Introduction
The power sector is challenged to keep the electricity network reliable. With an increasing number of renewables and smart electronics coming in the system, the power grid must cope with more and dynamic stress each year. To prevent outages from happening it is essential that grid operators can locate and respond to faults in the grid quickly and can eliminate weak spots before they fail. DNV GL has developed its upgraded Smart Cable Guard 2.0. Smart Cable Guard (SCG) is a system for online detection and location of partial discharges (PD’s) and faults (full breakdowns), of which its origins were developed together with the TU/e. PD’s are a phenomenon that is an indication for an upcoming power failure. Windfarms are struggling with the performance of their grid and worldwide the industry is looking for a solution to measure the health of the windfarm grids. SCG start to be recognized that it could be a solution for Windfarms, however the power electronics in a wind turbine is producing a lot of noise. This significantly hampers the detection sensitivity of Smart Cable Guard, thus reducing the effectiveness. Currently, noise pulses are also incorrectly recognized as PD pulses. This makes interpretation of the measurement results difficult, or even impossible. Extreme noise conditions not only happen in wind farms, but also in industry.
To increase the market potential and the performance of SCG we need to improve the noise filter to improve the detection sensitivity in extreme noise conditions and to classify detected pulses so that noise pulses can be rejected by the measurement system.
Further information on SCG and how it is being used at present can be found at: https://www.dnvgl.com/energy/transmission-distribution/scg/system.html
Project description
The project starts by gathering and analysing noise patterns from wind farms and industry using installed SCG systems. Existing models of power cable systems can be used to predict PD signals. Based on previous research and literature study potential noise reduction and signal classification methods must be selected. These methods must be evaluated using the gathered noise and signal patterns.
The goal of the project is to develop methods that can be implemented in the existing SCG 2.0 hardware.
The project will be done in a cooperation between the two groups EES (Electrical Energy Systems) and SPS (Signal Processing Systems) of the faculty Electrical Engineering of the Eindhoven University of Technology.
Candidate requirements
- Working towards a MSc. in Electrical Engineering
- Knowledge in digital signal processing
- Pro-active and self-motivated
- Good written and spoken command of English is mandatory.
- Previous internship experience not necessary
Contact For further information, please contact from the group EES: prof. Fred Steennis (email: e.f.steennis@tue.nl) or from group SPS: Dr. Alex Alvarado (a.alvarado@tue.nl)