Thermoelectric generator characterization at extra-low-temperature difference for building applications in extreme hot climates: Experimental and numerical study

Abstract

In locations subject to extreme climate and air-conditioned spaces, a temperature gradient is usually present in various parts of air-conditioned buildings. Building envelopes separate the temperature-controlled space indoor and the extreme outdoor climate and the temperature gradients between the two sides of the building envelope drive energy flow. Thermoelectric Generator (TEG) modules can be used to harvest energy due to this thermal gradient to power sensor nodes and other low power applications. At high-temperature differences, these modules perform well, however, many current and future sensors and Internet of Things (IoT) applications in buildings only have access to the Low-Temperature difference. The research described herein focuses on the development of the experimental setup and the numerical model of a characterization test rig in the low-temperature region for a building envelope integrated TEG application in the United Arab Emirates (UAE). This is a field lacking research especially when combined in consideration with the United Arab Emirates (UAE) climatic conditions. Actual experimental results are presented together with the corresponding simulation results. COMSOL Multiphysics® computer modelling software was used as a platform to develop and test the model. The model allows a detailed characterization of the TEG in open-circuit and loaded conditions. The experimental work includes Current-Voltage (I-V) tracing and Maximum Power Point Tracking (MPPT) tests on the subject TEG at 10 °C temperature deference which is relevant to the expected indoors to outdoors temperature difference in the extreme climate conditions considered in this work. Similar tests were simulated using the computer model and the results were compared to experimental results. Results showed a generation capability of about 18 mW matched load output at 10 °C temperature difference. The simulation results were in agreement with the experimental results, root-mean-square deviation was found to be below 5% which is within the acceptable range compared to the available literature.