Performance Evaluation of Carbon-Based Nanofluids for Direct Absorption Solar Collector

Abstract

In this study, carbon-based nanofluids (CBNFs) were prepared using a revised vortex trap method and applied in the direct absorption solar collector (DASC) to evaluate the feasibility of CBNFs in DASC. The thermal storage performance of water and different concentrations of CBNFs (0.01, 0.025 and 0.05 wt%) was assessed with a 1000 W halogen lamp as a simulated light source under different volumetric flow rates (1.5, 3.0, and 4.5 L per minute [LPM]) at a constant thermal storage load (2.4 kg of water) and ambient temperature of 26 °C. The thermal storage capacity, system efficiency factor (SEF), and heating rate of the CBNFs as the working fluid were higher than those of water in most cases. The thermal storage capacity and SEF of 0.05 wt% CBNF at a volumetric flow rate of 3.0 LPM were 10.36% and 9.36% higher than that of water, respectively. The relevant experimental results demonstrate the great potential of CBNFs in DASC.

1. Introduction

Nowadays, countries worldwide are committed to developing renewable energy and equipment to cope with global climate change and the greenhouse effect. Solar energy is one of the essential options among many renewable energy sources. In addition to using solar cells to generate electricity, solar thermal energy can also be used to generate electricity and obtain thermal energy [1,2]. Solar thermal collection systems can be broadly divided into two kinds: indirect and direct. The working fluid of indirect absorption solar collectors mainly focuses on the performance of heat conduction and convection. Aside from focusing on the heat conduction and heat convection performance of the working fluid, for direct absorption solar collectors (DASCs), the absorption characteristics of sunlight also need to be considered to achieve the optimal thermal collection efficiency [2,3,4].

With the development of nanotechnology, nanofluids, formed by adding nanoscale additives to traditional working fluids, have been widely studied for heat transfer, thermal collection, and thermal storage systems [5,6,7,8]. Carbon-based nanomaterials mainly include nanocarbon particles, carbon nanotubes, and graphene. Most of these materials have excellent chemical stability and thermal properties. In particular, carbon nanotubes and graphene have received great attention from researchers due to their large aspect ratio and special optical, mechanical, physical, and chemical properties. Carbon-based nanofluids (CBNFs) prepared from carbon-based nanomaterials have excellent heat transfer properties and are often used in heat exchange systems [9,10,11,12]. In addition, carbon-based nanomaterials have extremely broad spectrum absorption characteristics. Therefore, CBNFs have great potential for application in DASC [3,4,13,14].

Karami et al. [15] developed alkaline functionalized carbon nanotube nanofluids as absorbing fluids for DASCs and evaluated their dispersion and optical properties. They found higher extinction coefficients of the alkaline functionalized carbon nanotube nanofluids than the base fluids added at very low concentrations. The thermal conductivity of alkaline functionalized carbon nanotube nanofluid with 150 ppm concentration at 25 °C and 60 °C was 32.2% and 27.2% higher than that of water, respectively.

Gupta et al. [16] used different volume concentrations (0.001, 0.005, 0.01, and 0.05 vol%) of Al₂O₃-H₂O nanofluids to conduct thermal collection performance experiments in full-scale DASCs with a gross area of 1.4 m². They observed an instantaneous efficiency enhancement of 22.1%, 39.6%, 24.6%, and 18.75% for 0.001 vol%, 0.005 vol%, 0.01 vol%, and 0.05 vol%, respectively, which was comparable with water. The 0.005 vol% Al₂O₃-H₂O nanofluids had the optimal instantaneous efficiency and showed that the high concentrations of Al₂O₃-H₂O nanofluids did not necessarily have high instantaneous efficiency.

The photo-absorption properties and photothermal conversion performance of MWCNT-H₂O nanofluids were investigated by Qu et al. [17], who prepared 0.0015 wt% to 0.1 wt% multi-walled carbon nanotube (MWCNT)-H₂O nanofluids and conducted heating and light-irradiation cycling tests at temperatures below 90 °C. The temperature of the 0.01 wt% MWCNT-H₂O nanofluid was about 14.8 °C (or 22.7%) higher than that of deionized water after 45 min of light irradiation. The receiver efficiency decreased with an increase in the light irradiation time, and the highest value of the receiver efficiency of the 0.01 wt% MWCNT-H₂O nanofluid reached up to 96.4%.

Duan et al. [18] mixed Au nanoparticles of different shapes in water to form plasmonic blended nanofluids and studied them for direct solar absorption. Numerical and experimental investigation of the optical and thermal properties of plasmonic blended nanofluids revealed that the extinction spectrum of Au plasmonic blended nanofluids was broader than that of single-component nanofluids. The photothermal properties of three plasmonic blended nanofluids were experimentally measured. The blended nanofluid exhibited higher temperature rise and higher heat collection ability due to the higher extinction coefficient, and the numerical analysis results were consistent with the experimental results.

Wang et al. [19] studied the photothermal properties of titanium nitride nanofluids with thermal transfer oil as the base fluid under different solar radiation intensities. To overcome the low photothermal efficiency of DASCs, the reverse irradiation direct absorption solar collector (RI-DASC) was proposed to convert the heat transfer within the fluid from heat conduction to free convection; the temperature difference between the irradiated and non-irradiated surfaces using 0.005 wt% titanium nitride nanofluid was about 50 °C and 10 °C at 5000 kW/m² in DASC and RI-DASC, respectively. The photothermal conversion efficiency of the RI-DASC was about 10% higher than that of DASC, and the steady-state temperature of RI-DASC reached 165 °C.

Balakin et al. [20] prepared Fe₂O₃-H₂O nanofluids by dispersing 60 nm Fe₂O₃ nanoparticles in distilled water with concentrations ranging from 0.5 wt% to 2.0 wt%. Then, the Fe₂O₃-H₂O nanofluid was applied to the DASC equipped with a solenoid to study the heat collection efficiency under different magnetic field intensities. The 2.0 wt% Fe₂O₃-H₂O nanofluid obtained a velocity of about 5 mm/s and a maximum collector thermal efficiency of 65% under a magnetic field of 28 mT.

The above-mentioned studies show that nanofluids have been widely used as working fluids in solar thermal collection systems because of their excellent heat transfer performance. The nanofluids used in DASC were not necessarily CBNFs; however, in addition to considering the heat transfer performance of the working fluid, the optical absorption performance of the working fluid is also a key factor for the efficiency of solar thermal collection in the DASC. CBNFs have superior heat transfer performance and excellent optical absorption properties and thus should be the first choice as the working fluids of DASC. In this study, the revised vortex trap method (RVTM) was used to prepare CBNFs and used as working fluids for small-scale DASCs. The thermal collection and storage capabilities of DASC were determined at different flow rates of the working fluid to confirm the practicability of the CBNFs prepared in this study in a DASC.

2. Sample Preparation

CBNF was prepared using the manufacturing system with the RVTM (Figure 1) [21]. RVTM is modified from the vortex trap method (VTM) [11,22], and both use water mist to condense and collect carbon nanoparticles (NCs) produced by oxygen–acetylene combustion to prepare CBNFs. The main difference between RVTM and VTM is that RVTM is equipped with an external circulation device to improve the collection efficiency of NCs to improve the production rate of NCs. The flow rates of oxygen and acetylene were 0.5 L per minute (LPM) and 2.5 LPM, respectively. An external circulation device also promotes the mixing of the NCs and water to improve the dispersion performance of the CBNF. The morphology of NCs suspended in CBNF prepared by RVTM is shown in the transmission electron microscope (TEM; H-7100, Hitachi, Saitama, Japan) image in Figure 2. The mean particle size of NCs in CBNF was about 30 nm. The main component of NCs was amorphous carbon, and a small amount of graphene oxide and reduced graphene oxide were confirmed by X-ray diffraction analysis and Raman spectroscopy [22].

First, CBNF was prepared using RVTM, and 0.05 wt% CBNF was prepared via a concentration adjustment procedure. CBNF (0.05 wt%) was added with different concentrations of poly(sodium 4-styrenesulfonate) (PSS; Sigma-Aldrich, St. Louis, MO, USA) as a dispersant, and then the preliminary finished product was completed after continuous mechanical stirring and ultrasonic dispersion [21]. The stability test showed that CBNF prepared by RVTM had better stability than that of VTM, and only 0.01 wt% PSS needed to be added to maintain the static suspension performance for more than 30 days [11]. CBNF (0.05 wt%) was added to 0.01 wt% PSS, and the average suspended particle size and Zeta potential were measured using a dynamic light scattering (DLS) analyzer (SZ-100, Horiba, Japan). The average suspended particle size and Zeta potential were 198.4 nm and −87.2 mV, respectively, showing excellent suspension performance of the sample [23,24].

CBNF (0.05 wt%) with PSS (0.01 wt%) was mixed with water to finally prepare 0.025 wt% and 0.01 wt% CBNF, and then the samples were prepared by ultrasonic dispersion and mechanical stirring for subsequent DASC thermal storage performance tests. The weight ratio of NCs to PSS in these CBNFs was 5:1; sample numbers for water, 0.01 wt% CBNF, 0.025 wt% CBNF, and 0.05 wt% CBNF were S0, S1, S2, and S3, respectively. Figure 3 shows the actual image of the CBNFs after standing for 30 days, according to which these samples do not settle after 30 days of standing, indicating good stability of these samples. Figure 4 shows the absorption rate spectrum of water and CBNFs measured by a UV/VIS/NIR spectrometer (V670, Jasco, Tokyo, Japan) with a quartz cuvette (optical path: 1.0 mm) at wavelengths of 200–1800 nm. The range of wavelength covers more than 90% of the energy distributed in the solar spectrum. The absorption rate spectrum was obtained by 100%–T–R, where T and R are the transmittance and reflectance spectrums, respectively. Each sample was measured by the spectrometer with an integrating sphere (ISN-723, Jasco, Tokyo, Japan) to avoid the influence of the scattering effect of the sample on the correctness of the test results. The absorption rate was proportional to the concentration of NCs (Figure 4). The high absorption rate of the samples for the spectrum from the visible to near-infrared will help thermal storage in the DASC under solar radiation.

The fundamental characteristics of test samples such as viscosity (μ) and density (ρ) were measured using a viscosity meter (VL700-T15, Hydramotion, Malton, UK) with an isothermal bath (P20, Yotec, Taiwan) and a liquid density meter (DA-645, KEM, Kyoto, Japan), respectively. The μ and ρ of each sample were tested three times at different sample temperatures, and the average value was taken as the final experimental data.

Table 1. Fundamental characteristics of the test samples.

Items	Temp (°C)	S0	S1	S2	S3
ì (mPa-s)	30	0.77	0.80	0.88	0.90
	40	0.75	0.77	0.80	0.83
	50	0.67	0.70	0.72	0.75
	60	0.60	0.62	0.65	0.67
ñ (kg/m3)	30	995.78	995.83	995.89	995.91
	40	992.31	993.03	993.10	993.11
	50	987.14	988.08	988.21	988.22
	60	982.34	983.33	983.43	983.45

presents the measurement results of the μ and ρ of water and CBNFs at different sample temperatures. The differences between each sample were very small for μ and ρ due to the extremely low concentration of NCs and PSS in CBNFs. Generally, the μ of the solid–liquid mixture is higher than that of the base liquid. According to the theory of the solid–liquid mixture, if the ρ of the additive is higher than those of the base liquid, the ρ of the mixed liquid will be higher than that of the base liquid; otherwise, it will be less than the base liquid. Table 1 shows a higher value of the μ of CBNFs than that of water, and the μ increased with the increase in the added concentration of NCs, which is in line with the expectations. In addition, the addition of PSS to water also increased the μ of water. The specific gravity of most carbon-based materials was between 1.6 and 2.4. Therefore, the ρ of the CBNFs and PSS was higher than that of water, which led to the increase in the ρ of the CBNFs with an increase in the added concentration of the CBNFs and PSS. The test results showed a highly significant influence of the sample temperature on μ. However, the effect of the sample temperature on ρ was very slight.

3. Experiment Design and Procedures

3.1. Design of Direct Absorption Solar Collector

Figure 5 shows the detailed dimensions and actual images of the DASC. The DASC base was made of an Al-6061 aluminum alloy by a milling machine. The prepared DASC base was sandblasted to reduce the reflectivity of the surface and improve the absorption efficiency of sunlight. The abrasive used in sandblasting was brown alumina with a particle size of 106–125 μm. The surface of the DASC base before and after sandblasting was examined by a fiber optic spectrometer (BRC112E-V Quest^TMX, B&W TEK, Plainsboro, NJ, USA); the average reflectivity of the surface for the non-sandblasted and sandblasted DASC base was 94.0% and 34.2% at wavelengths of 350–1000 nm, respectively. Therefore, the average reflectivity of the sandblasted DASC base was 63.7% lower than the non-sandblasted. Next, the glass sheet (thickness: 1.10 ± 0.05 mm; FL055L4B, NIPPON SHEET GLASS, Tokyo, Japan) was bonded to the upper layer of the DASC base with silicone glue to form a flow channel that can absorb sunlight. Finally, five thermocouples were installed in the DASC to measure the temperature of the working fluid at different locations.

3.2. Experimental System and Procedures

Figure 6 and Figure 7 show the experimental system diagram and actual photo of the DASC for thermal storage performance. A 1000 W halogen lamp was used as the light source to simulate sunlight. There are some differences in the spectrum of halogen lamps and sunlight, and this issue will be discussed in detail at the end of this paper. The distance between the illuminating surface of the luminaire and the center of the upper surface of the DASC was 31 cm. The illumination intensity at the center of the upper surface of the DASC was set to 900 ± 10 W/m² and was calibrated using a pyranometer (MP-200, Apogee Instruments, Logan, UT, USA) in each experiment. The DASC was placed under the simulated light source at a horizontal slope of 23 degrees. The temperature of the working fluid was measured by installing a total of five thermocouples (T₁~T₅; T-type, accuracy: ±0.75%) at the inlet, outlet, and flow channel of the DASC. The thermal storage tank was made of polyvinyl chloride pipe (ID: 110 mm, length: 315 mm), and the outside was covered with a 1.2 cm thick polyurethane insulation material for thermal insulation. The interior of the thermal storage tank was filled with 2400 ± 1.0 g of water as a thermal storage load, and weighed using an electronic balance (GX6100, A&D, Tokyo, Japan). In each experiment, the water used as the thermal storage load was first controlled by a constant temperature water tank at 26 ± 0.5 °C to maintain the initial temperature of the thermal storage load. A coil (coil turns: 12, coil diameter: 80 mm) made of copper pipe (OD: 3/8”) inside the thermal storage tank provided a heat exchange between the working fluid and the thermal storage load. Two thermocouples (T₈ and T₉; T-type, accuracy: ±0.75%) were installed in the thermal storage tank to measure the water temperature at different positions; the average water temperature of the thermal storage load (T_L) was averaged. The filling volume of the working fluids for water (S0) and CBNFs (S1~S3) was 1200 mL. The working fluid was transported by a water pump (DC 12 V), and the flow rate of the working fluid was controlled at 1.5, 3.0, and 4.5 LPM by a flow controller (FCV-10, Regal Joint, Sagamihara, Japan) with a control voltage signal. The DASC heat storage performance experiment was in the environmental control room at room temperature (25 ± 0.5 °C), and a thermocouple (T₁₀; T-type, accuracy: ±0.75%) was set near the reservoir to monitor the change in ambient temperature. A datalogger (TRM-20, TOHO, Sagamihara, Japan) was used to measure and record the temperature and flow rate data, and a digital power meter (WT230, Yokogawa, Tokyo, Japan) was used to measure and record the power consumption of the water pump. The datalogger and digital power analyzer recorded data at 5 s intervals. The testing time for each experimental parameter was 40 min. The same experimental parameters of each sample were repeated thrice and the average values were considered as the final experimental data to improve the stability of the experimental results. The details of all of the measuring instruments, sensors, and equipment are shown in

Table 2. List of measuring instruments, sensors, and equipment.

Item	Brand	Model	Specifications	Note
Thermocouple	Jetec	T-type	Accuracy: ±0.75% (0–350 °C)	According to ASTM E230-1998, E988-1996
UV/VIS/NIR spectrometer	Jasco	V670	Wavelength range: 190–2700 nm; Accuracy: ±0.3 nm (at 656.1 nm) and ±1.5 nm (at 1312.2 nm); Repeatability: ±0.05 nm (UV–Vis) and ±0.2 nm (NIR)	Wavelength range of quartz cuvette: 200–2500 nm
Integrating sphere	Jasco	ISN-732	Wavelength range: 200–2500 nm; Typical spectral bandwidth: 5 nm (UV)/20 nm (NIR); Inside diameter of integrating sphere: 60 mm	This is a measurement accessory for V670
Viscosity meter	Hydramotion	VL700-T15	Accuracy: ±1.0%	The sample temperature is controlled by an isothermal bath (P-20)
Isothermal bath	Yotec	P-20	Accuracy: ±0.5 °C	Setting range deviation: 30 ± 0.5 °C (±1.67%)
Liquid density meter	KEM	DA-645	Accuracy: ±0.0005 g/cm3 (density) and ±0.03 °C (temperature)	Maximum density deviation: 0.98234 ± 0.0005 g/cm3 (±0.05%) Maximum temperature deviation: 30 ± 0.03 °C (±0.10%)
Fiber optic spectrometer	B&W TEK,	BRC112E-V QuestTMX	Wavelength range: 350–1050 nm; Optical resolution: 0.5 nm	Integrating sphere to measure reflectivity
Pyranometer	Apogee Instruments	MP-200	Spectral range: 360~1120 nm; Measurement repeatability: less than ±1%	Setting range deviation: 900 ± 10 W/m2 (±1.11%)
Electronic balance	A&D	GX6100	Capacity: 6100 g; Accuracy: ± 0.1 g; Resolution: 0.01 g; Repeatability: 0.01 g	Sampling deviation range: 2400 ± 1.0 g (±0.04%) Accuracy: 2400 ± 0.1 g (±0.004%)
Flow controller	Regal Joint	FCV-10	Flow range: 0–10 LPM; Accuracy: ±3.0%	0–10 V corresponds to 0–10 LPM
Environmental control system	—	—	Room temperature: 25 ± 0.5 °C (±2.0%)
Datalogger	TOHO	TRM-20	Accuracy: ±0.1%	Sensor: T-type thermocouple
Digital power meter	Yokogawa	WT230	Power accuracy: ±0.2%	Direct input

.

3.3. Performance Estimation of Direct Absorption Solar Collector for Thermal Storage

This study focused on the effect of CBNFs on the thermal storage performance of DASC and proposed three performance indicators such as thermal storage capacity (Q_TS, W), system efficiency factor (SEF, W/W), and heating rate (HR, °C/min) to evaluate the heat storage performance. The thermal storage load used in the above-mentioned performance indicators refers to the temperature difference (ΔT_L) between the average temperature of the thermal storage load in the initial three minutes and the average temperature of the thermal storage load in the three minutes before the end of the experiment. The test time for each experimental parameter was 40 min, and the test time (t, s) corresponding to the ΔT_L was 37 min (2220 s).

The Q_TS (W) is shown in Equation (1), where M is the load mass (2.4 kg of water), c_p is the specific heat of water (4.18 kJ/kg °C), and t is the test time (2220 s). The c_p of the thermal storage load (water) is regarded as a constant value (4.18 kJ/kg °C) because the temperature change in the thermal storage load has very little effect on its c_p within 15 °C.

Q_TS = 1000 × (M × c_p × ΔT_L)/t

(1)

The SEF can be expressed as Equation (2), where P_p is the power consumption of the pump (W). SEF represents the thermal storage provided by the unit P_p, which has the meaning of system efficiency. The P_p refers to the average value of the pump power consumption during the experiment.

SEF = Q_TS/P_p

(2)

The HR is shown in Equation (3). From Equations (1) and (3), the trend of HR was the same as that of Q_TS.

HR = 60 ΔT_L/t

(3)

The relative uncertainty analysis performed in this study involved the calculation of deviations (e) of the related facility for the experiment. The details for each of the measuring instruments, sensors, and equipment can be seen in Table 2. According to Equation (4) for the standard uncertainty analysis [25], the range of relative uncertainty (U_e) of the physical quantities and performance indicators of each experiment are described below.

The deviation of the μ measurement came from the viscosity meter and isothermal bath, and the U_e of μ was ±1.95%. The deviation of the ρ measurement came from the density measurement unit and temperature control unit of the liquid density meter, and the U_e of the ρ was ±0.12%.

The common deviations of all performance indicators in the thermal storage experiment included the setting range of the simulated light source, the pyranometer, the sampling range of the load weighing, the electronic balance, flow controller, and the ambient temperature. The U_e of the common deviation was 3.91%. The U_e of Q_TS was calculated as 4.05% by the common deviation plus the deviation of the two thermocouples and the corresponding datalogger. The U_e of P_p was calculated as 3.91% by the common deviation plus the deviation of the digital power meter. The U_e of SEF was calculated as 4.05% by the common deviation plus the deviation of the digital power meter, the two thermocouples, and the corresponding datalogger. The U_e of HR was calculated as 4.05% by the common deviation plus the deviation of the two thermocouples and the corresponding datalogger.

U_e = [(e₁)² + (e₂)² + … … … + (e_n)²]^1/2 × 100%

(4)

The difference in the thermal storage performance between CBNFs and water was more clearly presented by using the percentage difference (PD). The PD between the experimental data of water (D_w) and those of the CBNFs (D_CBNFs) is expressed as Equation (5):

PD = [(D_CBNF − D_w)/D_w] × 100%

(5)

4. Results and Discussion

Figure 8 shows the test results of the temperature difference (ΔT_DASC) for the working fluid at the inlet and outlet of the DASC at different flow rates. The larger the ΔT_DASC, the better the thermal collection capacity of the DASC. The ΔT_DASC of each test sample decreased with the increase in the flow rate, and the thermal collection capacity of CBNFs in the DASC was higher than that of water. The thermal collection capacity of higher concentrations of CBNFs (S2 and S3) was significantly higher than that of water, mainly due to the higher light absorption capacity of higher concentrations of CBNFs (refer to Figure 4). The lowest concentration of S1 was close to the thermal collection capacity of water due to its lower light absorption rate compared with that of S2 and S3.

Figure 9 shows the load temperature (T_L) of the thermal storage for the DASC at various flow rates. The experimental results of T_L show that the T_L of the CBNFs was higher than that of water, suggesting that CBNFs have a higher thermal storage capacity than water. Although the initial temperature of the thermal storage load was controlled at 26 ± 0.5 °C, there were still some differences. Therefore, the difference in the thermal storage capacity of each test sample needs to be evaluated by the performance indicators calculated using Equations (1)–(3).

Figure 10 shows the thermal storage capacity (Q_TS) of the DASC system at various experimental parameters. While the Q_TS of S1 at 1.5 LPM was slightly lower than that of water, the Q_TS of CBNFs was higher in other experimental parameters. In addition, Q_TS showed a similar change trend at the flow rate of 3.0 LPM and 4.5 LPM. However, the changing trend of Q_TS at 1.5 LPM is different from that at 3.0 LPM and 4.5 LPM. This phenomenon could be caused by the fact that the flow controller is already close to the lower limit of the flow control capability at 1.5 LPM, which makes the flow control less stable. The percentage differences (PD) between the Q_TS of S1, S2, and S3 compared with water at a flow rate of 1.5 LPM were −1.42%, 6.42%, and 4.79%; the PD with water at a flow rate of 3.0 LPM was 2.09%, 7.33%, and 10.36%; the PD with water at a flow rate of 4.5 LPM was 2.64%, 3.37%, and 6.42%, respectively. Overall, the CBNFs exhibited better thermal collection capacity and Q_TS in the DASC system.

Figure 11 shows the power consumption of the pump (P_p) of the DASC system at various experimental parameters. While the P_p of S1 at 4.5 LPM was slightly lower than that of water, the P_p of CBNFs was higher in the other experimental parameters. In general, the higher viscosity and density of CBNFs resulted in higher P_p when circulating in the DASC system. Therefore, the slightly lower P_p of S1 than that of water at 4.5 LPM could be due to the experimental deviation. Furthermore, the density and viscosity of S3 were higher than those of S2, but the P_p under each flow rate was slightly lower than that of S2, which was mainly due to the experimental deviation and the characteristics of the pump itself. The PD between the P_p of S1, S2, and S3 compared with that of water at a flow rate of 1.5 LPM was 0.24%, 0.70%, and 0.59%; the PD with water at a flow rate of 3.0 LPM was 0.47%, 1.10%, and 0.93%; the PD with water at a flow rate of 4.5 LPM was −0.38%, 1.04%, and 0.80%, respectively. The above-mentioned results showed that the use of CBNFs instead of water as the working fluid in the DASC system only increased the P_p very slightly.

Figure 12 shows the system efficiency factor (SEF) of the DASC system at various experimental parameters. While the SEF of S1 at 1.5 LPM was lower than that of water, the SEF of CBNFs was higher than in other experimental parameters. The PD between the SEF of S1, S2, and S3 compared with that of water at a flow rate of 1.5 LPM was −1.65%, 5.68%, and 4.17%, respectively; the PD with water at a flow rate of 3.0 LPM was 1.61%, 6.17%, and 9.34%, respectively; the PD with water at a flow rate of 4.5 LPM was 3.03%, 2.30%, and 5.57%, respectively. Overall, the CBNFs exhibited better SEF in the DASC system.

Figure 13 shows the HR of the load for the DASC system at various experimental parameters. While the HR of S1 at 1.5 LPM was lower than that of water, the HR of the CBNFs was higher in the other experimental parameters. The PD between the HR of S1, S2, and S3 compared with water at a flow rate of 1.5 LPM was −1.42%, 6.42%, and 4.79%, respectively; the PD with water at a flow rate of 3.0 LPM was 2.09%, 7.33%, and 10.36%, respectively; the PD with water at a flow rate of 4.5 LPM was 2.64%, 3.37%, and 6.42%, respectively. Overall, the CBNFs exhibited higher HR due to the higher thermal collection capacity in the DASC system.

The above-mentioned performance indicators such as Q_TS, SEF, and HR show that except for a slightly lower S1 than water at a volumetric flow rate of 1.5 LPM, the CBNFs were higher than for water under other experimental parameters. CBNFs were only required at a very low concentration in the DASC systems to have a higher thermal collection and storage performance than water. Therefore, the CBNFs have great potential to replace water as the working fluid of DASC.

Figure 14 shows the normalized spectra of the sun and halogen lamps [26]. It can be seen from the figure that the energy ratio of the solar energy distributed in the range of wavelengths less than 700 nm was higher than that of the halogen lamps. The absorption rate (refer to Figure 4) of the CBNFs in the wavelength spectrum less than 700 nm was relatively high. Therefore, if the CBNFs in this experiment are used under sunlight, their performance can reasonably be considered better than that under the radiation of the halogen lamp.

5. Future Work

In this study, many percentage differences (PD) between water and CBNFs for performance indicators were within uncertainties, which reduced the credibility of these data. In the uncertainty analysis, it was found that the flow controller is the main factor leading to high uncertainty. Therefore, the use of higher-accuracy flow controllers should be considered in follow-up studies to further reduce the uncertainty of the experiment. Furthermore, the simulated light source still had considerable differences from the solar spectra. A simulated light source closer to the solar spectra will be used for follow-up research to improve the practicability of the experimental results in the future.

6. Conclusions

The CBNFs in this study were prepared by RVTM, and the practicability of CBNFs in the field of solar thermal collection was evaluated through characteristic measurements and the DASC system experiments. The relevant research results can be summarized in six points as follows.

• CBNFs have higher viscosity and density compared to water. In terms of the overall trend, the viscosity difference between the CBNFs and water decreased, while the density difference between CBNFs and water increased with the increase in the sample temperature.
• The optical absorbance is proportional to the concentration of CBNF. Furthermore, except for the spectrum range of 200–250 nm, the absorbance decreased with the increase in wavelength.
• The Q_TS, SEF, P_p, and HR of CBNFs as working fluids were higher than those of water in most cases.
• The use of CBNFs in the DASC system will have a slightly higher P_p than that of water, and the increase in P_p will not exceed 1.10%.
• The Q_TS and SEF of CBNFs under the optimal conditions were 10.36% and 9.36%, respectively, higher than that of water at a volumetric flow rate of 3.0 LPM.
• CBNFs are needed at a very low concentration to have a higher thermal collection and storage performance than those of water in the DASC systems. Therefore, CBNFs have great potential to replace water as the working fluid of a DASC.