IMPROVING CORROSION PENETRATION RATE ESTIMATES THROUGH META-ANALYSIS

Corrosion Penetration Rate (CPR) is a critical parameter in the oil and gas industry, as it directly impacts the safety, reliability, and operational costs of pipeline systems. In recent years, numerous studies have proposed various predictive models, including Artificial Neural Networks (ANN), Fuzzy Logic (FIS), Optimized ANN (LM), Hybrid ANN-FL, and Adaptive Neuro-Fuzzy Inference Systems (ANFIS), to estimate CPR under different operational conditions. This meta-analysis aims to synthesize the findings of 5 key studies, providing a comprehensive assessment of the predictive accuracy of these models. The analysis included a total of 166 data points, with sample sizes ranging from 28 to 40. Effect sizes (Cohen's d and Hedges' g) were calculated for each model to quantify the magnitude of the predictive power. The results indicate that the Optimized ANN (LM) model demonstrated the highest effect size (Hedges' g = 2.20), suggesting superior predictive accuracy, while the ANFIS model, despite its smaller sample size, also exhibited strong predictive performance (Hedges' g = 1.75). The overall effect size across all studies was found to be significantly large, confirming the robust impact of these models on accurate CPR prediction