Energy Transition - New Studies for new challenges in System Operation

In recent times, experts in the energy systems industry have been directing efforts to understand the new challenges and possible new solutions that emerge in the ongoing irreversible process of energy transition.

 

Classical problems such as the reduction of system-synchronous inertia are objects of constant attention. On this topic, our SC.C2, together with SC.C4, proudly announce the recent publication of Technical Brochure nº 851, the result of the completion of the work of JWG C2/C4.41.

 

We continued our efforts to anticipate further (technical) challenges to be overcome and, because of the complexity surrounding the energy transition process, we increased the variety of study focuses with two new Working Groups starting their activities.

 

The Impact of Electricity Market Interventions by System Operators during Emergency Situations

 

The Joint Working Group C2/C5.06 aims to study the impact of market interventions by System Operators during emergency situations. On one hand the interventions in the market during emergency situations may help System Operators in restoring the system back to normal state, which is positive, but on the other hand, Market Parties will be impacted by the interventions which is seen as negative. It will be investigated what the impact is both for System Operators and Market Parties.

 

It is common practice for System Operators to have the option to intervene in energy and reserve markets during emergency situations. The objective of these interventions is to prevent ongoing market activities to further deteriorate the situation and/or adversely influence the restoration process. Many System Operators may have limited experience with intervening in markets and therefore it is important to learn which practices exist and what can be learned from actual situations where it was applied.

 

The impact on Market Parties is of a non-technical nature and is expected to be to a large extent financial due to imbalances and/or missed trading opportunities. This might in turn impact the System Operators in case such financial impacts on Market Parties should be compensated post event. Such interventions by the System Operators may also be subject to subsequent scrutiny by the regulators and other stakeholders. An even important question is whether the market intervention actually does support the system restoration process and if so, then to what level.

 

To get a better understanding, the Joint Working Group will identify existing practices, methodologies and procedures and how they work. Apart from reviewing previous work on the topic also a worldwide questionnaire is planned to collect information and insights from actual experiences. There will be focus too on how the markets should be re-started after the emergency situation is solved. This could result in an overview of pros and cons.

 

The found effects and impacts as well as the lessons learned will be compiled. This should lead to recommendations on how and when System Operators could use market interventions and also when not to intervene. The work of this JWG is expected to support rule and policy makers in making their decisions when designing rules for market interventions.

 

The Joint Working Group started its work late 2021 and plans to publish a Technical Brochure mid-2023.

 

The impact of the growing use of machine learning based Artificial Intelligence in the operation and control of Power Networks from an Operational perspective

 

The latest Working Group in C2 with the title above aims to understand where this growing area of advanced yet complex AI-based computations intersect with power systems and what this means for operating power systems.

 

During the last decade, AI has indeed shown new promises for augmenting and assisting industrial system operations. Data center cooling, large and coordinated warehouse logistics, large scientific machine operations are such examples of pre-existing systems that integrated AI in operations. Autonomous vehicle, drone or balloon fleet management are also emerging system operations with advanced built-in AI developments. This is happening given the digitalization of such systems, the availability of larger computational resources but also the accessibility to new advanced models for Natural Language Processing, Computer vision, Recommender Systems and Automatic Control.

 

As the complexities of operating a modern power system grow, more and more parties within the end-to-end energy lifecycle are also turning to either machine learning or AI to assist in managing these complexities. This is true from consumers and large-scale users being offered services to manage their energy usage to suppliers and generators utilizing advanced algorithms to optimize their portfolio. Also new entrants such as Tesla are more likely to be from non-traditional energy background who are more familiar with utilizing advanced computational techniques. This utilization is not just limited to the downstream operations, distribution and transmission utilities are also exploring the usage of these new technologies to optimize their demand and generation forecasting to outage planning or to help operators with situational awareness and decision-making support tools in congestion management (as showcased in “Learning to run a power network” competitions).

 

This Working Group will first seek such examples of emerging “new” AI uses for system operations in the power system community. It will distinguish those applications where AI only provides predictive information for augmented observability, to those where it actually makes recommendations for decision or those where it acts autonomously. It will also gather and describe successful examples or failures from other industries using AI for system operations, as well as lessons learnt, as they all possibly come with their pros and cons in terms of operator acceptance. Considering those different and mixed examples, this WG will highlight in particular what this implies in terms of requirements for digitalization, data and IT processes or organizational practices. These examples and recommendations will be reported in the Technical Brochure. 

 

As with all new and emerging technologies and the rush to embrace the advantages these bring, it is important to understand any risks associated with their adoption. This is especially true for the energy industry as the financial and societal impact of getting it wrong are substantial. Other industries like the financial markets, in which the use of Machine Learning has become the norm, both the World Economic Forum (WEF) and the British Financial Stability Board have issued concerns about the growing usage of the technology, and its elimination of human judgement. The loss of the human in the loop is reducing operational understanding of the impact of potential outcomes recommended by the AI. Therefore, it is becoming critical and operationally important to all to understand the level of maturing and adoption of these technologies, their requirements, potential future growth areas and the risks and impact of their usage. We need to better understand which information we are using to making operational decisions while assisted with AI.

 

The Working Group will hence further develop an impact assessment of this usage from System Operations and Operators perspective, and the identification of any associated risk for system operations. We indeed aim for more transparent and trustworthy AI that enables synergetic human machine collaboration instead of opaque and possibly brittle predictions or recommendations. A specific focus will be dedicated on the needs of such a collaborative assistant for control room operators, illustrated through realistic scenarios. These will be included in the technical brochure, and guidelines for Human-centric AI to meet human acceptance will eventually be shared.

 

This exciting new Working Group will be working over an accelerated 2-year span, as this is the relevant timescale to be up to date with the fast-pace development of AI today.  We are looking for a wide range of contributors from those with a deep understanding and practical experience of AI/machine to those with operational experience of operating power systems at all voltages who want to understand the impact of this growing field.