A Spectrally Resolved Model for a PV + PCM System

A Spectrally Resolved Model for a PV + PCM System

de Mooij, Cas (TU Delft Electrical Engineering, Mathematics and Computer Science)

Isabella, O. (mentor)
Ziar, H. (graduation committee)
Ortiz Lizcano, J.C. (graduation committee)

Delft University of Technology

2020-11-27

As the world is trying to rely less on fossil fuels for electrical energy. To do this, the alternatives should be as efficient and durable as possible. The alternative that has been growing the fastest in recent years is solar energy. Apart from it’s growth in the market share, the efficiency of photovoltaic (PV) cells has also steadily grown. Recently, this trend has started to stagnate, since most avenues of efficiency improvement have been exhausted. However, the possibilities of thermal management (i.e. cooling) of PV cells are relatively unexplored. An increase in cell temperature causes a decrease in efficiency, in other words, less sunlight is converted into electrical energy. Moreover, high operating temperatures cause damage in solar cells, reducing their lifetime by months (or even years). One promising technology that can keep cell temperatures low is the Phase Change Material (PCM) based heat sink. A PCM takes a lot of energy to melt and during phase change it remains at a stable temperature. PV + PCM systems have shown to be able to increase cell efficiency by up to 8%. In order to get a better grip on the workings and the potential of PCMs as a heat sink for PV cells, this work presents an implicit, transient, spectrally resolved, MATLABbased model that can simulate module temperature and phase change within the PCM. The model uses SMARTS in combination with irradiance data to create spectral irradiance data. Using this spectral data together with GenPro4 provides the amount of energy absorbed by each layer of the PV cell. The PV part of the model is validated for a summer week in the Netherlands, with a Mean Absolute Error (MAE) of 1.83 and a Mean Bias Error (MBE) of 0.56. For a winter week in January the MAE was 1.76 and the MBE 1.44. When simulating only a PV module (so no PV + PCM system), the model can simulate 45 hours of data per hour, with one minute time step. When validating the PV + PCM model for the same weeks, the MAE was 2.96 and the MBE 1.91 for the summer week and a MAE of 1.37 and MBE of 1.15 for the winter week. When simulating a PV + PCM system the simulation time drops to 18 hours of simulated data per hour, due to a larger mesh and the requirement of several iterations in the PCM model. With this model a wide range of PV + PCM systems (e.g. with different thicknesses and melting temperatures) can be simulated for any location, provided that there is access to the following weather data: temperature, wind speed and irradiance.

PV-PCM
Thermal modelling
Thermal management
PV System
MATLAB
GenPro4
SMARTS

http://resolver.tudelft.nl/uuid:1ae70ca5-8a8e-4f5b-8838-a106de39c18c

Student theses

master thesis

© 2020 Cas de Mooij