ENERGYTICS
Maintenance and reinvestment decisions
The network operators spend substantial amounts of money on maintenance each year, and it is expected that the electricity grid will require substantial upgrades in the years to come to meet an increasing electricity demand and higher power peaks. It is therefore important that they know which components in the grid should be repaired and which should be replaced to achieve an optimal balance between reinvestments and maintenance. This process is not automated today, but ENERGYTICS will look at possibilities for this to become possible.
In the distribution network, the total value of the infrastructure is particularly high, while each component has a relatively low cost. This is in opposition to the transmission network and the production sector. In addition, the instrumentation in the distribution network is very low. Compared to other projects, ENERGYTICS will be able to compensate for this situation by taking advantage of big data sets from participating network operators, a powerful Big Data infrastructure through a cooperation with Microsoft, and the project partners' expertise in Big Data analytics and data science methods.
The focus will be to obtain knowledge about the value of applying data driven technologies like predictive maintenance and predictive analysis on AMS data and other data from the grid infrastructure. In addition, the project will strive to exploit technologies like text analysis, segmentation and classification to collect information from the large amounts of unstructured data associated with network components which the partners in the project possess. This includes event logs and condition observations. Correct decisions of reinvestment will be based on actual measured data, inclusion of external data (e.g. temperature), and techniques for forecasting, clustering and characterization of components in the electricity network.
Demonstrators: