Computational Simulation and Analysis of Major Control Parameters of Time-Dependent PV/T Collectors

Jimeng Shi, Cheng-Xian LinarXiv 2021 · arXiv

Introduction

Photovoltaic (PV) cells experience a drop in efficiency as their operating temperature rises. Photovoltaic/Thermal (PV/T) systems mitigate this by circulating a fluid, such as water, to cool the cells while simultaneously capturing thermal energy for secondary use. Despite their potential, accurately predicting the real-world performance of these systems remains a challenge due to the rapid fluctuations in environmental conditions like solar irradiance.

This research addresses two significant gaps in existing PV/T modeling:

  1. Temporal Resolution: Traditional models often rely on 1-hour time steps, which are too coarse to capture the transient "jitter" of real-time solar irradiance.
  2. Simplified Physics: Many earlier simulations neglected edge loss coefficients and used simplified radiative heat transfer calculations between the solar cell and the sky.

Methodology

The researchers developed a comprehensive dual-layered model combining thermal and electrical dynamics.

Improved Thermal Model

Energy balance equations were derived for each distinct layer of the collector: the glass cover, solar cells, Tedlar backing, and the water channel. A key advancement in this model is the integration of an edge loss coefficient ($U_e$) and a refined radiative heat transfer coefficient ($h_{rad}$), which are critical for determining the overall loss coefficient ($U_L$).

Cross-sectional view of the PV/T collector layers used in the thermal model.

High-Resolution Simulation and Electrical Modeling

To match the granularity of experimental records, the simulation operates at a 1-minute time step. Data from historical experiments were digitized to provide high-fidelity inputs. On the electrical side, the model incorporates series and shunt resistances, allowing for precise current-voltage (I-V) and power-voltage (P-V) curve predictions under varying thermal conditions. The entire system was implemented in MATLAB and validated against the landmark experimental data from Huang et al. (1999).

Datasets and Validation

The study utilizes experimental data from Huang et al. (1999) as a benchmark. This dataset includes real-time measurements of solar intensity, ambient temperature, solar cell temperature, and outlet water temperature from a 45L storage tank. By using this "ground truth," the authors could directly compare the accuracy of their high-resolution model against traditional, coarser simulations.

Key Insights and Results

The transition to a 1-minute time step allowed the simulation to accurately mimic the fluctuating, "spiky" nature of experimental cell temperature curves. Coarser models typically smooth out these peaks, leading to significant errors in efficiency estimation.

Real-time experimental data showing the fluctuating nature of solar irradiance and corresponding temperatures.

The inclusion of edge losses proved essential; these parameters are vital for calculating the total energy loss of the system. Furthermore, the electrical model reinforced the principle that while irradiance primarily dictates current output, cell temperature is the primary driver of voltage. Effective thermal management is therefore crucial to prevent voltage "sag" and maintain high efficiency.

Performance Benchmarks

The improved model demonstrated a marked increase in predictive accuracy:

  • Water Temperature Accuracy: Root Mean Square (RMS) deviation improved from 5.87% to 2.01%.
  • Cell Temperature Accuracy: RMS deviation improved from 7.22% to 5.09%.
  • Efficiency: For the MSX-60 module, the model calculated a thermal efficiency of 40.78% and a photoelectric efficiency of 15.54%.

Limitations and Future Work

While the 1-minute time step significantly improves accuracy, the model still assumes a "quasi-steady state" for energy balance, which might overlook extremely rapid transient behaviors. Additionally, the current 1D heat conduction model could be expanded to 2D or 3D Computational Fluid Dynamics (CFD) for more complex geometries. Future research could also focus on developing active control algorithms, such as variable-speed pumps, to dynamically optimize flow rates based on high-resolution real-time data.

Summary updated

BibTeX
@misc{shi2021computational,
  author = {Jimeng Shi and Cheng-Xian Lin},
  title = {{Computational Simulation and Analysis of Major Control Parameters of Time-Dependent PV/T Collectors}},
  year = {2021},
  eprint = {2105.05358},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2105.05358},
}
Figure