Increased storm activity and extreme temperatures have broken records around the world this year, leading our industry to rethink what it means for the grid to be resilient. This makes it a particularly complex moment in time for the grid. Now is the time for utilities to give a hard look at resiliency and what measures they have in place to ensure a rapid return to normal operations after a catastrophic event.
As it stands today, much of the current training and preparation in the energy industry focuses on those in the field. In reality, software is the key to facilitating situational preparedness and, ultimately, resiliency. Big data, machine learning and predictive analytics are the key to navigating extreme weather patterns and volatility.
According to a recent Accenture study of utilities executives, “99% believe that artificial intelligence (AI) will be used routinely in decision support in the control room and in network planning by 2025,” and predictive analytics should be a key focus for those investments.
The Opportunity in Data
Power plants, wind farms, grids, substations and energy management systems generate terabytes and petabytes of data a day, and yet, less than 3% to 6% of this data is analyzed. That underutilized data has the potential to unlock valuable insights and to help regulate variables that impact the grid via predictive maintenance.
For many utilities today, the damage assessment process is often conducted and managed in isolation from the outage management system (OMS) and geographic information systems (GIS), with many systems focused solely on collecting data on network damage caused by major storm events. Predictive analytics and machine learning can bridge the Industrial Internet of Things (IIoT) divide by enabling utilities to leverage, predict damaging events, and reduce grid loss and unplanned downtime.
In a way, this solution creates a crystal ball for digital electricity. Software can now proactively alert utilities when something may be headed towards a breakdown and help jumpstart responses.
Predictive analytics also assist with team response time and protocol. Field crews conducting damage assessments can remotely access existing outage data stored within the utility’s OMS and GIS to identify damaged portions of the network.
Data enables a higher fidelity on where the storm will likely impact the grid most from a physical asset perspective. This precise data allows teams to direct vehicles to those specific locations — resulting in not only less downtime, but also less safety risk for response teams. Repair crews will no longer need to brave the storm in search of damages but could already be stationed nearby potentially vulnerable areas.
Utilities Leading the Charge
Consider Exelon, one of America's leading energy providers in the US. It partners with GE to tap into the power of data to outpace the weather. Together, they are creating storm-prediction technology that anticipates outages and deploys response teams to keep the electricity flowing and turn the lights back on quickly.
Another utility, based in the storm-laden U.S. south, uses GIS to look at predicted storm surges along the coast, and historical data on what supplies are likely to be needed, such as poles, wires and crossarms. Armed with this data, the utility can move workers to potentially impacted areas before a storm hits.
During 2017’s Hurricane Irma, another southern utility used drone teams across the state. The drones provided detailed visuals of otherwise inaccessible elements of their badly damaged grid. This modern process produced more actionable data, and teams were able to more accurately plan for recovery.
Protecting the Energy Future with Software
Response speed and efficiency will be critical in the digital energy future. Energy providers must actively leverage technology to enable a more proactive grid — one that can accurately predict volatile threats. Software will be the ultimate safeguard for the grid during this season and in the future — whatever nature may bring.