Solar and wind power cannot substitute for dispatchable generation. Although the U.S. added 200 gigawatts “nameplate capacity”[7] of solar and wind from 2015 to 2025, the average output of these resources is only 21% and 33% of nameplate, respectively. When this nameplate capacity is adjusted for actual output, only 55 gigawatts remain.[8] Moreover, solar and wind are intermittent resources. Snow cover is a surprisingly effective barrier to photovoltaic (PV) operation. The Germans refer to unfavorable renewable conditions as “Dunkelflaute,” meaning “dark doldrums.” During such conditions, the amount of installed renewable capacity is meaningless; there’s no electrical output at all. Proponents say offshore wind is more reliable, but its construction is expensive and slow. Moreover, the Trump Administration began revoking offshore wind permits in April 2025.

Most electricity in the U.S. is consumed in metro areas on the East and West Coasts, but wind turbines work best in the middle of the continent, where the wind is strongest. Plenty of people are in favor of massive, high-voltage transmission lines to transport wind power from flyover country to power-hungry, coastal consumers.

However, three significant impediments exist to this solution: 1) legal challenges, 2) NIMBY opponents, which together have nearly halted construction of new interstate transmission lines, and 3) extremely long lead times to acquire their key components. The third is most significant, because it cannot be resolved by court orders or legislation. Large power transformers at transmission substations are 400-ton behemoths that are not built in the United States and have lead times of three years or more. Additionally, during the past five years, the price of such transformers has increased by 40%; tariffs will only exacerbate inflating prices.

The viability of high-capacity batteries, such as those in BESS, depends on the application. When BESS are installed on customer premises, they can have significant reserve capacity compared to facility load. Accordingly, BESS can materially improve the economics of facility power in areas with high time-of-use tariffs or high demand charges. However, in utility-scale configurations with solar and wind generation (“hybrid systems”), BESS capacity is typically limited to four hours or less of discharge; therefore, renewables plus batteries cannot substitute for dispatchable generation. So, while BESS can alleviate electricity shortages at peak hours, these batteries are too costly to solve the more general problem of providing backup power during days-long wind and solar droughts over wide areas.

Consider Minnesota as an example. According to 2024 EIA data, Minnesota’s average daily electrical consumption is 179 gigawatt-hours (GWh). Let’s assume in a hypothetical future that 50% of Minnesota’s electric power is generated by either wind or solar. Let’s also assume a three-day wind drought, heavy cloud cover, and prior snowfall on solar farms —output from these renewables would be zero. The battery capacity needed to supply Minnesota’s needs for three days would thus be 179 GWh/day *50% renewable share *3 days, or 269 GWh. This energy of 269 GWh is three times the entire installed BESS capacity of 88 GWh[9] in the entire United States at the end of 2024.

According to 2024 EIA data, Minnesota consumes approximately 1.65% of total United States electricity generation. Although this is an oversimplification, extrapolating Minnesota’s three-day need for backup to the entire United States results in required battery capacity of 16,000 GWh. This number far exceeds available financing and manufacturing capacity. It means that the role of BESS will remain limited—much like the role of batteries in a hybrid vehicle on a cross-continental trip.

Nuclear power looks like a solution, and should be a solution, but won’t be a solution in the short term. The U.S. has 94 operating nuclear reactors generating about 20% of total demand. Twenty-one commercial power reactors are undergoing decommissioning but only three of these are candidates for restart: Palisades Nuclear Plant, Three Mile Island Unit 1, and Duane Arnold Energy Center.

Large nuclear plants of conventional design require more than a decade to build. In fact, it took 15 years to permit and build Vogtle Units 3 and 4, the last United States nuclear reactors commissioned. These units cost twice as much to build as originally budgeted.

While new nuclear technologies such as small modular reactors (SMR) are being developed, their timeline is also long. The only U.S. Nuclear Regulatory Commission (NRC)-approved design for SMR, NuScale’s VOYGR, started its formal design process in 2007.

Other SMR companies are entering the race. Most of them are betting on a new type of nuclear fuel called high assay, low-enriched uranium (HALEU), versus the low-enriched uranium used by all current U.S. nuclear plants and NuScale. The combination of new reactor designs and new fuel type makes it possible that their NRC certifications will take even longer and cost much more than NuScale endured. Thus, the prospect of widespread SMR adoption in the next 5 or even 10 years is remote.

Of the dispatchable generating plants currently in interconnection queues, nearly all are gas-fired. However, this new capacity also faces impediments: limits on opposition from environmental groups, slow permitting, and long lead times for construction —natural gas power plants typically take four years for permitting and another four years to build. New gas-fired plants cannot be built without sufficient pipeline capacity. Unfortunately, the interstate gas transmission system has little slack capacity. Court challenges by environmental groups and NIMBYs have halted construction of new pipelines. The U.S. Congress had to pass special legislation to overcome court challenges impeding completion of the last major pipeline, the Mountain Valley Pipeline.

According to the industry newsletter CTVC, manufacturing constraints will also hamper new plant construction:

Another source, Heatmap, says rather than pushing for massive expansion, manufacturers have begun to limit investments to protect high equipment margins. They fear overexposure if the expected growth of data center demand fails to come to fruition.[11]

Will delaying retirement of dispatchable generating plants, especially coal-fired plants, save the day? Not so fast. Plant owners are increasingly declaring forced outages during energy emergencies. Steam turbine generators in these facilities were designed for constant base-load operation. The grid in 2025, however, includes enough renewable generation to force large thermal plants to run intermittently.

For decades, coal plants delivered highly dependable, low-cost power. In April 2025, the Trump administration signed an Executive Order to keep coal plants open, but this lifeline fails to address two fundamental problems. First, deferred maintenance costs to refurbish coal-fired plants would be monumental. Why do more than minimal maintenance on an asset that has effectively been banned? Second, start-stop operation causes thermal cycling for which plant boilers were not designed. Twenty years’ worth of stress can be accumulated in a fraction of that time.

Thermal stress amplifies the risk of boiler explosions, like the one that killed three workers at the Salem Harbor Power Station in 2007, although the Salem Harbor boiler was inspected annually. This boiler was 50 years old, which also approximates the average age of grid assets in 2025.

Most coal plant operators have reacted to two decades of societal antagonism quite logically—they have chosen to run these assets into the ground. As a result, a growing number of experts estimate that most U.S. coal-fired generation (today 17% of supply) will be retired by 2030.

Absolutely …there are grounds for optimism! A potential source of dispatchable generation sits fully fueled and mostly idle. It is standby power at C&I enterprises. The combined capacity of C&I standby generators installed in the United States is between 130 and 200 gigawatts[12], with about 45 gigawatts of that capacity located at hyperscale data centers.

This installed base of standby generators could replace retiring dispatchable generation much faster than building new generating plants and transmission lines for their interconnection. If generator installations and data centers continue to expand, which they will, this capacity could provide reserve power during peak usage periods, negating much of the need for new plants. Minimal upgrades to distribution systems would be required. One hundred gigawatts of standby generators would add about 10% dispatchable capacity to the U.S. electric grid—enough for 3-5 years of load growth. Unfortunately, this proposal faces three obstacles: