During the latest years, climate change and the influence it will have on our lives has been one of the most discussed problems in the whole world. In order to face it, Europe set a plan with long-term goals among which many are energy related, with a general trend leading to a greener future. This is going to progressively decrease the number of power plants fueled by programmable energy sources, which are the main agents of the Dispatching Service. It's aim is to guarantee the security of the Electrical System through real time energy tradings by means of different complicated procedures. This thesis analyses one of them, namely the extit{opposite call}, that happens whenever a power reserve generated by a plant is immediately used in real time. We know that all power reserve generations are just precautionary, hence in most cases the power reserve is not used. Needless to say, this leads to a lot of extra costs for the Transmission System Operator (TSO), and consequently for the consumers, that can be drastically reduced if this phenomenon was predicted. Moreover, less plants powered by programmable sources would be needed to manage this particular service. The first goal reached in this thesis was to build a rich dataset that describes the Ancillary Services Market (ASM), both with market-related data and external context variables. Then, we classified and then predict the presence of an opposite call using a Random Forest model, with the final aim of reducing the costs of the Dispatching Service.
Negli ultimi anni si sono sempre più frequentemente e attentamente discussi i cambiamenti climatici e le conseguenze che questi avranno sul nostro modo di vivere. In risposta all'emergenza, l'Europa si è posta degli obiettivi a lungo termine tra i quali molti in ambito energetico, promuovendo un netto spostamento verso le energie pulite. Ciò porterà ad una progressiva diminuzione degli impianti programmabili, ad oggi i principali agenti del Servizio di Dispacciamento. Quest'ultimo si pone l'obiettivo di garantire la sicurezza del sistema elettrico programmando scambi energetici in tempo reale. Questo lavoro di tesi tratta un fenomeno presente nel Dispacciamento, ovvero la extit{chiamata discorde}, che si verifica quando un impianto viene chiamato a generare un margine di riserva che viene immediatamente utilizzato in tempo reale. Essendo rare le situazioni in cui la chiamata discorde si verifica, si osserva che la maggior parte dei casi in cui un margine di riserva viene creato non viene poi utilizzato. Questo genera degli alti costi aggiuntivi per il Gestore della Rete, e di conseguenza per gli utenti, che potrebbero essere ridotti se il fenomeno fosse previsto. Un primo obiettivo raggiunt in questa tesi è stato costruire un dataset ricco di informazioni che descriva l'andamento del Mercato per il Servizio di Dispacciamento (MSD), sia con informazioni del mercato stesso che con variabili esterne di contesto. Successivamente, abbiamo classificato e fatto predizione per la chiamata discorde utilizzando un modello di Random Forest, con l'obiettivo di ridurre i costi del Servizio di Dispacciamento.
Statistical models and algorithms for the analysis of the electricity market and the economic dispatch service
Innocenti, Pietro
2019/2020
Abstract
During the latest years, climate change and the influence it will have on our lives has been one of the most discussed problems in the whole world. In order to face it, Europe set a plan with long-term goals among which many are energy related, with a general trend leading to a greener future. This is going to progressively decrease the number of power plants fueled by programmable energy sources, which are the main agents of the Dispatching Service. It's aim is to guarantee the security of the Electrical System through real time energy tradings by means of different complicated procedures. This thesis analyses one of them, namely the extit{opposite call}, that happens whenever a power reserve generated by a plant is immediately used in real time. We know that all power reserve generations are just precautionary, hence in most cases the power reserve is not used. Needless to say, this leads to a lot of extra costs for the Transmission System Operator (TSO), and consequently for the consumers, that can be drastically reduced if this phenomenon was predicted. Moreover, less plants powered by programmable sources would be needed to manage this particular service. The first goal reached in this thesis was to build a rich dataset that describes the Ancillary Services Market (ASM), both with market-related data and external context variables. Then, we classified and then predict the presence of an opposite call using a Random Forest model, with the final aim of reducing the costs of the Dispatching Service.| File | Dimensione | Formato | |
|---|---|---|---|
|
Innocenti Pietro - Master's Thesis.pdf
accessibile in internet per tutti
Descrizione: Master's Thesis from Innocenti Pietro
Dimensione
1.84 MB
Formato
Adobe PDF
|
1.84 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/10589/167537