This article was originally written in French for the Annales des Mines review.
A moment of truth has arrived for electrical networks: a new future beckons with the digitalization of energy and the energy transition accelerating, just as the new tools to control systemic risks and uncertainty of this future are being developed and deployed. Powerful technological capabilities will be or already are in the hands of energy actors to help them drive short-, medium- and long-term grid transformation, and measure the significant impact their decisions will have on the world around us.
On February 15, 2021, extreme weather in Texas, USA, demonstrated the growing impact of climate change on grids. Faced with plummeting temperatures, snowstorms, and a cascade of power plant outages, Texas nearly lost much of its electrical grid for weeks. This energy crisis also highlighted the consequences of the Texas power grid’s lack of connection to the rest of the US grid.
This event firmly underscores the systemic challenges and the level of uncertainty facing energy executives. Weather events increase the level of financial risk for grid companies, while they concurrently face three other major challenges:
Complexity, systemic risk, and uncertainty are the new normal. Gaining visibility, anticipating the future, and enabling a faster, more agile response are absolute necessities to manage this new reality. This situation can be worrisome, but it is also a source of immense opportunity for value creation. Technologies are now ready and have a great role to play in helping organizations make more confident and effective decisions, as well as in strengthening their ability to adapt in real-time. In other words, they are the new source of resilience.
Many players have already invested in big data and the Internet of Things (IoT). Consumer data is available through smart meters, and network data is also more accessible.
The use of artificial intelligence (AI) is also progressing in the management of the electrical system, with the development of AI assistants to simplify the work of operators for thousands of repetitive tasks (for example, testing a switchgear) by automating the sending of work orders and the retrieval of information.
The digital twin ecosystem is growing rapidly in the energy sector. The new generation of Simulation Digital Twins incorporate three fundamental advances to address the systemic challenges of electrical networks:
Simulation Digital Twins today provide three levels of understanding, described in the diagram below: the description of what exists and what is happening, the prediction of possible futures, and finally the prescription of optimization and action paths.
By combining these levels, Simulation Digital Twins help industry to move from reaction to anticipation, then to better control and even achieve a certain level of automation of overall performance.
These Simulation Digital Twins have a holistic approach to the network and consider all the diversity and heterogeneity of the real system such as assets, teams, environmental and regulatory policies, investment budget constraints, as well as external events such as weather conditions, external supplier constraints or the evolution of prosumer1 demands.
The more the digital twin allows companies to understand and optimize their decision-making (horizontal axis) in a systemic way, considering all the complexity of the considered reality and its constraints (vertical axis), the more the understanding and the holistic value provided will be important.
1.A prosumer refers to a network user with a decentralized electricity production facility that is likely to inject and withdraw electricity from the network.
Simulation Digital Twins take a different approach than existing AI solutions. The real-world modeling, symbolized in the diagram below, is not only based on historical data, but also on network processes and constraints determining the interdependencies and causality rules between the elements of the simulated system. The structure and dynamics of the organization are thus replicated in all its complexity, to give a truly 360° view of the organization and simulate its future behavior, even under conditions that have never occurred before.
In the digital twin above, strategic and operational network decision-makers can run thousands of simulations to predict the future state of the system and test both What-if or How-to scenarios:
These simulations can project the evolution of the system for the next 30 minutes or 30 years, depending on the needs of the decision-maker and the specific use case.
The holistic simulation and optimization capacity of Simulation Digital Twins represent a significant advance for network managers over traditional tools in determining optimal network management and transformation plans, considering their feasibility and robustness to changing conditions.
Conventional optimization tools can already determine optimal power system management policies but provide no assurance that these policies are feasible. An optimal policy might be to replace all aging assets in the same year, for example, although this is not practically possible when you consider outage capacity on the grid.
Simulation Digital Twins also predict more reliable results than machine learning tools that are based exclusively on past data, especially as the timeline for action extends further into the future.
For example, TenneT, the Dutch and German Transmission System Operator (TSO), used Cosmo Tech’s simulation digital twin technology to explore how to optimize the maintenance and life cycle of its towers. In this case, the modeling integrated the life cycle status of the towers, including corrosion development factors and all mission-critical elements (including work schedules, financial projections, and planned service interruptions). The simulations demonstrated that the TSO could achieve a reduction in network operational risk of 12% while maintaining the same resource allocation.
Since they reproduce the behavior of the network, Simulation Digital Twins offer the advantage of testing the possible evolutions of the network and optimizing its robustness to changes.
RTE, the French electricity TSO, has been using Cosmo Tech’s simulation digital twin technology for several years to simulate and optimize its asset management strategy across several use cases. RTE tests the effect of its maintenance and investment policies at different time scales according to the scenarios considered, notably identifying the impacts on the health of the network, the quality of service, and the network’s environmental performance. The results of these simulations have informed RTE’s teams about their ability to invest and maintain their assets in an optimal way for the years to come and to assess the robustness and resilience of the power system in line with different risk scenarios.
RTE has used this technology to justify its business plan to the French Energy Regulatory Commission (CRE). By sharing objective data and showing that increasing the maintenance budget would result in substantial savings on the overall budget in the medium term, RTE was able to obtain a 15% budget increase for its asset management from the CRE. In its public report, CRE encourages the dissemination of this approach.
Thanks to simulation digital twin technology, uncertainty can be controlled and robustness optimized by quantifying the consequences of that uncertainty. For example, the impacts of coastal weather conditions on the aging of metal towers as well as the impact of storms can both be considered in scenario simulations.
The robustness of maintenance and asset renewal strategies can also be optimized according to potentially large variations in the evolution of the network. Depending on the transformation of the network, electricity flows can change considerably and a power line that is secondary today may be essential in fifteen years to supply three or four wind farms, or to ensure the stability of a city’s power supply when a nuclear plant has closed.
Simulation Digital Twins can be connected to the ecosystem they represent to trace the past, monitor the present, predict the future, and point to the optimal way path forward.
Significantly, this connection to the real world is bidirectional.
On the one hand, the digital twins are up to date with reality to serve as an accurate representation of the current operational state of the system. The organization is thus assured that the decisions taken for the future correspond to the realities of the moment.
On the other hand, a feedback loop, symbolized by the control arrows in the diagram above, sends the optimization results back to the real-world system. This connection allows a company to automate certain decisions or to autogenerate action recommendations for validation, thus accelerating the agility of the organization and allowing teams to focus on tasks with higher-value tasks.
This dual connection further enhances the capabilities already at work in Simulation Digital Twins to break down organizational silos and merge data from previously disparate devices and systems, so capturing the full value of the IoT and artificial intelligence solutions already implemented.
Simulation Digital Twins bring a new way of understanding and managing organizations and electricity systems. They provide unique visibility into the effects of decisions, whether applied to day-to-day operations or strategic considerations, considering the cascading effects of unforeseen events.
With the digitalization of energy, the necessary acceleration of the energy transition, and the rise of uncertainty, energy players will increasingly need to simulate their system, as well as their ecosystem, in all its complexity. As a result, Simulation Digital Twins will play a central role in enabling those players to react in a faster and more agile way, and in providing the visibility and understanding to make optimized choices at each stage of the network transformation.
These technologies are not intended to replace operator expertise. On the contrary, they will augment it, providing operators with additional scope to redirect their actions more easily or innovate in response to changing needs. They will provide considerable support to public or private energy managers to become agile drivers of change and to manage the complexity, systemic risks, and uncertainty that await us.