Active Grid Response: Real-Time Visibility Drives a More Effective Energy System

After decades of consistent demand, the electric grid is entering a period of tremendous load growth driven by AI data centers, industrial electrification, and electric vehicles (EVs). Navigating this challenge often seems like trying to steer a car by looking into the rear-view mirror: Grid operators have plenty of data about what has happened and what has worked. But to progress through a landscape of weather extremes, consumer engagement, and distributed energy resources, grid operators need a better view of what’s happening right now on the grid and, ideally, what’s likely to happen next.

Visibility of activity at the edges

For decades the industry discussed the promise of additional data streams to understand conditions on the grid. We have deployed smart meters at most premises, for example, and this is helping to not only bill customers more accurately but to also analyze the nature and extent of outages more precisely—after they occur.

Such automation at the edge is evolving to deliver some real-time control of how people and businesses consume energy. Smart thermostats, the most common example, have seen rapid market growth in the last decade, giving consumers and utilities (with permission) means to control demand as well as energy costs. Now we are witnessing wider adoption of in-home sensors and intelligence from companies like Sense and new smart breakers from Span.

But what about the grid itself? To improve efficiency and capacity of the whole electric infrastructure, operators need a path to improve real-time visibility of conditions on the energy system. Rather, utilities today are driving blind, an increasingly untenable situation in the face of unprecedented energy demand.

Early grid-level automation improves efficiency

Active grid visibility saw its first steps through industrial deployments within large infrastructure such as substations, large transformers, and power generation units. They have been directly integrated with some level of automation, most obviously SCADA (Supervisory Control and Data Acquisition), that made remote monitoring possible.

Meanwhile, the current wave of visibility and automation solutions are improving operational visibility. Examples include advanced distribution management systems (ADMS), outage management systems (OMS), and even more complex outage solutions like fault location isolation and system restoration (FLISR). These systems improve asset management and speed up response to events, both of which improve customer experience.

New platforms to overcome hazard awareness delay

While those legacy systems are important to operations, they are still responding to the “latest bump in the road.” To continue meeting oncoming demand increases and customer expectations, grid operators need to look ahead and spot problems before they jolt the system. This time between when a hazard occurs on the line and when the line worker identifies it, is what we define as hazard awareness delay.

The key to the future of the grid is to significantly reduce hazard awareness delay. To meet the moment, a new generation of visibility solutions—empowered by low-cost hardware, ubiquitous connectivity, and AI-accelerated analysis—are giving utilities a clear view of where service on the grid is now and where it is going. Operating largely outside the legacy IT/OT infrastructure, this new way of managing the grid is called “Active Grid Response,” an emerging category that enables utilities to quickly assess, prioritize, and dispatch line workers directly to the hazard, significantly accelerating restoration efforts while eliminating the hazard awareness delay. In other words, it empowers utilities to quickly respond to issues when they are occurring or are about to occur.

Here are a few examples:

• Wildfire sensing – Gridware and other companies deploy sensors directly at the pole to detect in real time suspicious events on the distribution grid. This detection can drive faster response to quickly unfolding major event days, such as those caused by wildfires, before they threaten communities and the electric system. This will save lives and prevent untold billions in property and environmental losses. 
• Image recognition – Drone and advanced imaging solutions such as Skydio allow rapid survey of operating conditions, with image recognition automation that can detect issues and initiate repairs before they lead to a disruption.
• Line sensing – Need to know how a transmission line segment hundreds of miles away is performing right now? Prisma Photonics puts fiber optics directly into the power line and uses advanced analytics to deliver real-time sensing. Other companies like Pitch Aeronautics add sensors to the system that allow operators to know what is happening on transmission lines in real time.
• Power controls – Intelligent power flow controls and solutions from innovators like Switched Source can balance the three phases of the distribution circuit remotely in real time. This streamlines integration of various resources to expand future energy availability.
• Real-time forecasting – As severe weather becomes more frequent and extreme, new AI-driven solutions improve forecasting precision and accuracy with highly local and up-to-the-minute data. Companies like Amperon and Synoptic Data help energy and utility companies to understand where the greatest risks of fires, storms, and floods are and proactively deploy resources.
• Asset monitoring – With the extremely long backlog of substation parts and transformers, grid operators need insight into the health of operating assets, many of which are decades old. New predictive monitoring can tell when a key asset may be near the end of its life, giving utilities plenty of advance notice for procurement.

AI helps the grid help AI

Most of these solutions use some form of AI or machine learning to make sense of real-time data made available by new sensors. This will become even more important as these sensors look deeper into behavior of power on the grid. For example, phasor measurement units (PMUs) collect detailed transmission line data at 60 times per second. Platforms like Ping Things, Utilidata, and ThinkLabs AI have arisen specifically to process these vast amounts of data effectively to operators for more effective decision making.

The flip side of this trend: AI-based tools are helping utilities manage a broad range of generation sources needed to meet the demand for AI process power. NVIDIA and its graphic processing units (GPUs) in large data center operations are working to deploy supercomputers that will help AI to automate sophisticated grid controls. Grid operators will be able to look forward at how our electric grid can meet the demand for AI and other expanding loads.

With these solutions, can we automate the response for grid operations in real-time given actual conditions? The short answer is yes, as these tools and technologies mature. We are excited to see how these solutions will continue to progress so the most advanced machine of the 20th century can also become one of the most sophisticated in the 21st.