An honor to be the opening keynote at the AESP #AESPSummerCon plenary session this morning. Attendees were world class. Great to be in the same room with such a high caliber audience!
Topic was on the intersection between #DemandFlexibility, #GridoftheFuture and #AI. Sharing some of my main talking points:
1. In the demand world we structure around the 3P's - pricing, programs, and procurements. To innovate beyond existing fixed structures, we need to peel back and look at how the grid works, and where the grid is going.
2. Megatrends can be summarized as the energy trilemma (energy security, sustainability and affordability), 3D + E (decarbonization, decentralization, digitalization, electrification), grid as the platform (push/pull effects with the edge), and time (we're running out!).
3. Electrification is driving change in particular, as exemplified by an expected tripling of demand (data centers, electrification of transportation and heating), while global grid investments need to quadruple from today's level to $4T every year to 2030 in order to hit net zero by 2050.
4. The grid of the future can be described by the 3 dimensions of flexibility, automation and value.
5. Flexibility = variability + an actor/driver/mover, whether that is a person, control system, a contract, or price signal. This vastly complicates analysis from a quasi-static time series problem to a huge multi-dimensional optimization problem.
5. Automation is required to overcome the insurmountable volume and complexity of analysis required (e.g. NWAs, 8760 analyses, pre-processed interconnection assessments, generative recommendations of what/how much/where to connect).
6. Flexibility analysis will unlock a new level of dynamic, locational, stacked value that will evolve the 3P's.
7. The only way we can achieve this is through AI. Not just back office productivity AI, but teach AI how to do power flow engineering, which is a foundational underpinning of how utilities make decisions in planning and operations.
8. Analogy with Google Maps, smart watches, ChatGPT, and Spot - continuous, automated, generative analytics that can be deployed centrally and distributed.
9. Priority use cases we've identified includes Dynamic Planning, Model Validation, Grid Orchestration, Digital Assistants, and Distributed Intelligence.
10. Amara's law - as humans we tend to overestimate the short-term impacts of technology advancements while underestimating their long-term implications. AI is a hype today, but if we laser focus, we'll find truly transformative use cases that will shape the energy sector for the next decade to a level that Thomas Edison will never recognize.