The Influence of Al2O3- ZnO-H2O Nanofluid on the Thermodynamic Performance of Photovoltaic- Thermal Hybrid Solar Collector System

A thermodynamic analysis was conducted after alternatively testing different working fluids to particularly assess the effect of Nanofluid as heat transfer agent on improving thermal, electrical, and exergetic efficiencies of photovoltaic/thermal (PV/T) hybrid solar collector system under State of Qatar climate, using experimentally and computationally obtained data. The investigated fluids were water, and Al2O3-ZnO-H2O Nanofluid mixed with Ethylene Glycol as surfactant. Mass fractions of Nanoparticles were 0.05 wt.% Al2O3 with particle size of 5nm and 0.05 wt.% ZnO with particles size of 10-30 nm. The results showed that in comparison between water as absorption medium and the mentioned Nanofluid, an increase in thermal and hence total efficiency of the system was recorded for the latter absorption medium scenario, despite considering concentrations of the Nanoparticles relatively low. A very general trend of dramatic decrease in the positive effect of Nanofluid in consistency with the ambient temperature increase was also observed after fitting the data to segments of declining straight lines. For 0.05 wt.% mass fraction of Nanoparticles the averaged increment in total efficiency was 4.1%, and in total useable energy (exergy) efficiency was 4.6%.


General Introduction
Hybrid photovoltaic/thermal (PV/T) technology, through photovoltaic and photo thermal interaction, generates electricity and thermal power simultaneously from absorbed solar energy. Besides the higher overall energy performance, the merit of a PV/T system lies in the reduction of the demands on physical space and the equipment cost through the use of common frames and brackets as compared to the separated PV and solar thermal systems placed side-by-side. Its façade integration provides architectural uniformity and facilitates building thermal load reduction [1].
The assessment of solar photovoltaic/thermal (PV/T) hybrid collector system performance is commonly tackled from viewpoint of thermodynamics, as [1] conducted a study of appropriateness of glass cover on a thermo syphon heating PV/T system from the first law point of view and from the exergy analysis point of view. Ozgoren M, et al. [2] concluded that the conversion efficiency of the photovoltaic (PV) system with water cooling might be improved by an average of 10%, while Debbarma et al. [3] studied the thermal, electrical and exergetic characteristics of PV/T hybrid air collector system. Nanofluids are skillfully prepared colloidal suspensions of nanoparticles (which is of size 1-100 nm) dispersed in a base fluid [4]. Many authors investigated the effect of Nanofluids on the performance of solar collectors, such as Yousefi et al. [5] who experimentally studied the effect of Al 2 O 3 -water nanofluid on the efficiency of a flat-plate solar collector. Nagarajan et al. [6] prepared Nanofluids containing Al 2 O 3 , ZnO, and MgO nanoparticles. The amount of forced convective heat transferred through the as-prepared nanofluids in tubular solar collector showed ZnO-water nanofluid with 0.2% concentration by volume as the optimal option for the collector. Other authors like Li et al. [7] reviewed the fields of application for solar collectors using nanofluids as the working fluid. The effect of Nano fluids on PV/T hybrid collector performance was studied by Cui and Zhu [8] who applied the Nanosolution on top of the PV module. Although the electrical efficiency of the PV/T system was lower than the bare PV module, the overall energy conversion efficiency of the PV/T system was higher.
The use of finite element simulation is of interest in this study, so to simulate the temperature distribution on top of the solar collector or on the PV panel surface and then evaluate its performance. Yang et al. [9] followed the same path of investigation by measuring and simulating the temperature distribution, and the resulting data of experiments and simulation were very consistent. Analysis of a conceptual photovoltaic thermal (PV/T) solar panel design was performed by Fontenault and Gutierrez-Miravete [10] using COMSOL Multiphysics software and combinations of water flow rates and reservoir thicknesses were analyzed to determine which one produced optimal PV/T total efficiency. The work of Elmir et al. [11] included studying by numerical simulation the cooling of a solar cell by forced convection in presence of a nano fluid, and concluded to an increase in the rate of heat transferred in comparison with absence of nanoparticles.
In the present work, the effect of Nanofluid in the performance of thermally biased solar PV/T hybrid collector system is analyzed in terms of the relevant efficiencies and in comparison, with the normal water as working fluid, taking advantage of the collected experimental data from an available set of PV/T collectors at GORD's Techno Hub research facility in Qatar and with the aid of finite element modeling COMSOL Multi physics software. It was of interest in this work to investigate a Nanofluid composed of two different nanoparticles with low mass fraction for each one dispersed by manual stirring in a mixture of distilled water and Ethylene Glycol serves as surfactant.

Materials and Set up
Experimental investigations were conducted on two working fluids: water and Al 2 O 3 -ZnO-distilled water (H 2 O) Nanofluid mixed with Ethylene Glycol. The Nanofluid composition is 0.05 wt.% ZnO 99+% purify nearly spherical Nanoparticles (particle size 10-30 nm), 0.05 wt.% Al 2 O 3 99.99% purify nearly spherical Nanoparticles (particle size 5 nm) [12], ethylene glycol as natural surfactant for dispersion of Nano-powders, and a distilled water as the base fluid. Gentle stirring was applied to reassure dispersion of Nanoparticles in the base fluid. Figure 1 shows the prepared mixture.
The following Table 1 [13] summarizes thermo-physical properties and quantities of Nanofluid structural components.
Experiments were conducted on a ready setup consists of: an outdoor set of five solar PV/T hybrid collectors tilted 25°S and connected in series within a closed hydronic circuit in where the volumetric flow rate is maintained constant at 3.25 L/min, indoor 500 Liter storage tank with integrated heat exchangers in where the collectors' hydronic circuit is connected, variable speed circulating pump, electromagnetic flow meter, and a simple ball valve to control the flow. The ball valve is connected to the inlet (downstream) of the collectors' set. Two K-type Thermocouples and their data loggers were attached to the inlet and outlet pipes of the collectors' set to record the relevant temperatures. The local weather data of ambient temperature and wind speed was provided by Vantage Pro 2 weather station model manufactured by Davis ® weather monitoring. The solar irradiation at the plane of the PV/T collectors was measured using an ISO 9060 first class complaint Kipp & Zonen CMP 6 pyranometer. The following Figure 2 shows a schematic diagram of the full setup and Figure 3 shows the Photographic view of solar PV/T hybrid collector system.
The following Table 2 presents specifications of VOLTHER ® POWERTHERM hybrid PV/T collector as provided by the manufacturer.

Test Procedure and Condition
The closed hydronic circuit of the solar PV/T hybrid collector was firstly injected with the Nanofluid, and a static pressure of 0.4 bar and flow rate of 3.25 L/min was maintained. Figure 4 shows a photographic view of Nanofluid insertion process. The experimental tests took place in March 2017, solar irradiation, wind speed, inlet, outlet, and ambient temperatures were all recorded on hourly basis starting from 9:00 AM till 15:00 PM. The experiments have been repeatedly conducted so to obtain nearly matching or fixed surrounding atmospheric conditions and hence the comparison analysis becomes as realistic and accurate as possible. Selected two days data representing each different heat carrier fluid scenario were used for the analysis. The following Figure 5 compares solar irradiation, wind speed, and ambient temperatures for the two days of interest climatic conditions.

Thermo-physical properties of nano-fluid
The following available in the literature set of equations were utilized to quantify the thermo-physical properties of the resulting Nano-solution. Pak and Cho [14] mentioned in their work the wellknown equation of density which will represent the nano fluid as: Where ρ nf is density of nanofluid, ϕ is mass fraction of Nanoparticles, ρ bf and ρ np are densities of base fluid and Nanoparticles respectively. Density of the base fluid ρ bf represents the mixture of distilled water and Ethylene Glycol and is computed using the same Eqn (1).
For the heat capacity of the Nanofluid, Chow et al. [1] has quoted the following equation: Where nf p C is heat capacity of the Nanofluid, Thermal conductivity of the Nanofluid k nf is calculated from the empirical correlation derived by Corcione [15] Where thermal conductivity of the base fluid k bf represents the mixture of distilled water and Ethylene Glycol and is computed by Jamieson et al. [16]: Where k w and k Eth are thermal conductivities of water and Ethylene Glycol, ϕ w and ϕ Eth . are mass fractions of water and Ethylene Glycol respectively.
Re is the Nanoparticles Reynold's number and is defined by Where K b = 1.38066 × 10 -23 JK -1 is Boltzman constant, T nf is the Nanofluid temperature, d p is the Nanoparticle diameter, and dynamic viscosity of the base fluid μ bf represents the mixture of distilled water and Ethylene Glycol and is computed by Papaioannou et al. [17]: , T fr is freezing point of the base fluid which is -3.2°C for the 10 wt.% of Ethylene Glycol aqueous solution [18], and k np , x np are the thermal conductivity and volume fraction of the Nanoparticles respectively. Eqn (3) strictly represents a Nanofluid that contains solely one Nanoparticle substance in its mixture, hence the effective thermal conductivity of the Nanofluid will be the averaged thermal conductivities calculated for each Nanoparticles substance.
Dynamic viscosity of the Nanofluid μ nf is calculated from the empirical correlation derived by Corcione [15]: d bf is the equivalent diameter of the base fluid molecule given by Where M is the molecular weight of the base fluid, N= 6.022 × 10 -23 mol -1 is Avogadro number, and ρ bfo is the density of the base fluid calculated at T o =293K. Eqn (7) strictly represents a Nanofluid that contains solely one Nanoparticle substance in its mixture, hence the effective dynamic viscosity of the Nanofluid will be the averaged dynamic viscosities calculated for each Nanoparticles substance.

Computational model
Finite Element Modeling (FEM) and simulation was introduced in this work to help computing magnitudes of PV module surface temperature, structural part in the hybrid PV/T collector system, in accordance with day hours so to eventually evaluate thermodynamic and electrical efficiencies. Figure 6 illustrates a section drawing of Power Therm PV/T collector various components that are of interest in the computational analysis.
The conjugate heat transfer, laminar flow interface built in COMSOL Multiphysics ® FEM software allows analysis of the phenomenon of forced convection between a heat carrier fluid and the high-temperature PV cells as a heat source. The governing heat transfer equation for every previous and subsequent layer of the PV cells needs to be considered so to conclude to a model when entered and solved computationally by the software will accurately predict the PV cells temperature. Figure 7 shows the geometry considered for deriving the computational model and the major corresponding heat transfer modes in different directions.
The general heat energy balance equation would be: Heat transferred downwards in direction of thermal insulation layer Q = 0. The inward steady state heat transferred from the Sun is: Where G is the solar irradiation and A c is the PV/T collector surface area. The glazing radiated heat out of the system is formulated by [1] as: Where Q RAD,g-a is the heat transferred by radiation from the glazing surface to the surroundings, h RAD,g-a = 4σT a 3 (Ingersoll, 1986) is the radiative coefficient from the glazing, where     The glazing will also radiate heat inside the system and towards the PV cells again according to the formula [1]: Where Q RAD,g-c is the heat transferred by radiation from the glazing surface to the PV cell surface, and T g is the PV cells surface temperature.
PV cells surface will by its turn radiate heat in the upwards direction, represented by the equation (Oruc et al., 2016): Where Q RAD,pv is the upward heat transferred by radiation from the PV cells, and ϵ=0.85 is the average surface emissivity of silicon.
Summation of Q RAD,g-a , Q RAD,g-c , and Q RAD,pv will give the total heat transfer by radiation of the system Q RAD .
Heat transferred due to convection out of the system is [1]: Where Q CONV,g-a is the heat transferred by convection from the glazing surface to the surroundings, where v w is the wind speed.
Heat transferred due to convection towards the PV cells surface [1] , ( ) Where Q CONV,g-c is the heat transferred by convection from the glazing surface to the PV cells surface.
Summation of Q CONV,g-a , and Q CONV,g-c will give the total convection heat of the system Q CONV .
The general equation for steady state heat transfer within heat carrier fluid layer is stated by Yousefi et al.: Where m is the mass flow rate of the working fluid, C p is the heat capacity of the working fluid, T o and T o are the inlet and outlet temperatures, respectively. The forced convection is described using the continuity and momentum equations for laminar and incompressible fluid flow as follows [20]: The initial temperature of the PV cells was assumed to be derived from the equation [21]: So, to reduce number of iteration carried out by the software and to increase the accuracy of the resulting temperature.
Thermo-physical properties of each layer material and used in the COMSOL software are listed in Table 3.

Thermodynamic analysis
The conventional first law of thermodynamics thermal efficiency τ η as mentioned by Yousefi et al. [5] is: Where G is the solar irradiance and A c is the collector's surface area.
The electrical efficiency of the PV module mounted at the upper part of the PV/T hybrid collector can be represented by the simplified equation [22]: Where η e,ref is the PV module electrical efficiency at standard testing condition (STC), β temperature coefficient for cell efficiency and equals 0.0045°C -1 [23], and T ref is reference temperature at STC which is 25°C.
The total or overall efficiency of the PV/T hybrid collector η PV/T then is the automatic linear combination of the thermal and electrical efficiencies of the collector system according to Fujisawa and Tani [24]: The exergy term represents the convertible amount of low grade energy into useful one, and hence the exergetic efficiency (second law of thermodynamics efficiency) equation for the predefined quasi-static process would be [1]: Where T a is the ambient temperature, T o is outlet temperature of the fluid medium, and T SUN is the solar radiation temperature at 6000K.
For the electrical exergetic efficiency ε e of the PV part in the PV/T hybrid collector system the following equation is inclusive [1]: Thermal insulation GLASSWOOL +EPS ------------ And the total exergetic efficiency is:

Results
The following Figure 8 demonstrates magnitudes of PV cells temperature for each of the two heat transfer fluids obtained from the computational analysis.
Improvements in the cooling of PV cells due to use of Nanofluid is barely exists according to Figure 8 by an averaged reduction in temperature of 0.65°C. Figure 9 presents the electrical, thermal, and total efficiencies of the PV/T hybrid system for the heat transfer fluids used in this study. Figure 10 presents the electrical, thermal, and total exergetic efficiencies of the PV/T hybrid system for each of the two heat transfer fluids used in this study.
Figures 9 and 10 illustrates the variation in different efficiencies between the two heat transfer fluids. The effect of Nanofluid in the electrical efficiency is trivial when compared with the effect in the thermal efficiency, which complies with the general benefit of Nanofluid utilization which is the increase of thermal capability of the system at the first place. The following Table 4 summarizes the averaged increment in the relevant efficiencies.
Curve fitting was applied over the obtained data range to define the behavior of Nanofluid in compliance with the ambient temperature. Fig. 11 plots PV cells temperature difference between the two working fluids against the averaged ambient temperature during daytime testing. The points were splitted into two regions of interest and connected in form of descending lines.
The observed trend of Fig. 11 generally states a segmental decrease in the Nanofluid effect. The following possibilities were suggested to explain the illustrated behavior: Either a discontinuous function correlates the independent variables (PV cells temperature difference and ambient temperature) together, which splits around 12:00 PM to recontinue its dramatic descending direction, therefore the effect of Nanofluid is fluctuating with the increase in ambient temperature while keeps its declining manner.
Or it is a linear descending function, with some outliers which caused that noise in the graph.

Conclusions
The averaged PV surface temperature was obtained by means of finite element modeling software COMSOL Multiphysics for each working fluid scenario. The average increase in the electrical efficiency was less than 0.1% which is rational as it complies with the low concentrations of the Nanoparticles and it may considered neglectable. The traceable effect of Nanoparticles appeared in the averaged increase of thermal efficiency, which was around 4.1%. The increase in total exergy efficiency was also detectable by 4.6%, which presents a typical increase in the use of low grade energy.

Efficiency
Averaged increasment  Table 4: Averaged increment while using the Nanofluid.