Letter published in Physics Today , May 1999, pp. 11-12.
On 22 December 1998, as reported in your February 1999 story entitled "DOE Decides TVA is Cheapest, Most Flexible Option to Produce Tritium for Nuclear Weapons" (page 54), Secretary of Energy Bill Richardson selected the Tennessee Valley Authority's Watts Bar and Sequoyah nuclear reactors as preferred tritium production facilities for US nuclear weapons. While we commend Richardson for choosing the least expensive method, the US could reap even greater cost savings by reducing its nuclear arsenal in paral lel with Russia's retirement of obsolete and decrepit systems in its nuclear arsenal.
The US has not produced tritium since 1988, when it shut down the Savannah River Site's tritium production reactors for safety reasons. Given that tritium decays with a halflife of about 12 years, the US will have about half as much tritium in 2000 as it did in 1988, one-quarter as much in 2012, one-eighth as much in 2024, and so on. Since 1988, dismantled nuclear weapons have supplied tritium for weapons being kept in the US arsenal. Although the Department of Energy (DOE) has classified the amount of t ritium available for US weapons, one can still estimate the effects of additional arms reductions on the need for new tritium.
The current START I arsenal contains about 8400 strategic and tactical warheads in an operational stockpile and about 2300 warheads in a reserve stockpile. Although it is known that the operational warheads are all filled with tritium, it is not known -- in the open literature -- how many reserve warheads actually have tritium allocated to them.
We estimate the number of tritium-allotted warheads by beginning with the year 2016, when, according to DOE, the stockpiled tritium will dip below the requirements of the currently planned START II and tactical arsenals. Those arsenals are scheduled to c onsist of 3500 deployed strategic, 1000 tactical, 500 spare, and 2500 upload "hedge" warheads, for a grand total of 7500 warheads. Using the radioactive decay equation and working backward to 2010, when, according to DOE, insufficient tritium will be a vailable for the START I and tactical arsenals, we find that the current tritium requirements are to support about 10,500 warheads.
The current obstacle to agreed-upon reductions to START II levels is the Russian Duma's reluctance to ratify that treaty, which was signed in 1993. The US government insists that the Duma must ratify START II before negotiations on deeper cuts can be la unched. Regardless of ratification, however, most knowledgeable Russian analysts project that Russia's rapidly decaying strategic arsenal will drop below START II levels within a few years of the treaty's final implementation deadline at the end of 2007 a nd will continue to fall much further.
Nonetheless, for two reasons, the Pentagon insists on an upload "hedge" consisting of thousands of extra warheads filled with tritium. First, the US can redeploy these warheads if it feels that its "supreme national interests" require it to reverse the reductions process. Second, the US can use the upload "hedge" as a bargaining chip to compel Russia to negotiate reductions in its tactical arsenal during START III negotiations. It seems likely, however, that most of Russia's tactical warheads will also have to be scrapped within a decade. In all probability, Russia will be able to replace only a few hundred of those warheads.
We also question DOE's requirement to maintain a five-year reserve supply of tritium, which will force DOE to begin producing tritium in 2005 and 2011 for the START I and II arsenals, respectively. The five-year reserve is an anachronism resulting from t he time needed to restart the Savannah River Site's reactors. This lead time could be significantly shortened with DOE's new policy, which calls for using already operating power reactors as a tritium source.
Adhering to START II's agreed-upon number of 3500 deployed strategic warheads, the US could keep a total of 4500 warheads filled with tritium, including 1000 tactical warheads. DOE could delay tritium production until 2025.
Under START III, the total US arsenal could fall below 3500 warheads, thus delaying tritium production until 2029.
We believe that under longer-term reduction agreements, involving Russia, China, France, and the UK, the US arsenal could shrink to only about 200 warheads--still more than enough explosive power to destroy any nation. Such a drastic reduction could make possible a delay in the resumption of tritium production until 2080.
The fiscal 1999 Defense Authorization Act bars the US government from spending any money on tritium production. This one-year hiatus provides policymakers and other concerned parties with a period for serious study of tritium requirements under different scenarios, including START I, II, and III, as well as reciprocal unilateral arms reductions.
However, aside from reducing the five year reserve, DOE cannot implement any new production plans unless the administration and Congress break their stalemate. Specifically, although the Pentagon is interested in cost-saving reductions of US strategic fo rces even if START II remains stalled, Congress has enacted a law mandating that US nuclear forces will remain at START I levels as long as START II remains unratified. The president cannot enforce any Pentagon-proposed reductions without Congress acting first.
 W. M. Arkin, R. S. Norris, and J. Handler, Taking Stock: Worldwide Nuclear Deployments 1998, Natural Resources Defense Council, Washington, DC (1998), p. 1.
 Arkin, Norris, and Handler, p.11.
Federation of American Scientists
FRANK VON HIPPEL
Princeton, New Jersey