The ongoing energy transition is reshaping the structure and operation of electric power systems, particularly at the distribution level. The increasing penetration of Distributed Energy Resources (DERs) is transforming traditionally passive distribution networks into Active Distribution Networks (ADNs). While this evolution supports decarbonization goals, it also introduces significant operational challenges for Distribution System Operators (DSOs). In this context, Local Flexibility Markets (LFMs) have emerged as a promising market-based solution to efficiently procure active and reactive flexibility services from DERs, delaying conventional grid reinforcements. This PhD thesis investigates the role of LFMs as an enabling mechanism for secure, efficient, and sustainable operation of ADNs, providing insights for system operators and market participants. Particular attention is devoted to studying the viewpoints of both DSOs and Balancing Service Providers (BSPs), which are essential for the practical deployment of flexibility-based solutions. In this regard, after an extensive overview of LFMs in Europe, this thesis addresses the operational needs of DSOs through the development of advanced forecasting models aimed at identifying critical conditions in ADNs. A novel multi-layer machine learning framework is proposed to reconstruct load and generation profiles under limited observability in a transparent and scalable architecture. From the BSP perspective, this thesis introduces an original probabilistic formulation of Feasible Operating Regions (FORs) for DERs, explicitly accounting for uncertainty, reliability, and service-related costs, relying on historical power profiles rather than real-time measurements. The resulting FORs enable the quantitative assessment of heterogeneous portfolios of DERs, supporting risk-aware market participation and optimized bidding strategies. Finally, the thesis investigates the contribution of DERs to reactive power management in urban distribution networks, with a specific focus on the HV/MV interface. The results show that coordinated activation of DERs can significantly reduce unwanted reactive power exchanges and improve voltage profiles, offering a viable alternative to centralized compensation devices.
La transizione energetica sta profondamente trasformando la struttura e l’esercizio dei sistemi elettrici di potenza, specialmente a livello di distribuzione. La crescente penetrazione di risorse energetiche distribuite (Distributed Energy Resources - DER) sta convertendo le reti di distribuzione, tradizionalmente passive, in attive. Sebbene tale evoluzione supporti gli obiettivi di decarbonizzazione, essa introduce anche criticità per gli operatori di rete (Distribution System Operator - DSO). In tale contesto, i mercati locali di flessibilità (Local Flexibility Market - LFM) rappresentano una promettente soluzione per l’approvvigionamento di servizi di flessibilità locale forniti dalle DER, consentendo di posticipare gli interventi di rinforzo strutturale della rete. La presente tesi di dottorato analizza il ruolo dei LFM al fine di un esercizio sicuro, efficiente e sostenibile delle reti di distribuzione, fornendo indicazioni operative concrete per gli operatori di sistema e per gli attori del mercato. Particolare attenzione è dedicata all’analisi dei punti di vista sia dei DSO sia dei Balancing Service Providers (BSP). A tal proposito, dopo un’ampia ricerca sui LFM in Europa, la tesi affronta le esigenze dei DSO tramite lo sviluppo di modelli avanzati di previsione di condizioni operative critiche nelle reti di distribuzione, proponendo un innovativo approccio multilivello basato sul machine learning, al fine di ricostruire i profili di carico e di generazione in condizioni di osservabilità limitata. Dal punto di vista dei BSP, la tesi introduce una formulazione probabilistica delle regioni di funzionamento ammissibili delle DER (Feasible Operating Regions - FOR), esplicitando l’affidabilità e dei costi associati ai servizi, basandosi su profili storici di potenza piuttosto che su misure in tempo reale. Tali FOR consentono una valutazione quantitativa di portafogli eterogenei di DER, supportando una partecipazione consapevole al mercato e la definizione di strategie di offerta ottimizzate. Infine, la tesi analizza il contributo delle DER alla gestione della potenza reattiva nelle reti di distribuzione urbane, con particolare riferimento all’interfaccia alta/media tensione. I risultati dimostrano che l’attivazione coordinata delle DER può ridurre in modo significativo gli scambi indesiderati di potenza reattiva e migliorare i profili di tensione, rappresentando una valida alternativa ai dispositivi di compensazione centralizzati.
Local flexibility services in active distribution networks: a DSO and BSP perspective
DACCÒ, EDOARDO
2026
Abstract
The ongoing energy transition is reshaping the structure and operation of electric power systems, particularly at the distribution level. The increasing penetration of Distributed Energy Resources (DERs) is transforming traditionally passive distribution networks into Active Distribution Networks (ADNs). While this evolution supports decarbonization goals, it also introduces significant operational challenges for Distribution System Operators (DSOs). In this context, Local Flexibility Markets (LFMs) have emerged as a promising market-based solution to efficiently procure active and reactive flexibility services from DERs, delaying conventional grid reinforcements. This PhD thesis investigates the role of LFMs as an enabling mechanism for secure, efficient, and sustainable operation of ADNs, providing insights for system operators and market participants. Particular attention is devoted to studying the viewpoints of both DSOs and Balancing Service Providers (BSPs), which are essential for the practical deployment of flexibility-based solutions. In this regard, after an extensive overview of LFMs in Europe, this thesis addresses the operational needs of DSOs through the development of advanced forecasting models aimed at identifying critical conditions in ADNs. A novel multi-layer machine learning framework is proposed to reconstruct load and generation profiles under limited observability in a transparent and scalable architecture. From the BSP perspective, this thesis introduces an original probabilistic formulation of Feasible Operating Regions (FORs) for DERs, explicitly accounting for uncertainty, reliability, and service-related costs, relying on historical power profiles rather than real-time measurements. The resulting FORs enable the quantitative assessment of heterogeneous portfolios of DERs, supporting risk-aware market participation and optimized bidding strategies. Finally, the thesis investigates the contribution of DERs to reactive power management in urban distribution networks, with a specific focus on the HV/MV interface. The results show that coordinated activation of DERs can significantly reduce unwanted reactive power exchanges and improve voltage profiles, offering a viable alternative to centralized compensation devices.| File | Dimensione | Formato | |
|---|---|---|---|
|
Thesis_Complete.pdf
non disponibili
Licenza:
Tutti i diritti riservati
Dimensione
12.96 MB
Formato
Adobe PDF
|
12.96 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/382759
URN:NBN:IT:POLIMI-382759