Statistical Lifetime Modeling of Fe-Ni-Cr Alloys Subject to High-Temperature Corrosion in Waste-to-Energy Production Units

• Published in: Corrosion (Houston, Tex.) Vol. 71; no. 11; pp. 1360 - 1369
• Main Authors: Camperos, Sheyla, Brossard, Jean Michel, Floquet, Pascal, Monceau, Daniel
• Format: Journal Article
• Language: English
• Published: Houston NACE International 01.11.2015; National Association of Corrosion Engineers
• Subjects: Chemical and Process Engineering; Chemical engineering; Chemical Sciences; Combustion; Corrosion; Corrosion products; Corrosion tests; Economics; Electric power; Electric power generation; Electric power sources; Engineering Sciences; Ferrous alloys; Gases; Heat resistant alloys; High temperature; Incineration; Iron; Life assessment; Materials selection; Mathematical models; Modelling; Municipal solid waste; Municipal waste management; Nickel; Optimization; Prediction models; Predictive maintenance; Preventive maintenance; Principal components analysis; Regression analysis; Regression models; Solid waste management; Statistical analysis; Statistical methods; Temperature; Variables; Waste to energy; USA, Texas, Houston; Uniform corrosion; Corrosion rate; Statistical modeling; Hot corrosion
• ISSN: 0010-9312; 1938-159X
• DOI: 10.5006/1808

The incineration of municipal solid waste as the main process of waste-to-energy (WtE) plants is often associated with high-temperature corrosion problems. With the idea of further increasing the efficiency of electric power generation and reducing the total cost of WtE units, it is important to develop preventive maintenance strategies based on accurate predictive methods resulting in economic savings and resource optimization. The main purpose of this study is to propose a statistical methodology for lifetime prediction modeling over a wide range of conditions of these complex environments and discuss the results regarding the mechanisms described in the literature. In order to create a quantitative tool to evaluate material corrosion performances based on adapted corrosion tests and the definition of accurate criteria for life assessment, a database with 1,595 test results has been built from several published high-temperature corrosion studies. The data distribution was analyzed by descriptive statistic approaches; the procedure of principal components analysis was applied to determine the most important parameters that govern the corrosion process. The statistical results were compared with the experimental findings of the authors to create a model by multiple linear regression analysis whose accuracy and physical interpretation are discussed.