Smart Energy Management Through Air-Fuel Ratio Optimization in a Coke Oven Gas-Fired Boiler: A Case Study

Abstract

Industrial boiler performance depends on accurate balancing of fuel and combustion air. In the examined case, a coke oven gas (COG)-fired boiler operated under manual control showed unstable combustion, higher specific gas consumption, and a risk of unburnt carbon deposition. This paper presents a smart energy management approach based on real-time oxygen measurement, PLC logic, VFD-controlled air fan adjustment, and fuel trimming. The study followed a case-study design with root-cause validation, before-and-after performance comparison, and operational benchmarking at JSW Steel Ltd., Dolvi. Validated causes showed that poor air-fuel mixing, insufficient air flow, and fuel flow mismatch were the main contributors to inefficiency, while several suspected causes were ruled out by inspection and testing. After implementation, specific gas consumption reduced from about 165 Nm³ per ton of steam to about 156 Nm³ per ton, while process stability improved and unburnt carbon deposition reduced. The project also demonstrated managerial value through lower recurring fuel cost, operator standardization, and safer boiler operation. The case shows that combining instrumentation, feedback control, and disciplined problem solving can convert a manual combustion process into a measurable and sustainable efficiency system.

Digram about conceptual framework of the proposed smart air-fuel optimization model. Measurement -> decision -> actuation -> imporved efficiency — *Figure 1. Conceptual framework of the proposed smart air-fuel optimization model.*

Introduction

Boilers are central to industrial utility systems because they convert fuel energy into useful steam. Even small deviations in air supply or fuel supply can produce measurable losses through incomplete combustion, excess flue gas heat, soot formation, unstable operation, and unplanned maintenance. In gas-fired utility systems, these losses become more visible when fuel composition changes, load fluctuates, or operators rely on manual judgment instead of measured feedback. In a continuous plant environment, the difference between a fixed setting and a measured control loop can translate directly into recurring fuel cost and process variability.

The present case study focuses on a 24-tonne/hour steam boiler firing coke oven gas (COG), where the original control practice depended on manual excess oxygen checks and operator response. Such a method is practical for short-term operation, but it is weak when the process is continuous and variable. A modern energy-management approach therefore needs continuous measurement, closed-loop control, and clear operating standards. In this case, the control logic was built around oxygen feedback so that air flow could be corrected in real time rather than after visible combustion disturbances had already developed.

Diagram showing CO gas analysis leading to real-time data flow to a PLC at the boiler, enabling automated air flow control for efficient combustion in the chamber.

The objective of this study is to demonstrate how smart air-fuel ratio optimization can improve boiler efficiency, stabilize combustion, and reduce recurring fuel loss. The paper is organized around a literature-based justification, a field-based methodology, root-cause validation, and a before-and-after comparison of operating performance. The practical emphasis is on converting a manually managed boiler into a measurable, repeatable, and standardized energy system.

Literature Review

Research on boiler efficiency consistently shows that heat loss through flue gas and excess air is one of the most common sources of energy waste. Improvement studies frequently combine heat recovery, oxygen trim control, and automatic combustion regulation to lower fuel consumption while maintaining safe operation. The sample paper on boiler efficiency improvement illustrates how flue gas heat recovery and automatic combustion control can produce measurable gains in efficiency and fuel savings [1]. The broader literature also supports the idea that energy savings are most durable when they are linked to measurement, standard operating practice, and clear control limits.

The literature also emphasizes the value of direct control of combustion air. Oxygen-based feedback is widely used because residual oxygen in flue gas is a practical indicator of whether combustion is too rich or too lean. Earlier studies on boiler controls and intelligent combustion systems show that better feedback reduces both fuel loss and process instability [2]-[6]. In practice, the benefit is not only higher efficiency; it is also better repeatability and fewer operator-dependent variations.

Another recurring theme is standardization. Energy savings are most durable when the new method is embedded into operating practice, instrumentation checks, and maintenance routines. For that reason, a technically sound improvement must also be a managerial one, linking measurement, training, and control discipline. This case therefore treats the boiler upgrade as both an engineering intervention and an operating-system improvement.

Methodology

The study follows a case-study approach based on an industrial boiler system using COG as fuel. The baseline condition was a manual operating method in which excess oxygen was measured periodically, and adjustments were made by operator experience. The improved condition introduced real-time oxygen measurement, PLC-based logic, VFD control of the air fan, and fuel trimming according to measured operating signals. A zirconia-based oxygen analyzer continuously measured flue-gas oxygen, and the PLC used that signal to maintain the air-fuel balance.

Root-cause analysis was used to separate assumptions from verified causes. The initial list of possible losses included poor mixing of air and fuel, incorrect air flow, incorrect fuel flow, poor quality of COG, pressure fluctuation, valve malfunction, feedback problems, insufficient residence time, and insufficient temperature. These causes were then checked by training review, inspection, scientific reasoning, and flue gas analysis. The purpose was to distinguish between symptoms and actual operating causes before changing the control system.

The improvement logic was simple but effective: measure the actual combustion condition, compare it against the desired operating band, and correct the air-fuel balance in real time. This made the process more responsive to variations in load and fuel quality, while also reducing the chance of unburnt carbon deposition in the flue gas path. Operators were trained on the new setpoint-based method, and manual override remained available for safety and maintenance situations.

Performance was evaluated using specific gas consumption, operating stability, and qualitative evidence of cleaner combustion. The improvement was also translated into annual cost savings based on the assumed operating conditions used in the project. The comparison therefore captured both technical performance and managerial value.

Two people in office chairs monitor multiple computer screens displaying technical data and control panels in a control room setting.

Results and Discussion

Figure 1 presents the conceptual framework of the proposed control model. The framework moves from oxygen measurement to PLC decision logic, then to VFD fan control and fuel trimming, and finally to more stable combustion. This is a closed-loop improvement because the process continually corrects itself based on feedback rather than relying on a fixed manual setting. The result is a clearer operating band and less dependence on operator judgment.

Table 1 summarizes the root-cause validation. The most important validated causes were poor mixing of air and fuel, inadequate air flow, and fuel flow mismatch. Several suspected causes such as improper operation, insufficient residence time, insufficient temperature, poor fuel quality, pressure fluctuation, valve malfunction, and feedback problems were found to be invalid after checking. This validation step was important because it prevented the team from correcting the wrong variables.

Table 2 shows the before-and-after operating comparison. The key shift was from manual oxygen checking to real-time oxygen measurement, which made air-fuel control more stable and repeatable. Specific gas consumption reduced from about 165 Nm³ per ton of steam to about 156 Nm³ per ton of steam, indicating a clear efficiency improvement. The operating data were also easier to review because the new system produced continuous feedback instead of occasional manual checks.

The practical effect of the intervention was not limited to lower fuel consumption. The process became more standardized, operator dependence decreased, and the risk of unburnt carbon deposition was reduced. The report also indicates annual savings of roughly 89 lakh rupees under the stated assumptions, although actual savings will depend on fuel price, load pattern, and operating discipline. The result is therefore both a technical improvement and a recurring cost benefit.

This result aligns with earlier boiler-control studies that emphasize excess-air management, oxygen trim, and automatic combustion optimization as high-value efficiency measures [1]-[6]. The case therefore supports the broader conclusion that measurement-driven control is more reliable than experience-based adjustment in continuous utility systems. It also shows that a small instrumentation upgrade can have a large operational effect when it is paired with proper operating discipline.

Table analyzing root causes with columns: "Probable root cause," "Validation method," "Evidence/remark," and "Status." Valid causes noted by gas analysis. — *Table 1. Validation of probable root causes.*

*Table 2. Before-and-after operating comparison.*

Table compares baseline and improved metrics: specific gas consumption, operating control, carbon deposition risk, and annual benefit. — *Table 3. Main performance outcome.*

Conclusion

This paper demonstrates that smart energy management can significantly improve a COG-fired boiler when measurement and control are integrated into one closed-loop system. The project moved the boiler from a manual, operator-dependent regime to a more stable and measurable operating condition. The practical effect was a more disciplined operating environment, not only lower fuel use.

The validated root causes showed that the major losses were related to air-fuel imbalance rather than the less likely causes that were initially suspected. Once real-time oxygen measurement, PLC logic, and VFD control were introduced, specific gas consumption decreased and combustion stability improved. This also reduced the likelihood of unburnt carbon deposition and supported smoother boiler operation.

The study shows that the value of an energy project lies not only in saving fuel but also in making the process easier to control, easier to standardize, and safer to operate. For similar industrial boilers, the same approach can be extended by maintaining analyzer calibration, reviewing operating limits periodically, training operators, and embedding the logic into standard operating practice. That combination makes the improvement more durable over time.

Author:

Ketankumar Patel
Energy Management Department
JSW Steel Ltd., Dolvi, Maharashtra

References

[1] Suntivarakorn R, Treedet W. Improvement of Boiler’s Efficiency Using Heat Recovery and Automatic Combustion Control System. Energy Procedia. 2016;100:193-197.

[2] Liao Z, Dexter AL. The Potential for Energy Saving in Heating Systems Through Improving Boiler Controls. Energy and Buildings. 2004;36:261-271.

[3] Gibbs BM. Boiler Fuel Savings by Heat Recovery and Reduced Standby Losses. Heat Recovery Systems and CHP. 1987;7:151-157.

[4] Wen S, Yongcheng J. The Research and Application on Intelligent Control in Boiler Combustion System. Procedia Engineering. 2011;23:167-173.

[5] Prokhorenkov AM, Sovlukov AS. Fuzzy Models in Control Systems of Boiler Aggregate Technological Processes. Computer Standards & Interfaces. 2002;24:151-159.

[6] Saha PK, Shoib M, Kamruzzaman J. Development of a Neural Network Based Integrated Control System of 120 ton/h Capacity Boiler. Computer & Electrical Engineering. 1998;24:423-440.

[7] World Steel Association. Energy Use in the Steel Industry. Fact Sheet. 2021.

[8] Jiang D, Wang Z, Li K, Zhang J, Ju L, Hao L. Predictive Modeling of Blast Furnace Gas Utilization Rate. Metals. 2022;12(4):535.

[9] Deng L, Adams TA. Comparison of Steel Manufacturing Off-Gas Utilization Methods via Life Cycle Analysis. Fuel. 2023.

FAQs

What is air-fuel ratio optimization in a COG-fired boiler?

Air-fuel ratio optimization is the process of maintaining the correct balance between coke oven gas (COG) and combustion air inside a boiler. In this case study, real-time oxygen measurement, PLC-based logic, and VFD-controlled air fan adjustment were used to automatically maintain the ideal combustion condition. This reduced fuel waste, improved combustion stability, and lowered the risk of unburnt carbon formation.

How does real-time oxygen monitoring improve boiler efficiency?

Real-time oxygen monitoring continuously measures flue-gas oxygen levels and allows the control system to instantly correct air and fuel imbalance. Unlike manual oxygen checks, this closed-loop system responds immediately to load changes and fuel variations. The study showed that this approach reduced specific gas consumption from approximately 165 Nm³/t of steam to 156 Nm³/t, improving overall boiler efficiency and reducing recurring fuel costs.

Why was PLC and VFD integration important in this boiler optimization project?

The PLC (Programmable Logic Controller) processed oxygen analyzer signals and automatically adjusted the VFD-controlled combustion air fan to maintain the desired air-fuel ratio. This automation reduced operator dependency, stabilized combustion, improved repeatability, and created a more standardized energy management system for continuous industrial operation.

What were the main causes of boiler inefficiency identified in the study?

The root-cause analysis confirmed that poor air-fuel mixing, insufficient air flow, and fuel flow mismatch were the primary causes of inefficient combustion. Other suspected causes such as poor fuel quality, pressure fluctuation, valve malfunction, and operator error were investigated and ruled out through inspection, scientific checks, and instrumentation validation.

What business and operational benefits were achieved after implementing smart combustion control?

The project delivered both technical and managerial benefits. Fuel consumption decreased, combustion stability improved, and unburnt carbon deposition risk was reduced. The system also improved operator standardization, simplified process monitoring, and enhanced boiler safety. Based on project assumptions, the improvement generated an estimated annual savings of approximately 89 lakh INR through lower recurring fuel costs and more efficient operation.