Assessing remaining life of reclaimed transformer oil with magnesium oxide absorbent for biorefinery sustainability

The disposal of aged transformer insulating oils presents a significant environmental challenge due to the presence of diverse contaminants. This study proposes a novel biorefinery approach for reclaiming three aged transformer insulating oils, each with service periods ranging from 20 to 25 years, utilizing magnesium oxide (MgO) as a potent absorbent. The reclamation process involves a synergy of centrifugation, dehydration, and absorption with MgO, resulting in the substantial restoration of aged transformer insulating oils, imbuing them with desired insulating characteristics. Additionally, an adaptive neuro-fuzzy inference system (ANFIS) algorithm is developed to estimate the remaining life of transformers based on the properties of the reclaimed oil. This entails analyzing input variables, selecting the optimal combination based on their correlation with the degree of polymerization, and identifying optimal membership functions for the ANFIS model. The performance of the proposed ANFIS model is evaluated, demonstrating training and testing accuracies of 91.2% and 88.6%, respectively. Subsequently, the model is applied to assess transformer insulation and estimate expected life, offering valuable assistance to experts in transformer condition assessment. Overall, this algorithm presents a sustainable alternative to traditional methods for assessing transformer conditions and estimating expected life, thereby enhancing transformer maintenance practices and promoting environmental sustainability.