In this week’s edition of Let’s Talk Energy, Arun Nimmala, Oracle’s Global Head of Grid Operational Technology Products and Services and Tom Eyford, Global Industry Specialist, Utility Operations Solutions for Oracle look at the rising stresses to power grids from the growing frequency and severity of extreme weather and wildfires alongside rapidly rising demand from data centers. 

Stream the full episode now: How data centers and natural disasters are pushing the grid to its limits 

NB: How is AI changing the way utilities predict, manage, and respond to extreme weather events? 

Arun Nimmala (AN): "AI can help utilities to become more reactive, more prepared in response to these as well as making the decisions faster. We cannot put humans in a place which is not safe. That's where the software comes into picture. For example, the Oracle ADMS [advanced distribution management system] solution gives the telemetry and the SCADA [supervisory control and data acquisition]. Now, that information can be an input to how to react to these storm events for those control room operators, or even the field group. Coming back to AI, AI can actually make the utility operators more anticipatory. AI has improved prediction models, combined with weather forecasts, even the vegetation LiDAR data, asset age, and the historical patterns. We can actually generate a score, which helps us to understand the risk of each asset. We have seen where the deep learning frameworks have shown up to 35% reduction in load shedding during extreme weather events."

 NB: How does managing a large data center's power needs differ from a typical storm-restoration problem? 

Tom Eyford (TE): "I'd probably separate it out a little bit because, from a traditional storm restoration perspective, a large data center is probably not the thing keeping the distribution operators up at night because typically they're so big that they are not served by the distribution system and they generally speaking are going to have their own facilities served directly on the transmission system.  

 I think where it really becomes an issue for the system operations is the fact that they're just tremendously big loads. A single large data center can represent hundreds of megawatts or more and we're increasingly talking about campuses of these. That [load] showing up and going away on its own can have major impacts on the grid. We spent more than a century thinking about what happens when we lose a large generator - a data center represents that size of impact to the grid as well. It's definitely an event that can have its own stability issues. We have to think about that large scale, but also the rate of change. So if it all of a sudden drops off, that's a problem for the grid. We think ‘Oh hey, that's just load going away. We should be fine.’ No - we have to match generation and load. Suddenly whatever we just lost, we have to immediately remove that from the generation mix. So the more warning that we have, the easier we can ramp that down. We're working with those data centers so they don't lose large chunks of load immediately. 

 NB: Should data centers be treated differently from other kinds of grid load? 

AN: "Most of the utilities or most of the grid operators look data centers as a different load. Data centers definitely are a different load, not just from the energy consumption point of view, but they also require finer grain visibility. That's why we always say data centers, particularly any energy demand coming from data centers, they should be looked as grid participants, not just as a passive unit."

 NB: Could facilities with their own on-site power actually help support grid stability? 

TE: "Absolutely. We hear people talk about behind the meter (power) - it doesn't necessarily mean disconnected from the grid. In fact, most of the distributed generation being added today is interconnected with the grid. It just happens to sit on the customer's side of the meter. Now that wasn't always the case. I think people still kind of have in their head the concept of a plug-in portable generator that can't be connected to the grid. That's really not the case for anything new, and anything that's permanently installed. The biggest value we can get out of those, both from the grid perspective and from the customer's own perspective, is to have that interconnected full time. That unlocks the value for the grid because I can then leverage that to enhance the grid stability, and when there are issues, I can potentially call upon it. And the customer then has access to provide that service and sell their excess energy, as needed, to the grid. And that's value for both sides."

 NB: Is AI-driven, autonomous grid management a reality today, or still a story for the next few years? 

AN: "I would say it is a reality today. Maybe not every piece of it, but there are lots of building blocks which are already there. For example, we have tools like switching, voltage optimization, FLISR [fault location, isolation, and service restoration], which are already part of the ADMS application. That's the primary building block. Now, additionally, we also have the AMI [advanced metering infrastructure]-based outage detection tools, and large-scale commercial VPP [virtual power plant] programs coming from DERMS [distributed energy resource management systems]. These are all already being implemented by utilities currently. That's why I say it's part of the story today. The building blocks are there."

 NB: What's different about wildfires compared to other extreme weather events utilities have to prepare for? 

TE: "The response to them can be pretty similar, but there's a huge difference these days, and it's really about liability. This is pretty much the one major event that they [utilities] potentially could be responsible for, so this changes everything in terms of how they prepare, how they operate, and how they interact with the public. I would say it's unique within the electric industry in that we operate 1000’s of miles of energized equipment outdoors through forests, through communities exposed to wind and vegetation. We can't practically engineer that risk to zero. We used to, as society, accept that as a community risk because wildfires - even utility-caused wildfires - have been happening since we've had utilities. What's changed is society has moved from ‘Yeah, it's just something that happens’ to ‘Hey utility, you should have prevented this, you should have fixed this.’ We've now seen in several cases now billion-dollar lawsuits to the point where this is potentially even an existential risk to the utility. We've seen even bankruptcies as a result of this." 

Stream the full episode now: How data centers and natural disasters are pushing the grid to its limits.