In the last decades, developing countries are constantly increasing their energy demand due to different factors like economic growth, industrialization and population growth. In developing contexts, international electricity masterplans play a pivotal role in assuring a sustainable development of the countries and assure the goal of universal access to electricity by 2030. In this context, robust electricity masterplans shall rely on appropriate models for long-term electricity demand forecasting. The company CESI S.p.A. is at the forefront of providing consultancy services to developing countries in the field of electricity planning. To support CESI’s activities in this field and to contribute to the same effort of other researchers focused on the demand forecasting modelling, the objective of this thesis work is to develop a software for time-series and econometric analysis for the forecast of the electricity demand in developing countries. First, an extensive analysis of the forecasting methods and procedures was performed. Second, the obtained information was organized and schematized in a novel flowchart to guide the modeler thorough the necessary statistical tests and inferences. Third, on the basis of the flowchart the FRED tool (Forecasting Electricity Demand) was developed in Excel VBA, which is able to automatize the forecasting procedure, guide the modeler through the entire analysis, providing robust statistical tests, facilitate the application to real cases and collaborate with the final analysis of the results. Three study cases (i.e. Bolivia, Nigeria, Cambodia) were analyzed, considering three geopolitically different nations and the results indicate that the tool is flexible to various situations, in addition to being intuitive. The tool will represent a useful supporting instrument for automatizing the demand forecasting procedures, facilitating the selection of appropriate and statistically robust scenarios of demand, and contributing to more sustainable electricity masterplans for developing countries.
Negli ultimi decenni la richiesta di energia elettrica nei paesi in via di sviluppo è aumentata considerevolmente a causa di diversi fattori quali la crescita economica, l’industrializzazione e l’aumento della popolazione. Nei contesti in via di sviluppo, i masterplan elettrici internazionali hanno un ruolo fondamentale per assicurare uno sviluppo sostenibile dei paesi e per raggiungere l’obbiettivo dell’accesso universale all’elettricità entro il 2030. In tale contesto, i masterplan più solidi si basano su modelli appropriati per la previsione della domanda di elettricità a lungo termine. La società CESI S.p.A. è in prima linea nel fornire servizi di consulenza ai paesi in via di sviluppo nel campo della pianificazione elettrica. Lo scopo di questa tesi consiste nello sviluppare un software innovativo per l’analisi econometrica e delle serie storiche per la previsione del consumo di energia elettrica nei paesi in via di sviluppo, al fine di supportare le attività svolte dalla CESI in questo settore, ma anche altri ricercatori impegnati sulla modellizzazione della previsione della domanda. In primo luogo, è stata svolta un’intensa analisi dei metodi di previsione e delle relative procedure. In secondo luogo, tali informazioni sono state schematizzare e organizzare in un nuovo diagramma di flusso al fine di guidare il modellatore attraverso i test statistici e le assunzioni necessarie. In terzo luogo, utilizzando come base il diagramma di flusso ottenuto è stato sviluppato il programma FRED (Forecasting Electricity Demand) in Excel VBA, capace di automatizzare il processo di previsione, guidare l’utilizzatore durante l’intera procedura, fornire robusti test statistici, facilitare l’applicazione a casi reali e collaborare nell’analisi finale dei risultati. Sono stati realizzati tre casi studio (i.e. Bolivia, Nigeria, Cambodia), prendendo in considerazione tre nazioni geopoliticamente differenti, mostrando come il programma sia flessibile per diverse situazioni oltre ad essere alquanto intuitivo. Il programma rappresenterà un utile strumento di supporto per automatizzare le procedure di previsione della domanda, facilitando la selezione di scenari appropriati e statisticamente robusti, e contribuendo nel progettare masterplan elettrici più sostenibili per i paesi in via di sviluppo.
Forecasting electricity demand in developing countries : a review of statistics and development of FRED software for time-series and econometric analyses
De SOUZA VILLARROEL, DAIANE FERNANDA;PESCE, DAMIANO
2019/2020
Abstract
In the last decades, developing countries are constantly increasing their energy demand due to different factors like economic growth, industrialization and population growth. In developing contexts, international electricity masterplans play a pivotal role in assuring a sustainable development of the countries and assure the goal of universal access to electricity by 2030. In this context, robust electricity masterplans shall rely on appropriate models for long-term electricity demand forecasting. The company CESI S.p.A. is at the forefront of providing consultancy services to developing countries in the field of electricity planning. To support CESI’s activities in this field and to contribute to the same effort of other researchers focused on the demand forecasting modelling, the objective of this thesis work is to develop a software for time-series and econometric analysis for the forecast of the electricity demand in developing countries. First, an extensive analysis of the forecasting methods and procedures was performed. Second, the obtained information was organized and schematized in a novel flowchart to guide the modeler thorough the necessary statistical tests and inferences. Third, on the basis of the flowchart the FRED tool (Forecasting Electricity Demand) was developed in Excel VBA, which is able to automatize the forecasting procedure, guide the modeler through the entire analysis, providing robust statistical tests, facilitate the application to real cases and collaborate with the final analysis of the results. Three study cases (i.e. Bolivia, Nigeria, Cambodia) were analyzed, considering three geopolitically different nations and the results indicate that the tool is flexible to various situations, in addition to being intuitive. The tool will represent a useful supporting instrument for automatizing the demand forecasting procedures, facilitating the selection of appropriate and statistically robust scenarios of demand, and contributing to more sustainable electricity masterplans for developing countries.| File | Dimensione | Formato | |
|---|---|---|---|
|
Thesis Daiane F. De Souza Villarroel and Damiano Pesce.pdf
accessibile in internet solo dagli utenti autorizzati
Dimensione
6.47 MB
Formato
Adobe PDF
|
6.47 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/175556