Enhancing the Security of the North American Electric Grid

A secure and reliable supply of electric power is a key component of modern economies. Not only are other energy sources often poor substitutes, but essentially every industrial and commercial process in the United States requires its use, and nearly all homes rely on it. Even short-term interruptions in the delivery of electric power result in economic losses or inconveniences for consumers and businesses. Longer outages can result in spoilage of food and other perishables, forgone sales, the idling of resources in production processes, disruptions to the supply of water and fuels, and other threats to health and safety.

This study by CBO examines a range of threats that could cause widespread, long-lasting disruptions for the electric grid, including ones beyond historical experience. The study discusses a range of illustrative approaches to enhance the security of the electric grid and some considerations for policymakers to take into account.

The North American Power Grid and Major Threats It Faces

The power grid is a collection of generating plants, power transformers, transmission lines, and other equipment that helps move large quantities of electricity over long distances; components that distribute smaller quantities to end users; and collections of customers that use the power. The delivery of power to customers is usually highly reliable. Though the grid faces a wide range of threats, the vast majority are localized and are handled by grid operators with minimal disruption for customers.

But the grid also faces a number of larger but rare threats that have the potential to cause regional disruptions that last longer. Naturally occurring threats include a burst of solar particles—referred to as a solar storm—that interact with Earth’s magnetic field and create a geomagnetic disturbance that could overload certain critical grid components; a hurricane that could affect the supply of power along an entire coastal region; and an earthquake that could damage or disrupt generating plants, transmission lines, and other equipment and, thereby, the power supply of extended areas. Human-made threats include a high-altitude electromagnetic pulse (EMP)—most likely created by the detonation of a nuclear weapon at high altitude—which, like a severe solar storm, could overload and disable key components; a cyberattack targeting generating plants or grid control systems; and a physical attack against certain critical components.

The likelihood of wide-ranging and long-lasting outages is small, but the consequences could be severe. Some estimates suggest that losses in the economy could be in the hundreds of billions of dollars or even more than a trillion dollars in some scenarios. Losses could also be considerably less depending on the extent of the disaster or attack; the condition of the system, including whether the grid retained enough power to handle emergencies; and the effectiveness of existing protections and recovery measures, among other factors.

Approaches to Reduce the Costs of Major Threats

The utility industry has a number of operational and procedural protections that it uses to enhance the security of the electric grid and prevent or limit power outages. Most are day-to-day protections. But events like the 2015 cyberattacks in Ukraine, which targeted that country’s grid control systems, and a cyberattack in the western United States in early 2019, which briefly disrupted communications at several small generating sites, have increased awareness about risks and heightened concerns.

CBO identified a number of approaches for boosting the security of the grid—approaches to either prevent or mitigate damage or to improve recovery after the damage has occurred. The approaches identified are not an exhaustive list but, rather, illustrate the wide span of possibilities for reducing the risks of a large, long-lasting outage. The approaches include improving information sharing, enhancing cyber protections, and improving physical security. They also include two approaches that CBO examined in relative detail: one to prevent or mitigate damage—deploying space-based sensors to monitor solar activity—and one to improve recovery—increasing the stock of replacement transformers, which are critical in allowing large amounts of electricity to flow throughout the grid.

Space-Based Sensors. One option for monitoring solar activity—a dedicated satellite placed in orbit between Earth and the sun—would provide early warnings of a solar storm. It would carry a coronagraph to provide images of the sun that would allow forecasters to provide long-term warnings (one day to four days in advance) of a solar storm that might strike Earth. It would also carry instruments to measure the solar wind, which would allow forecasters to provide short-term warnings (15 to 60 minutes in advance) with more accurate estimates of a solar storm’s likely arrival time and the severity of its effects on Earth and on the grid. The United States has satellites that provide such warnings today, but they are old and are expected to stop functioning within several years. At a cost of about $500 million to purchase two satellites (one that would be launched in 2024 and another that would replace the first roughly five years later) and another $500 million to launch and support the satellites through 2029, this option would replace the current system and improve the reliability and quality of the data for more accurate forecasts of solar weather.

Two other options—placing coronagraphs on the next generation of weather satellites or on the International Space Station—would cost significantly less and maintain some capability for monitoring solar storms when the current space weather satellites fail. Building and deploying those chronographs might cost $100 million to $150 million over 10 years. But by themselves, neither of those two options would provide the data necessary for the accurate short-term warnings of an impending solar storm that grid operators rely on to take steps to protect their systems—warnings that are provided today and that would continue under the first option. More accurate warnings might also avoid the cost to operators of taking unnecessary steps to prepare for storms that end up having little effect on Earth.

The National Oceanic and Atmospheric Administration (NOAA) has published plans to deploy a dedicated satellite between Earth and the sun but without a follow-on spare satellite (as under the first option) and place a coronagraph on the next weather satellite (as under the second option) so that the agency would have two coronagraphs in orbit. But it has not yet secured most of the funding to implement that plan.

Replacement Transformers. Large power transformers can take a long time to manufacture, leaving portions of the grid vulnerable if they become disabled and need to be replaced. One option for boosting the stock of transformers would be to provide subsidies—in the form of funds or tax credits—to suppliers of electricity that they could use to buy and hold transformers in reserve. This option would leave various technical decisions in suppliers’ hands, but setting the appropriate subsidy level would be difficult, and a significant share of the federal costs would only reduce the utilities’ net cost of units that they would have purchased anyway. Another option would be for the federal government to own a stockpile of transformers, which, as necessary after a disaster or attack, it could sell or give to suppliers. By the Department of Energy’s estimate, the stockpile would need to consist of at least 100 transformers, at a cost of $2 million to $9 million each. Yet another option would be for the federal government to require suppliers to hold private reserves of a specified size. That option would have negligible costs for the government, but determining the appropriate size of such a requirement, like setting an appropriate subsidy level, would be difficult.

Some Key Considerations for Policymakers

One consideration for policymakers is the appropriate role for the federal government in improving the security of the electric grid. To what extent would the private sector acting alone take the full range of potential benefits into account when deciding what to invest in protection or recovery?

The benefits of a new class of space-based sensors dedicated to monitoring and evaluating space weather, for example, would extend beyond the electricity sector. Other industries, too, such as the telecommunications and transportation industries, could benefit from early warnings about potentially damaging solar storms. Because the benefits would be widespread and difficult, if not impossible, to limit only to parties that paid for them, it is unlikely that the private sector would invest in space weather sensors on its own and more likely that the approach would depend on federal support, similar to the federal role in providing Earth weather satellites.

Private-sector electricity suppliers have a greater incentive to pursue some of the benefits associated with investing in reserve transformers. As a result, suppliers hold their own reserves, both individually as part of their business planning and collectively in reservesharing arrangements. But in making decisions from the perspective of their business and their geographic area, suppliers may not fully account for some benefits of avoiding outages, such as ensuring economic stability or public safety.

Another consideration is just which factors to weigh. CBO’s analysis focused primarily on the potential loss of national economic output (gross domestic product, or GDP) resulting from major outages and on the budgetary costs of policy alternatives. But GDP does not capture all the costs of an outage—such as inconvenience, personal discomfort, or even loss of life—and policymakers could take those or other factors into account. Some threats to the electric grid also threaten military security and public health, so policymakers might weigh the benefit of avoided damage to those sectors—even in circumstances when the avoided loss of GDP would be relatively small or the costs of the policy would be high. Other potential factors are the possibility of inefficiencies that subsidies or regulations could impart to the economy and the effects that policy-induced changes in prices might have on households with different amounts of income or people who live in different regions of the country.

Still another consideration is the advantages and disadvantages of federal intervention amid the uncertainty surrounding estimates of them. Avoiding a loss of GDP is one benefit of improving the security of the grid and reducing the chance of a widespread, enduring power outage. But estimates of the size of potential losses are highly uncertain, as are estimates of their likelihood, suggesting a large range of possible outcomes and complicating decisions about investing in the security of the grid. Deploying new solar satellites, for example, which would probably cost about $1 billion over 10 years, could offer some protection from solar storms. Without it, the economic costs of a severe solar storm could be large, but the likelihood of such infrequent storms is uncertain. Moreover, the degree to which the early warning from a satellite would reduce the damage from a severe solar storm is also uncertain.