Among readers’ comments on last week’s note regarding electrification, several included an oft-heard concern that solar and wind power may be too variable to be principal energy sources; that sunless days, in particular, when solar power is limited, may also be cold days when the energy needed for heating buildings is high. In that note, mea culpa, I focused almost entirely on solar power as a non-fossil energy option while barely mentioning wind or nuclear – and making no mention of geothermal energy or hydropower. My bad.

Let me do some redressing of the balance, beginning with nuclear. Even in the USA, where nuclear power has struggled against headwinds of controversy and cost, the nuclear share is still a positively surprising (to me, at least) 18% of total electricity produced. That USA share still seems a lot smaller than it should be: it’s within a rounding error of Russia’s, despite Russia’s far worse history of nuclear accidents (it’s not just about Chernobyl). Greater shares of lesser national outputs are seen in South Korea (32% of country total, about one-quarter of USA nuclear output) or France (67% of country total, about one-half of USA nuclear output), just to pick some arbitrary examples.

The outlook for nuclear is improving, though, with a focus on smaller reactors, “deployable either as a single or multi-module plant, and…designed to be built in factories and shipped to utilities for installation as demand arises…avoiding the need to commit initially to a large power output, as is the case with conventional large reactors.” Following regrettably various and erratic pathways toward development and demonstration, the “small modular reactor” approach may mitigate many obstacles of planning and delivery – which have long postponed the arrival of what people projected in 1954 as “too cheap to meter” electrical power.

Geothermal power is even more in need of a better PR team, with only about 1% of USA homes using geothermal heat pumps for heating and cooling. The math of a geothermal home is a perfect demonstration of long- versus short-term benefits and costs: in the long run, what’s not to like about having a huge chunk of thermal inertia under your home that stays (typically) between 45 and 75 degrees Fahrenheit year round, ready to absorb heat during hot months or give up heat during cold months using a simple heat pump mechanism? “It's the most efficient HVAC technology available, four times more efficient than a furnace and twice as efficient as an air source heat pump,” contended Mitch Ratcliffe at Earth911.com. Further, the typical lifetime of these systems is several decades, on the order of twice the lifetime of more familiar systems. What’s the catch? Geothermal’s upfront cost of $15,000 to $38,000 is a hurdle that home builders don’t want to face when bringing a new house to market, even though it’s typically going to break even and start yielding savings within a decade or less after initial installation. Larger installations at the scale of a campus or a major utility are also increasingly proving their worth.

Overarching all such discussions is the question of balancing the times of day when energy is most cheaply produced against the times when it’s most needed. This is not a discussion that’s only required for variable energy sources like solar and wind. The so-called “duck curve,” named for its whimsically noted resemblance to a waterfowl’s profile, is a consideration in leveling the load on power grids regardless of energy source – but it’s especially relevant to solar power, whose output is most available at mid-day and in summer months while some demands are greatest on winter nights.

Efficient storage of peak-generation output, for use at later times, should therefore be part of any conversation about running more of the world on electrical power. The TL;DR on this is the need to recognize widely varying needs, ranging from storage over long periods of time to storage for only tens of hours.

Iron-air batteries, using a process that’s been called “reversible rusting,” will probably never show up in your car – but as MIT’s Technology Review noted, in naming iron-air batteries as a “Climate Tech to Watch” for 2026, this chemistry may be suited to “multiday supply-demand imbalances, like those that crop up during heat waves, deep freezes, or periods with little sun or wind.” These batteries may achieve a cost that would make them competitive with having peak-load capacity fueled by natural gas – and as Technology Review further observed, “AI’s thirst for energy has already delayed closure of aging coal plants, led to the building of new natural gas plants, and caused price hikes for consumers. Cheap energy storage could help renewables meet more of the AI-driven growth in electricity demand.”

A notably critical comment on my note of last week asked why, if the math is as straightforward as I asserted, fossil fuel use is not already plummeting. A downhill plunge has not yet begun, but it’s projected as of November of last year that “coal is now at or near its peak, while oil use will crest around 2030 and gas will top out before 2035.” The inertia of infrastructure, and the distortions of suddenly surging need for AI-enabling energy, may prolong the endgame – but as noted here in January, the forces driving a fossil-fuel decline are clearly already in play.

What’s perhaps more important, though, is that we use questions like that as definitions of the problems to be solved, instead of letting them become rebuttals to the best solution: we need to recognize that “This paradox—more renewables but persistent fossil support—stems from the disconnect between generation growth and system flexibility. Without complementary infrastructure and exceptionally diverse and integrated energy storage, the effective utilization of variable renewable energy remains fundamentally constrained.” It’s about treating the difficulties as directives, not as discouragements.