In this project we will develop a physics inspired minimal stochastic model for partial discharges from first principles. We will begin by modeling PD activity originating from a single void with a fixed geometry. The initial goal is to generate partial discharge distributions that correspond to the measured data. A challenge is that the distributions are not stationary in time.
The final model should reproduce key statistical features of measured PD data and provide a physical interpretation of observed variations over time
Potential extensions of the project if time permits:
Thermal effects: temperature affects the material properties of the insulation material including the inception threshold for partial discharges and reaction rates creating feedback loops.
Void creation: voids are created in insulation material due to several mechanism such as mechanical stress and waxing. This formation in itself can also be modeled as a stochastic process.
Identify partial discharge events in raw partial discharge waveform measurement data.
Objectives
Perform a scan of the scientific literature on this topic
Build a minimal model that generates partial discharge distributions from a given geometrical void in a cable insulation.
Perform simulations using for example Monte Carlo techniques to generate partial discharge distributions and compare against measurement results.
Background in mathematics physics or electrical engineering
Knowledge of numerical modeling techniques of power grids (stochastic processes electrostatics etc.).
Experience with Python
Stagebureau Alliander
Alliander operates energy networks which distribute gas and electricity to large parts of the Netherlands. With our work we facilitate businesses, homes, transport and recreation.