pool
and 70% in contracts, among others.
Indicado [0 nifio; 1 no nifio] (Arriba) APORTES [GWh-mes] (abajo)
g
ZB 10
a
a | UU Lf
3
& 00
3
=o
2,000 2,002 »=-2,004 2,006 += 2,008.» 2,010» 2,012,014
Tiempo
6,00
5,00
8 4,00
B 3.00
2,00
1,00
2,000 2,002 2,008 2,006 2,008 2,010 2,012 2,014
Tiempo
Figure 11. Occurrences of the ENSO phenomenon (top) and aggregated inflows of the system
(bottom). Baseline scenario.
Figure 12 shows the pool price, not only monthly but also the average (the historical prices are
included to calculate the average). It is show the influence of the phenomenon ENSO over the
prices, where the occurrence of the ENSO means a considerable increases in the prices. However,
the increase in the prices is higher or lower according to the difference between the offer and the
demand.
PRECIO DE BOLSA [US/MW]
= Precio_Bolsa
404
—z-Promedio PB
2,000 2,002 2,004 2,006 2,008 2,010 2,012 2,014
Tiempo
L
Figure 12. Monthly and average pool price,. Baseline scenario.
The feedback in the module of expansion establishes that the higher the price implies higher
incentives to invest. Many investors can observe the signal of high prices and decide to invest.
These investments increase the installed capacity and a reduction in prices is perceived with a delay.
This dynamic is show in Figure 13, where the installed capacity, the demand and the margin (ration
between the difference of capacity minus demand over demand) are presented. Here, there are
some cycles of sub and over installation, which agrees with the economy theory. First, from 2000
to 2003, some projects under construction are finished, hence are part of the installed capacity.
However, the demand growth is not as large as was expected, which means over capacity in the
system and low prices, inclusive during the occurrence of the ENSO. It is not attractive to investors
and the entrance of new projects is stopped.
During the ENSO between 2006 and 2008 the price increases considerable, which is observed by
the investors as a market signal. It means that there are new power plants in the system, but the
investment makes again a reduction in the prices and the cycles are continued during the rest of the
simulation period.
DEMANDA Vs CAPACIDAD (arriba); MARGEN DE POTENCIA (abajo)
15,000} = se
————<—<
ae —-Demanda_maxima
Capacidad
5,000} =
2,000 2,002 2,004 2,006 2,008 «2,010 2,012 2,014
Tiempo
Eo.
gos
04
& 03
9 0.2
gov
~ 00
2,000 2,002 2,004 2,006 2,008 2,010 2,012 2,014
Tiempo
Figure 13. Demand and installed capacity (top), margin (bottom). Baseline scenario.
Figure 14 shows the technology composition of the system, no only the capacity but also the
percentage. In general, the initial composition is conserved (aprox. 70% hydro and 30% thermo),
despite of some differences in a few periods. In addition, the model shows the new projects in the
system and the entrance date (see Figure 15).
CAPACIDAD DEL SISTEMA: EN CANTIDAD (arriba); COMPOSICION (abajo)
=z Capacidad_Hidro
5,004 — —2--Capacidad_Térmica
2500}
o
2,000 2,002 2,004 2,006 2,008 2,010 2,012 2,014
Tiempo
10
08
8 —-Porcentaje_Hidro
04 —2- Porcentaje_Térmic
2,000 2,002 2,004 2,006 2,008 2,010 2,012 2,014
Tiempo
Figure 14. Technology composition of the system, capacity (top) and percentage (bottom).
Baseline scenario.
‘Aho En Proceso
2000 Fy unRAs
2001. Ponce
2002 Fy MeL!
2000 1 PieDaas
1999 Fy PayaniTo
1999 Fy poLones
2001 F9 sousow
2003 Fy ewcnus
2001 F9 TennosienRA
opopaizm
F snrana
T saujorce
Tmsienon
TF Ternocesan
°
°
°
°
0 Prema
0 Paraw
© TTerMocaranaar
oP sans
© FTeRMocusjina
© PF ceneacauce
‘Bho
2008
2014
2013
o
2008
2014
2013
o
2008
2014
2013
o
2008
2014
2013
0
0
2008
2013
2014
2014
2013
‘Gas Ciclo Combinado
I ataree-cc-100 aw
FF nicenewcc-100 mv
FF Mags e.cc-200
FF catce-o0 mw
FF Atarae-cc-150 aw
1 Teawoaaus
I Tenworsones nv
> Tenwowmar
Fe Tenwovanicues
We Terwonewa
tenwonio
‘Gas Ciclo Abierto
Fe Aasca choo
F viveencoc-200 ay
I age Meaicaaoo me
F catca-soo mw
PF anaca-cha50
ve
Fo Mags Meacaaso my
> age Hea.ca200 me
TF catca.zo0 mw
> asca-ck300
TF veeencoca-300 ae
FF Mage Hea.caa00 me
TF catca-soo mw
WF rernouran
Fr Tersiosanranoen
Fe Puento Benno
I TenwovoRs04
TF Temmica oe. cxré
‘Aho Hidréulico
OF meLism vesy,
0 caumaw
© TF socamoso 1
0 Pata
0 macHow
0 T Fonce
OT Guavaserat
oT casnera
0 ames
OF qverane
0 Force m
0 PF utcasnene
OF atnene
0 camera
0 nece x
0 samana meoio
Oo Pramas
2005 fy muanco
0 Fetcuaco
2005 IF saw Francisco
200€ fe MONTARTAS
2005 9 encimaoas
2007 fe cAmAVERAL
Figure 15. New power plants in the system, name and entrance year. Baseline case.
An investor in the Colombian electricity market is exposed to different risks; one of them is the
price volatility. The model calculates the risk associated to the price using volatility. The volatility
is an indicator of the price’s variability. It is presented in Figure 16, where the total accumulated
monthly volatility is 113%.
.
Volatilidad Mensual Anualizada
Lo.
08
06
8
a
~ 04
SS
0.2.
0.0.
0.2
2,000 2,002» 2,004 = 2,006 = 2,008» 2,010 2,012 2,014
Tiempo
Volatilidad mensual total acumulada y anualizada [%]
< J
Figure 16. Monthly volatility, annualized and total. Baseline scenario.
6.2 Case: an standard 150 MW CCGT under the baseline scenario
The analysis of a standard 150 MW CCGT is an example of the use of the investment analysis
platform. The project is located in the region of Magdalena Medio in Colombia. The main features
of the project are summarized in the Table 1, and some details are in UPME (1999).
Table 1. Main information about the project: 150 MW CCGT.
Project
CCGT
Location Magdalena Medio
Capacity 150 [MW]
Investment (without taxes)* 121.82 *10° [US$]
Investment (with taxes)* 126.74 *10° [US$]
Average Energy Cost (with taxes)* 37.53 [US$/MWh]
*Constant US Dollar of December (1997), own calculus using 2% of deflation rate (ANIF, 2000)
Source: UPME (1999)
Assuming that the period of construction of the project is 3 years, the investor decides that the
project will operate in July, 2007 (it implies that he decides to invest in July, 2004). Some other
aspects about the investment are in Arango (2000). Using the platform we get the indicators of the
project, which are shown in Figure 17. This figure presents the net present value, the retum
interest rate, the capital recuperation period and the cash flow, according either the project
with or without taxes or the investor (after taxes). The project without taxes has a net present value
of MUSD 30 and the retum interest rate of 14%; but, after taxes, it is reduce to 10% of return
interest rate and MUSD 1 of net present value. Finally, the investor has a better net present
value after taxes than the project by itself, it is because of the debt that the investor can
have.
INDICADORES FINANCIEROS
ENTRADA EN OPERACION DEL PROYECTO ANO: 2007 MES: 7
VPN TIR PRC FF
Del proyecto antes de impuestos (PAI) 30 13.50 23
Del proyecto después de impuestos (PDI) 1 10.00 34
Del inversionista después de Impuestos (IDI) 2 10.00 33
mill uss} 0) [aitos)
(se Worresnenee TH Tose Rene PRC) PeiodeRecencinae nal) [aves]
Figure 17. Financial indicators of a CCGT of 150 MW in the Magdalena Medio, Colombia.
Baseline scenario.
Each indicator is estimated using its corresponding cash flow, according either is of the project with
or without taxes or the investor. The buttons in the right hand allows the view of the annual cash
flow. For example, the Figure 18 shows the cash flow of the project after taxes, during the
commercial life of the project.
FLUJ 0 DE FONDOS DEL INVERSIONISTA, DESPUES DE IMPUESTOS
[mill US$]
oq
rt
=) 1
}
12.345 6 7 @ 9 1011 1213 14 25 1617 18 19 20 21 22 23 24 25 26 27 28 29 30 3] 32 33 34.35 36 37 3839 AD
NOTAS:
El proyecto comienza su operacién comercial a partir del afio 11
El iujo de fondos a pari del periodo 40 permanece constante hasta el aio 60 |
)
Figure 18. Cash flow of the investor after taxes of a 150 MW CCGT. Baseline scenario.
The user has to be careful about results because of the context. In this example, the risk associated
to the price is the total accumulated monthly volatility (113%). He/she should have seen that the
project started to operate while the system has high prices, then the user should also ask what has
happed if the project starts operation one year later (before). Indeed, he/she has to know the risk
and the uncertainty associated to the scenario. He/she has to check the robustness of the investment
under different scenarios.
7 FINAL COMMENTS
The outcome of the project is a microworld that focuses on investments in electricity
generation in Colombia. Its basic function is to estimate the project cash flow. The analysis
is carried out through a System Dynamics model that simulates the evolution of the most
representative system variables and their relationships. The model, properly validated, is
supporting the microworld in which the user has the possibility to “play” and be “trained”
for better understanding risk and power investment.
This tool is totally new in the Colombian electricity market, and it provides the bases for
the coming project to extend this modelling approach to some Latin-Americans electricity
markets. The next steps are the use of it with real subjects and evaluate the useful of the
microworld.
The model properly represents cycles of under and over capacity. They depend on variables
such as demand, hydrologic expectations and investment incentives, among others. The
model confirms, once again, the importance of hydrology issues in the Colombian
Electricity Sector. This fact is reflected in the consequences of water inflows over
investment decisions.
The basic problem behind the developed tool is investment in new capacity. This problem
has more components involved that the ones being included in the model, such as:
— Restrictions of the transmission network
— The fuels market
— Impacts of possible new regulation
— Influence of the load curve over dispatch
This is an undergoing project which is now focusing on further validation and on
developing more user-friendly interfaces. Research is also focusing on the generalization
of the tool to other Latin-American countries.
8 REFERENCES
ARACIL, J., GORDILLO, F. 1997. Dinamica de Sistemas. Ed. Alianza, Madrid, Espaiia, 198 p.
ARANGO, S, 2000. Micromundo para inversion en generacién en el Mercado Eléctrico
Colombiano. Posgrado en Aprov. de Rec. Hidr. Universidad Nacional de Colombia,
Medellin, Colombia.
BLANK, |. T. Y TARQUIN, A. J., 1996. “Ingenieria Economica”. Bogota: ed. Mc Graw Hill. 3ra
edicion. Colombia.
BUNN, D. W., DYNER, I., LARSEN, E., 1997. “Modelling Latent Market Power Across Gas and
Electricity Market. System Dynamics Review, Winter,Volume 13, Number 4. Printed in
Great Britain by Unwin Brother. Ltd.
BUNN, D., W. & LARSEN, E., 1997. Systems Modelling for Energy Policy. Ed. Wiley, England,
331 p.
CONGRESO DE COLOMBIA, 1994a. Ley de Servicios Publicos: Ley 142. Julio 11, Bogota,
Colombia.
CONGRESO DE COLOMBIA, 1994b. Ley de Servicios Publicos: Ley 143. Julio 11, Bogota,
Colombia.
DIXIT, A. K. y PINDYCK, R. S., 1994. “Investment Under Uncertainty”. Princenton: Princenton
University Press. New Jersey, United Estates of America. p. 468.
DYNER, I., 1993. Dinamica de Sistemas y Simulacidn Continua en el Proceso de Planificacion.
Universidad Nacional de Colombia, Medellin. [s.n.], 160 p.
DYNER, I., 1995. System Dynamics Platforms for Integrated Energy Analisys. PhD Thesis.
London Business School , University of London, UK. 339 p.
DYNER, I., 2000. Energy Modelling Platforms for Policy and Strategy Support. Joumal of the
Operational Research Society, Vol. 51, N°2, Febrero, 2000, p. 136-144.
DYNER, I., SMITH, R., FRANCO, C. J., ARANGO, S., BEDOYA, P. y OCHOA, P., 2000.
Microworld For Training Traders In The Colombian Electricity Market. Systemn
Dynamics Conference, Bergen, Norway.
DYNER, R. SMITH, C. J. FRANCO, S. ARANGO, L. BEDOYA, P. OCHOA, 2000. Microworld
For Training Traders In The Colombian Electricity Market. System Dynamics Conference,
Aug, 2000. Bergen, Norway.
DYNER, I. and LARSEN, E., 2001. “From Planning to Strategy in the Electricity Industry”.
Energy Policy, Volume29, Number 13, November 2001.
FORD, A., 1997. “System Dynamics and the Electric Power Industry”. System Dynamics
Review, Spring,Volume 13, Number 4. Printed in Great Britain by Unwin Brother. Ltd.
FORD, A., YABROFF, I. W., 1979. “Technical Documentation of the Electric Utility Policy and
Planning Analysis Model. LA-7773-MS Informal Report. University of California, Los
Alamos Scientific Laboratory. Los Alamos, New Mexico, EU. 74 p.
GENTA, P. J., NEVILLE S., 1996. A microworlds of Exploration and Development: Creating a
Leaming Laboratory for the Oil & Gas Industry. International System Dynamics
Conference. Cambridge, Massachusetts. p.p. 174.
HIRST, E., & HADLEY, S., 1999. Generation Adequacy: Who Decides?. En: The Electricy
Journal, Elservier Science Inc., Octuber, p.p 11-21.
INTEGRAL-UN-COLCIENCIAS, 2000. “Plataforma en Dindmica de Sistemas para Analizar
Posibilidades de Inversion en Generacién Eléctrica”. Informe Final. Universidad Nacional
de Colombia - Integral S.A - COLCIENCIAS. Medellin, Colombia.
ISA - UNIVERSIDAD NACIONAL DE COLOMBIA, 2000. “Micromundo para Comercializacion
de Energia en el Mercado Eléctrico Colombiano”. Informe final. Medellin, Colombia.
ISA S.A. E.S.P., 1997. Anilisis del Mercado Mayorista de la Electricidad en Colombia. Primera
Edicion. Editorial Colina. Medellin.
ISA S.A. E.S.P., 1999. Analisis del Mercado Mayorista de Electricidad en Colombia 1998 - 1999.
Gerencia de Servicios de Informacion. Primera edicién, Medellin: Colombia. p. 101.
LANGLEY, P. etal, 1996. Beefeater Restaurant Microworld. User guide.
LARSEN, E., ACKERE, A., WARREN K., 1995. The growth of service and the service of growth.
London Business School. April 1995.
LEE, T.H., BALL, B.C. and TABORS, R. D. (1990). Energy aftermath, how we can learn from
blunders of the past to create a hopeful energy future. Harvard Business School. Press:
Boston, Massachusetts, US.
MACHUCA, J. A., DEL POZO, R., 1997. The Beer Game, with the possibility of the Use of EDI,
through the Intemet. Intemational System Dynamics Conference. Pp. 383. Instambul
Turkey.
MONTOYA M., S. F., 1997. “Estudio de la Evoluci6n Futura de Sistemas de Generacién en
Ambiente de Competencia”. Medellin, Universidad Nacional de Colombia, Facultad de
Minas. Colombia, 231 p.
NAIL, R. F., 1992. “A System Dynamics Model for National Energy Policy Planning. System
Dynamics Review, Winter,Volume 8, Number 1. Printed in Great Britain by Unwin
Brother. Ltd.
OSORIO, S., 2000. Tesis de Maestria en Proceso. Posgrado en Aprovechamiento de Recursos
Hidrdulicos. Universidad Nacional de Colombia, Medellin.
PARK, C.S. SHARP-BETTE, G.P. 1990. Advance Engineering Aconomics. Ed. Jhon Wiley &
Sons, Inc. USA. p. 740.
PAUL, L., ERIK, L., 1995. “Edutainment, leaming and system dynamics”. System Dynamics
Review Vol 11, No. 4. June.
SALAZAR, J. E., 1994. Aplicacion de dos modelos No Lineales al Estudio de Series Temporales
en Hidrologia. Medellin: [s.n.] Tesis de Maestria. Posgrado en Aprovechamiento de
Recursos Hidraulicos, Universidad Nacional de Colombia, 114 p.
SENGE, P., LANNON C., 1992. Managerial Microworlds. Technology Review.
SIMONS, K., 1990. “New Technologies in simulation games”. System Dynamics Group, MIT.
April.
SMITH, R., DYNER, I., et al, 1999. Simulador de la comercializacion de energia en el mercado
eléctrico colombiano. COCIER, Bogota, Colombia.
STERMAN, J. D., 2000. “Business Dynamics: Systems Thinking and Modeling for a Complex
World”. McGraw-Hill Higher Education, USA, 982 p.
UMPE, 1995. Estudio de Proyeccion de la Demanda de Energia del Pais. Ministerio de Minas y
Energia, Santa Fe de Bogota, Colombia.
UN-COLCIENCIAS-ISA, 2000. Opciones de Manejo del Recurso Hidrico en el Mercado Eleéctrico
Colombiano. Proyecto de investigacion del Posgrado en Aprovechamiento de Recursos
Hidrdulicos, Universidad Nacional de Colombia. Patrocinado por Interconexion Eléctrica
S.A. y COLCIENCIAS.
UNIVERSIDAD NACIONAL DE COLOMBIA, COLCIENCIAS, 1997. “Planeamiento Integrado
de la Expansion de un Sistema Hidrotérmico Interconectado de Generacién de Energia
Eléctrica”. Tomos I y II. Posgrado en Aprovechamiento en de Recursos Hidraulicos”.
Facultad de Minas. Medellin, Colombia.
UPME, 1999. Plan de Expansién Referencia Generacién Transmisién 1998 - 2000. Unidad de
Planeacion Minero Energética, Ed. Estrategy, Santa Fé de Bogota D.C., Colombia, 176 p.
VARELA, R., 1993. Evaluacién Econémica de Inversiones. Bogota: Ed. Norma. Colombia. p.
512.
VLAHOS, K. 1998. The Electricy Markests Microworld. Version 1.0. LBS, UK.
Acknowledgement
The authors sincerely acknowledge the financial support of COLCIENCIAS (the
Colombian research council) and Integral
Back to the Top
CC BY-NC-SA 4.0