AI and Electric Power: Five Grand Challenges collaborate to move the dial across two transformative industries
- Aug 29, 2022 7:14 pm GMT
When EPRI’s Five Grand Challenges for the AI and electric power industries were launched last year, it planted a seed that continues to grow into a community of experts focused on raising awareness; developing, testing, and deploying algorithms; and documenting early deployment successes to accelerate widescale adoption of all those solutions across the energy industry. Hundreds of experts across government agencies, universities, tech companies, national laboratories, utilities, and more are coming together and collaborating within these five areas.
The grand challenge working groups are looking to execute on a number of highly impactful collaborations. For example, AI algorithms that could maximize and coordinate distributed energy resources (DERs) while minimizing customer’s electricity bills, an AI-powered platform for digital twins – virtual replicas of physical assets – that could reduce outages for generation plants, AI solutions that could help decarbonize the power sector by identifying optimal locations for renewables as well as EV charging stations, and energy storage facilities, among others.
These Grand Challenge working groups will meet later this year as a part of EPRI’s AI and Electric Power Summit to continue building and funding collaborative projects around AI in the energy industry. Collaborative work continues across these five grand challenges as detailed below.
Grand Challenge #1 – Grid-Interactive Smart Communities
This working group is exploring and testing AI for improving building-to-grid integration like:
- home energy management systems or local controllers to enable energy savings for customers faced with time-of-use type variable energy rates;
- aggregation of behind-the-meter flexible loads and DERs to provide distribution and bulk system services; and
- advanced AI/ML methods (e.g., reinforcement learning) with real-world data from field demonstrations.
The team is working closely with University of Texas at Austin and several other academic and industry partners to enable a pilot demonstration with EPRI’s behind-the-meter data to develop and benchmark advanced AI/ML methods. The team is also working with a sustainability community in Texas piloting a cloud-based data visualization and analytics platform that monitors building loads, geothermal infrastructure, and behind-the-meter DERs.
Grand Challenge #2 – Energy System Resilience
This group is cataloging AI applications for outage prediction, vegetation management, and equitable energy system resilience investments, among others. This year’s work includes:
- data extraction from thousands of infrared images, using neural networks with extracted features fed to a random forest machine learning model;
- geostatistical machine learning models leveraging tens of trillions of hourly weather data points to improve understanding of historical events and fire spread risk;
- modeling of complex grid-edge environments and psychographics for bottom-up scenario modeling based on incomplete datasets; and
- advanced data analytics to assess risk and customer access differences to inform equitable resilience investment decisions.
International stakeholders are collaborating to understand current resilience challenges and near-term AI opportunities. Long-term efforts include identifying long-run potential cost-savings to justify strategic investments, and improving integration of AI-based approaches into existing planning processes.
Grand Challenge #3 – Environmental Impacts
This group is looking at AI solutions to optimize decarbonization technologies for transportation, buildings, homes, and the grid.
“This group is among the first of its kind to facilitate connections and large-scale information sharing across industries, academia, and utilities,” said lead and Assistant Professor at the University of Colorado Boulder Kyri Baker. “We had a unique webcast in March that brought together representatives from the EPRI working group, Climate Change AI and the Global Power System Transformation Consortium. Many of these groups had never interacted with transmission system operators on how they are using AI solutions.”
The team is exploring use cases for improved forecasts for renewable generation; AI-enabled digital inverters connected to home energy management systems; reinforcement learning for demand-response simulation to capture people’s homes preferences; and learning travelers’ behaviors to better plan public transportation routes, and placement of EV charging stations.
Grand Challenge #4 – Intelligent and Autonomous Power Plants
While EPRI has done work in the digital twin space for nuclear plants, a new effort is being launched in collaboration with this working group to create a common platform for all generation digital twin developments. The aim is to provide a one-stop shop for generation operators where AI-enhanced digital twins will aid in component failure detection, performance monitoring, and maintenance forecasting. The research project team aims to have the first common framework version available by end of the year.
The team is also working to expand a natural language processing (NLP) dictionary for nuclear plants to all generation facilities and capture a broad range of industry knowledge and language. First initial trials of the NLP dictionary are planned for the end of November. NLP is also being looked at for automated asset information sorting.
Grand Challenge #5 – AI-Enhanced Cybersecurity
This working group is developing a comprehensive AI-enhanced cybersecurity strategy that energy system operators and utilities can deploy and scale across their service territories. Use cases include semi-automated security monitoring and response, event correlation and predictive analytics, vulnerability and patch management, intelligent threat hunting, sophisticated malware and supply chain malware detection, and advanced persistent threat detection.
Intelligent threat hunting, for example, leverages AI advanced analytics to gather cyber threat intelligence, analyze it, and determine likelihood of an attack, its impact and response measures.
Short-term, the team will develop use cases around security monitoring and response solutions. Patch management and detecting malware in supply chains are among longer-term efforts.
“AI’s time for the energy industry is here. We are all in a global energy transformation, and the time is now to make investment decisions in these game-changing solutions. It takes critical analysis and evaluation to determine the precise solutions for the unique environments where they will be deployed,” said Senior Vice President of Energy System Resources, Neil Wilmshurst. “Leaders and innovators from over 75 companies, utilities, and collaborating agencies will be sharing their use cases, the value they are seeing, and lessons learned in October at our AI and Electric Power Summit. Forums like these will enable widescale AI adoption.”
No discussions yet. Start a discussion below.
Get Published - Build a Following
The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.
If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.