Ars TechnicaThe increased awareness of the potential consequences of humanity's carbon emissions has generated intense interest in renewable energy sources. One nonrenewable technology that has also received significant attention is nuclear power. Although nuclear isn't truly renewable—there are finite sources of usable uranium—it has the significant advantage of being emissions-free once the construction and fuel isolation steps are completed. Although nuclear power does carry significant baggage in terms of safety and proliferation concerns, a significant barrier to its adoption remains the long-term storage of nuclear waste, some of which will remain a health threat for millions of years. Now, in a Policy Forum published in this week's Science, two former members of the US Geological Survey argue it's time to start addressing that issue by opening a long-term storage facility to pilot studies.
The problem the authors address is unlike anything humanity has ever faced. Some of the waste from nuclear plants will retain harmful levels of radioactivity for tens of thousands to millions of years. Beyond basic issues of securing and identifying it in a way that will persist even if our current culture doesn't, we will also have to encase it in a way that will be stable on geologic time scales. In the US, a proposed solution to the storage problem was to use areas in the desert Southwest where the water table remains hundreds of meters below the surface of geologically consolidated and stable mountains.
According to the authors, the trouble started with the selection of the site for the US storage facility. Initial legislation called for the full evaluation of three potential sites before choosing a final one; instead, five years later, Congress short-circuited the process and selected Yucca Mountain in Nevada. This action, according to the authors, obscured the fact that Yucca Mountain was on the list of finalists for a variety of well-documented technical reasons.
From there, the article discusses how the inherent uncertainties of science and engineering have left the public and legal system with a poor picture of our understanding of long-term storage. Science is poorly equipped to provide the certainty that everyone would like to see for a project of this nature, and engineering faces clear limits when predicting the behavior of structures over periods that are longer than human civilization has existed. "There is unlikely to be complete closure," the authors write. "Nor will honest disagreements among scientists and engineers regarding some YM [Yucca Mountain] issues likely ever cease."
Despite the uncertainties, the authors argue that there are very real reasons to start using Yucca Mountain: 60,000 metric tons of waste, currently stored in 72 sites, "many adjacent to metropolitan areas and all next to rivers, lakes, or the ocean." It's easy to default to inertia while waiting for greater certainty about Yucca Mountain or hoping something better comes along, but the authors argue that the current storage system creates far too much risk for this to be an acceptable path.
The paper argues that storage in the facilities at Yucca Mountain is not irreversible; if problems arise, the waste could be temporarily removed, or adjustments to the structural properties could be made. In fact, the authors argue, experience with pilot programs may be the best way to start reducing some of the outstanding uncertainties that are making the current debate so difficult. Without this knowledge, we may never be able to refine long-term models of waste storage.
What is perhaps most striking about the discussion is that the message of the authors focuses on helping the public understand that science is actually not a method of establishing certainty—"There need be no embarrassment to admit to the limitations of our explanatory and predictive capabilities," they write. Getting the public to realize that science can make the very best predictions possible despite residual uncertainties remains a significant challenge for nearly every case where science has to be translated to policy.
