Maintenance planning of transmission assets under uncertainty for long-term horizon

Abstract

Outage scheduling for maintenance in asset management of electrical transmission and distribution system plays an important role in power system reliability. For instance, failure probability of transmission line may change from time to time due to exogenous conditions. Impact of failure or service interruption can be described by using uncertainty condition while designing the long-term model for outage scheduling. Within a year horizon, a transmission system operator needs to schedule the maintenance outage of set of transmission lines due for maintenance. This is important because transmission line maintenance schedule ought to minimize the total maintenance cost and transmission provider’s loss of revenue while satisfying the reliability and the network requirement. It is expected that in coming years, there will be substantial increase in renewable energy in-feed to the primary grid. Combined with increase in demand, high level of uncertainty from both renewable as well as demand can be predicted in the system. Definitely, transmission system operators (TSOs) have to tackle such increase in demand and generation while addressing security of supply (SoS); thereby transmission assets will play an important role since TSOs are not in favour of new investments. In order to maintain such reliable system with SoS, TSOs ought to have a proper and flexible maintenance scheme for their transmission assets. The scheduling scheme should be able to determine the exact transmission assets in the cluster of network which can be brought out of service for maintenance. The scheduling scheme should be accurate and fulfil the required constraints, both maintenance and network, while keeping the system reliable throughout the maintenance horizon. To solve such maintenance scheduling problem, benders decomposition technique is adapted incorporating uncertainties. Uncertainty plays a crucial role in formulation of the scheduling problem and has been given due consideration. Stochastic programming provides an adequate modelling framework in which problems of decision making under uncertainty are properly formulated. Optimization under uncertainty, spanning two-stage stochastic programming approach is used in this research study. For validation, small (RBTS 6-bus), medium (modified IEEE RTS-24 bus) and large (modified IEEE-118 bus) systems are studied, all in the GAMS environment.