Nuclear power in the future won’t be ‘too cheap to meter’

What will nuclear power cost if new plants are built in the US? Back in 1954, Lewis Strauss, chairman of the Atomic Energy Commission, predicted that nuclear power would be “too cheap to meter”. Reality turned out differently. Construction costs for nuclear plants turned sharply upward beginning in the early 1980s (see graph below), just the opposite of the direction one would expect from “learning by doing”. The causes of this escalation are difficult to pinpoint with certainty, but longer and longer construction times (due to non-standard designs, increased scale and complexity, and added regulation following the Three Mile Island accident) were important contributors. 

But the question on many peoples’ minds today is not what the last nuclear power plant cost, but rather what the next nuclear plant will cost to build. By far the largest contributor to the generating cost for a new nuclear power plant is the capital investment, so this is a critical question. Unfortunately, we don’t actually know what it will cost to build a new nuclear power plant in the US today, because no new plants have been constructed in recent decades.  This is a key motivation for the Department of Energy’s loan guarantee program: facilitate financing of a few plants (by lowering the cost of borrowing) to help find out what the cost of building them is today.

Until these few plants are built, we have only paper studies that estimate costs, including three prominent ones in the past couple of years.


Estimated construction costs (including interest during construction) per unit of generating capacity for individual US nuclear power plants. Note the sharply rising costs starting in the 1980s.