Project

Synapse: innovative operational planning enabled by Advanced Analytics

Context

Johan Maricq

Artificial Intelligence

Building tomorrow’s infrastructure and maintaining Elia’s grid and its equipment is crucial to guarantee the security of electricity supply for several million consumers. The energy landscape is in full transition and brings with it many challenges, such as the emergence of new constraints. These are the result of decentralized and intermittent production, aging assets and an interconnected European grid. The consequences of these challenges are double in terms of operational planning:

  • More interventions are required on electricity grids.
  • Fewer opportune moments are available to plan these outages and interventions.

The Synapse program aims to thoroughly rethink how our teams will plan, prepare and execute works.

Approach

Synapse uses a decision-support optimization tool, Opso, in order to propose a more agile planning of works when possible, but also a more stable planning when necessary.

Opso will propose an optimal outage schedule, while taking into account the different planning constraints. In the current process, teams work using annual outage plans: Opso will allow optimization of the whole outage plan when needed in order to be more flexible as soon as new information is available, such as delays in the maintenance of facilities, new work to be added, unplanned line outages, etc.

 

Description

In the solution, the different constraints encountered in grid activity planning are modelled as a large optimization problem. Different objectives (e.g. minimize energy at risk, maximize the number of planned activities) can be combined and optimized.

Due to the size of the problem to be solved (e.g. hundreds/thousands of works/outages to be scheduled), Constraint Programming has been selected and combined with Large Neighborhood Search as the ideal mix of technologies to solve the optimization problem.

Indeed, the choice of those two technologies allows for a great degree of flexibility in the definition of nonlinear constraints and objectives while keeping a reasonable execution time, which is key to ensuring frequent use of the tool by the end users.

Furthermore, with the tool, the grid planning operator is able to perform what-if analyses by changing the weightings of each sub-objective in the optimization function and by observing the impact of changing a parameter on the Key Performance Indicators.

Results

A proof of concept has been performed by the National and Regional Control Centers (NCC/RCC) of Elia, supported by the N-SIDE team. This proof of concept showed the ability of the tool to increase the number of grid activities planned while significantly reducing the stress on the grid (reducing the number of outage days).

During summer 2020, the tool was even used to help the operational planning teams reschedule the works canceled due to the COVID-19 pandemic. This field test illustrated one of the main benefits of the tool: being able to quickly reschedule activities affected by external factors.

Thanks to the success of this proof of concept, the optimization tool, Opso, is going to be industrialized based on an SaaS approach. The operational team of Elia will start using the solution during the first semester of 2021.

Partners

N-Side

Visit website

Interested in becoming our next partner?

Partner with us