A parametric assessing and intelligent forecasting of the energy and exergy performances of a dish concentrating photovoltaic/thermal collector considering six different nanofluids and applying two meticulous soft computing paradigms

Highlights

• Six engine oil based Nanofluids were examined as the coolant.

• Two soft computing paradigms used to predict the first and second law efficiencies.

• CuO-engine oil Nanofluid provides highest useful thermal energy.

• MWCNT-engine oil Nanofluid has the highest first and second law efficiencies.

• The highest and lowest electrical performances belong to CuO and SiO2, respectively.

Abstract

In the present study, the application of six engine oil-based Nano fluids (NFs) in a solar concentrating photovoltaic thermal (CPVT) collector is investigated. The calculations were performed for different values of nanoparticle volume concentration, receiver tube diameter, concentrator surface area, receiver length, receiver actual to the maximum number of channels ratio, beam radiation, and a constant volumetric flow rate. Besides, two novel soft computing paradigms namely, the cascaded forward neural network (CFNN) and Multi-gene genetic programming (MGGP) were adopted to predict the first law efficiency (${\eta }_I$) and second law efficiency (${\eta }_{II}$) of the system based on the influential parameters, as the input features. It was found that the increase of nanoparticle concentration leads to an increase in ${\eta }_I$ and a decrease in ${\eta }_{II}$. Moreover, the rise of both the concentrator surface area (from 5 m² to 20 m²) and beam irradiance (from 150 W/m² to 1000 W/m²) entails an increase in both the ${\eta }_I$ (by 39% and 261%) and ${\eta }_{II}$ (by 55% and 438%). Furthermore, it was reported that the pattern of changes in both ${\eta }_I$ and ${\eta }_{II}$ with serpentine tube diameter, receiver plate length, and absorber tube length is increasing-decreasing. The results of modeling demonstrated that the CFNN had superior performance than the MGGP model.

Introduction

The thermal performance of solar collectors is important due to its significant effect in decreasing the collector size and capital investment. More specifically in the Photovoltaic (PV) solar panels with low solar to electrical efficiency and high capital costs, the cooling down of PV modules has proven to be an efficient alternative to enhance the PV efficiency. Several techniques have been introduced and examined to enhance the PV efficiency such as spraying water on the PV surface ([[1], [2], [3]]), using phase change materials ([[4], [5], [6], [7]]), thermoelectric ([4,[7], [8], [9]]), microchannel-jet impingement ([[10], [11], [12]]), or using the air or other fluids micro-channel cooling. The hybridization of solar heating and electricity generation through photovoltaic thermal (PVT) technology leads to enhance the total efficiency by cooling the PV panel and producing electricity at higher electrical efficiency while generating thermal energy. Numerous research studies have been conducted on the application of NFs as the coolant in the PVT or concentrated photovoltaic thermal (CPVT) technologies. Sheikholeslami et al. [13] performed a comprehensive review on the application of NFs in PV/flat plate solar collectors. Based on that study, the researches on NF-based PVT systems are mostly associated with the application of 28% AL₂O₃, 12% MWCNT, and 10% CuO nanoparticles in the mentioned PVT system. The observations indicated that nearly 29%, 20–30%, and 9% improvement in the efficiency of PVT are obtained by using Al₂O₃, MWCNT, and CuO with concentrations ($\phi$) of 0.2%, 1%, and 0.05%, respectively. A numerical heat transfer and thermodynamic research performed by Ould-Lahoucine et al. [14] showed that the application of water/TiO₂ NF in a PVT unit with a rectangular channel has no significant advantage over the pure water coolant. Jia et al. [15] performed a numerical analysis to evaluate the thermal and electrical performance of a PVT with water/Al₂O₃ and water/TiO₂ NFs. The results of that study revealed the higher performance of PVT with the water/Al₂O₃ over the other NF. In addition, the authors showed that the thermal and electrical efficiencies of the PVT system are notably increased by increasing the NF mass flow rate from 0.0005 kg/s to 0.03 kg/s. Sardarabadi et al. [16] experimentally investigated the effect of metal/water NF on PVT system using three NFs of Al2O3, TiO2, and ZnO. The results revealed that the total exergy efficiency for TiO2 and ZnO increased by 15.93% and 15.45%, respectively, compared with that for the PV system without cooling. In addition, the highest entropy generation was obtained for the Al2O3. Fudholi et al. [17] investigated the efficiency of PVT collector by using the TiO₂ considering the NF concentrations($\phi$) of 0.5% and 1%, solar irradiance of 500–900 W/m²as well as the NF mass flow rate of 0.012 kg/s to 0.0255 kg/s. They showed that the TiO₂ with $\phi$ of 1% enhances the energy efficiency of PVT by 17% over the system with pure water coolant. Moreover, they showed that the effect of increasing $\phi$ is more prominent on the thermal efficiency rather than the electrical efficiency. Al-Shamaniet al. [18] performed a theoretical and experimental study on the influence of using NFs in increasing the efficiency of a PVT system considering three different NFs of water/CuO, water/SiO₂, and water/ZnO. The results indicated that water/SiO₂enhances the thermal and electrical efficiencies of the PVT by 5.76% and 12.70%, respectively, owing to a 30% decrease in the PV temperature and a 25% increase in the NF outlet temperature. Al-Waeliet al. [19] numerically examined the effect of three Nano-materials SiC, CuO, and, Al₂O₃ in three different base fluids to investigate the thermal efficiency of a PV/T system. They concluded that the base fluids and nanoparticles significantly affect the pressure drop and thermal efficiency. Besides, the highest heat transfer coefficient and pressure drop were obtained for SiC and glycerin, respectively. Rubbi et al. [20] proposed a novel method for better optimization of the PVT performance employing the Soybean oil/MXene (SO/Ti₃C₂)NF considering $\phi$ of 0 $0$.025–0.125%. They showed that a 55% increase in the thermal conductivity of SO/Ti₃C₂ was observed at $\phi$ = 0.125%. In addition, the thermal and electrical efficiencies of the PVT system were improved by 24.49% and 15.44%, respectively, as compared to water/Al₂O₃ NF considering the solar irradiance of 1000 W/m² and mass flow rate of 0.07 kg/s. Abadeh et al. [21] considered five various scenarios to evaluate the different aspects of using three NFs of Al₂O₃, TiO₂, and ZnO from energy and exergy efficiency points of view. The results showed that with applying the PVT/TiO₂the size of the PVT system is decreased and the energy and exergy efficiencies of the PVT system are enhanced by 33% and 7%, respectively, as compared to the PV system. Hissof et al. [22] examined the use of water/Al₂O₃ and water/Cu NFs in the PVT. The authors showed that the PV temperature is reduced by 3.6% by using the NF in the PVT. In addition, they found that the water/Cu provides higher efficiency over the water/Al₂O₃. The thermal and electrical efficiencies of PVT have observed an enhancement of 4.1% and 1.9%, respectively when Cu with $\phi$ of 2% in pure water is used as the PVT coolant. Jidhesh et al. [23] performed a series of experiments to evaluate the effect of using water/Cu NF in a semitransparent PVT collector. They showed that the temperature of the PV module reduces by 9 $°C$ and 12$°C$, respectively, for the pure water and water/Cu are coolants. The results also demonstrated that the energy and exergy efficiencies of the PVT with water/Cu are respectively 43% and 108% higher than that obtained for the PVT with the pure water coolant. Rahbar et al. [24] performed a numerical analysis on the performance of a CPVT with parabolic trough reflector and triple junction PV modules with water/Ag NF integrated with an organic Rankine cycle. Their results demonstrated that the CPVT with NF has thermal, electrical, and overall efficiencies of respectively 3.3%, 1.8%, and 5.1% higher than that with pure water coolant considering the CPVT concentration ratio of greater than 7. An experimental study was conducted by Rafiei et al. [25] to evaluate how Al₂O₃, Cu, CuO, TiO₂, and MWCNT oil-based nanoparticles affect the performance of a concentration parabolic dish used as the thermal source of a humidification dehumidification desalination system powered by a PVT. In fact, the PVT in that study comprises a flat plate thermal collector and PV module. The thermal energy required by the desalination unit is supplied on both parabolic dish concentrator and PVT. The experiments were performed for a constant NF volumetric flow rate. The results showed that the pressure drop is increased as $\phi$ increases from 0% to 5%; more specifically for the diathermic oil/CuO NF. The lowest pressure drop was observed for the NF with MWCNT nanoparticles. In addition, the overall efficiency of the system increased by increasing $\phi$ and the lowest exergy efficiency was obtained for the oil/Cu NF.

The soft computing prediction methods, which predict the output performance of a system based on different input parameters, are promising techniques to avoid the expensive experimental tests of the system ([26,27]). Mojumder et al. [28] applied three soft computing techniques (Extreme Learning Machine (ELM), Genetic Programming (GP), and Artificial Neural Networks (ANNs)) to present a model for a typical PVT performance based on the experimental data. Their results demonstrated that the ELM model is faster and more accurate than the two other models for predicting the electrical and thermal efficiencies of the PVT. Ammar et al. [29] applied the ANN model to develop the thermal energy distribution of a PVT and predict the optimum power production point. Varol et al. [30] used three soft computing models of ANN, Adaptive-Network-Based Fuzzy Inference System (ANFIS), and Support Vector Machines (SVM) to predict the useful thermal energy and efficiency of a solar collector filled with phase change material. They concluded that the SVM model is a highly efficient regression technique that is superior to ANN and ANFIS models.

Based on the literature, most of the studies are immersed with the application of NFs inside of the PVT systems. However, to the best knowledge of the authors, the CPVT dish concentrator with NF coolant has been rarely addressed in the previous studies. The CPVT dish concentrator provides high temperatures of up to 160 $°C$, which is suitable to derive organic Rankine cycle, double-effect absorption chillers, desalination, and so forth. The efficiency enhancement of CPVT leads to a decrease in its size and capital costs. To this end, the first-law and second-law thermodynamic modeling of a specific CPVT dish concentrator are performed to evaluate the effect of using five different NFs in a CPVT dish concentrator on the thermal, electrical, first-law, and second-law efficiencies. Buonomano et al. [31] have previously performed the detailed thermodynamic modeling of the so-called CPVT with diathermic oil coolant fluid. The idea of using NF in this CPVT type is novel and presented for the first. In addition, two robust soft computing techniques including the cascaded forward neural network (CFNN) and multi-gene genetic programming (MGGP) were developed to predict the first and second-law efficiencies considering six input design parameters of nanoparticle concentration ($\phi$), receiver tube diameter ($d_t$), concentrator surface area ($A_{conc}$), receiver length ($L_r$), receiver actual to the maximum number of channels ratio ($f_{ch}$), and beam irradiance ($I_b$). The applied soft computing techniques require low training time because of their quick learning process and robust performance. These soft computing techniques, which try to obtain the small training errors and norms of the weights, are useful to forecast the first and second law efficiencies of the CPVT against different input variables for which they are not trained on.