1. Introduction
In response to the global decarbonization strategy and the urgent need for renewable energy development, solar photovoltaic systems have attracted significant research attention. The international energy landscape has been transformed considerably over the past decade, with renewable energy sources accounting for an ever-increasing share of global power generation. According to the BP Statistical Review of World Energy 2024, global primary energy consumption increased by 2% in 2023 compared to 2022, with renewable energy comprising 14.6% of primary energy consumption. Meanwhile, fossil fuels still contributed to 60% of global electricity generation. In China, coal consumption has remained a dominant yet problematic energy source, emphasizing the critical importance of accelerating renewable energy adoption.
Solar energy stands out as one of the most promising renewable resources due to its abundance, sustainability, and environmental friendliness. China possesses vast solar energy resources, with regions categorized into four classes based on annual solar radiation levels, ranging from 4200 MJ/m² to 8400 MJ/m². Table 1 summarizes the distribution of these solar resource regions. Given the substantial solar energy potential, enhancing the efficiency of solar energy conversion devices becomes essential.
| Region Category | Resource Level | Regions | Annual Total Radiation (MJ/m²) |
|---|---|---|---|
| 1 | Very Rich | Tibet, Xinjiang, Ningxia, etc. | 6680–8400 |
| 2 | Relatively Rich | Inner Mongolia, Qinghai, Gansu, etc. | 5850–6680 |
| 3 | Rich | Shaanxi, Shandong, Henan, etc. | 5000–5850 |
| 4 | Moderate | Jiangsu, Hunan, Anhui, etc. | 4200–5000 |
While solar energy offers clear advantages as a renewable resource, its practical application faces constraints including low energy density, intermittency, and system costs. Currently, solar utilization technologies can be classified into photothermal conversion, photovoltaic (PV) conversion, photovoltaic/thermal (PV/T) combined utilization, and photochemical conversion. Among these, the PV/T system has demonstrated considerable promise because it simultaneously generates electricity and captures thermal energy, thereby improving the overall solar energy conversion efficiency.
One of the fundamental challenges limiting PV system performance is the temperature dependence of photovoltaic conversion efficiency. Experimental studies have shown that crystalline silicon solar cells lose approximately 0.5% of their relative efficiency for every 1°C rise in temperature. Passive cooling and active liquid cooling are commonly employed to mitigate this thermal issue. In particular, liquid-cooled cold plates attached to the backside of solar panels have proven effective. To further boost heat transfer, nanofluids—colloidal dispersions of nanoparticles in base fluids—have emerged as a highly effective cooling medium due to their superior thermal conductivity. This study aims to systematically investigate the coupled effects of a spiral-coil vortex-generator cold plate and Al₂O₃-MWCNT hybrid nanofluids on the thermal and electrical performance of a PV/T system.
2. Experimental System and Methodology
2.1 Overview of the PV/T Experimental Test Platform
I constructed a complete indoor PV/T experimental system comprising a solar simulator, photovoltaic module, liquid cooling plate, thermal storage tank, circulation pump, flow meter, stop valves, data acquisition unit, temperature sensors, and piping. The schematic architecture of the system is described here. The system integrates two functional loops: a water circulation loop for thermal energy collection and a photovoltaic generation loop for electrical energy output. During operation, the cooling fluid circulates through the cold plate attached to the backside of the solar panel, absorbing heat and transporting it to a storage tank, thereby cooling the panel while storing thermal energy.
The photovoltaic module was a laminated structure: a 2.5 mm thick glass cover, a PV layer with an effective cell area of 0.2358 m², a 0.5 mm thick thermally conductive silicone pad with a thermal conductivity of 13 W/(m·K), an aluminum liquid cooling plate, and a 10 mm thick insulation layer. The PV module has a rated maximum power of 50 W, with an operating voltage of 18 V and a current of 2.77 A. A constant amount of cooling water was maintained in the storage tank to ensure stable thermal capacity during testing. All experiments were conducted in the evening to eliminate the influence of natural sunlight.
2.2 Solar Simulator Design
To simulate solar radiation, I selected three 1000 W tungsten-halogen lamps with a color temperature of 3000 K, positioned at a distance of approximately 0.7 m from the PV panel surface. Table 2 reports the measured radiation intensity at different distances. For the standard testing condition (STC) specified by the International Electrotechnical Commission (IEC), the irradiance should be 1000 W/m². By adjusting the distance from the lamp array and the applied voltage, I was able to continuously regulate the radiation intensity from 0 to 1200 W/m², which accurately replicates the solar conditions typical of Changchun summer daytime irradiance.
| Distance from lamp (cm) | Radiation intensity (W/m²) |
|---|---|
| 100 | 684 |
| 90 | 806 |
| 80 | 921 |
| 70 | 1003 |
| 60 | 1129 |
| 50 | 1356 |
2.3 Design of the Spiral-Coil Vortex Generator Cold Plate
To enhance heat exchange performance, I designed a liquid cooling plate integrated with internal spiral coil vortex generators. The cold plate was fabricated from aluminum, with dimensions of 260 mm × 160 mm × 13 mm, and contained four circular flow channels each with an inner diameter of 7 mm. The spiral coils inserted into the channels served as turbulence promoters, converting laminar flow into turbulent flow and thus enhancing convective heat transfer. Two cooling plates were connected in series to adequately cover the backside area of the solar panel. The key structural parameters of the cold plate are listed in Table 3.
| Parameter | Value |
|---|---|
| Length × Width × Height (mm) | 260 × 160 × 13 |
| Flow channel cross-section | Circular |
| Inlet/outlet outer diameter (mm) | 11 |
| Flow channel inner diameter (mm) | 7 |
| Thermal conductivity (W/(m·K)) | 15 |
| Number of flow channels | 4 |
2.4 Instrumentation and Measurement
All experimental data were acquired using calibrated instruments. Type K thermocouples (−200 to 1200 °C, accuracy ±0.2°C) were employed for temperature measurement. Six thermocouples were uniformly arranged on the back side of the solar panel to measure its surface temperature distribution. Three additional thermocouples were installed inside the thermal storage tank at upper, middle, and lower levels. Two thermocouples each were placed at the inlet and outlet of the pipeline system to record the cooling fluid temperatures. Anemometer measurements confirmed that the ambient air velocity remained below 1 m/s during experiments. Solar radiation intensity was recorded using a pyranometer with a range of 0.1–1999.9 W/m². Electrical parameters were measured using a digital display ammeter-voltmeter with current accuracy of ≤±1 mA and voltage accuracy of ≤±1 V. Fluid flow rate was measured using a turbine flow meter covering a range of 0.04–0.25 m³/h with an accuracy of ≤1%.
The average temperature of the thermal storage tank was calculated using Equation (1):
$$\overline{T} = \frac{T_1 + T_2 + T_3}{3} \tag{1}$$
where $\overline{T}$ is the average temperature of the storage tank, and $T_1$, $T_2$, and $T_3$ are the measured temperatures at the upper, middle, and lower points, respectively.
2.5 Performance Evaluation Criteria
I employed the following metrics to evaluate the PV/T system performance. All tests were carried out following the Chinese National Standard GB/T 4271-2007 for solar collector performance testing. The acceptable fluctuation ranges are listed in Table 4.
| Parameter | Allowable deviation |
|---|---|
| Radiation intensity | ±20 W/m² |
| Ambient temperature | ±1 °C |
| Coolant volume flow rate | ±1% |
| Coolant inlet temperature | ±0.1 °C |
The total heat collection quantity was calculated as:
$$Q = m c_p (T_f – T_i) \tag{2}$$
where $m$ is the mass of the working fluid in the tank, $c_p$ is the specific heat capacity, $T_i$ is the initial tank temperature, and $T_f$ is the final tank temperature.
The cumulative heat collection efficiency is:
$$\eta_{th} = \frac{Q}{H A_c} \tag{3}$$
where $H$ is the total solar irradiance and $A_c$ is the collecting area of the PV module.
The instantaneous electrical power output was computed using:
$$P = U I \tag{4}$$
The instantaneous PV efficiency is:
$$\eta_{PV} = \frac{U I}{G A_{PV}} \tag{5}$$
The total solar energy utilization efficiency was defined as:
$$\eta_{total} = \eta_{th} + \zeta \eta_{PV} \tag{6}$$
where $\zeta$ is the coverage factor defined as the ratio of total PV module area to the collecting area.
To comprehensively compare the combined photovoltaic and thermal performance, I used the dimensionless PV/T efficiency:
$$\eta_f = \eta_{th} + \frac{\zeta \eta_{PV}}{\eta_{power}} \tag{7}$$
where $\eta_{power}$ is the conventional electric power generation efficiency, typically 38%.
2.6 Nanofluid Performance Evaluation
To compare the liquid cooling performance of different nanofluids relative to their base fluid, I defined a performance enhancement factor:
$$B = \left( \frac{\Delta T”}{\Delta T’} – 1 \right) \times 100\% = \frac{T_2” – T_2′}{T_2′ – T_2} \times 100\% \tag{8}$$
where $B$ is the cooling performance enhancement of the nanofluid, $\Delta T’$ is the temperature drop achieved by the base fluid, $\Delta T”$ is the temperature drop achieved by the nanofluid, $T_2$ is the final solar panel temperature without cooling, $T_2’$ is the final temperature with base-fluid cooling, and $T_2”$ is the final temperature with nanofluid cooling.
When the cooling time required to reach a predetermined temperature is used as the evaluation criterion, Equation (8) is transformed into:
$$B = \frac{\tau_2” – \tau_2′}{\tau_2′ – \tau_2} \times 100\% \tag{9}$$
where $\tau_2$ is the time required without liquid cooling, $\tau_2’$ is the time required with base-fluid cooling, and $\tau_2”$ is the time required with nanofluid cooling.
2.7 Uncertainty Analysis
I performed a systematic uncertainty analysis following standard propagation of error methods. The uncertainty of directly measured quantities was calculated as:
$$u_i = \left( \frac{w_i^2}{3} \right)^{0.5} \tag{10}$$
and the combined uncertainty for indirectly measured quantities was obtained from:
$$u_R = \left( \sum_{i=1}^{n} u_i^2 \right)^{0.5} = \left( \frac{1}{3} \sum_{i=1}^{n} w_i^2 \right)^{0.5} \tag{11}$$
Table 5 presents the calculated uncertainties. All relative uncertainties are within acceptable ranges, confirming that the measurement precision satisfies the experimental requirements.
| Parameter | Instrument accuracy | Uncertainty |
|---|---|---|
| Radiation intensity | ±5 W/m² | 2.9 W/m² |
| Ambient temperature | ±0.2 °C | 0.11 °C |
| Coolant volume flow rate | ±3.5 × 10⁻⁴ kg/s | 2 × 10⁻⁴ kg/s |
| Output current | ≤±0.1 A | 0.05 mA |
| Output voltage | ≤±1 V | 0.58 V |
| Heat collection | — | 0.03 W |
| Heat collection efficiency | — | 2.88% |
| Electricity generation | — | 0.05 W |
| PV efficiency | — | 2.9% |
3. Effects of Cold Plate Configuration on PV/T Performance
3.1 Comparative Study of Three Configurations
I conducted a comparative experiment using three configurations: (a) without cold plate, (b) with a conventional flat cold plate, and (c) with the spiral-coil vortex generator cold plate. The experiments were performed under a radiation intensity of 950 ± 25 W/m², an ambient temperature of 24.5 °C, a collector tilt angle of 0°, deionized water as the working fluid, a flow rate of 1.1 L/min, and an ambient wind speed of 0.5 m/s. Each test lasted for 540 minutes.
3.2 Effect on Solar Panel Temperature
Figure 1 illustrates the temporal variation of the solar panel temperature for the three configurations. In the absence of a cold plate, the panel temperature rose continuously and finally stabilized at approximately 70 °C. With the conventional flat cold plate, the steady-state temperature was approximately 61 °C. In contrast, when the spiral-coil cold plate was used, the panel temperature reached a stable value of approximately 57 °C. The maximum temperature reduction achieved with the spiral-coil cold plate relative to the uncooled case was 13.8 °C, with an average temperature drop of 10.74 °C. For 74.07% of the experimental duration, the temperature reduction exceeded 10 °C. Compared with the conventional flat cold plate, the spiral-coil cold plate achieved a maximum temperature reduction of 3.6 °C and an average reduction of 2.16 °C, maintaining more than 2 °C reduction for 66.67% of the test duration.
3.3 Effect on Heat Collection Performance
The total heat collection quantities and corresponding collection efficiencies are summarized in Table 6. Without a cold plate, the total heat collected was essentially zero. The conventional flat cold plate achieved a total heat collection of 3386.25 kJ with a collection efficiency of 40.04%. The spiral-coil cold plate attained a total heat collection of 3901.8 kJ, corresponding to a collection efficiency of 46.6%, representing a 15.22% improvement in total heat collection relative to the conventional cold plate.
| Configuration | Total heat collection (kJ) | Collection efficiency (%) |
|---|---|---|
| Without cold plate | ≈0 | ≈0 |
| Conventional flat cold plate | 3386.25 | 40.04 |
| Spiral-coil cold plate | 3901.80 | 46.60 |
3.4 Effect on Power Generation Performance
The temporal profiles of instantaneous power output showed that the spiral-coil cold plate consistently maintained the highest power output throughout the experiment. Table 7 summarizes the total power generation and PV efficiency. The uncooled system produced 455.9 kJ with an efficiency of 6.82%. The conventional flat cold plate improved total generation to 500.76 kJ with an efficiency of 7.44%. The spiral-coil cold plate achieved the highest total generation of 564.05 kJ, corresponding to an efficiency of 8.46%. Compared with the uncooled configuration, the spiral-coil cold plate improved total electricity generation by 23.72%, and relative to the conventional flat cold plate, it improved generation by 12.64%.
| Configuration | Total electricity generation (kJ) | PV efficiency (%) |
|---|---|---|
| Without cold plate | 455.90 | 6.82 |
| Conventional flat cold plate | 500.76 | 7.44 |
| Spiral-coil cold plate | 564.05 | 8.46 |
3.5 Effect on Combined Efficiency
Table 8 presents the total solar utilization efficiency and the PV/T combined efficiency for the three configurations. The spiral-coil cold plate achieved the highest total solar utilization efficiency of 53.36%, compared to 45.99% for the conventional flat plate and 5.45% without cooling. More importantly, the PV/T combined efficiency reached 64.41% with the spiral-coil cold plate, representing a 15.64% improvement over the conventional flat cold plate (55.7%) and a 348.85% improvement over the uncooled configuration (14.35%).
| Configuration | Total solar utilization efficiency (%) | PV/T combined efficiency (%) |
|---|---|---|
| Without cold plate | 5.45 | 14.35 |
| Conventional flat cold plate | 45.99 | 55.70 |
| Spiral-coil cold plate | 53.36 | 64.41 |
These findings strongly demonstrate that the spiral-coil cold plate structure can effectively improve the overall performance of a solar panel by simultaneously lowering operating temperature and enhancing thermal energy recovery.
4. Effect of Radiation Intensity on System Performance
Using the spiral-coil cold plate configuration, I further examined the influence of radiation intensity by testing three levels: 800 W/m², 1000 W/m², and 1200 W/m². Table 9 summarizes the key results for these radiation levels.
When the irradiation intensity was 1200 W/m², the final panel temperature was about 67 °C, and the total electricity generation was only 459.91 kJ with a PV efficiency of 5.23%. At 1000 W/m², the panel stabilized at about 60 °C, generating 567.82 kJ with a PV efficiency of 7.79%. At 800 W/m², the panel temperature was around 52.5 °C, producing 499.28 kJ and achieving the highest PV efficiency of 8.52%.
The temporal data indicate that while higher radiation increases the thermal energy input, it simultaneously raises the solar panel temperature, which in turn degrades the PV conversion efficiency. The optimal radiation intensity in this study was therefore 1000 W/m², where the photovoltaic and thermal efficiencies were optimally balanced.
| Radiation intensity (W/m²) | Steady-state panel temperature (°C) | Total heat collection (kJ) | Heat collection efficiency (%) | Total electricity generation (kJ) | PV efficiency (%) | Total solar utilization efficiency (%) | PV/T combined efficiency (%) |
|---|---|---|---|---|---|---|---|
| 800 | 52.5 | 2947.35 | 40.06 | 499.28 | 8.52 | 46.88 | 58.00 |
| 1000 | 60.0 | 3853.50 | 42.08 | 567.82 | 7.79 | 48.31 | 58.48 |
| 1200 | 67.0 | 4779.60 | 43.25 | 459.91 | 5.23 | 47.43 | 54.26 |
The maximum instantaneous power difference between 1000 W/m² and 1200 W/m² reached 5.3 W, with an average difference of 3.33 W, confirming that high radiation intensity without adequate cooling causes a net degradation in electrical output. These observations highlight the critical importance of thermal management for solar panels operating under high irradiance.
5. Preparation and Thermophysical Characterization of Nanofluids
5.1 Materials and Preparation Procedure
I prepared nanofluids using the two-step method. Multi-walled carbon nanotubes (MWCNT) with 98% purity, tube diameters of 3–15 nm, tube lengths of 15–30 μm, and a specific surface area of 250–275 m²/g, together with Al₂O₃ nanoparticles (99.9% purity, average diameter 30 nm), were selected as the nanoparticles. Three types of surfactants were evaluated: cetyltrimethylammonium bromide (CTAB), sodium dodecylbenzenesulfonate (SDBS), and polyvinylpyrrolidone (PVP). Deionized water and a thermal fluid (a 1:1 water–ethylene glycol mixture) served as base fluids.
The preparation procedure involved dissolving the surfactant in deionized water under magnetic stirring for 10 minutes at 25 °C, followed by adding a precise amount of nanoparticles and further stirring for 20 minutes. The resulting suspension was then subjected to ultrasonic dispersion for a predetermined duration. To prevent temperature rise from damaging the surfactant, the sonication was applied intermittently with 15-minute sonication cycles followed by 15-minute cooling in a water bath. The detailed preparation route is summarized as follows:
| Step | Operation | Duration |
|---|---|---|
| 1 | Dissolve surfactant in base fluid with magnetic stirring at 25 °C | 10 min |
| 2 | Add nanoparticles and continue magnetic stirring | 20 min |
| 3 | Ultrasonic dispersion (intermittent, with cooling intervals) | 1–4 h |
5.2 Stability and Uniformity — Surfactant Selection
To quantify the dispersion stability, I measured the average optical transmittance of the nanofluids using a UV-3600 spectrophotometer under the 200–1500 nm wavelength range. Lower transmittance values indicate better particle dispersion. Table 11 reports the optimal surfactant types and ratios found for each nanofluid system. For the MWCNT nanofluid, PVP at a nanoparticle-to-surfactant mass ratio of 1:2 produced the lowest transmittance of 46.2%. For the Al₂O₃ nanofluid, SDBS at a ratio of 1:3 yielded a transmittance of 51.6%. For the Al₂O₃-MWCNT hybrid nanofluid, CTAB at a 1:2 mass ratio achieved the lowest transmittance of 45.9%.
| Nanofluid | Optimal surfactant | Mass ratio (nanoparticle : surfactant) | Average transmittance (%) |
|---|---|---|---|
| MWCNT | PVP | 1:2 | 46.2 |
| Al₂O₃ | SDBS | 1:3 | 51.6 |
| Al₂O₃-MWCNT hybrid | CTAB | 1:2 | 45.9 |
Sedimentation observations confirmed that the prepared nanofluids remained visually uniform for at least 24 hours without significant precipitation, and the long-term stability was adequate for the cooling experiments. The choice of surfactant type had a minor effect on thermal conductivity but a significant effect on long-term suspension stability.
5.3 Thermal Conductivity Measurements
The thermal conductivity of the nanofluids was measured using the transient hot-wire method with a DRE-2A thermal conductivity analyzer. The governing equation for this method is:
$$\theta = T – T_0 = \frac{-q}{4\pi\lambda} E_i\left( -\frac{r^2}{4\alpha t} \right) \tag{12}$$
where $\theta$ is the temperature rise of the probe, $q$ is the heating power, $\lambda$ is the thermal conductivity, $\alpha$ is the thermal diffusivity, $r$ is the radial position, $t$ is time, and $E_i$ is the exponential integral function. Under the condition $r^2/(4\alpha t) \ll 1$, the thermal conductivity can be obtained from the slope of temperature versus logarithmic time:
$$\lambda = \frac{q}{4\pi} \Big/ \frac{d\theta}{d\ln t} \tag{13}$$
The theoretical models describing nanofluid thermal conductivity include the Hamilton-Crosser model:
$$\frac{k_{eff}}{k_{bf}} = \frac{k_{np} + (n-1)k_{bf} – (n-1)\phi(k_{bf} – k_{np})}{k_{np} + (n-1)k_{bf} + \phi(k_{bf} – k_{np})} \tag{14}$$
where $k_{bf}$ and $k_{np}$ are the thermal conductivities of the base fluid and nanoparticles, respectively, $\phi$ is the volume fraction, and $n$ is the particle shape factor ($n = 3/\psi$). Avsec extended this model by incorporating the interface layer thickness $h$ and the particle radius $r$ through an effective volume fraction:
$$\phi_e = \phi \left( 1 + \frac{h}{r} \right)^3 \tag{15}$$
5.4 Effect of Dispersant on Thermal Conductivity
To isolate the effect of the dispersant, I prepared Al₂O₃-MWCNT/deionized water hybrid nanofluids with a fixed 0.1% total mass fraction and three different SDBS-to-nanoparticle ratios of 1:1, 1:2, and 1:3. The measured thermal conductivities were 0.668, 0.673, and 0.667 W/(m·K), respectively. When rounded to two decimal places, all values are identical at 0.67 W/(m·K). Therefore, I conclude that while the dispersant is essential for stabilizing the suspension, its influence on thermal conductivity itself is negligible.
5.5 Effect of Ultrasonic Dispersion Time on Thermal Conductivity
To investigate the influence of ultrasonic processing on the thermal conductivity, I prepared Al₂O₃-MWCNT/deionized water hybrid nanofluids with a 0.1% mass fraction using 30 minutes of magnetic stirring followed by 1–4 hours of ultrasonic dispersion without any surfactant. The results are reported in Table 12. The thermal conductivity of pure deionized water was 0.605 W/(m·K). After 30 minutes of magnetic stirring, the thermal conductivity increased to 0.652 W/(m·K), corresponding to a 7.77% enhancement. When an additional 1 hour of ultrasonic dispersion was applied, the thermal conductivity increased further to 0.663 W/(m·K). Extending the sonication time from 1 to 4 hours produced only marginal changes, with the maximum value reaching 0.664 W/(m·K) after 4 hours, corresponding to a 9.75% enhancement over the base fluid. These results indicate that the additional benefit of prolonged sonication beyond 1 hour is limited.
| Preparation condition | Thermal conductivity (W/(m·K)) | Enhancement relative to water (%) |
|---|---|---|
| Deionized water | 0.605 | — |
| 30 min magnetic stirring | 0.652 | 7.77 |
| 30 min stirring + 1 h sonication | 0.663 | 9.59 |
| 30 min stirring + 2 h sonication | 0.663 | 9.59 |
| 30 min stirring + 3 h sonication | 0.662 | 9.42 |
| 30 min stirring + 4 h sonication | 0.664 | 9.75 |
5.6 Effect of Nanoparticle Concentration on Thermal Conductivity
At a constant temperature of 24 °C, I measured the thermal conductivity of Al₂O₃-MWCNT/deionized water hybrid nanofluids at mass fractions of 0.05%, 0.1%, 0.15%, 0.2%, and 0.25%. Table 13 presents the results.
| Mass fraction (%) | Thermal conductivity (W/(m·K)) |
|---|---|
| 0.05 | 0.635 |
| 0.10 | 0.670 |
| 0.15 | 0.674 |
| 0.20 | 0.702 |
| 0.25 | 0.712 |
Thermal conductivity increased monotonically with increasing nanoparticle concentration. The largest incremental gain occurred at the lowest concentration step (from 0.605 to 0.635 W/(m·K)), while the growth rate gradually diminished at higher concentrations, exhibiting a distinct non-linear, diminishing-marginal-return behavior.
5.7 Effect of Temperature on Thermal Conductivity
I further measured the thermal conductivity of Al₂O₃-MWCNT nanofluids prepared with three different base fluids—deionized water, thermal fluid, and dimethyl silicone oil—at temperatures of 10 °C, 20 °C, 30 °C, 40 °C, and 50 °C. The nanoparticle mass fraction was fixed at 0.25%. Table 14 reports the measured values.
| Base fluid | 10 °C | 20 °C | 30 °C | 40 °C | 50 °C |
|---|---|---|---|---|---|
| Deionized water | 0.611 | 0.656 | 0.713 | 0.735 | 0.764 |
| Thermal fluid | 0.393 | 0.420 | 0.487 | 0.523 | 0.571 |
| Dimethyl silicone oil | 0.152 | 0.153 | 0.155 | 0.156 | 0.158 |
For all three systems, thermal conductivity increased with temperature. However, the effect was much more pronounced for the water-based and thermal-fluid-based nanofluids than for the silicone-oil-based system. This contrast can be attributed to the high viscosity of dimethyl silicone oil, which promotes nanoparticle aggregation and reduces the effective nanofluidic heat transfer enhancement. This finding highlights the important role of base fluid viscosity in determining the thermal performance of the nanofluid.
5.8 Density Measurements
The density of nanofluids was both theoretically estimated and experimentally measured. The theoretical density follows a simple mixing rule:
$$\rho_{nf} = (1 – \phi)\rho_{bf} + \phi \rho_{np} \tag{16}$$
I validated the density measurement system using deionized water as a standard, confirming that deviations between measured averages and theoretical values were within 0.3%.
At 25 °C, the density of Al₂O₃-MWCNT/deionized water nanofluids increased monotonically with mass fraction. The highest density of 1.0002 kg/m³ was observed at the 0.25% mass fraction. Furthermore, the density of the nanofluids decreased with increasing temperature, consistent with the behavior of the pure base fluid. However, the temperature sensitivity varied with concentration: the density of the 0.1% nanofluid decreased by only 0.4% from 10 °C to 50 °C, whereas the 0.25% nanofluid decreased by 1.49% over the same temperature range. This concentration-dependent thermal response is linked to enhanced particle aggregation at elevated temperatures.
6. Cooling Performance of Al₂O₃-MWCNT Nanofluids for Solar Panels
6.1 Experimental Design
I evaluated the liquid cooling performance of various working fluids under three distinct radiation scenarios that simulate different periods of a sunny day: (1) a low-intensity scenario at 700 W/m², designed to simulate morning and afternoon periods; (2) a high-intensity scenario at 1100 W/m², simulating midday conditions; and (3) a cool-down scenario at 200 W/m², simulating the evening period when the solar panel cools from 40 °C to 25 °C. The working fluids included deionized water, thermal fluid, and Al₂O₃-MWCNT/thermal fluid nanofluids with mass concentrations of 0.1%, 1%, and 2%. Two flow rates were tested: 0.9 L/min and 1.8 L/min.
6.2 Cooling Performance of Base Fluids
Table 15 summarizes the cooling performance of deionized water under the three radiation conditions. At 700 W/m², the final panel temperature was 38.1 °C at 1 L/min and 37.2 °C at 2 L/min. At 1100 W/m², the final temperatures were 51.5 °C and 50.2 °C, respectively. In the cool-down test at 200 W/m², the cooling time from 40 °C to 25 °C was 32 min 24 s at 1 L/min and 27 min 42 s at 2 L/min.
| Radiation (W/m²) | Flow rate (L/min) | Final temperature (°C) | Cooling time (min) |
|---|---|---|---|
| 700 | 1 | 38.1 | — |
| 700 | 2 | 37.2 | — |
| 1100 | 1 | 51.5 | — |
| 1100 | 2 | 50.2 | — |
| 200 | 1 | — | 32.4 |
| 200 | 2 | — | 27.7 |
The thermal fluid exhibited slightly lower cooling performance than deionized water because of its higher viscosity and lower specific heat. At 700 W/m², the thermal fluid reached final temperatures of 38.6 °C and 37.8 °C at 0.9 L/min and 1.8 L/min, respectively. At 1100 W/m², the corresponding values were 52.3 °C and 51.2 °C. In the cool-down test at 200 W/m², the cooling times were 37 min 54 s and 32 min 35 s for the two flow rates.
6.3 Effect of Nanoparticle Concentration on Cooling Performance
Table 16 reports the cooling performance of the Al₂O₃-MWCNT/thermal fluid nanofluids at 700 W/m² radiation with a flow rate of 0.9 L/min. At the end of the 100-minute experiment, the panel temperatures for the 0.1%, 1%, and 2% mass concentrations reached 38.25 °C, 37.19 °C, and 37.62 °C, respectively, corresponding to cooling performance enhancements of 2.69%, 10.85%, and 7.54% relative to the pure thermal fluid.
| Working fluid | Final temperature (°C) | Temperature reduction (°C) | Cooling enhancement (%) |
|---|---|---|---|
| Thermal fluid (base) | 38.63 | — | — |
| 0.1% nanofluid | 38.25 | 0.35 | 2.69 |
| 1% nanofluid | 37.19 | 1.44 | 10.85 |
| 2% nanofluid | 37.62 | 0.98 | 7.54 |
At 1100 W/m² radiation with a flow rate of 0.9 L/min, the corresponding final temperatures for the three concentrations were 52.15 °C, 51.59 °C, and 51.77 °C, yielding cooling enhancements of 2.46%, 11.64%, and 8.69%, respectively. At 200 W/m², the cooling times from 40 °C to 25 °C were 36 min 33 s, 35 min 53 s, and 37 min 39 s, representing improvements of 6.45%, 9.65%, and 3.31%, respectively, compared to the base thermal fluid.
Across all three radiation intensities, the cooling performance of the Al₂O₃-MWCNT/thermal fluid nanofluids followed a consistent pattern: it first increased with nanoparticle concentration and then decreased beyond an optimal concentration of 1%. At 1100 W/m², the improvement was most pronounced, with a clear peak at 1% mass fraction. At 200 W/m², the improvement was comparatively limited, and the benefit of increasing particle concentration was less evident.
Notably, the 2% mass concentration nanofluid did not enhance cooling as much as theoretically expected. Under identical flow conditions, its cooling performance was inferior to that of the 1% nanofluid and sometimes even lower than that of the pure thermal fluid. This finding is attributable to the high viscosity of the thermal fluid combined with the tendency of nanoparticles to agglomerate at elevated concentrations. When nanoparticle clusters form, the micro-convection and high-surface-area effects that enhance heat transfer are substantially diminished, causing the nanofluid to behave more like the base fluid. Although the circulating pump action prevented sedimentation, the aggregated nanofluid had largely lost its enhanced thermophysical properties.
6.4 Effect of Flow Rate on Cooling Performance
The flow rate had a significant influence on the cooling performance of the nanofluids. Table 17 compares the cooling enhancements at 1100 W/m² for two flow rates. At 0.9 L/min, the cooling enhancements for the 0.1%, 1%, and 2% nanofluids were 2.46%, 11.64%, and 8.69%, respectively. At 1.8 L/min, the enhancements decreased to 1.72%, 6.66%, and 1.67%. The reduced benefit at the higher flow rate suggests that high-velocity flow increases the likelihood of nanoparticle agglomeration, thereby diminishing the nanofluid’s heat transfer advantage.
| Mass concentration (%) | Cooling enhancement at 0.9 L/min (%) | Cooling enhancement at 1.8 L/min (%) |
|---|---|---|
| 0.1 | 2.46 | 1.72 |
| 1 | 11.64 | 6.66 |
| 2 | 8.69 | 1.67 |
6.5 Inlet-Outlet Temperature Differences
I also analyzed the temperature difference between the inlet and outlet of the cooling system, which serves as an indicator of heat extraction effectiveness. Table 18 reports the final inlet-outlet temperature differences for the different working fluids at 700 W/m² and 0.9 L/min. The 1% nanofluid achieved the largest inlet-outlet temperature difference of 3.9 °C, consistent with its superior cooling performance. In contrast, the pure thermal fluid achieved only 2.62 °C. A similar trend was observed at 1100 W/m², where the maximum temperature differences for the 0.1%, 1%, and 2% nanofluids were 5.86 °C, 6.26 °C, and 6.07 °C, respectively, compared to 5.6 °C for the thermal fluid.
| Working fluid | Maximum inlet-outlet temperature difference (°C) |
|---|---|
| Thermal fluid (base) | 2.62 |
| 0.1% nanofluid | 2.83 |
| 1% nanofluid | 3.90 |
| 2% nanofluid | 3.57 |
At 200 W/m², the inlet-outlet temperature differences decreased over time as the panel cooled down. For the 0.1% nanofluid, the difference dropped from 4.8 °C to 1.2 °C at 0.9 L/min, while for the 2% nanofluid it dropped from 4.6 °C to 1.4 °C. These data confirm that nanofluids enhance the heat extraction capacity of the cooling loop, although the micro-scale effects responsible for the improvement become less effective at higher concentrations due to particle agglomeration.
7. Conclusions
In this work, I designed and fabricated a spiral-coil vortex-generator cold plate and investigated the cooling performance of Al₂O₃-MWCNT hybrid nanofluids for solar panel thermal management using a carefully constructed PV/T experimental platform. The main conclusions are as follows:
(1) The spiral-coil cold plate structure significantly improved the thermal and electrical performance of the solar panel. Compared with the uncooled configuration, it achieved a maximum temperature reduction of 13.8 °C and an average reduction of 10.74 °C, with the reduction exceeding 10 °C for 74.07% of the experimental duration. Compared with the conventional flat cold plate, it achieved a maximum temperature reduction of 3.6 °C and an average reduction of 2.16 °C. The total heat collection reached 3901.8 kJ with a collection efficiency of 46.6%, and the total electricity generation reached 564.05 kJ with a PV efficiency of 8.46%. The PV/T combined efficiency reached 64.41%, representing a 15.64% improvement over the conventional flat cold plate.
(2) The radiation intensity significantly influenced the system behavior. At 1000 W/m², the system achieved the best overall solar utilization efficiency of 48.31% and the highest PV/T combined efficiency of 58.48%. At 1200 W/m², the increased panel temperature degraded the PV conversion efficiency despite the higher energy input, highlighting the critical role of thermal management for solar panels under high irradiance.
(3) The two-step method was successfully used to prepare stable Al₂O₃-MWCNT hybrid nanofluids. The optimal surfactant was CTAB with a nanoparticle-to-surfactant mass ratio of 1:2, yielding the lowest transmittance of 45.9% and excellent suspension stability. The thermal conductivity of the nanofluids increased with nanoparticle concentration and temperature, but followed a nonlinear, diminishing-return pattern. The dispersant had a negligible effect on thermal conductivity while being essential for suspension stability.
(4) The Al₂O₃-MWCNT/thermal fluid nanofluids exhibited optimal cooling performance at a mass concentration of 1% across all tested radiation intensities. The cooling enhancement was most pronounced at 1100 W/m², where the 1% nanofluid improved cooling performance by 11.64% at 0.9 L/min. High concentrations (2%) were less effective because nanoparticle agglomeration, aggravated by the high viscosity of the base fluid and high flow rate, suppressed the heat transfer enhancement. These findings provide useful design guidelines for the application of hybrid nanofluids in solar panel liquid cooling systems.
Future research should explore alternative nanoparticle systems, including environmentally friendly and organic-inorganic hybrid materials, systematically investigate the influence of different dispersants on nanofluid thermophysical properties, and validate the present findings under real outdoor conditions. Such studies will further advance the practical application of nanofluid-based cooling for solar photovoltaic systems.