Real-time optimization of an industrial steam-methane reformer under distributed sensing

• Control-relevant empirical low-order model for steam methane reformers (SMR).
• Model calibrated using distributed measurements from array of infrared camera.
• Optimization of SMR temperature distribution is carried out based on the model.
• Industrial results confirm excellent performance of proposed framework.

Industrial hydrogen production takes place in large-scale steam methane reformer (SMR) units, whose energy efficiency depends on the interior spatial temperature distribution. In this paper, a control-relevant empirical reduced-order SMR model is presented that predicts the furnace temperature distribution based on fuel input to a group of burners. The model is calibrated using distributed temperature measurements from an array of infrared cameras. The model is employed to optimize in real-time the temperature distribution and increase the energy efficiency in an industrial furnace. Experimental results confirm that the proposed framework has excellent performance.