This work consists in the development of a mathematical model focused on predicting temperature and salt concentration profiles in the non-convecting zone (NCZ) of a salinity gradient solar pond. The one-dimensional transient model proposed is based on the heat and mass diffusion equations, using energy and mass balances for the upper convective zone (UCZ) and the lower convective zone (LCZ) as boundary conditions, respectively. The model was then numerically solved using the finite difference method and the Crank-Nicolson scheme in MATLAB R2012a. Time-variable weather and radiation parameters for a specific location (Monterrey, N.L., Mexico) were taken into account, along with heat losses from the surface and from the sides of the pond and with variable thermophysical properties for the brine. Due to the complexity of the problem, different versions of the model were developed, in hope to better understand the effect of the different parameters on temperature and salinity profiles. For each version of the model temperature and salinity were calculated at 22 different depths of the pond, for 22 days, with a time-step of 5 minutes, starting July 18th 2014. Results from the different models were compared between them, and a subsequent comparison was carried between between models' results and experimental data collected from a small-scale solar pond built in the "Casa Solar" of the Tecnológico de Monterrey, Monterrey, N.L., Mexico. Overall, the different versions of the model did show an increase in the LCZ temperature as the simulations progressed, and at the end of the 3 weeks time all presented a qualitatively good shape in temperature and salinity profiles. However, the versions of the model developed to take into account evaporation heat losses, the wind power law and the shading effect for radiation failed in describing the real effect of these parameters on temperature and salinity profiles of a solar pond. Moreover, comparison with experimental data highlighted the discrepancy in temperature values obtained from the models. Those versions of the model that took into account radiation as a source term for all the 3 zones of the pond showed better agreement with experimental data than the others, demonstrating the importance of using an adequate model for radiation parameters. Finally, comparison with experimental data highlighted the effects of the hypothesis of fixed coordinates on temperature and salinity results.
Questa tesi consiste nello sviluppo di un modello matematico per il calcolo di profili di temperatura e di concentrazione salina che si formano nello strato non convettivo di uno stagno solare a gradiente di salinità. Il modello proposto si basa sull'equazione del calore unidimensionale in regime transitorio, e sulla corrispettiva equazione per il trasporto di materia diffusivo. Le condizioni al contorno per tale modello sono state ricavate da bilanci energetici sui due strati convettivi, superiore ed inferiore, dello stagno. Il modello è stato poi risolto numericamente secondo il metodo delle differenze finite e lo schema di Crank-Nicolson, con il software MATLAB R2012a. Le simulazioni sono state effettuate a partire dalle condizioni climatiche giornaliere della città di Monterrey, N.L, Messico, prendendo in considerazioni le perdite termiche caratterizzanti sia la superficie che i lati dello stagno, e assumendo come variabili le proprietà termofisiche della miscela acqua-sale. A causa della complessità del problema, sono state sviluppate diverse versioni del modello, con l'obiettivo di studiare l'effetto delle diverse variabili sui profili di temperatura e salinità. Per ciascuna versione del modello, sono stati calcolati valori di temperatura e salinità per 22 punti a diversa profondità, per 22 giorni (a partire dal 18 di luglio 2014), ogni 5 minuti. Al confronto tra i risultati dei vari modelli, è seguita una comparazione tra questi e i risultati sperimentali ottenuti da uno stagno solare di piccola scala costruito alla "Casa Solar" dell'università Tecnológico de Monterrey, Monterrey, N.L., Messico. Nel complesso, le diverse versioni del modello hanno tutte portato ad un aumento della temperatura nella zona convettiva inferiore, durante le simulazioni, e alla fine delle 3 settimane simulate, i profili di temperatura e di salinità ottenuti si sono dimostrati in accordo con i risultati sperimentali, a livello qualitativo. Tuttavia, le versioni del modello sviluppate per studiare alcuni fenomeni particolari, quali le perdite per evaporazione, oppure l'effetto d'ombra causato dalle pareti dello stagno, si sono rivelate fallimentari nel descrivere tali effetti. Al contrario, le versioni del modello in cui la radiazione è stata presa in considerazione come termine sorgente nell'equazione del calore, per tutti e tre gli strati, hanno mostrato una maggiore concordanza con i risultati sperimentali.
Numerical simulation of the development of temperature profiles for a salinity gradient solar pond
GATTI, ILARIA ALESSANDRA
2015/2016
Abstract
This work consists in the development of a mathematical model focused on predicting temperature and salt concentration profiles in the non-convecting zone (NCZ) of a salinity gradient solar pond. The one-dimensional transient model proposed is based on the heat and mass diffusion equations, using energy and mass balances for the upper convective zone (UCZ) and the lower convective zone (LCZ) as boundary conditions, respectively. The model was then numerically solved using the finite difference method and the Crank-Nicolson scheme in MATLAB R2012a. Time-variable weather and radiation parameters for a specific location (Monterrey, N.L., Mexico) were taken into account, along with heat losses from the surface and from the sides of the pond and with variable thermophysical properties for the brine. Due to the complexity of the problem, different versions of the model were developed, in hope to better understand the effect of the different parameters on temperature and salinity profiles. For each version of the model temperature and salinity were calculated at 22 different depths of the pond, for 22 days, with a time-step of 5 minutes, starting July 18th 2014. Results from the different models were compared between them, and a subsequent comparison was carried between between models' results and experimental data collected from a small-scale solar pond built in the "Casa Solar" of the Tecnológico de Monterrey, Monterrey, N.L., Mexico. Overall, the different versions of the model did show an increase in the LCZ temperature as the simulations progressed, and at the end of the 3 weeks time all presented a qualitatively good shape in temperature and salinity profiles. However, the versions of the model developed to take into account evaporation heat losses, the wind power law and the shading effect for radiation failed in describing the real effect of these parameters on temperature and salinity profiles of a solar pond. Moreover, comparison with experimental data highlighted the discrepancy in temperature values obtained from the models. Those versions of the model that took into account radiation as a source term for all the 3 zones of the pond showed better agreement with experimental data than the others, demonstrating the importance of using an adequate model for radiation parameters. Finally, comparison with experimental data highlighted the effects of the hypothesis of fixed coordinates on temperature and salinity results.| File | Dimensione | Formato | |
|---|---|---|---|
|
Gatti_Tesi.pdf
solo utenti autorizzati dal 20/09/2017
Descrizione: Testo della tesi
Dimensione
6.3 MB
Formato
Adobe PDF
|
6.3 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/10589/125921