Conversion of non-edible Baobab oil as an alternative cooling medium in transformers and practicability analysis of ageing using Euler’s method and artificial neural network

The use of natural esters as an insulation medium in transformers is well known. This paper proposes to use Baobab oil (Adansolia digitata) as insulator. The Baobab oil is modified using the additives such as hydrocarbon -C₁₂H₂₆-C₁₅H₃₂ (AD1) and aluminium-II-oxide (AD2) of size 10 nm. The modified natural ester (NE) called nano natural ester (NNE) shows good dielectric strength at 7.5 mL/L of AD1 and 0.75 g/L of AD2. NNE is prepared by blending AD1 and AD2 using ultrasonic waves. Here, in NNE, AD1 serves as thinner, and AD2 acts as electron scavenger. The NNE is then subjected to accelerated heasting at 100 °C for 720 h. The dielectric properties like AC breakdown voltage, fire point, kinematic viscosity (at 30 °C), acid number and water content are measured at 120 h. The data is then used as an input to Euler’s method (EM) for predicting the ageing rate for 1440 h. The prediction results of EM are confirmed using the artificial neural network (ANN); ANN shows significantly less deviance. The prediction results of NNE, NE and conventional insulating fluid (CIF) like mineral oil are compared, which indicates that NNE ages at a slower rate than CIF. Under lower temperature loading, NNE ages slower than the predicted results and shows endurance beyond 2880 h. Hence, the use of NNE is a cost-effective and suitable solution to insulating fluids in power and distribution transformers.