Comprehensive modelling framework for a low temperature gradient thermoelectric generator

Sergei Vostrikov, Andrey Somov, Pavel Gotovtsev, Michele Magno

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)


Thermoelectric energy harvesting (TEH) is a promising technology enabling the self-sustainable operation of ultra low-power sensing devices performing monitoring tasks. Conventional TEH research mainly focuses on the improvements of the material properties (i.e thermoelectric figure of merit ZT) and thermoelectric module (TEM) geometry. However, enhanced performance of the thermoelectric generator (TEG) can not be practically achieved without optimal integration of the novel TEM with the DC-DC converter. The power converter, having its own input impedance requirements and conversion efficiency curve, may keep the operation point of TEG far from the maximum power point of TEM, leading to power loss. Furthermore, since the physical processes in the thermoelectric transducer are coupled with the electrical processes in the DCDC converter, both components should be analyzed together to achieve the best harvesting performance. Given the system complexity, an accurate simulation model is required to estimate the output power and make an optimal system design. We report on a simulation approach for low temperature gradient TEG that accounts for both the thermoelectric and the power conversion processes. The proposed approach was validated on the TEG testbed by imitating the realistic ambient conditions for thermoelectric transducer connected to a commercial off-the-shelf DC-DC converter. Experimental results are characterized by 9% mean percentage error of power estimation which outperforms the methods based on conventional TEG representations.

Original languageEnglish
Article number114721
JournalEnergy Conversion and Management
Publication statusPublished - 1 Nov 2021


  • DC-DC power converters
  • Energy harvesting
  • System modelling
  • Thermoelectric devices


Dive into the research topics of 'Comprehensive modelling framework for a low temperature gradient thermoelectric generator'. Together they form a unique fingerprint.

Cite this