CTBT, SSP, and U.S. National Security American Physical Society, May 1, 2000 Sidney Drell I want to address three questions: Why in the year 2000 is there so much interest in the Comprehensive Test Ban Treaty? Secondly, what strategic assumptions underlie U-S- signing the Comprehensive Test Ban Treaty (still to be ratified)? Thirdly, what do I see as the prospect, on technical grounds, of meeting the challenge to maintain reliable nuclear deterrence for US security under the CTBT? First two comments: Efforts to end all nuclear tests are not new. They commenced more than four decades ago during the Eisenhower administration* Upon leaving office President Eisenhower commented that not achieving a nuclear test ban "would have to be classed as the greatest disappointment of any administration - of any decade - of any time and of any party.," And second: If the U.S. is to ratify and adhere to a CTBT, it must first understand the technical implications well enough to conclude that its restraints are consistent with our national security. Question 1: Why is there so much interest today in the Comprehensive Test Ban Treaty? Why have the U-S. and a total of 155 nations altogether, including all 5 declared nuclear powers, signed the CTBT following the U.S. lead as first -signatory on September 24, 1996? The reason is simply that it is broadly recognized that a CTBT banning all nuclear explosive tests is essential to accomplishing broadly shared non-proliferation goals. This was made clear in the debate at the United Nations in May 1995 when 181 nations signed on to the indefinite extension of the nuclear Non-Proliferation Treaty (NPT) at its fifth and final scheduled five-year review. A commitment by the nuclear powers to cease testing became an essential part of the NPT bargain in 1995. It reflects the growing importance of reducing and eventually eliminating the discriminatory nature of the NPT in the post-Cold War world. The NWS would have to cease testing if the NNWS were to agree to forego all such work. [Positive and negative security assurances by NWS to NNWS, and Article VI on reductions.] The CTBT will do more than help limit the spread of nuclear weapons through the nonproliferation regime, particularly if current diplomatic efforts succeed in strengthening the provisions for verifying the NPT and for imposing appropriate sanctions for noncompliance. It will also dampen the competition among nations who already have nuclear warheads, but who now will be unable to develop and deploy with confidence more advanced ones at either the high or the low end of destructive power. The CTBT would also force rogue states seeking even a primitive nuclear capability to place confidence in untested bombs. As of this year, 187 nations have agreed to the indefinite extension of the NPT. There are only 4 non-signatories: India, Pakistan, Israel, Cuba. 155 nations have now signed the CTBT ending all nuclear explosions, anywhere and without time limit, This comes after more than 2,000 nuclear tests extending over the past 50 years. For the Treaty to come into force all 44 nations deemed nuclear capable, that is who have nuclear weapons or nuclear reactors for research or for power, must ratify the Treaty. 28 nations of the 44 required have ratified the CTBT. Altogether 53 nations have ratified it. The Russians, the Chinese, and the U.S. have signed but not ratified the Treaty. Eleven NATO nations, including France and UK, have ratified it. India, Pakistan (they await us), North Korea have not signed it. The US Senate voted the CTBT down in October in a highly charged political confrontation between the White House and the Senate- This action generated a world-wide torrent of criticism of US "unilateralism" (London Economist), since we led the right for it, generated a strong world-wide consensus for a true ban on all nuclear explosive testing as the NNWS insisted, and then blinked at the moment of truth. Turning next to Question 2: What are the strategic assumptions underlying interest in the CTEV. This is the context for considering the CTBT. 1, The US intends to maintain a credible nuclear deterrent for the foreseeable future. 2. The US remains committed to support a worldwide non-proliferation effort. (Figure 1) 3. Most critical is that the current US policy, first announced by President George Bush in 1992, of not developing any new nuclear weapons designs for deployment, will stay in effect; i.e. we will not encounter new military or political circumstances in the future causing us to abandon this policy, This reflects a Judgment that the U.S. does not need to develop new warheads with either larger or smaller capabilities than those currently in our arsenal. Our present large yield warheads are quite destructive enough. (up to 100 times the yields of the A-bombs dropped on Hiroshima and Nagasaki). At the lower end of the scale it would be better to seek conventional weapons with improved accuracy and penetration. Indeed, against a hard military target, each factor of two improvement in accuracy translates into an effective increase of 23 , or a factor or 8, in the effectiveness in destroying that target. This judgment could change in the future, were we to encounter new military or political circumstances, and therefore the U.S. has to maintain a readiness to respond appropriately. Let me then come to the third question, and that is: What is the technical situation with respect to testing vs. a comprehensive test ban and US national security'? To start with I want to review four important facts. These are: 1. Today the nuclear weapons that are designated to remain in our enduring stockpile are, and will remain for the near-term future, meaning for the next 5 - 10 years, safe, effective, and reliable. This is a finding of all technical reviews of the stockpile that I am aware of - including my own official studies. Our confidence in the stockpiles is based on understanding gained from 50 years of experience and analysis of more than 1,000 nuclear tests, including more than 150 nuclear tests of modem weapons types over the past 25-30 years. 2. The overwhelming majority of our nuclear tests during the Cold War were devoted to developing for deployment new and more advanced warheads and weapon-, systems. Only a very small percentage, well under 10% of the 150-200 underground nuclear explosive of modem weapons from 1972 to the end of testing in 1992 by the U.S., were so-called stockpile confidence tests; i.e. tests conducted on currently deployed weapons to confirm our confidence in them. That is well less than 1 year for the whole arsenal of many thousands -
4. Finally, the CTBT is of unlimited duration, but ten years after it enters into force, a conference of the states parties will be held to review the operation and effectiveness of the treaty. Each party has the right to withdraw at any time, with 6 months notice, if it decides that extraordinary events are changing the strategic situation, or unanticipated technical findings are jeopardizing our "supreme national interest". The U.S. highlighted the importance of this safeguard at the time of its signing. I will divide this discussion of the technical issues into three parts: A. Can the U.S. maintain confidence in the reliability and effectiveness of our forces over the long term under the treaty'? This is what the Stockpile Stewardship program (SSP) is all about- What does it do? B. Can we verify the treaty effectively; i.e., consistent with U,S. security requirements? C. Can we maintain safety of our deterrent without underground yield producing tests? These three questions were all raised in the Senate hearings last October and were the subject of some controversy and disagreement. What I will be describing are my own views, based upon more than a decade of extensive involvement in the nuclear weapons program in several capacities. Most relevant, during the 1990's I led a number of independent technical studies on the CTBT and stockpile stewardship for the US government. Most of these were done with scientists of the so-called JASON group, a group of independent scientific consultants, largely academic, with long experience working on technical problems as requested by U.S. government deparments. For a critical study on the specific role of what we learn from low yield tests, we were joined by four senior scientists who made more contributions to developing our modem irsenal. I also chaired for more than 6 years, up until March, 1999, the oversight council for the labs managed by the University of California, in particular, Live more and Los Alamos, i.e. the physics labs for developing and maintaining our nuclear arsenal. First as to reliability and effectiveness. The careful management and stewardship of the U.S. nuclear arsenal, that is its evaluation, surveillance, and maintenance has, been going on for more than 40 years. This program, led and implemented by outstanding scientists at the weapons labs, has given this country a nuclear deterrent that today is reliable and safe. We have a good statistical basis for high confidence in the deterrent. Since 1994 in response to the continuing moratorium on testing, and in anticipation of the signing of' a comprehensive test ban, a more sophisticated and technically demanding and comprehensive stockpile stewardship program has been developed. - This program is being implemented today with Strong and critically important bipartisan support including financial support of Congress and the White House. It is providing the needed data - which is the coin of the realm - for deepening our scientific understanding of nuclear weapons and what goes on during the explosion process, and of signatures and effects of the aging of the warheads that will be remaining in the stockpile longer than the earlier generations. For example, a variety of dynamic and static tests, including the important sub-critical experiments being pursued with very sophisticated equipment underground at the Nevada Test Site, are revealing detailed features of the crystal structure of plutonium, and whether its aging affects its strength and integrity under the enormous pressures and temperatures during the implosion. In addition we are doing detailed forensics on each weapon type in the stockpile. In particular each year 11 copies of each type are removed from the stockpile and evaluated for evidence of changes. One of each type is destructively disassembled and inspected in every detail for signs of cracks or defects developing as the warhead ages due, for example, to the radiation environment created by the plutonium, or due to defects in its production. These are areas in which the data are now available - for example on properties of plutonium and high explosives - and in which findings so far have shown that the weapons are not noticeably aging. With improved diagnostics in above ground experiments we are able, literally, to x-ray the behavior of a warhead up to when the nuclear chain reaction would be ignited, had the plutonium not been replaced by a non-fissioning substitute. (Over two years, such analysis gives 90% confidence of detecting a flaw present in 10% of the weapons in the stockpile). We also have the ability now, with the world's most powerful computers acquired by the Accelerated Strategic Computation Initiative (ASCI), to make detailed analyses and simulation-, using quantitative three-dimensional explosion codes. With these advanced codes and computers we can model imperfections due to cracks or voids that may develop in the structure and calculate to what extent they would degrade the performance of a warhead. The codes can be benchmarked against old data, and in particular by study of previous anomalous observations, as well as against new data on plasmas under extreme conditions of pressure and temperature anticipated in the future -from NEF. Overall this is a very sophisticated and technically challenging program* We are getting data not heretofore available. And with the two device labs, Livermore and Los Alamos, peer-reviewing each others work, there is no room for slack. (This is the best argument to keep both labs active in weapons work). This will tell us if, when, and what we have to do to refurbish or remanufacture warheads as needed, and most importantly it will enable us to hear whatever warning bells may ring signaling evidence of deterioration due to the aging, no matter how unanticipated, and enable us to make the necessary fixes. in a timely fashion. 4 Any scientist will always welcome more quality data, but the question is not what makes the job easier, or what is useful, but what is necessary. It was the unanimous conclusion of our studies -including four of the major creators of our present arsenal - that underground nuclear explosions have little to contribute, and nothing essential, relative to what we are presently learning from a multi -faceted, well supported stewardship program. There are qualified people who disagree with that conclusion, and the most serious arguments come in the following forms: Without testing, our cadre of nuclear weapons scientists will inevitably lose competence. We are already losing confidence in our deterrent because of the lack of testing since 1992. We have in the past found by testing that weapons already deployed in the stockpile had to be withdrawn and refurbished or redesigned, because of problems that developed and were only detected subsequently, some on the basis of test explosions. Clearly these questions have to be better addressed than they have been by the political process before a consensus for the CTBT can he established, It is a failing that we have. not done that well enough so far; but let me give here briefly my answers to those three concerns. My response to the first concern about retaining competence is based on my own research career at SLAC, also on my oversight of the programs at Livermore and Los Alamos for U.C. , also my interactions in my various technical studies, and most importantly as a member of the Congressionally mandated Commission chaired by Admiral Hank Chiles, former CINCSAC, on Maintaining United States Nuclear Weapons Expertise. I am confident that a strong science based stockpile stewardship program that challenges the scientists to develop a deeper understanding of what goes on during an explosion, and also provides them with the latest diagnostic and computer tools, will challenge, attract, and sustain. top notch weapons scientists at Livermore and Los Alamos much more effectively than blowing out the side of a mountain with an already existing warhead. They may not test and prove new designs but they can of course think about them too. This is a big management challenge, and a major responsibility of government to provide sustained support. As far as losing confidence in our deterrent, I said in my Congressional testimony on October 7, that I, today, have more confidence in the long-term credibility of our stockpile than was possible five years ago. This conclusion is based upon what has been learned from the stockpile stewardship program over the last five years, and the formal reviews that require the laboratory directors to certify annually, to the President, the Secretary of Energy, and the Secretary of Defense, that our weapons are meeting stated military requirements. The data being derived from our stockpile stewardship program is far more important for understanding our enduring arsenal, and maintaining confidence in its performance, than continued underground, very-low-yield testing. We now know much more, based on real data, how Pu, high explosives, etc. age. They aren't! We have looked at this with care.
As to the allegations that in the past we have found that many of our weapons introduced into the stockpile had subsequently developed problems and were withdrawn, I have two comments to make. One statement in congressional testimony claimed that 75% of our weapons had faced such post-deployment problems. In fact, most of those cases refer back to the period 1958 to 1961, during a testing moratorium with the Soviet Union when a number of new weapons were introduced into the stockpile. without tests. Furthermore what is now a mature subject was then in a much more speculative, developmental stage, As to more recent history, numbers like 33% have been bandied about as the percentage needing post-deployment fixes. But in our 1995 JASON study on Nuclear Testing, we went through this in detail, working with the four weapons designers mentioned previously, I will quote our unclassified conclusion. This is the unclassified summary of a very detailed, several hundred page long classified study full of technical details: "For the weapon types planned to remain in the enduring stockpile we find that the device problems which occurred in the past, and which either relied on, or required, nuclear yield tests to resolve, were primarily the result of incomplete or inadequate design activities. In part, these were due to the more limited knowledge and computational capabilities of a decade, or more, ago. We are persuaded that those problems have been corrected and that the weapon types, in the enduring stockpile are safe and reliable in the context of explicit military requirements."
There is a basic point in all this when we compare now with 10 or more years ago: The end of the Cold War has signaled a dramatic change in the nuclear weapons program of the U.S. The continuous cycle of new nuclear weapons development, testing and deployment has ended. We have replaced it by a SSP that I support and view to be both necessary and appropriate to ensure that the nuclear weapons in the arsenal remain safe and reliable into the indefinite future without nuclear testing. Let me turn next to the second part of the technical case. Can we verify effectively a CTBT? This again proved a contentious point in the discussions and was not satisfactorily resolved, but has to be in the future. Let me give the reasons I state that the CTBT can be effectively verified. The word "effective" is critical here: Will the kinds of low-yield, covert tests that we may not detect or identify have any impact on our security? When it comes into force, the CTBT will improve America's current ability to detect and identify very low yield nuclear testing that may be done in violation of the provisions of the CTBT. In particular the Treaty will add a significant number of seismic, hydroacoustic, radionuclide, and infrasound sensors that are part of the CTBT's International Monitoring System to supplement the existing system, including our national monitoring capabilities, Furthermore with the CTBT in force we will have the ability to request short-notice, on-site inspections of suspicious events. Added together these will make it much more difficult for other nations to believe that they might get away with very low-yield clandestine testing. Having said that, let me add: I consider cooperation, with appropriate transparency and confidence building measures, to be essential ingredients of a successful CTBT regime. I support every serious effort to achieve such cooperation with other weapon states - and Russia and China in particular. I know that there are serious questions about what the Russians are doing at their test site in Novaya Zemlya. I am aware of extensive work there as part of the Russian stewardship program and have been extensively involved in reviewing it for the DCI. I am not persuaded by the evidence - and we do get very good data from a seismic array in the Scandanavian area - that nuclear-yield producing testing has occurred in violation of the CTBT_ Previous allegations that the Russians carried out nuclear yield producing tests at Novaya Zemlya in August 1997 were demonstrated to be false. Concerning our detection limits and possible value to the Russian military of very low yield testing I can make two comments: We will be able to strengthen our ability to monitor and detect nuclear explosions when the CTBT's full international monitoring network and challenge on-site inspection regime is in force, as I already remarked. And secondly, I see no threat to U.S. security in what is going on at present at Novaya Zemlya, This is, an issue that will require resolution in a satisfactory treaty regime for which improved information exchange and transparency with the Russians will he required. I support adding specific requirements to our record of ratification of the CTBT to ensure more transparency, and mutual access with sensors at the test sites, on a direct bilateral basis with Russian and China. As in the debate leading to ratification of the Chemical Weapons Convention in 1999, this should not require reopening the Treaty for re-negotiation. The International Monitoring System of the CTBT will also greatly strengthen the ability to monitor and interpret activities by rogue states seeking to develop a new weapon capability. The Treaty will of course not affect those who choose to ignore it and it will be up to the international community to deal with such states through sanctions and political pressure, aided by whatever added credibility and power of dissuasion that is gained from establishing a no-testing norm. However for those countries -seeking to enter the nuclear club clandestinely, the task will be made |