Thanks to economies of scale and steady improvements in technological reliability, Li-ion Battery Energy Storage Systems (LiBESS) are now an effective solution for improving the energy efficiency of industrial power generation systems. This thesis assesses the techno-economic feasibility of integrating a LiBESS into the cogeneration unit of an Italian Oil & Gas facility. The objective is to reduce the spinning reserve currently provided by three partially loaded gas turbine generators coupled with a thermal-following steam turbine, thereby reducing CO2 emissions while maximizing plant flexibility and revenues from energy arbitrage. LiBESS sizing is therefore affected by the variability of electric load, steam demand, ambient temperature and electricity market prices. To model uncertainty and operational constraints, the thesis proposes a data-driven Stochastic Unit Commitment (SUC) optimization formulated as a Mixed-Integer Linear Programming (MILP) model. The objective is to minimize the daily operating cost, including LiBESS investment and replacement costs as a function of its capacity fade due to degradation. Daily profiles of the stochastic variables are generated by estimating their hourly probability density functions via Kernel Density Estimation (KDE) and consequently combined into scenarios using elliptical copulas in the normalized principal components space via Principal Component Analysis (PCA). The generated scenarios are then reduced to a representative and computationally tractable set through the Fast Forward Selection (FFS) algorithm. Comparing the stochastic solution with the deterministic dispatch, the Value of the Stochastic Solution (VSS) is approximately 5,562 EUR/day in savings. The analysis performed with commercial Lithium Iron Phosphate (LFP) technology, suitable for Oil & Gas applications, quantifies a Net Present Value (NPV) of the investment up to 28.51 MEUR over the project lifetime, yielding to 6.83% increase in cogeneration efficiency and CO2 emissions reduced by 22,050 ton/year compared with the current plant configuration.
Grazie alle economie di scala ed al miglioramento dell'affidabilità, i sistemi di accumulo di energia a batterie ioni di Litio (LiBESS) rappresentano oggi una soluzione concreta per l'efficientamento energetico dei sistemi di generazione di potenza industriali. La tesi valuta la fattibilità tecnico-economica dell'integrazione di un LiBESS nell'unità di cogenerazione di un impianto di produzione Oil & Gas italiano. L'obiettivo è ridurre la riserva rotante fornita dai tre turbogas a carico parziale accoppiati ad una turbina a vapore in contropressione, riducendo così le emissioni di CO2 e massimizzando al tempo stesso la flessibilità dell'impianto ed i ricavi dall'export di energia. Il dimensionamento del LiBESS è pertanto influenzato dalla variabilità del carico elettrico, della domanda di vapore, della temperatura ambiente e dei prezzi del mercato elettrico. Per modellare l'incertezza ed i vincoli operativi, la tesi propone un approccio data-driven di ottimizzazione stocastica (Unit Commitment) in un modello di programmazione lineare intera mista (MILP). L'obiettivo è minimizzare il costo operativo giornaliero, includendo i costi di investimento e sostituzione del LiBESS in funzione della sua riduzione di capacità dovuta alla degradazione. I profili giornalieri delle variabili sono generati stimando la loro densità di probabilità oraria tramite Kernel Density Estimation (KDE) ed accoppiati in scenari usando copule ellittiche nello spazio normalizzato delle componenti principali (PCA). Gli scenari generati sono ridotti ad un set rappresentativo e trattabile mediante l'algoritmo di Fast Forward Selection (FFS). Confrontando la soluzione stocastica con il dispacciamento deterministico, il Valore della Soluzione Stocastica (VSS) ammonta a circa 5.562 EUR/giorno di risparmio. L'analisi svolta con tecnologia commerciale Litio Ferro Fosfato (LFP), adatta ad applicazioni Oil \& Gas, quantifica il valore attuale netto (VAN) dell'investimento a 28.51 MEUR lungo la vita del progetto, aumentando l'efficienza cogenerativa del 6,83% e riducendo le emissioni di CO2 di 22.050 ton/anno rispetto alla configurazione attuale dell'impianto.
Stochastic and ageing-aware optimal sizing of a Li-ion Battery energy storage system in an italian oil and gas facility
Pajer, Alessandro
2025/2026
Abstract
Thanks to economies of scale and steady improvements in technological reliability, Li-ion Battery Energy Storage Systems (LiBESS) are now an effective solution for improving the energy efficiency of industrial power generation systems. This thesis assesses the techno-economic feasibility of integrating a LiBESS into the cogeneration unit of an Italian Oil & Gas facility. The objective is to reduce the spinning reserve currently provided by three partially loaded gas turbine generators coupled with a thermal-following steam turbine, thereby reducing CO2 emissions while maximizing plant flexibility and revenues from energy arbitrage. LiBESS sizing is therefore affected by the variability of electric load, steam demand, ambient temperature and electricity market prices. To model uncertainty and operational constraints, the thesis proposes a data-driven Stochastic Unit Commitment (SUC) optimization formulated as a Mixed-Integer Linear Programming (MILP) model. The objective is to minimize the daily operating cost, including LiBESS investment and replacement costs as a function of its capacity fade due to degradation. Daily profiles of the stochastic variables are generated by estimating their hourly probability density functions via Kernel Density Estimation (KDE) and consequently combined into scenarios using elliptical copulas in the normalized principal components space via Principal Component Analysis (PCA). The generated scenarios are then reduced to a representative and computationally tractable set through the Fast Forward Selection (FFS) algorithm. Comparing the stochastic solution with the deterministic dispatch, the Value of the Stochastic Solution (VSS) is approximately 5,562 EUR/day in savings. The analysis performed with commercial Lithium Iron Phosphate (LFP) technology, suitable for Oil & Gas applications, quantifies a Net Present Value (NPV) of the investment up to 28.51 MEUR over the project lifetime, yielding to 6.83% increase in cogeneration efficiency and CO2 emissions reduced by 22,050 ton/year compared with the current plant configuration.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_07_Pajer_Tesi.pdf
accessibile in internet per tutti
Descrizione: testo tesi
Dimensione
39.85 MB
Formato
Adobe PDF
|
39.85 MB | Adobe PDF | Visualizza/Apri |
|
2026_07_Pajer_Executive_Summary.pdf
accessibile in internet per tutti
Descrizione: testo executive summary
Dimensione
14.64 MB
Formato
Adobe PDF
|
14.64 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/10589/261627