Russia’s invasion of Ukraine, sending global oil and gas prices sky high, is the latest shock to the energy and utilities industry. Barely two years ago, the COVID-19 recession severely depressed electricity demand to threaten the financial health of energy providers. In between, a stream of climate change events, from blackouts in California and winter storms in Texas to flash floods in Western Europe and wildfires in Turkey, caused untold hardships, including major power outages.
Wars and pandemics are unpredictable, one-off events. But these days, the environmental crisis is a certainty, and it is creating bigger, more frequent catastrophes all over the world. Climate change is disrupting every industry, arguably, none more than utilities. So, it is no surprise that in a June 2021 study conducted in the United States, Australia, Germany, France and Indonesia, 88% of the 500 utility executives surveyed said they were highly concerned about the impact of disasters on the power grid.
Improving disaster preparedness so they can keep the lights on, literally, during a catastrophe, is therefore, a top priority for utility providers. However, as one of the biggest actors on the climate stage, the industry must also play a leading role in checking the progress of the climate crisis. Reducing energy consumption and losses, cutting emissions by replacing fossil fuels with renewable sources and improving grid reliability and performance are the biggest goals. Digital technology, particularly artificial intelligence (AI), can help utility companies achieve all of them.
Improving Disaster Preparedness
At its peak, Hurricane Ida disrupted power supply to 1.2 million customers in eight U.S. states last year. But experts believe that inadequate disaster planning is a big part of the problem. All utility companies maintain extra capacity so the grid is unaffected if some equipment fails. But the redundancy strategy doesn’t work during a weather catastrophe when large areas go down all at once. Utility providers need a totally different approach to building resilience, starting with factoring a higher risk of extreme weather events in their operational plans. They should also consider sustaining the most important services required by consumers, rather than the entire grid, in the event of a disaster. Digital tools and technology, such as sensors, IoT and AI, can facilitate this by providing data and visibility into consumer needs, usage patterns, grid status and more.
The challenge is that resilient grids are not only hard to sustain, but they also come at a sizeable expense. Fortunately, some new digital solutions help improve network resilience at lower cost. By leveraging these solutions, electric utilities can maintain services even during extreme weather and restore outages faster. A cooperative utility in Colorado has established an AI-based, automated, self-healing grid in the most hazardous terrains that it serves. The grid can maintain supply during both summer and winter storms, and in the event of a failure, restores power almost instantly.
Saving Energy Consumption and Loss
Transitioning to non-fossil fuel resources is of course the way forward for the utilities industry; in the meantime, providers should also look at cutting energy consumption, wastage and loss. A Brazilian electricity distribution company that was suffering big losses due to power theft addressed the problem with a combination of digital technologies. Digital meters sent usage data to a single device that connected to the company remotely. AI-based tools analyzed this information to understand “normal” usage patterns based on the customers’ profiles and alert the company to any unusual activity. This insight also enabled the utility to predict which customers were likely to have unauthorized power connections so it could take action to reduce leakage.
Utility providers can also reduce energy consumption by motivating customers to participate in demand side management programs (DSM). In the past, providers used domestic usage data and online surveys to give customers a comparison of their energy bills versus that of other households, as well as suggestions for saving costs (for example, switching to smart thermostats or energy-efficient appliances). But in the absence of personalized data, and with only a slim minority of consumers responding to the survey, the programs were not very effective. Worse, the utilities had no way of identifying the best targets — those that had the most to gain from the program.
Now, AI tools have solved the problem. Intelligent sensors connected to domestic households provide granular data, even at appliance level, about how, and how much, energy they use. With this insight, utility companies know exactly which consumers to target, without the consumers telling them anything.
The United States will retire a large part of its coal generation capacity by 2025. Many utilities have also made plans to slash, or even eliminate, carbon emissions by 2050. While decarbonization is imperative, its economics is quite challenging at higher levels. The view is that utility providers will be able to cut the bulk of emissions with current technologies, and slight additional investments. However, beyond that, they will encounter diminishing returns from renewable resources, which will make substituting fossil fuels very expensive.
Integrating distributed energy resources (DERs), such as solar, wind and hydro power, into the grid is also operationally challenging. Because power now flows in both directions, it is subject to variability. Studies show that providers are very concerned about power quality, backfeed and voltage surges in the case of renewable energy sources.
The good news is that as more DERs are introduced on the grid, utilities are not only seeing their emissions going down but even reaping other benefits, such as lower system losses, and more stable demand. Deploying distribution intelligence strategies and other digital technologies improves utilities’ visibility and control of DERs and allows them to realize their full potential. Energy companies can further reduce greenhouse emissions by using AI solutions to track and control leakages from pipelines and equipment.
Improving Grid Response, Resilience, Reliability and Performance
Reliance on AI will steadily increase as power generation and transmission passes into the hands of a complex system of microgrids, domestic solar plants, batteries and wind farms. One estimate says that 36 million devices, including electric vehicles, solar panels and batteries, will be added to the grid in Europe within three years. At that time, the principal role of a central utility will be to finely balance the grid to prevent it from breaking down. That will only be possible with AI.
An AI solution can orchestrate the various parts of the grid such that the excess electricity that is sent to the grid or is stored for later by millions of DERs is directed to where it is needed, when it is needed (in real-time or nearly that). What’s more, AI can process grid data to make forecasts in seconds, to make the system responsive, resilient and reliable. Here is a real example: on a wind farm, operators leveraged AI software to monitor data and alert them to anomalies early; this allowed them to fix turbines, generators and other equipment before they failed.
A combination of cloud computing, AI and analytics will be indispensable for the uninterrupted functioning of the grid of the future. Even digital twins — virtual replicas of physical systems — find wide application in the energy asset lifecycle. For example, energy and utility companies can create a digital twin of a planned grid to simulate various scenarios and understand how it responds before building it physically.
Many of the recent electrical system failures in the United States occurred because of the grid’s inability to cope with soaring temperatures. For example, when temperatures rose to 115 ° F in Portland, Oregon, last summer, power cables melted, leaving thousands of people in the dark. At exactly the same time, New York City also faced a heat-related power outage, but of a different kind—the grid was simply unable to meet the spike in energy demand as New Yorkers suffered sweltering conditions.
With extreme climate events, from heat waves to hurricanes to winter storms, coming fast and furious, utilities need to factor disaster preparedness in demand planning, infrastructure creation, asset maintenance and every other operational area.
In the past, planning required a team of planners, designers and engineers to manually gather data during field inspections, and combine that with past information, to extrapolate future demand, grid capacity, peak load etc. The risk of an extreme climate event was rarely, if ever, part of the exercise. Today, it must be put at the top of the agenda.
Estimating what part of the power generation and transmission system will be hit by weather, and how hard, is impossibly complex to do manually. The alternative, which is building something based on assumptions, can be inaccurate, expensive, or very likely, both. A digital twin lays all these problems to rest. Utility company personnel can test assumptions, forecast events and model scenarios iteratively until they are satisfied, before building out. Once the system is up and running, the digital twin continues to take data from sensors and smart equipment to provide a real-time view of the grid. This can make all the difference to its resilience and performance in the face of climate change.
With increasing digitization comes higher cybersecurity risk. The energy sector has suffered several attacks, with as many as 42% of enterprises being the target of phishing. Memories of a previous crisis in Ukraine, when its power grid was hacked a few years ago, are a warning not to take cybersecurity lightly. Utilities should invest in AI encryption tools, along with video surveillance and other digital technologies to secure their assets and processes from end to end.
Think-tank, Electric Power Research Institute, is exploring using drones for inspecting transmission and distribution infrastructure. AI software can quickly analyze the images captured by the drones, to spot any hazards; this is not only more efficient but also much safer than deploying staff in the field.
Climate change is among the biggest threats that human beings have ever faced. It is also taking an enormous toll on the energy and utilities industry. Utilities must quickly learn to cope up with weather events that are becoming more frequent and also more ferocious. Disaster preparedness is important, but it is mostly, a reactive, damage control type of strategy. Providers also need to prevent such events in the future. That means a long, hard battle against climate change.
Fortunately, utility companies have a powerful ally in digital technology, specifically AI. AI can help utilities withstand calamity better and bounce back faster; beyond that, its insights can guide utilities on how to reduce fossil fuel consumption (and hence, greenhouse gas emissions) as well as facilitate the integration of renewable energy sources with existing grids. AI can not only help secure the grid, but it can also enable its transformation to support two-way flow of energy for every service point — thus, accelerating the rise of a “prosumer.” A prosumer is any energy customer who produces energy as well as consumes it. In fact, AI can actually help utilities focus on the right set of consumers who can make a difference to the supply/demand equation.