Back to search

NAERINGSPH-Nærings-phd

Dynamic risk management for smart grids in large-scale interconnected power systems

Alternative title: MeasurEGrid: målbar sikkerhet og persornvern for tjenester i smarte energinett

Awarded: NOK 1.4 mill.

For most power grid operators, it is challenging to have an accurate and up to date overview of the condition of their grid, equipment, and power lines. In fault situations and for preventive and corrective maintenance of power grid infrastructure, power grid companies have traditionally relied on manual ground-based inspections (crews walking the lines and inspecting transformer substation) and helicopter-based inspections. This industrial PhD project aimed at investigating the use of the new generation of information technologies, available on both the hardware and software sides. These innovations enable, from a general perspective, automatic management and real-time processing of large quantities of data (big data) using machine learning-based solutions. In the field of power grid management, they also enable to develop advanced tools for dynamic risk analyses applicable on large-scale interconnected power systems. State-of-the-art methods regarding the evaluation of risk levels in industrial infrastructures and the way to model how this risk level evolves over time have thus been studied for the present project. A review of information enabling to support this project has also been done. In addition to information gathered from power grid operators, multiple sources of data have been analyzed and reported. These include satellite-based images, LiDAR point clouds, photogrammetry point clouds, aerial images, forest inventory from NIBIO, among others. The PhD work then focuses on how to manage datasets originating from those data sources based on the quality of the information they provide. In particular, we developed an approach enabling to compare the value of the datasets prior to their integration in the risk analyses. Furthermore, we have also developed several processes to combine the insights provided by the different data sources, enabling us to increase the level of knowledge an operator can obtain for a power grid under review, de facto supporting optimization of decision-making.

The main contributions of this thesis can be summarized are follows: (1) We diversify the panel of exploitable data sources for risk analysis and fully explore the analysis level scale. (2) We augment conventional risk assessment frameworks to enable efficient, large-scale heterogeneous data processing. (3) We provide multiple solution development propositions enabling power grid operators to make better risk-based decisions. The propositions are based on various perspectives and enable to find an adequate trade-off between global and local analyses of the grid, by always keeping the user-needs at the centre of the solution definition. (4) We make multiple recommendations usable by power grid operators to optimize the exploitation of historical data and the planning of future data capture. (5) We indicate how vegetation management along power lines may be improved using the different other contributions of this thesis. From a risk perspective, this Ph.D. first contributes to the understanding and clarification of basic risk-related concepts. The findings of this work then enable risk analysis processes to be more resilient to data capture issues. They especially enable more robust decision making by reducing uncertainties relative to data integration, therefore “better knowing how well we know”. In addition, the results strongly contribute to a better quantifiability of problems at scope in risk analysis, de facto enabling more objective decision-making. The thesis is also particularly valuable from a power grid management perspective. It first provides a familiarization opportunity with the notion of risk for the stakeholders requiring further insights in that field. It then shows how this knowledge can be used in combination with news data capture and processing solutions to enable the emergence of innovative tools supporting power grid operators in their daily operations. The final results are discussed, and different evolution opportunities are reported along with the provided contributions, such as executing risk quantification or analysing other hazards. The provided suggestions represent as many possibilities to reinforce and further extend the results of this doctoral project. They are also an indication that further development is required to facilitate more robust decision-making when practical implications and currently existing technical limitations are faced. Eventually, this thesis is a good illustration of the benefits of braking silos and encouraging cross-disciplinary cooperation. It stimulates power grid operators to further investigate the advances made in the academical world. At the same time, it also favours the communication of constraints faced in real-world situations but maybe too often excluded from the research scope in fundamental research.

For most power grid operators, it is challenging to have an accurate and up to date overview of the condition of their grid, equipment, and power lines. In fault situations and for preventive and corrective maintenance of power grid infrastructure, power grid companies have traditionally relied on manual ground based inspections (crews walking the lines and inspecting transformer substation) and helicopter based inspections. This industry PhD project aims at investigating the use of the new generation of information technologies based on big data, machine learning, and real-time processing, to develop advanced tools for dynamic analysis of risks related to smart-grids in large-scale interconnected power systems.

Funding scheme:

NAERINGSPH-Nærings-phd