National Technical University of Athens (NTUA)
Στοχαστική μοντελοποίηση του παγκόσμιου ενεργειακού συστήματος και εφαρμογές
Abstract
dc:descriptionThe objective of the current thesis concerns the design and development of an innovative integrated energy-economy model that projects the evolution of the global energy system by 2050, which systematically incorporates explicit representation of uncertainty for model parameters. The most important novelty of the model is the fact that it is the first energy-economy in literature that incorporates full uncertainty for all model parameters, while most other model-based assessments assume uncertainty for only a small subset of input data. The theoretical background of the model combines energy-economy, mathematical modeling and statistics, in particular econometrics (advanced statistical estimation techniques) and stochastic simulation. Modelling concerns behavior of energy consumers and producers as well as formulation of economic balance in energy markets at national and international level. The balance between supply and demand of energy products is achieved in the model by means of market prices, which are affected by the cost of production, the amount of energy resources and the development of energy technologies. Decisions of energy system actors are dynamic over time and are modelled based on readjusted myopic expectations. Mathematical algebraic forms simulating decisions for technology and energy equipment are based on discrete choice theory that is suitable for the representation of decisions when there is significant heterogeneity among decision makers. Dynamic model equations simulate the time evolution of energy technology equipment that can be dynamically replaced based on economic criteria. The projection of international fossil fuel prices is endogenous and depends on the evolution of global demand and supply and hydrocarbon resources. The dynamic evolution of the energy system is affected by energy and climate policies, such as carbon tax, availability of new forms of energy, RES promotion policies, energy efficiency standards and construction of infrastructure. Model projections are also affected by the development of technology costs which are endogenous to the model and depend on the evolution of R&D expenditures and learning-by-doing (economies of scale with increased technology deployment in the market).The model is solved in PC environment using parallel processing to perform stochastic Monte Carlo experiments. The model was developed and evaluated statistically in a specific software environment (EViews) which provides the most modern statistical estimation algorithms. The model is numerically solved successfully in full scale to provide future projections of the global energy system and has been used in a variety of real large-scale applications by the European Commission. In model applications, emphasis was placed on exploring the importance of hydrocarbon resources and the evolution of international prices in the framework of long-term energy system development. The model was applied in numerous studies to explore global energy system transformation and climate change mitigation by drastically reducing carbon dioxide emissions. The model represents interactions between climate and energy policies, international energy prices and technological progress and their impact in all energy sectors by country or region, covering the entire globe and horizon to 2050 in annual steps.The main effort of the research conducted is the systematic incorporation of uncertainty into all factors which determine the projection of the system to the future. The comprehensive and detailed coverage of the uncertainty in a large scale integrated global energy system model allows for the quantification of all endogenously produced results and projections in the form of probability distributions, which do not receive pre-defined form. The interesting result of the model is exactly the form of the probability distributions that the model produces for endogenous variables. This form can be evaluated by the user of the model in order to derive conclusions about the range of uncertainty for each projection. For example, the model produces probabilistic estimates for the evolution of important endogenous variables (e.g. the possibility that global carbon emissions will exceed a certain threshold in 2050) as a result of the operation of the energy system under uncertainty. The information on the variability of projections is a critical parameter for the design of effective energy and climate policies and therefore integrated risk quantification provided the model analysis is of strategic importance.The model incorporates explicitly and without bias the main categories of uncertainties identified in the literature and in particular the lack of knowledge about the evolution of exogenous assumptions of the model (e.g. GDP and population), the uncertainty due to incomplete knowledge of the system and the value of the parameters (parametric uncertainty) and the variance caused by variables that have not explicitly been incorporated in model equation (residuals). The above are introduced in the form of probability distributions that are used by the model to perform "Monte Carlo" analysis providing probability distributions for all endogenous variables. Econometric estimations are widely used for the production of appropriate distributions for the parameters and assumptions entered into the model. The statistical analysis of historical data provides the basis for the selection of the distributions and the variance-covariance analysis. The estimators of the parameters are based on unbiased assessment methods using appropriate regression methods and statistical time series transformation. Special attention is given to the statistical dependence of the parameters introduced in the model, since their covariance plays an important role in determining the distribution of endogenous variables and especially aggregate ones.As part of the thesis, the model database has been developed in full scale, which is very extensive as it includes all the global energy system data, energy technologies of all kinds, but also historical data used for econometric estimations. Furthermore large-scale applications were performed and model results were extensively used by the European Commission. In the model applications, special emphasis was given to the impact of alternative energy and climate policies on the development of the global energy system and to the assessment of the importance of assumptions for global hydrocarbon resources and reserves for the formulation of international fossil fuel prices and the projection for fuel mix. The importance of uncertainty was extensively analyzed and unbiased probability distributions for all endogenous model variables were produced, which can provide probabilistic estimates of alternative outcomes. The analysis was extended to stochastic scenarios that investigate the role of uncertain climate policies and worldwide exploitation of shale gas and assessed differences that arise in probability distributions of variables and probabilistic estimates of future outcomes. Finally, the optimal allocation of investments in power generation technologies was examined based on risk minimization paradigm assuming that investors are risk averse. Modeling was based on methods mixed integer linear programming, while data on the uncertainty for costs of technologies, climate policies and electricity prices were provided by the stochastic energy model.The main advantage of the model compared to the literature lies in the comprehensive and systematic representation of the uncertainty surrounding the development of the main components of the global energy system and the integrated modeling of the interactions between energy and climate policies, the evolution of the global energy system and projections for international oil, natural gas and coal prices. The modelling also represents feedbacks and interactions of the global energy system with the development of the necessary infrastructure and with technological progress mainly with regard to low and zero-carbon technologies.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2016
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Fragkos, Panagiotis
- Φράγκος, Παναγιώτης
Subjects
dc:subject × 16- Μοντέλο ενεργειακής οικονομίας
- Ενσωμάτωση αβεβαιότητας σε μοντέλα ενέργειας
- Αξιολόγηση ενεργειακών πολιτικών
- Κλιματική αλλαγή
- Μακροχρόνιος ενεργειακός σχεδιασμός
- Προσομοιώσεις Monte Carlo
- Κλιματικές πολιτικές
- Climate change economics
- Energy - economy modelling
- Long - term energy planning
- Monte Carlo experiments
- Integration of uncertainty into world energy modelling
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/37233