Stoichiometric-Based Fan Flow Optimization Using PI System to Improve Boiler Efficiency and Ensure Reliable Control of Complete Combustion

Abstract

Efficient combustion air control is essential to maintain optimal performance in coal-fired power plants. However, the reliance on flue gas oxygen (O₂) sensors is often hindered by frequent sensor failures, contamination, and calibration drift. This study proposes a sensor-independent method utilizing real-time stoichiometric airflow recommendations derived from coal ultimate analysis and predefined excess air targets. The method was implemented using the PI System (PI AF and PI Datalink) to generate dynamic airflow setpoints for operators. Field validation was conducted at PLTU Muara Jawa, a 2×27.5 MW circulating fluidized bed (CFB) power plant, under 14 MW and 20 MW load conditions. Performance metrics included residual O₂, CO concentration, excess air, total airflow, fuel flow, flue gas temperature, boiler efficiency, and auxiliary power. Results demonstrated reductions in O₂ (up to 0.86%), CO (up to 7.5%), and excess air (7%), along with a 0.43% improvement in boiler efficiency. The strategy also achieved annual savings of 644 tons of coal and 156,000 kWh of auxiliary power. Based on the plant’s typical annual output of 300,000,000 MWh and a coal cost of USD 40/ton, the total estimated economic benefit reaches approximately USD 46,040. This approach offers a reliable, cost-effective solution for combustion optimization.