Electricity's Dirty Secret

For the decade of the 1980s I worked in the Corporate Communications department of the Pacific Gas & Electric Company, PG&E, one of the largest energy companies in the country, with a service territory covering most of Northern California. One of the biggest things I learned from this time—aside from the fact that your local utility is made up of good people who support their community—is that there are many ways to generate electricity and the key to choosing among them is economics rather than technology.

By a quirk of geography and history, PG&E had—and still has, for all I know—one of the most diversified generating systems in the country, although some of that generating capacity has since been spun off to private owners and suppliers. The company inherited a network of dams and flumes in the Sierra Nevada that provided powerful water jets for hydraulic gold mining in the 19th century, and these were easily converted to run turbines in hydroelectric powerhouses up and down the mountains. It had four large thermal power plants—steam boilers driving turbine generators—that drew on the company’s abundant natural gas supplies for fuel. PG&E also operates smaller units that burn the gas to directly drive turbines, similar to a grounded jet engine attached to a generator. It built a major nuclear power plant at Diablo Cove in San Luis Obispo County, and built almost two dozen generating units drawing on The Geysers geothermal steam field in Sonoma and Lake counties. It draws power from the world’s largest photovoltaic power plant, on the Carrizo Plain, also in San Luis Obispo County, and from the Shiloh wind-power farm in the Montezuma Hills along the Sacramento River in Solano County, among others. The company buys electricity from the Western System Power Pool (WSPP) and the California Independent System Operator (CaISO).

With all of this diversity, PG&E’s energy cost is relatively low, depending on factors like snowfall in the Sierras to feed those dams and the state of the aquifer feeding the steam fields. The company does not draw on enough renewable energy—yet—to be much affected by variations in the wind and sundown over the state.

If the state ever fulfills its promise to get rid of all fossil fuels and provide all power from renewables like wind and solar, the remaining nuclear and geothermal assets will not be able to make up the difference from those abandoned gas-fired power plants.¹ There is talk of making up the difference from windless days and dark nights with some kind of energy storage: batteries, compressed air in underground chambers, or superconducting materials that let an electrical charge chase round and round in a donut-like torus. None of these technologies has been tried or proven at any scale needed to supply a utility grid. There is also talk of mandating solar powered roofing in all new housing and retrofitting existing roofs, with transformers to convert the electricity to household current and with batteries to supply energy on dark days and at night. Aside from the initial cost and payback time, generally measured in tens of years, these plans are intended—at least in the promoters’ dreams—to put the local utility entirely out of business.²

The dream of “free electricity” without fuel costs or emissions, using wind and solar power, runs into some basic engineering realities involving energy efficiency and capital cost.

In making these technologies work, the engineer has to move from conceptual design—linking up components, energy flows, and costs in back-of-the-envelope calculations and drawings—to detail design—putting the components in place at the right scale, establishing the true cost of each component, and accounting for variables like heat loss and line losses.³

Engineers constantly work with another variable set, too. For them, there is no such thing as perfection, no solution that is best under all conditions. Everything is a tradeoff in the engineer’s world. Instead of “good” and “bad,” the engineer thinks in terms of “better” and “worse.” You can make electricity with a gasoline generator—if the EPA and county authorities will approve it—or with a hand crank, or by rubbing a silk scarf on a glass rod. The question is always—and this is what I learned at PG&E—at what site, with what investments, and using what fuel supply at what cost? How attractive or interesting or politically correct the technology might be is not a factor.

Solar photovoltaics—generating an electric current by using the energy in sunlight to pass an electron through a semiconductor substrate—is about 20% to 22% efficient, even in cells and panels of the highest quality. This means that three-quarters of the solar energy that falls on them is lost to heat or reflection. And how that efficiency is affected by dust or a layer of snow and ice is still undetermined in large-scale applications, although probably not to good effect. Perhaps, in time, research into new materials can boost that efficiency up to maybe 30%, but much farther than that doesn’t seem to be in the cards.

Wind turbines have an efficiency of about 50% to 59%. This is comparable to the energy efficiency of a gas turbine or thermal power plant. But wind farms require the right conditions, a place with strong, steady, and predictable winds. Like a geothermal steam field, such locations are a resource that can’t be established by fiat or political rezoning. And wind turbines, like any machine dealing with strong forces, are subject to mechanical stresses on the blades and shafts. Although their energy resource is free, the capital investment to harvest it is expensive, not easy to maintain—that is, a heavy generator on a tall tower, sometimes sited on a hilltop, is harder to fix than a turbine generator under cover in a power plant—and subject to depreciation and eventual replacement.

Either of these fuel-free, renewable resources would require the participating utility to maintain a commensurate amount of “spinning reserve”—an alternate, dispatchable generating resource all fired up and ready to come on line instantly to meet the system load dropped when the wind dies or the sun goes behind a cloud. In most cases, this reserve power would have to come from fossil fuels, because the small amounts of electricity available from hydro and geothermal power, and the supply from an operating nuclear plant, would already be spoken for. And some form of “battery backup” on a systemwide basis is not currently technically or economically feasible.

And finally, fusion—the dream of limitless energy by harvesting hydrogen isotopes from sea water and compressing them with laser blasts or electromagnetic fields—is still ten years away. Always ten years away. Yes, we can make deuterium and tritium fuse with either compression technology; we just can’t make them give off more energy than we must put into the reaction. For now, it seems, the only way to fuse hydrogen into helium reliably is to compress it in a steep gravity field, like the inside of a star. Until we find some magical gravity-manipulation technology, utility-scale fusion is going to remain a dream.

All of these renewable technologies—except for fusion—have their place in a diversified system. None of them is ready, yet, to satisfy all of our energy needs. And a modern economy runs on ready availability of energy the way ancient economies ran on resources of clean water and food. Maybe in a few hundred or a thousand years, when we have run out of conveniently obtained fossil fuels, we will develop efficient and low-cost solar⁴ or fusion power. But for right now, we run on bulk carbon energy.

And no amount of wishing will make it otherwise.

1. Of all the fossil fuels, natural gas is the most efficient in terms of high energy output with low carbon dioxide emissions. This is because the methane molecule (CH₄) burns completely, breaking all of its hydrogen bonds in turning methane into carbon dioxide and water. Other carbon sources like coal and oil either burn incompletely or tend to put soot particles and other contaminants into the atmosphere along with greater amounts of carbon dioxide.

2. Of course, manufacturing plants that need large amounts of electric power to run their operations—more than rooftop solar can supply, like steel mills, auto factories, shipyards, and other heavy industries—can either run their own generating stations or leave the state.

3. Building a solar- or wind-power farm—whose energy resource and efficiencies are generally be weaker than a thermal plant’s, and which will generally have to be sited some distance from the end user—must take into account energy lost to resistance and heat on a transmission line. This is usually accounted as 5% to 15%, depending on distance traveled.

4. Probably from orbit, as in my novel Sunflowers, where sunlight has an energy density of 1,300 watts per square meter instead of the 130 W/m² that strikes the Earth’s surface.