Electric utilities are no strangers to intelligent technology within the grid (think sensors on the lines and advanced analytics that can forecast potential outages due to upcoming storms). But there’s a lot more scope to bring intelligence in the control room to improve operations, customer satisfaction and enable cost savings and potential new revenues. So what technologies are on the table, and how can they be used to transform the way control rooms work?
The reality check: the reactive control center
Before joining Accenture, I spent 20 years (give or take) in and around utility control centers. And the truth is, since I started, not much has changed in terms of how they operate. Control rooms are heavily dependent on human operators working in shifts, looking at screens, and interpreting data from multiple tools without much context to identify and solve real-time problems.
It’s a fundamentally reactive approach: watch and wait…see a problem occur…now find its root cause and solve it as fast as you can. That problem might be a substation outage or a fire beginning to take hold. And it’s a challenging and stressful job, requiring fast reflexes and the ability to analyze multiple inputs and derive a rational decision.
Meanwhile, those challenges are only increasing, with extreme weather events now more frequent and severe, and likely to escalate further, the increase in penetration of renewable energy on the distribution grid, and the increase in threats by foreign adversaries to the security of the grid.
<<< Start >>>
<<< End >>>
Electric utilities are already acutely aware of the grid’s sensitivity. And in response, they’re adding smart technology onto, e.g., transmission and distribution lines to identify faults faster and reduce the likelihood and magnitude of an outage. So the grid is getting smarter, but is that translating to the control room?
Added to that, operators are generally siloed, with established information and process flows but limited collaboration. Example: the maintenance group is working with assets and, in many cases, possesses intelligence that provides information on the sensitivity of those assets to failure. However, the only collaboration the control center operator may have with that group is to provide clearance for people to work on those assets in high-voltage stations, which in many cases is only via voice.
Ready for reinvention?
What if I told you that adding digital technologies, like Artificial Intelligence (AI), could watch the grid for you? And not only alert you when something might happen but pre-empt it too using data-driven analytics? So let’s say you’re preparing for an event, like a storm that’s on its way. Or when a fire is moving closer to substations. AI and analytics can take the information already available in the control center systems and prompt, proactive actions to prevent an outage.
For example, one of my clients is already working on an advanced visualization tool to improve operators’ situational awareness. This tool will inform about potential circuits and stations that will be susceptible to an upcoming storm, and shows the location of available field crews in the area that could assess possible outages. The system will predict where outages may happen and bring audible warnings into play that inform the root cause of outages (before the operator has to spot it).
Meanwhile, some tasks can be automated, again removing pressure from the operator. And when AI is doing the grunt work, imagine the headspace that gives the operator to focus on outcomes in a holistic way (rather than watching a screen for 12 hours straight). And advanced collaboration tools like a mobile application, a virtual control room, augmented reality from a HoloLens application, or a collaboration table that shows a substation digital twin make real engagement with others possible. All of which reduces risk, improves productivity, crew safety, and reduces outage handling times.
Here’s an example: going back to that storm that’s now hitting. The operator can work much better with the maintenance and engineering teams if it’s possible to model: What assets are with the highest risk of failure? And what might the outage of those assets look like? How would this impact the grid? Then engineering can run a study to examine the impact in more detail and prevent the grid from being affected.
And we’re only just hitting the cusp of what’s possible. For example, today, it takes several minutes to understand how the grid is loaded (and you need that information to optimize the grid and serve customers at a lower cost). But if you can shortcut that process, you can de-risk and save cost. Imagine if you could use sonar to send a signal through the grid and detect the location of a fault that way, or imagine if you could use AI to solve the complex power flow of a grid faster than today’s power algorithms. It’s something being examined by academia and industry and shows just how much potential there is.1
The overall idea is this: by using AI and related digital technologies, you not only improve the work control center operators do, but also the outcomes they generate. And their ability to de-risk the grid at large, improving customer outcomes. AI has the potential to give transmission and distribution operators super-powers!
It’s not pie in the sky
Much of this is already happening, particularly in oil and mining. In those industries, intelligence is integrated into control centers but there is space to grow for electric utilities.
Utilities know how to bring intelligence into their field devices. Think back to the intelligent grid with sensors on the line. But there is still more opportunity to bring that intelligence into the control room and make it a force for transformation—even in the absence of regulatory pressure on this topic.
And let’s consider the workforce a moment. As control centers necessarily become more technology-driven, a new type of operator is needed. One that is a digital native and can work alongside AI for better outcomes. But to attract those people, you need a modern control center (not the ones I knew in my youth—back then, the control centers with the biggest wallboard were considered the best). A modern control center is one with AI and other digital technologies where operators increase their ability to focus on customer outcomes by 40-50% of the time. That doesn’t mean the experiential insights of the traditional operator are obsolete—far from it. It’s about augmenting those insights with technology, and with complementary skills and people.
Now add the black swan
Over this past year, COVID-19 has changed all our assumptions about the roles that can be done remotely, how co-located teams must work, and how systems can support them.
That also means electric utilities must improve the resilience of the control room systems for onsite/offsite working and all those new complexities. Example: one electric utility has been modeling what happens if an operator has COVID-19 and the control room has to be shut down. That might mean a rapid switch to another physical control room location, or even a virtual reality control room.
The same concepts apply for possible security threats. If a critical threat emerges, the idea is that you can switch to a disaster recovery site that is isolated from the security risk, with recovery tools that can transition operations fast. And naturally, AI and related technologies are at the core.
With extreme weather + security imperatives + the black swan on the horizon, now is the time to capitalize on everything AI and other digital technologies can bring to the control center. It’s a path to de-risking the grid, protecting crews, and resilience and reliability for customers. Contact me to find out more.
1 Deep Learning Algorithm development for RTE (France)