Identify the Optimal Sensor Location: Analysis of Flue Gas Behavior in Stack by Using CFD Simulation

Abstract

Industrial flue gas emission consists of harmful substance such as carbon monoxide (CO), carbon dioxide (CO₂), nitrogen oxides (NOₓ), sulfur dioxide (SO₂), which contribute to air pollution, climate change, and health risks. Effective emission monitoring is essential to reduce environmental impact and comply with air quality regulations. However, the accuracy of in-situ gas sensors largely depends on proper placement within the exhaust stack, as poor positioning in turbulent or unstable flow regions can result in inaccurate readings. This study aims to analyze flue gas flow behavior within an L-shaped industrial exhaust stack and determine the optimal sensor placement using Computational Fluid Dynamic (CFD) simulation. The geometry was created in SolidWorks and imported into ANSYS Fluent 2024 R1 for steady state analysis using the standard k-epsilon turbulence model. Key flow parameters such as velocity, pressure, and gas interaction with sensor geometry were examined. The stack consists of a 4-meter horizontal section, a 12-meter vertical section, and a 2-meter diameter. Four sensor locations 2 m, 4 m, 6 m, and 8 m from the outlet were tested to evaluate flow stability and suitability for sensor installation. Simulation results showed that the sensor location at 10 meters from the bend which is 2 meters before the outlet offered the most stable flow conditions, with an average velocity of 10.75 m/s and a pressure difference of only 59.5 Pa across the sensor. This minimal pressure drop indicates low flow disturbance caused by the sensor, helping maintain continuous and reliable contact between the flue gas and the sensor surface. These favorable flow characteristics support the accuracy of gas detection and confirm this location as optimal for in-situ sensor deployment. Overall, the findings demonstrate the effectiveness of CFD in guiding sensor placement strategies for enhanced emission monitoring in industrial stack systems