Studies and press reports have been warning that our electricity supply is becoming less reliable, which means we are facing a greater chance of power outages in the future. This reliability crisis will come to a head within the next 5 years, we estimate, unless meaningful steps are taken to head off the crisis. The mounting reliability problem has been highlighted in reports by electricity experts as far back as 2018, but the problem has continued to grow more serious.[i] In some ways, this situation could be characterized as problem admiration, which means describing and analyzing a problem but not really solving it. We propose a handful of solutions later in this blog.
Electricity Growth
The Washington Post ran a front-page article on March 10 titled “Amid explosive demand, America is running out of power.” The article points out that economic and technology trends are causing an enormous increase in the demand for electricity. The article quotes the chairman of the Georgia Public Service Commission as saying that the increase in electricity demand “has created a challenge like we have never seen before.” The article also points out that “Northern Virginia needs the equivalent of several large nuclear power plants to serve all the new data centers planned and under construction. Texas, where electricity shortages are already routine on hot summer days, faces the same dilemma.”
The Post article adds to a host of other clear warnings such as those from the North American Electric Reliability Corporation (NERC), members of the Federal Energy Regulatory Commission (FERC), grid operators responsible for keeping the lights on in large regions of the country, and public utility commissioners. One FERC commissioner warned of “potentially catastrophic consequences” unless these warnings are taken more seriously.[ii]
Electricity demand has been almost flat over the past decade, but the electricity picture is now changing rapidly because of data centers to support artificial intelligence, manufacturing of solar panels, electric car batteries, and computer chips, charging electric vehicles, and powering new energy facilities such as green hydrogen plants, to name just a few. Overall, some $630 billion of investment in new facilities is driving electricity growth forecasts.[iii] Data center growth is responsible for more than $150 billion of this investment. Boston Consulting Group estimates that data centers could consume 7.5% of all U.S. electricity by 2030. Extreme weather is also driving up peak demand growth in some regions of the U.S.
Grid Strategies, a power sector consulting firm, published an overview of the challenges associated with increasing electricity demand. The report indicates that the nationwide forecast of electricity demand over the next five years has almost doubled. This translates into the need for an additional 38,000 megawatts (MW) of electric generating capacity through 2028, but this figure likely understates the need for additional generating capacity. For perspective, 38,000 MW is roughly equivalent to the electricity supply of New York State. Moreover, new generating capacity will require more transmission lines which are expensive and notoriously difficult to build because of permitting and other obstacles. The 15-state Midwest Independent System Operator (MISO), for example, has proposed to build 18 transmission projects at a cost of $10.3 billion plus up to $23 billion for a second round of transmission projects.[iv] Last year, PJM (a 13-state grid operator) approved $6.6 billion in new transmission projects, a substantial increase over the $2.2 billion of projects PJM’s board approved in 2022.[v] (Besides their enormous expense, new transmission lines are taking ten years or more to complete.[vi]) Transmission costs were cited recently as one of the reasons why electricity prices over the past 12 months have increased faster than the rate of inflation.[vii] Government policies were mentioned as another reason for the increase in electricity prices.
More than 60% of data centers are expected to locate in MISO, CAISO (most of California), PJM, and the southeast by 2027. However, neither MISO nor CAISO appear to have included substantial data center growth in their forecasts, according to Grid Strategies. On the other hand, PJM is projecting load growth that will increase the grid operator’s winter peak demand by more than 28,000 MW over the next 10 years.[viii]
Wrong Direction
In the meantime, the factors that influence our ability to satisfy the fast-growing demand for electricity are moving in the wrong direction. For example, the premature retirement of power plants continues, depriving the grid of dependable generating capacity and attributes, such as fuel security, that are necessary to operate the grid reliably. Utilities have announced plans to retire 60,000 MW of coal-fired generating capacity during 2024-2028, the same period during which Grid Strategies estimates that at least 38,000 MW of new electric generating capacity will be needed. That leaves a gap of at least 98,000 MW of generation that needs to be filled by the end of 2028.
Utilities in MISO and PJM have announced plans to retire 32,000 MW of coal-fired capacity during 2024-2028, but this likely understates coal retirements. For example, PJM’s independent market monitor estimates that as much as 58,000 MW of thermal capacity (30% of the PJM’s total installed capacity) are at risk of retiring by 2030.[ix] It is worth noting that utilities in 13 states have postponed the retirement of almost 14,000 MW of coal (31 coal-fired generating units) because of concerns about load growth and grid reliability.[x]
Projects that are intended to replace traditional sources of electricity are stuck in long interconnection queues waiting for approval.[xi] As of 2022, almost 10,300 projects representing 1.35 million MW of capacity plus 680,000 MW of battery storage were seeking approval to connect to the grid.[xii] (MISO and PJM have almost 4,800 projects totaling 637,000 MW in their interconnection queues.) More than 90% of proposed generation projects are solar and wind facilities. Completion rates are 20% for wind projects and 14% for solar projects. The average time projects spent in queues before being completed has increased. The typical project that was built in 2022 took 5 years to complete the interconnection process and connect to the grid.
Solutions
We strongly urge decision makers to take the following steps to help prevent a reliability crisis and electricity shortages:
- Before dispatchable capacity retires, the replacement source(s) of electricity should be built and in operation. This would prevent capacity gaps that could lead to electricity shortages
- The replacement source(s) of electricity should have at least the same accredited capacity and other reliability attributes as the retiring capacity. This would maintain resource adequacy and grid reliability.
- Additional transmission that is needed to accommodate the replacement capacity (mostly solar and wind facilities) should be connected to the grid, not merely lined up in an interconnection queue or under construction. (Also, the cost of new transmission should be considered in deciding whether to retire existing capacity.)
- In addition to the three steps above, grid operators should identify and value all attributes that are necessary for grid reliability.
- EPA rules should be designed so they do not cause reliability problems. EPA should work with grid experts to conduct careful, not perfunctory, reliability analysis. For example, Quanta Technology conducted recent analysis showing that the PJM grid could face reliability violations and electricity shortages in the future because of coal retirements.[xiii] This is the type of thorough analysis that EPA should conduct for its rules.
Final Word
Robust electricity growth forecasts are a positive trend because they indicate technology development and economic growth. The coming years could be a Golden Age for electricity, but only if we move quickly from admiring the reliability problem to implementing solutions, especially the five steps listed above.
For additional information regarding the nation’s coal fleet, grid reliability, and related topics, please visit www.AmericasPower.org.
———————–
[i] See, for example, NERC’s 2018 “Long-Term Reliability Assessment.”
[ii] https://www.rtoinsider.com/73336-ferc-christie-warns-very-dark-place/
[iii] “The Era of Flat Power Demand is Over,” Grid Strategies, December 2023.
[iv]https://cdn.misoenergy.org/Executive%20Summary%202024%20Reliability%20Imperative%20report%20Feb.%2021%20Final631825.pdf and https://www.utilitydive.com/news/miso-midcontinent-tranche-2-lrtp-transmission-expansion/709576/
[v] S&P Capital IQ, “PJM approved $6.6B of incremental baseline power transmission projects in 2023,” March 11, 2024.
[vi] See, for example, “Transmission in the United States – What Makes Developing Electric Transmission So Hard?,” Scott Madden, June 2021. The report is available at www.AmericasPower.org.
[vii] US electricity prices outpace annual inflation | Utility Dive, March 13, 2024,
[viii] https://www.pjm.com/-/media/library/reports-notices/load-forecast/2024-load-report.ashx
[ix] https://www.utilitydive.com/news/pjm-coal-gas-power-plant-risk-retirement-market-monitor/710518/
[x] States with deferred coal retirements are DE, GA, IN, KS, KY, MO, NE, ND, NM, NV, SC, WI, and WV.
[xi] Grid operators require projects to undergo studies before they can be built and added to the grid. The lists of projects that have applied to connect to the grid are known as interconnection queues.
[xii] “Queued Up: Characteristics of Power Plants Seeking Transmission Interconnection As of the End of 2022,” Lawrence Berkeley National Laboratory, April 2023.
[xiii] “Ensuring Reliability: A Case Study of the PJM Power Grid,” Quanta Technology, February 2024