Cumulative damage-based diagnostics and control of oil-immersed power transformers

Abstract

Relevance: oil-immersed power transformers are among the most expensive and critical components of power supply systems, and their failures cause significant economic losses and long outages. Existing online monitoring devices usually track a single physical process (the thermal regime or gases dissolved in oil) and cannot express, in one indicator, the damage accumulated from short-circuit currents, vibration and thermal ageing.

Aim: to develop a mathematical model, an algorithm and a real-time diagnostic device with operating mode control that account for cumulative damage caused by short-circuit currents, vibration acceleration and hot-spot temperature, and to evaluate their effectiveness.

Methods: the Palmgren–Miner linear damage accumulation rule, the Arrhenius equation, Fisher’s criterion, the exponential reliability model, simulation in MATLAB/Simulink and a one-year experimental study on a 20 MVA, 110/6 kV transformer.

Results: after introducing control, the switching damage share decreased from 8.75 % to 4.86 %, vibration damage from 4.0 % to 1.6 % and thermal degradation from 3.28 % to 0.30 %; the total annual damage fell from 16 % to 6.8 % (by 57.5 %). The one-year probability of failure-free operation increased from 0.852 to 0.934, the annual economic benefit amounted to 219 million UZS, and the payback period of the device is less than one month.