The ongoing nuclear crisis at Japan’s Fukushima Daiichi nuclear power plant has brought to mind nuclear disasters from the past, specifically, the accidents at Chernobyl in 1986 and Three Mile Island in 1979. While experts will work in the coming years to determine what actually happened in Japan and how it compares to the events at Chernobyl and Three Mile Island, there is an interesting lesson to be drawn from the Fukushima Daiichi disaster by comparing it to a nuclear plant that wasn’t even built: New Jersey’s Atlantic Generating Station. The Atlantic Generating Station (AGS) was extensively covered by John McPhee in The New Yorker in an article published May 12, 1975. That article was also included in McPhee’s 1979 collection of essays, Giving Good Weight. The proponent of the AGS, Public Service Electric & Gas Co. (PSEG), dropped its plans for the plant in 1978, but the story of the AGS nonetheless can provide a valuable lesson.
The most notable feature of the AGS was that it would be located on a floating barge in the ocean several miles off the Jersey shore. To protect the plant, PSEG planned to build a large breakwater that would extend seven fathoms to the sea floor around the site where the barge would be anchored. The site that was chosen for the AGS was off the South Jersey Shore, 2.8 miles from Little Egg Inlet and 11 miles northeast of Atlantic City. According to McPhee, the primary reason PSEG wanted to build a floating nuclear plant in the ocean was that it had run out of sites on land where plants could be located. Richard Eckert was the PSEG engineer who dreamed up the AGS one day while in the shower: as McPhee writes, “a large part of his job was to seek out new sites for power plants—an assignment that had once, in a less complicated era, called for a person who could look at a map and find a river. Now difficulties had thickened to a point near desperation, and plant siting required scientific and diplomatic talents undreamed of in the age of the pluming smokestack and the good five-cent cigar.”
Because of the great harm that a nuclear plant can do if something goes wrong, engineers must design a plant to minimize the risk so that the expected harm from the plant is reduced to a satisfactory level. This kind of risk management is part of the cost side of a cost-benefit analysis: take the harm that could happen, multiply it by the chance that it will occur, and you will have the expected harm. In addition to the standard expected harm that comes along with any nuclear plant, the AGS had the potential to do greater harm, and was exposed to greater risks. A worst case scenario might have seen nuclear waste contaminating some sizable portion of the New Jersey coastline—perhaps the shores of neighboring states and even the Gulf Stream. Floating in the ocean, the AGS would be exposed to risks of hurricanes, high tides, and out of control ocean tankers. PSEG thought it could respond to this greater possible harm and greater risk profile by building greater safety measures, namely the breakwater that would surround the plant. Taking measures to reduce the risk would have resulted in a proportional decrease in the expected harm from the plant.
However, such efforts cannot eliminate the expected harm altogether; they can only reduce it to what is thought to be an acceptable level. The designers of the Fukushima Daiichi plant knew there was a risk that a tsunami could hit the plant, and, accordingly, they built the plant to withstand a 5.7 meter tsunami—preparing for a tsunami greater than 5.7 meters would have made the plant safer, but it apparently would have been extraneous according the cost-benefit analysis. When the 14 meter tsunami of March 11 hit, the expected harm came home to roost. In some ways, the Fukushima Daiichi disaster can be summarized by the cliché, ‘accidents happen.’ Seeing the results of that accident has reinforced the conclusion that PSEG’s decision to drop plans for the AGS was the right choice.
Even if PSEG was right that it could reduce the risk of a catastrophic accident from occurring at the AGS to an acceptable level in a statistical sense, it is possible that the actual harm that could have been done by the AGS in a worst case scenario was so large that it should not have been taken on. In a casino, the house is big enough and plays often enough that it can afford to be risk-neutral. When regulators are called upon to consider a project that gambles the shoreline of the nation’s most densely populated state, they should adopt a more risk-averse posture. Real property is not a liquid asset, and irreversible harms should not be lightly risked. This is not to say that nuclear power is altogether unacceptable; however, siting decisions should be made to minimize potential harm, not to ease the bureaucratic process.