Waste Heat Recovery System Applied to a High-Performance Video Card

J. Alfredo Jimenez-Bernal,Pedro Quinto-Diez,Arturo Reyes-Leon,Carlos A. Garcia-Rodriguez,L. Annette Romero-De Leon

doi:10.1109/access.2020.2964207

J. Alfredo Jimenez-Bernal, Pedro Quinto-Diez + Show 3 more

Open Access

https://doi.org/10.1109/access.2020.2964207

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2020
License type: CC BY 4.0

Affiliation: Instituto Politécnico Nacional

Abstract

This paper presents results given by a waste heat recovery (WHR) system applied to a high-performance video card, as well as average energy generated per hour according to emulation of computer graphics requirements demanded by the user while the card is working. A WHR system includes three phases: (1) waste heat collection, (2) energy conversion and (3) signal conditioning. The analysis of the WHR system is presented. The emulation of waste heat has been generated using electrical resistors as if they were the main components that generate waste heat, mainly the GPU (graphics processing unit), and DDR3 memories. This WHR system has considered the MSI-R4850 video card as a reference, operation temperature of which has an overall range between 60°C- 90°C. Thermoelectric generator modules (TEG) are based on the Seebeck effect, and the thermoelectric array used is an important part of the WHR system, which has been constructed based on the locations of the main components to convert waste heat into electrical power. The waste heat recovery process has two treatments: First, once the operating conditions, per GPU and DDR3 memories have been emulated, the energy recovered is measured per component and whole WHR system; the second one measures energy recovered considering the output signal conditioning of the WHR system, which was converted to 5V output through a DC-DC boost converter, while the input voltage operates within a range (0.9V- 5V). The energy recovered may be applied to low-power electronic devices, which is a contribution to energy efficiency.

Highlights

The increasing demand for electricity around the world has diversified the generation technologies in their primary sources, to permit coverage of the projected worldwide demand according to World Energy Outlook 2015
POWER CURVE OF WASTE HEAT RECOVERED The thermoelectric generator (TEG) array’s positive and negative terminals are connected to a variable load resistor (RL ), Which has been chosen to ensure the occurrence of maximum power transfer, while the emulation of the video card is operating within the same range of T defined previously for the waste heat recovery (WHR) system; data were acquired and recorded
ENERGY RECOVERED For the first experiment set up, the parameters of the WHR system to convert the thermal power into electrical are measured and given in Table 1, where T0 is the temperature of the standard heat sink, which is given as room temperature, T1 is the temperature emulated of graphics processing unit (GPU), T2 is the temperature emulated of DDR3 memories(1), T3 is the temperature emulated of DDR3 memories(2), and VOUT, IOUT and POUT are the variables measured of the WHR system

Summary

Introduction

The increasing demand for electricity around the world has diversified the generation technologies in their primary sources, to permit coverage of the projected worldwide demand according to World Energy Outlook 2015. According to the new policies scenario, global primary energy demand will increase by nearly one-third between 2013 and 2040 to reach 17 900 Mtoe [1] in a sustainable way. This electrical energy demand projection includes fossil fuel usage and renewable and clean sources. One important tool related to certain waste heat recovery systems is the thermoelectric generator (TEG), used as a device to generate clean-energy [5]–[7]. Some studies show that a new model of a compact thermoelectric generator has been developed to improve the electric performance per unit volume and maximize the energy conversion efficiency [8]

Objectives

Results

Conclusion