Reservoir sedimentation is one of the most challenging problem that affects hydroelectric production, since it overall causes a reduction of the reservoir capacity that overcomes the annual increase in storage volume and implies a dangerous net loss of energy. For this reason, the following master thesis focuses on the evaluation and estimation of sediment yield in data scarce catchment. All the parameters used for the definition and the validation of the model (temperature, rainfall, AET - actual evapotranspiration) are obtained from open-sources databases in the web. The case study hereby presented is the Arror Multipurpose Dam in Kenya. The overall study methodology adopted for this project includes analytical and numeric modeling. Once the input data are obtained both from literature reviews and technical reports, the hydrological model is created and then used in order to estimate sediment loads entering the reservoir from the catchment area. The Soil and Water Assessment Tool (SWAT), a semi-distributed and open source hydrological model, was applied to the study area along with the sufi-2 optimization using the SWAT-CUP tool for calibration and validation of the model results. A good performance was observed in both calibration period (1985 – 1995) and validation period (1996 – 2005) for the actual evapotranspiration (AET), with Nash-Sutcliffe model Efficiency coefficient (NSE) of 0.66 and 0.61 respectively. Future scenarios have also been investigated for two RPC_s (4.5 and 8.5) using different models (CM, CMS and CESM) obtained from open-source database too. For both current and future situations, sediment yield has been evaluated and solutions are proposed for long term scenarios. Results have been compared with the outcomes both from technical reports and theoretical background, showing good results in line with the ones of previous studies. In conclusion, this project shows the potentiality of open-source data for the hydrological modelling of catchments in scarce data regions, giving the opportunity to achieve good results even when few data are available. The innovative and distinguishing feature of this thesis therefore lies on the possibility to exploit remote sensing data (i.e. AET) to successfully calibrate hydrological models in scarce data regions. Information are also provided for easily extracting NetCDF data and importing them into SWAT by means of customized MATLAB programs. Both discharge and sediment current and future results will help in the next project stage of Arror dam. The versatility and the user-friendly interface of the SWAT model is also proven in data scarce catchment management.
La sedimentazione nei bacini è una delle problematiche maggiormente sfidanti che possono inficiare la produzione idroelettrica. Essa infatti comporta nel complesso una riduzione della capacità del bacino che può superare l’incremento annuo dei volumi gestiti e implica conseguentemente una pericolosa perdita di energia. La seguente tesi di laurea ha come obiettivo la valutazione e la stima della produzione di sedimenti in bacini con pochi dati a diposizione. Tutti i parametri utilizzati nel modello (temperatura, precipitazioni, AET – evapotraspirazione effettiva) sono ottenuti da database nel web accessibili liberamente. Il caso di studio presentato è la diga di Arror in Kenya. La metodologia di studio generale adottata include la modellazione analitica e numerica del bacino. Una volta ottenuti i dati di input, sia da report tecnici che della letteratura, viene generato il modello idrologico che verrà quindi utilizzato per stimare i carichi di sedimenti prodotti. Il software Soil and Water Assessment Tool (SWAT), che consiste in un modello idrologico semi-distribuito e liberamente accessibile, è stato utilizzato per il caso in esame insieme all’algoritmo di ottimizzazione sufi-2 utilizzato in SWAT-CUP per la calibrazione e la validazione del modello realizzato. Ottimi risultati sono stati osservati nei periodi di calibrazione (1985–1995) e validazione (1996–2005) utilizzando come grandezza di riferimento l’evapotraspirazione (AET), con valori del Nash-Sutcliffe model Efficiency coefficient (NSE), rispettivamente pari a 0.66 e 0.61. Gli scenari futuri sono stati investigati per due RCP_s (4.5 e 8.5) utilizzando diversi modelli (CM, CMS e CESM), anch’essi open-source. Sia per la situazione attuale che per gli sviluppi futuri, la produzione dei sedimenti è stata valutata unitamente alla proposta di diverse soluzioni per gli scenari a lungo termine. I dati sono stati confrontati con quelli ottenuti da report tecnici e letteratura di bacini limitrofi e i risultati ottenuti si dimostrano essere in linea con quanto da altri ottenuto. Si confermano le potenzialità relative all’uso di database open-source per la modellazione idrologica di bacini con carenza di dati, fornendo la possibilità di ottenere buoni risultati anche in tali situazioni. La novità della tesi consiste dunque nella possibilità di sfruttare remote sensing data (es. AET) per calibrare con successo modelli idrologici, laddove non fossero disponibili sufficienti dati. Si forniscono informazioni relative alla stesura di programmi in MATLAB, che possano facilitare l’estrazione dellle informazioni (formato NetCDF) da utilizzare nella stesura del modello in SWAT. Si evidenziano infine la versatilità e la facilità di utilizzo dello stesso. I risultati ottenuti, con le relative previsioni negli anni a venire, saranno utilizzati nelle prossime fasi dei processi di costruzione e gestione della diga di Arror.
Investigating the effects of climate variability on reservoir sedimentation : Arror dam, Kenya
BALDIN, ANDREA
2018/2019
Abstract
Reservoir sedimentation is one of the most challenging problem that affects hydroelectric production, since it overall causes a reduction of the reservoir capacity that overcomes the annual increase in storage volume and implies a dangerous net loss of energy. For this reason, the following master thesis focuses on the evaluation and estimation of sediment yield in data scarce catchment. All the parameters used for the definition and the validation of the model (temperature, rainfall, AET - actual evapotranspiration) are obtained from open-sources databases in the web. The case study hereby presented is the Arror Multipurpose Dam in Kenya. The overall study methodology adopted for this project includes analytical and numeric modeling. Once the input data are obtained both from literature reviews and technical reports, the hydrological model is created and then used in order to estimate sediment loads entering the reservoir from the catchment area. The Soil and Water Assessment Tool (SWAT), a semi-distributed and open source hydrological model, was applied to the study area along with the sufi-2 optimization using the SWAT-CUP tool for calibration and validation of the model results. A good performance was observed in both calibration period (1985 – 1995) and validation period (1996 – 2005) for the actual evapotranspiration (AET), with Nash-Sutcliffe model Efficiency coefficient (NSE) of 0.66 and 0.61 respectively. Future scenarios have also been investigated for two RPC_s (4.5 and 8.5) using different models (CM, CMS and CESM) obtained from open-source database too. For both current and future situations, sediment yield has been evaluated and solutions are proposed for long term scenarios. Results have been compared with the outcomes both from technical reports and theoretical background, showing good results in line with the ones of previous studies. In conclusion, this project shows the potentiality of open-source data for the hydrological modelling of catchments in scarce data regions, giving the opportunity to achieve good results even when few data are available. The innovative and distinguishing feature of this thesis therefore lies on the possibility to exploit remote sensing data (i.e. AET) to successfully calibrate hydrological models in scarce data regions. Information are also provided for easily extracting NetCDF data and importing them into SWAT by means of customized MATLAB programs. Both discharge and sediment current and future results will help in the next project stage of Arror dam. The versatility and the user-friendly interface of the SWAT model is also proven in data scarce catchment management.File | Dimensione | Formato | |
---|---|---|---|
2020_04_Baldin.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
9.89 MB
Formato
Adobe PDF
|
9.89 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/10589/164508