Master Thesis: Fault Tolerance in Smart Grids: Simulation and Recovery Approaches


With the increasing number of renewable energy producers connected to the power grid, the necessity to modernise the current grid to a Smart Grid increases. A Smart Grid consists of a power grid and a communication network. The purpose of the commu- nication network is to be able to control the large amount of small production and consumption units. The extended usage of the communication network implies that a lot of messages are sent through the network in order to reach every Smart Grid par- ticipant which increases the probability of communication faults to occur. Such faults can have a high impact on the stability of the power grid, due to its strong connection to the communication network. In order to be able to investigate the consequences of faulty communication and test possible fault countermeasures, we want to develop a Smart Grid simulation framework, that is able to simulate faulty communication, such as: corrupt messages, manipulated messages, Smart Grid component failure, etc. We simulate several Smart Grid scenarios and we present an evaluation tool that displays the performance of the tested fault handling strategies. Our simulations show that even basic strategies that go beyond a simple history lookup, can recover 80% of the faulty messages. Additionally these strategies perform a stabilization of the power grid that is up to 64% better than the standard baseline strategy that takes the last value from the history. Furthermore the provided simulation and evaluation infrastructure allows the users to easily simulate their own communication faults and fault handling strategies.

Author: Luc Weiler, 2015

Master Thesis Luc Weiler (English)