In the last decade, the methods used by Transmission System Operators (TSOs) to size electrical reserves have changed substantially, driven by the growing penetration of renewable generation in the energy mix. The transition has not been uniform: differences in generation mix, regulatory framework, and energy market structure have led synchronous areas to develop their own approaches, and today a wide variety of methods coexist. Comparing these methods directly is not straightforward, as each has been developed on a specific system. This thesis addresses this gap by developing a Comparative Reserve Sizing Framework. Eight reserve sizing methods, drawn from both TSO practice and academic research, are applied under identical input conditions to a common set of operating scenarios derived from a real power system dataset. The results show that the choice of method has substantial consequences. In the reference system considered, the spread between the lowest and highest reserve estimates is significant across all reserve categories: primary reserve ranges from approximately 400~MW to 1013~MW, while the secondary plus tertiary reserve spans from approximately 424~MW to over 1700~MW depending on the method. Within probabilistic approaches, methods that retain the dimensioning incident as a binding floor tend to produce higher requirements than those that derive the reserve purely from the imbalance distribution, with the magnitude of this difference depending on the size of the dimensioning incident relative to the system. Methods also differ in how they respond to changing operating conditions: both system load and renewable penetration affect the reserve requirements of probabilistic methods, with load being the dominant driver, while deterministic methods produce fixed requirements regardless of system state.
Nell'ultimo decennio i metodi di dimensionamento delle riserve di bilanciamento hanno subito una trasformazione significativa, trainata dalla crescente penetrazione delle fonti rinnovabili. La transizione non è avvenuta in modo omogeneo: il mix generativo, il quadro normativo e la struttura del mercato elettrico hanno spinto ciascuna area sincrona a sviluppare il proprio approccio, e oggi la pratica operativa varia considerevolmente da un sistema all'altro. Confrontare questi metodi non è immediato, poiché ognuno è stato sviluppato e validato su un sistema specifico. Questa tesi affronta questa lacuna sviluppando un Comparative Reserve Sizing Framework. Otto metodi di dimensionamento delle riserve, tratti sia dalla pratica operativa dei TSO sia dalla letteratura accademica, vengono applicati in condizioni di ingresso identiche ad un insieme comune di scenari operativi derivati da un sistema elettrico reale. I risultati mostrano che la scelta del metodo ha conseguenze rilevanti. Nel sistema di riferimento considerato, lo scarto tra le stime più basse e più alte è significativo in tutte le categorie di riserva: la riserva primaria varia da circa 400~MW a 1013~MW, mentre la riserva secondaria più terziaria si estende da circa 424~MW a oltre 1700~MW a seconda del metodo. All'interno degli approcci probabilistici, i metodi che mantengono l'incidente di dimensionamento come vincolo fisso tendono a produrre requisiti più elevati rispetto a quelli che derivano la riserva dalla distribuzione statistica degli sbilanciamenti, con una differenza che dipende dalla dimensione dell'incidente di dimensionamento rispetto alla taglia del sistema. I metodi si differenziano inoltre nella capacità di adattarsi alle condizioni operative: sia il livello di carico sia la penetrazione rinnovabile influenzano i requisiti dei metodi probabilistici, con il carico come driver dominante, mentre i metodi deterministici producono un requisito costante indipendentemente dallo stato del sistema.
Electrical reserve sizing: a comparative analysis of methodological approaches
COLPANI, ALESSANDRA
2025/2026
Abstract
In the last decade, the methods used by Transmission System Operators (TSOs) to size electrical reserves have changed substantially, driven by the growing penetration of renewable generation in the energy mix. The transition has not been uniform: differences in generation mix, regulatory framework, and energy market structure have led synchronous areas to develop their own approaches, and today a wide variety of methods coexist. Comparing these methods directly is not straightforward, as each has been developed on a specific system. This thesis addresses this gap by developing a Comparative Reserve Sizing Framework. Eight reserve sizing methods, drawn from both TSO practice and academic research, are applied under identical input conditions to a common set of operating scenarios derived from a real power system dataset. The results show that the choice of method has substantial consequences. In the reference system considered, the spread between the lowest and highest reserve estimates is significant across all reserve categories: primary reserve ranges from approximately 400~MW to 1013~MW, while the secondary plus tertiary reserve spans from approximately 424~MW to over 1700~MW depending on the method. Within probabilistic approaches, methods that retain the dimensioning incident as a binding floor tend to produce higher requirements than those that derive the reserve purely from the imbalance distribution, with the magnitude of this difference depending on the size of the dimensioning incident relative to the system. Methods also differ in how they respond to changing operating conditions: both system load and renewable penetration affect the reserve requirements of probabilistic methods, with load being the dominant driver, while deterministic methods produce fixed requirements regardless of system state.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_07_Colpani_Tesi.pdf
accessibile in internet per tutti
Descrizione: Testo Tesi
Dimensione
4.57 MB
Formato
Adobe PDF
|
4.57 MB | Adobe PDF | Visualizza/Apri |
|
2026_07_Colpani_Executive_Summary.pdf
accessibile in internet per tutti
Descrizione: Executive Summary
Dimensione
913.09 kB
Formato
Adobe PDF
|
913.09 kB | 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/10589/260931