System identification-based analysis of dual PV-TEG units with nano-enhanced cooling and oscillating inlet temperatures

Abstract

Advanced cooling solutions are needed for photovoltaic (PV) panels to increase their effectiveness, to reduce the local damage and improve the lifespan of the products. In this work, impact of employing a pulsing inlet coolant temperature in the channel on the efficiency of two photovoltaic units installed on a single nano-enhanced cooling channel is investigated. Additionally, there is a thermo-electric generator (TEG) connected to each PV unit. The finite element method (FEM)-based numerical investigation is carried out for a range of values of Re (between 100 and 500), nanoparticle solid volume fraction (between 0 and 3%), and pulsing inlet temperature amplitude (Ap between 0 and 0.03). The Nusselt number (Nu) exhibits oscillatory behavior in the horizontal channel (section A) with varying Re and Ap. The PV cell temperature drop between the highest and lowest Re cases in unit A is 0.7 degrees C, but in unit B, it is 2.7 degrees C. For inclined channel (section B), temperature reduction of 1 degrees C is accomplished between different time steps. The largest temperature reductions are achieved by using inlet temperature pulsations among the different approaches, especially at the lowest Re, which is roughly 8.7 degrees C at Re=100 and 4.2 degrees C at Re=500 for unit A, while these values drop to 7.4 degrees C and 5.3 degrees C for unit B. Inlet temperature pulsation is shown to be an effective method for reducing PV-cell temperature in channel cooling for multi-PV systems with TEG modules. While no-pulsation with BF at Re=100 produces the greatest PV-cell temperature readings, NF-pulsed PV cells at Re=500 in both units produce the lowest PV-cell temperature values. For units A and B, the PV-cell temperature decreases under the best and worst conditions are 9.4 degrees C and 15.7 degrees C, respectively. Both linear and nonlinear system identification (SYS-ID) techniques are employed to develop dynamic models for the spatially averaged Nu under varying oscillation amplitudes. The model outputs can be coupled with the PV-TEG combined units in both channels to calculate the energy and exergy efficiency of the PV module.