Elsevier

Renewable Energy

Volume 151, May 2020, Pages 575-588
Renewable Energy

Thermal characteristics of evacuated tube solar collectors with coil inside: An experimental study and evolutionary algorithms

https://doi.org/10.1016/j.renene.2019.11.050Get rights and content

Highlights

  • Experimental evaluation of the evacuated tube solar collector with coil inside.

  • Using Multi-variate Adaptive Regression Spline method for performance prediction.

  • Utilizing Gene-Expression Programming method for the thermal performance prediction.

  • Applying Model Tree method to performance prediction of the tubular solar collector.

  • Presenting new formulas for the energy efficiency and water temperature difference.

Abstract

In this paper, the thermal characteristics of an evacuated tube solar collector for different volumetric flow rates of the fluid (10, 30 and 50 l/h) was experimentally improved by using copper oxide/distilled water (Cu2O/DW) nanofluid, and parabolic concentrator. Moreover, the effect of different volume fractions of the utilized nanofluid on the fluid properties, such as convective heat transfer coefficient, Nusselt number, and the useful gain of the collector was experimented. Finally, three artificial intelligence (AI) techniques namely, multi-variate adaptive regression spline (MARS), model tree (MT) and gene-expression programming (GEP) have been employed to predict the energy efficiency (ηІ) and inlet-outlet water temperature difference (ΔT). The input variables were volume of the storage tank (V), volume fraction of the nanofluid (VF), and mass flow rate of the fluid (M˙). The proposed AI methods presented robust formulations for prediction of ηІ and ΔT with an acceptable level of precision. The statistical results of AI models demonstrated that the MARS method can make a more accurate prediction of the collector performance than GEP and MT. It was also concluded that increase in both flow rate, and concentration of the nanofluid, lead to an increase in the thermal performance of the solar collector.

Introduction

“Nano” and “Energy” are two nationwide compelling keywords in the modern world. On the other hand, renewable energy has become of great significance due to considerable plunge in fossil fuels. Solar energy, as one of the most well-known renewable energy resources, can be substituted for many thermodynamic industrial applications, such as solar collectors and batteries [1,2]. Although there exist some limitations in terms of stability, and cost of using nanofluids, it is extensively applied to a large number of heating and cooling solar systems, such as combined cooling, heating and power (CCHP) stations, water heating systems, heat pumps, desalination systems, in order to enhance the efficiency level of solar systems [3,4].

Nowadays, the use of nanofluids, such as aluminum oxide (Al2O3) and copper oxide (Cu2O) in thermal systems has become a reliable approach so as to enhance the thermo-physical characteristics of the working fluid inside the thermodynamic systems [5,6]. Furthermore, Sadeghi et al. [7] improved the thermal efficiency level of an evacuate tube solar collectors (ETSC) by means of Cu2O/distilled water nanofluid and parabolic concentrator up to 11%. MA Sharafeldin et al. [8] used WO3/water nanofluid to enhance efficiency of an ETSC experimentally. The results demonstrated that the useful heat gain of the collector rised up to 23%, and the efficiency of the proposed ETSC was reported as 72.83%. H. Kaya et al. [9] applied ZnO/Etylene glycol-pure water nanofluid at different volume concentrations to a U-Tube solar collector. It was reported that this type of nanofluid can increase the thermal efficiency of the solar collector up to 5.2% compared to using the base fluid as the working fluid. I.M. Mahbubul et al. [10] increased the thermal efficiency of an ETSC up to 10% using carbon nanotube nanofluid for 0.2 vol fraction of nanoparticles (VF). Jowzi et al. [11] enhanced the efficiency level of an ETSC of about 12% by means of a bypass tube connecting the bottom of the thermal storage tank to the bottom of the evacuated tube in order to eradicate the stagnant zone at the bottom of the evacuated tube. Additionally, Sadeghi et al. [12] carried out experimental studies to investigate the influence of argan and air gases between cover and absorber coil in a tubular solar collector at various rates of mass flow. The results indicated that the optimum mass flow rate obtained 3.5 kg/h and argon gas presented higher level of efficiency due to less Prandtl number. McEnaney et al. [13] modelled a new form of tubular receiver using aerogels between the two glasses. They concluded that the modified type of solar receiver performs better than the evacuated tubes. In 2014, Mahian et al. [14] mathematically analyzed the effect of volume fraction of nanoparticles in the base fluid, and the shape of nano materials on second law efficiency of thermal solar systems. Goudarzi et al. [15] studied the effects of nanofluids pH on the performance of solar collectors. It was found that changing pH of nanofluids can tangibly influence thermo-physical properties of the fluid, and they could report the 52% efficiency.

According to the optimum design of solar systems, obtaining pieces of information pertained to thermal properties is necessary. This is highly recommended to evaluate efficiency level of various kinds of solar systems by means of thermal characteristics. Basically, it can be noted that there are a few mathematical expressions to assess efficiency level of solar systems. In the case of experimental studies, empirical equations may produce inaccurate predictions due to lack of facilities and limited ranges of variables. In contrast to regression based equations, another type of predictive tools, known AI approaches, have been employed in order to assess thermal performance of solar systems. AI techniques were applied to appraise the performance of various solar systems. According to application of AI technique, artificial neural networks (ANNs), adaptive neuro-fuzzy inference system (ANFIS), least square-support vector machine (LS-SVM), M5 model tree (MT), and multivariate adaptive regression splines (MARS) were applied to assess operation of solar systems. Results of previous studies proved that applications of AI models have provided more accurate predictions in comparison with regression based equations. Longo et al. [16] predicted the thermal conductivity of the group of oxide-water nanofluids using artificial neural networks technique. They found that the proposed ANN model could forecast precise the amount of thermal conductivity of the nanofluids regarding the average size of the nanoparticles.

In this study, the use of AI techniques is considered to perceive mechanism of ETSCs. In this way, a laboratory analysis is conducted on the thermal characteristics of a constructed ETSC with parabolic concentrator and additionally, the use of copper oxide/distilled water nanofluid for different mass flow rates is considered. Afterwards, energy efficiency and inlet-outlet water temperature difference for the ETSC are formulated by using the three well-known AI methods, namely gene-expression programming (GEP), MARS, and M5MT. Furthermore, the accuracy level of formulations is compared with the experimental observations. Ultimately, the physical meaning of experimental observations in terms of the effect of parabolic concentrator on the diffuse irradiance is examined.

Section snippets

The preparation procedure

In this case, in order for preparation of the Cu2O/DW nanofluid with 0.04 and 0.08 vol concentrations, the two-step technique is more efficient compared with the one-step one in dwindling sedimentation, and in rising the dispersal behavior [17,18]. In the present procedure, polyvidone (PVP–K90) has been considered to the prepared solution used as the surfactant, and the nanoparticles were scattered by the magnetic stirrer. In the next step, the nanofluid was set under an Ultra-Sonicator device,

Experimentation procedure description

A scheme of the adopted experimental procedures with a concise explanation of each part is indicated in Fig. 7. The experimentations were carried out in Kermanshah at longitude 34.3 oE and latitude 46.7 oN in August in 2018. The nanofluids were prepared for different volume concentrations of 0.4 and 0.08 VF. As illustrated in Fig. 8, the ETSC setup contains three WGETSC evacuated tubes, seven storage tanks with various volumes of 10, 20, 25, 30, 40, 50, and 60 l, a curved mirror as the

First law efficiency

The first law efficiency of the proposed collector could be obtained via Eqs. (1), (2) [20].ηІ=QuAcG=M˙Cp(Tf,outTf,in)AcGηІ=FR(τα)effFRUL(Tf,inTaG)where (τα)eff is the multiplication of collector effective transmittance-absorptance. The specific heat of the nanofluid at different volume fractions was obtained from Ref. [21]:Cp,nf=[ϕρnpCp,np+(1ϕ)ρbfCp,bf]ρnfρnf is the nanofluid density derived from Ref. [21]:ρnf=ϕρnp+(1ϕ)ρbf

In addition, Table 2 lists the thermodynamic characteristics of the

An overview of MARS algorithm

Multivariate adaptive regression splines (MARS) technique formulated as a non-parametric mathematical model is able to transform the governing equations among Input-output systems. In MARS model, pattern recognition is performed by means of setting a number of basis functions (BFs) and their corresponded weighted coefficients [26,27]. Mathematically, the MARS model is defined as,|xt|+=max(0,xt)={Δxtx>tΔ0xt|tx|+=max(0,tx)={Δtxx<tΔ0xtin which t is indicative of knot. BFs are occasionally

Accuracy of the measurement devices

Since, there are many errors while conducting the experiments, such as data mining, the uncertainty investigation of all parameters should be carried out to verify the applicability of the experimentations. The uncertainty equations related to validation of thermal characteristics (energy and exergy analyses) of the constructed ETSC are represented as follows:ΔηІ=[(ηІρfΔρf)2+(ηІTiΔTi)2+(ηІTfiΔTfi)2+(ηІGΔG)2]12ΔηІІ=[(ηІІTaΔTa)2+(ηІІTiΔTi)2+(ηІІTfiΔTfi)2+(ηІІGΔG)2]12

In Eq. (19), ρf

Results and implementations

The experiments have been conducted for different fractions of nanoparticles, but the results are presented just for 0.04 and 0.08 VF of the nanofluid for summarizing the findings, and extracting the trend. The results have been divided into two parts; experimental and numerical investigations. In this section, first of all the effects of nanofluid on the thermal characteristics of the fluid have been studied, then the performance of proposed AI techniques are compared and eventually the

Conclusion

This paper presented some findings about effect of using Cu2O/DW nanofluid on improving the thermodynamic efficiency of the ETSCs. The experimentations were carried out for three flow rates of the fluid passing through the coil. Moreover, three powerful AI-based formulations have been utilized to assess thermal performance of the ETSC system. The predictive AI models were fed by three input variables as, tank volume, nanofluid concentration, and mass flow rate of the fluid. Additionally, two

Author contribution

I declare that all the authors had a significant scientific contribution to the paper, and all the contents of this paper have been shared with all authors. The roles of all authors are listed as follows:

• Gholamabbas Sadeghi: constructing the ETSC, preparation of the nanofluid, conducting the experiments, data mining, writing the entire paper.

• Dr. Mohammad Najafzadeh: conducting robust three AI techniques, extracting the AI-based results.

• Prof. Mehran Ameri: verification of the results,

Declaration of competing interest

I declare no conflict of interest, and my agreement for submission of this manuscript and I claim that this work is novel and has not been submitted elsewhere.

References (34)

Cited by (50)

  • A parametric study on the efficiency of a solar evacuated tube collector using phase change materials: A transient simulation

    2022, Renewable Energy
    Citation Excerpt :

    So, some researchers attempted to reveal the influence of nanofluids on the ETC system operation [14–16]. The exergy and energy performance of a nanofluid-based ETC device is evaluated by Sadeghi et al. [17]. In this research, pure water, Cu2O/water (4% vol.), and Cu2O/water (8% vol.) are considered as the working fluid in the ETC system.

  • The spatial and temporal mismatch phenomenon in solar space heating applications: status and solutions

    2022, Applied Energy
    Citation Excerpt :

    High-efficient solar harvesting devices can reduce both land occupation and maintenance costs. For the non-concentrating solar collectors (such as flat plate solar collector [40] and evacuated tube solar collector [41]) that are commonly used in solar heating projects, the general method to improve their efficiency is to reduce the multiple kinds of thermal losses (such as the thermal convection loss, radiant loss, and reflection loss) in the whole solar harvesting process. In light of different methods for reducing the heat loss, it can be categorized as thermal loss recycling in the solar harvesting process and photothermal conversion enhancement via the source-directly measures.

View all citing articles on Scopus
View full text