The paper presents a logistic regression model developed to predict the need for corrosion inhibitors in coke plants based on real-time water quality monitoring. Given the critical implications of corrosion in industrial environments, efficient management using inhibitors is paramount to prevent equipment damage and operational disruptions. Traditional methods typically rely on scheduled maintenance and periodic chemical treatments which may not align with actual requirements, potentially leading to resource wastage and delayed responses. The proposal model integrates continuous data on key water quality indicators such as alkalinity, hardness, chloride, and phosphorus levels, to forecast the necessity for inhibitor application. This approach not only promises enhanced operational efficiency and reduced environmental impact but also provides a data-driven strategy to corrosion management. Through rigorous validation against observed events, the model demonstrates a good predictive capability, making it a valuable tool for plant operators. By optimizing the timing and quantity of inhibitor usage, the model helps in maintaining system integrity and avoiding excessive chemical use, thereby supporting sustainable industrial practices. The findings suggest significant potential for adopting logistic regression in dynamic and complex industrial settings for preventive maintenance.