Your typical city dweller doesn’t know just how much coal and uranium he burns each year. On Lake Shore Drive in Chicago—where the numbers are fairly representative of urban America as a whole—the answer is (roughly): four tons and a few ounces. In round numbers, tons of coal generate about half of the typical city’s electric power; ounces of uranium, about 17 percent; natural gas and hydro take care of the rest. New York is a bit different: an apartment dweller on the Upper West Side substitutes two tons of oil (or the equivalent in natural gas) for Chicago’s four tons of coal. The oil-tons get burned at plants like the huge oil/gas unit in Astoria, Queens. The uranium ounces get split at Indian Point in Westchester, 35 miles north of the city, as well as at the Ginna, Fitzpatrick, and Nine Mile Point units upstate, and at additional plants in Connecticut, New Jersey, and New Hampshire.

That’s the stunning thing about nuclear power: tiny quantities of raw material can do so much. A bundle of enriched-uranium fuel-rods that could fit into a two-bedroom apartment in Hell’s Kitchen would power the city for a year: furnaces, espresso machines, subways, streetlights, stock tickers, Times Square, everything—even our cars and taxis, if we could conveniently plug them into the grid. True, you don’t want to stack fuel rods in midtown Manhattan; you don’t in fact want to stack them casually on top of one another anywhere. But in suitable reactors, situated, say, 50 miles from the city on a few hundred acres of suitably fortified and well-guarded real estate, two rooms’ worth of fuel could electrify it all.

Think of our solitary New Yorker on the Upper West Side as a 1,400-watt bulb that never sleeps—that’s the national per-capita average demand for electric power from homes, factories, businesses, the lot. Our average citizen burns about twice as bright at 4 pm in August, and a lot dimmer at 4 am in December; grown-ups burn more than kids, the rich more than the poor; but it all averages out: 14 floor lamps per person, lit round the clock. Convert this same number back into a utility’s supply-side jargon, and a million people need roughly 1.4 “gigs” of power—1.4 gigawatts (GW). Running at peak power, Entergy’s two nuclear units at Indian Point generate just under 2 GW. So just four Indian Points could take care of New York City’s 7-GW round-the-clock average. Six could handle its peak load of about 11.5 GW. And if we had all-electric engines, machines, and heaters out at the receiving end, another ten or so could power all the cars, ovens, furnaces—everything else in the city that oil or gas currently fuels.

For such a nuclear-powered future to arrive, however, we’ll need to get beyond our nuclear-power past. In the now-standard histories, the beginning of the end of nuclear power arrived on March 28, 1979, with the meltdown of the uranium core at Three Mile Island in Pennsylvania. The Chernobyl disaster seven years later drove the final nail into the nuclear coffin. It didn’t matter that the Three Mile Island containment vessel had done its job and prevented any significant release of radioactivity, or that Soviet reactors operated within a system that couldn’t build a safe toaster oven. Uranium was finished.

Three Mile Island came on the heels of the first great energy shock to hit America. On October 19, 1973, King Faisal ordered a 25 percent reduction in Saudi Arabia’s oil shipments to the United States, launching the Arab oil embargo. Oil supplies would tighten and prices would rise from then on, experts predicted. It would take some time, but oil was finished, too.

Five months after Three Mile Island, the nation’s first energy secretary summed up our predicament: “The energy future is bleak,” James R. Schlesinger declared, “and is likely to grow bleaker in the decade ahead. We must rapidly adjust our economics to a condition of chronic stringency in traditional energy supplies.” Fortunately, some argued, the U.S. could manage on less—much less. Smaller, more fuel-efficient cars were gaining favor, and rising gas prices would curb demand. The nation certainly didn’t need any new giant electric power plants—efficiency and the development of renewable sources of power would suffice. “The long-run supply curve for electricity is as flat as the Kansas horizon,” noted one right-thinking energy sage.

In the ensuing decades, however, American oil consumption rose 15 percent and electricity use almost doubled. Many people aren’t happy about it. Protecting our oil-supply lines entangles us with feudal theocracies and the fanatical sects that they spawn. The coal that we burn to generate so much of our electricity pollutes the air and may warm the planet. What to do? All sober and thoughtful energy pundits at the New York Times, Greenpeace, and the Harvard Divinity School agree: the answer to both problems is . . . efficiency and the development of renewable sources of power. Nevertheless, the secretary of energy, his boss (now a Texas oilman, not a Georgia peanut farmer), and the rest of the country should look elsewhere.

The U.S. today consumes about 100 quads—100 quadrillion BTUs—of raw thermal energy per year. We do three basic things with it: generate electricity (about 40 percent of the raw energy consumed), move vehicles (30 percent), and produce heat (30 percent). Oil is the fuel of transportation, of course. We principally use natural gas to supply raw heat, though it’s now making steady inroads into electric power generation. Fueling electric power plants are mainly (in descending order) coal, uranium, natural gas, and rainfall, by way of hydroelectricity.

This sharp segmentation emerged relatively recently, and there’s no reason to think it’s permanent. After all, developing economies use trees and pasture as fuel for heat and transportation, and don’t generate much electricity at all. A century ago, coal was the all-purpose fuel of industrial economies: coal furnaces provided heat, and coal-fired steam engines powered trains, factories, and the early electric power plants. From the 1930s until well into the 1970s, oil fueled not just cars but many electric power plants, too. And by 2020, electricity almost certainly will have become the new cross-cutting “fuel” in both stationary and mobile applications.

That shift is already under way. About 60 percent of the fuel we use today isn’t oil but coal, uranium, natural gas, and gravity—all making electricity. Electricity has met almost all of the growth in U.S. energy demand since the 1980s. About 60 percent of our GDP now comes from industries and services that use electricity as their front-end “fuel”—in 1950, the figure was only 20 percent. The fastest growth sectors of the economy—information technology and telecom, notably—depend entirely on electricity for fuel, almost none of it oil-generated. Electrically powered information technology accounts for some 60 percent of new capital spending.

Electricity is taking over ever more of the thermal sector, too. A microwave oven displaces much of what a gas stove once did in a kitchen. So, too, lasers, magnetic fields, microwaves, and other forms of high-intensity photon power provide more precise, calibrated heating than do conventional ovens in manufacturing and the industrial processing of materials. These electric cookers (broadly defined) are now replacing conventional furnaces, ovens, dryers, and welders to heat air, water, foods, and chemicals, to cure paints and glues, to forge steel, and to weld ships. Over the next two decades, such trends will move another 15 percent or so of our energy economy from conventional thermal to electrically powered processes. And that will shift about 15 percent of our oil-and-gas demand to whatever primary fuels we’ll then be using to generate electricity.

Electricity is also taking over the power train in transportation—not the engine itself, but the system that drives power throughout the car. Running in confined tunnels as they do, subways had to be all-electric from the get-go. More recently, diesel-electric locomotives and many of the monster trucks used in mining have made the transition to electric drive trains. Though the oil-fired combustion engine is still there, it’s now just an onboard electric generator that propels only electrons.

Most significantly, the next couple of decades will see us convert to the hybrid gasoline-and-electric car. A steadily rising fraction of the power produced under the hood of a car already is used to generate electricity: electrical modules are replacing components that belts, gears, pulleys, and shafts once drove. Steering, suspension, brakes, fans, pumps, and valves will eventually go electric; in the end, electricity will drive the wheels, too. Gas prices and environmental mandates have little to do with this changeover. The electric drive train simply delivers better performance, lower cost, and less weight.

The policy implications are enormous. Outfitted with a fully electric power train, most of the car—everything but its prime mover—looks like a giant electrical appliance. This appliance won’t run any great distance on batteries alone, given today’s battery technology. But a substantial battery pack on board will provide surges of power when needed. And that makes possible at least some “refueling” of the car from the electricity grid. As cars get more electric, an infrastructure of battery-recharging stations will grow apace, probably in driveways and parking lots, where most cars spend most of their time.

Once you’ve got the wheels themselves running on electricity, the basic economics strongly favor getting that electricity from the grid if you can. Burning $2-a-gallon gasoline, the power generated by current hybrid-car engines costs about 35 cents per kilowatt-hour. Many utilities, though, sell off-peak power for much less: 2 to 4 cents per kilowatt-hour. The nationwide residential price is still only 8.5 cents or so. (Peak rates in Manhattan are higher because of the city’s heavy dependence on oil and gas, but not enough to change the basic arithmetic.) Grid kilowatts are cheaper because cheaper fuels generate them and because utility power plants run a lot more efficiently than car engines.

The gas tank and combustion engine won’t disappear anytime soon, but in the imminent future, grid power will (in effect) begin to top off the tank in between the short trips that account for most driving. All-electric vehicles flopped in the 1990s because batteries can’t store sufficient power for long weekend trips. But plug-in hybrids do have a gasoline tank for the long trips. And the vast majority of the most fuel-hungry trips are under six miles—within the range of the 2 to 5 kWh capacity of the onboard nickel-metal-hydride batteries in hybrids already on the road, and easily within the range of emerging automotive-class lithium batteries. Nationally, some 10 percent of hybrid cars could end up running almost entirely on the grid, as they travel less than six miles per day. Stick an extra 90 pounds—$800 worth—of nickel-metal-hydride batteries in a hybrid, recharge in garages and parking lots, and you can shift roughly 25 percent of a typical driver’s fuel-hungriest miles to the grid. Urban drivers could go long stretches without going near a gas station. The technology for replacing (roughly) one pint of gasoline with one pound of coal or under one ounce of uranium to feed one kilowatt-hour of power to the wheels is now close at hand.

So today we use 40 percent of our fuel to power the plug, and the plug powers 60 percent of GDP. And with the ascent of microwaves, lasers, hybrid wheels, and such, we’re moving to 60 and 80 percent, respectively, soon. And then, in due course, 100/100. We’re turning to electricity as fuel because it can do more, faster, in much less space—indeed, it’s by far the fastest and purest form of power yet tamed for ubiquitous use. Small wonder that demand for it keeps growing.

We’ve been meeting half of that new demand by burning an extra 400 million tons of coal a year, with coal continuing to supply half of our wired power. Natural gas, the fossil fuel grudgingly favored by most environmentalists, has helped meet the new demand, too: it’s back at 16 percent of electricity generated, where it was two decades ago, after dropping sharply for a time. Astonishingly, over this same period, uranium’s share of U.S. electricity has also risen—from 11 percent to its current 20 percent. Part of the explanation is more nuclear power plants. Even though Three Mile Island put an end to the commissioning of new facilities, some already under construction at the time later opened, with the plant count peaking at 112 in 1990. Three Mile Island also impelled plant operators to develop systematic procedures for sharing information and expertise, and plants that used to run seven months per year now run almost eleven. Uranium has thus displaced about eight percentage points of oil, and five points of hydroelectric, in the expanding electricity market.

Renewable fuels, by contrast, made no visible dent in energy supplies, despite the hopes of Greens and the benefits of government-funded research, subsidies, and tax breaks. About a half billion kWh of electricity came from solar power in 2002—roughly 0.013 percent of the U.S. total. Wind power contributed another 0.27 percent. Fossil and nuclear fuels still completely dominate the U.S. energy supply, as in all industrialized economies.

The other great hope of environmentalists, efficiency, did improve over the last couple of decades—very considerably, in fact. Air conditioners, car engines, industrial machines, lightbulbs, refrigerator motors—without exception, all do much more, with much less, than they used to. Yet in aggregate, they burn more fuel, too. Boosting efficiency actually raises consumption, as counterintuitive as that sounds. The more efficient a car, the cheaper the miles; the more efficient a refrigerator, the cheaper the ice; and at the end of the day, we use more efficient technology so much more that total energy consumption goes up, not down.

We’re burning our 40 quads of raw fuel to generate about 3.5 trillion kilowatt-hours of electricity per year; if the automotive plug-and-play future does unfold on schedule, we’ll need as much as 7 trillion kWh per year by 2025. How should we generate the extra trillions of kilowatt-hours?

With hydrogen, the most optimistic Green visionaries reply—produced by solar cells or windmills. But it’s not possible to take such proposals seriously. New York City consumes so much energy that you’d need, at a minimum, to cover two cities with solar cells to power a single city (see “How Cities Green the Planet,” Winter 2000). No conceivable mix of solar and wind could come close to supplying the trillions of additional kilowatt-hours of power we’ll soon need.

Nuclear power could do it—easily. In all key technical respects, it is the antithesis of solar power. A quad’s worth of solar-powered wood is a huge forest—beautiful to behold, but bulky and heavy. Pound for pound, coal stores about twice as much heat. Oil beats coal by about twice as much again. And an ounce of enriched-uranium fuel equals about 4 tons of coal, or 15 barrels of oil. That’s why minuscule quantities contained in relatively tiny reactors can power a metropolis.

What’s more, North America has vast deposits of uranium ore, and scooping it up is no real challenge. Enrichment accounts for about half of the fuel’s cost, and enrichment technologies keep improving. Proponents of solar and wind power maintain—correctly—that the underlying technologies for these energy sources keep getting cheaper, but so do those that squeeze power out of conventional fuels. The lasers coming out of the same semiconductor fabs that build solar cells could enrich uranium a thousand times more efficiently than the gaseous-diffusion processes currently used.

And we also know this: left to its own devices, the market has not pursued thin, low-energy-density fuels, however cheap, but has instead paid steep premiums for fuels that pack more energy into less weight and space, and for power plants that pump greater power out of smaller engines, furnaces, generators, reactors, and turbines. Until the 1970s, engineering and economic imperatives had been pushing the fuel mix inexorably up the power-density curve, from wood to coal to oil to uranium. And the same held true on the demand side, with consumers steadily shifting toward fuels carrying more power, delivered faster, in less space.

Then King Faisal and Three Mile Island shattered our confidence and convinced regulators, secretaries of energy, and even a president that just about everything that the economists and engineers thought they knew about energy was wrong. So wrong that we had to reverse completely the extraordinarily successful power policies of the past.

New York has certainly felt the effects of that reversal. In 1965, the Long Island Lighting Company (LILCO) announced plans to build a $75 million nuclear plant in Suffolk County, to come on line by 1973; soon after, it purchased a 455-acre site between Shoreham and Wading River. A bit later, LILCO decided to increase Shoreham’s size and said it wanted to build several other nuclear plants in the area. Public resistance and federal regulators delayed Shoreham’s completion. Then Three Mile Island happened. In the aftermath, regulators required plant operators to devise evacuation plans in coordination with state and local governments. In early 1983, newly elected governor Mario Cuomo and the Suffolk County legislature both declared that no evacuation plan would ever be feasible and safe. That was that. By the time the state fully decommissioned Shoreham in 1994, its price tag had reached $6 billion—and the plant had never started full-power commercial operation. To pay for it all, Long Island electric rates skyrocketed.

What scared many New Yorkers—and thus many politicians—away from nuclear power was what had originally attracted the engineers and the utility economists to it: nuclear facilities use a unique fuel, burned, in its fashion, in relatively tiny reactors, to generate gargantuan amounts of power. Do it all just right, end to end, and you get cheap, abundant power, and King Faisal can’t do a thing about it. But the raw material itself, packing so much power into so little material, is inherently dangerous. Sufficiently bad engineering can result in a Three Mile Island or a Chernobyl. And these days, there’s the fear that poor security might enable terrorists to pull off something even worse.

How worried should we really be in 2005 that accidents or attacks might release and disperse a nuclear power plant’s radioactive fuel? Not very. Our civilian nuclear industry has dramatically improved its procedures and safety-related hardware since 1979. Several thousand reactor-years of statistics since Three Mile Island clearly show that these power plants are extraordinarily reliable in normal operation.

And uranium’s combination of power and super-density makes the fuel less of a terror risk, not more, at least from an engineering standpoint. It’s easy to “overbuild” the protective walls and containment systems of nuclear facilities, since—like the pyramids—the payload they’re built to shield is so small. Protecting skyscrapers is hard; no builder can afford to erect a hundred times more wall than usable space. Guaranteeing the integrity of a jumbo jet’s fuel tanks is impossible; the tanks have to fly. Shielding a nuclear plant’s tiny payload is easy—just erect more steel, pour more concrete, and build tougher perimeters.

In fact, it’s a safety challenge that we have already met. Today’s plants split atoms behind super-thick layers of steel and concrete; future plants would boast thicker protection still. All the numbers, and the strong consensus in the technical community, reinforce the projections made two decades ago: it is extremely unlikely that there will ever be a serious release of nuclear materials from a U.S. reactor.

What about the economic cost of nuclear power? Wind and sun are free, of course. But if the cost of fuel were all that mattered, the day of too-cheap-to-meter nuclear power would now be here—nearer, certainly, than too-cheap-to-meter solar power. Raw fuel accounts for over half the delivered cost of electricity generated in gas-fired turbines, about one-third of coal-fired power, and just a tenth of nuclear electricity. Factor in the cost of capital equipment, and the cheapest electrons come from uranium and coal, not sun and wind. What we pay for at our electric meter is increasingly like what we pay for at fancy restaurants: not the raw calories, but the fine linen, the service, and the chef’s ineffable artistry. In our overall energy accounts, the sophisticated power-conversion hardware matters more every year, and the cost of raw fuel matters less.

This in itself is great news for America. We’re good at large-scale hardware; we build it ourselves and keep building it cheaper. The average price of U.S. electricity fell throughout the twentieth century, and it has kept falling since, except in egregiously mismanaged markets such as California’s.

The cheap, plentiful power does terrific things for labor productivity and overall employment. As Lewis E. Lehrman notes, rising employment strongly correlates with rising supplies of low-cost energy. It takes energy to get the increasingly mobile worker to the increasingly distant workplace, and energy to process materials and power the increasingly advanced machines that shape and assemble those materials.

Most of the world, Europe aside, now recognizes this point. Workers in Asia and India are swiftly gaining access to the powered machines that steadily boosted the productivity of the American factory worker throughout the twentieth century. And the electricity driving those machines comes from power plants designed—and often built—by U.S. vendors. The power is a lot less expensive than ours, though, since it is generated the old-fashioned forget-the-environment way. There is little bother about protecting the river or scrubbing the smoke. China’s answer to the 2-gigawatt Hoover Dam on the Colorado River is the Three Gorges project, an 18-gigawatt dam on the Yangtze River. Combine cheaper supplies of energy with ready access to heavy industrial machines, and it’s hard to see how foreign laborers cannot close the productivity gap that has historically enabled American workers to remain competitive at considerably higher wages. Unless, that is, the United States keeps on pushing the productivity of its own workforce out ahead of its competitors. That—inevitably—means expanding our power supply and keeping it affordable, and deploying even more advanced technologies of powered production. Nuclear power would help keep the twenty-first-century U.S. economy globally competitive.

Greens don’t want to hear it, but nuclear power makes the most environmental sense, too. Nuclear wastes pose no serious engineering problems. Uranium is such an energy-rich fuel that the actual volume of waste is tiny compared with that of other fuels, and is easily converted from its already-stable ceramic form as a fuel into an even more stable glass-like compound, and just as easily deposited in deep geological formations, themselves stable for tens of millions of years. And what has Green antinuclear activism achieved since the seventies? Not the reduction in demand for energy that it had hoped for but a massive increase in the use of coal, which burns less clean than uranium.

Many Greens think that they have a good grip on the likely trajectory of the planet’s climate over the next 100 years. If we keep burning fossil fuels at current rates, their climate models tell them, we’ll face a meltdown on a much larger scale than Chernobyl’s, beginning with the polar ice caps. Saving an extra 400 million tons of coal here and there—roughly the amount of carbon that the United States would have to stop burning to comply with the Kyoto Protocol today—would make quite a difference, we’re told.

But serious Greens must face reality. Short of some convulsion that drastically shrinks the economy, demand for electricity will go on rising. Total U.S. electricity consumption will increase another 20 to 30 percent, at least, over the next ten years. Neither Democrats nor Republicans, moreover, will let the grid go cold—not even if that means burning yet another 400 million more tons of coal. Not even if that means melting the ice caps and putting much of Bangladesh under water. No governor or president wants to be the next Gray Davis, recalled from office when the lights go out.

The power has to come from somewhere. Sun and wind will never come close to supplying it. Earnest though they are, the people who argue otherwise are the folks who brought us 400 million extra tons of coal a year. The one practical technology that could decisively shift U.S. carbon emissions in the near term would displace coal with uranium, since uranium burns emission-free. It’s time even for Greens to embrace the atom.

It must surely be clear by now, too, that the political costs of depending so heavily on oil from the Middle East are just too great. We need to find a way to stop funneling $25 billion a year (or so) of our energy dollars into churning cauldrons of hate and violence. By sharply curtailing our dependence on Middle Eastern oil, we would greatly expand the range of feasible political and military options in dealing with the countries that breed the terrorists.

The best thing we can do to decrease the Middle East’s hold on us is to turn off the spigot ourselves. For economic, ecological, and geopolitical reasons, U.S. policymakers ought to promote electrification on the demand side, and nuclear fuel on the supply side, wherever they reasonably can.