NRC Should Perform Non-Proliferation Assessment of Laser Enrichment Technology

Francis Slakey & Linda Cohen

Introduction

Uranium enrichment is a step in the process to convert uranium ore into fuel for nuclear reactors. Mined uranium ore is made up of roughly 99.2% U²³⁸ and 0.72% U²³⁵. Only the latter isotope is fissionable under bombardment by slow neutrons as in a reactor, and so in order to make reactor fuel, the U²³⁵ concentration must be increased. Enrichment services that increase the concentration are measured in separative work units, or SWUs. The number of SWUs required to produce a kilogram of reactor fuel depends on the percent of U²³⁵ in the final fuel, in the natural uranium feedstock, and in the depleted uranium stream or "tail". Numerous techniques of isotope separation have been developed for enriching uranium and their efficiency can be characterized by the number of SWUs produced per megawatt hour (SWU/MWh).

A uranium enrichment technology, SILEX (Separation of Isotopes by Laser Excitation), could significantly increase efficiency beyond existing centrifuge technologies and pose significant proliferation risks [1]. In this article we argue that an examination of the economics indicates that those risks would not be outweighed by a public benefit to the consumer in the form of lower electricity bills, as some have suggested. Consequently, under the Atomic Energy Act, the Nuclear Regulatory Commission (NRC) should carry out a non-proliferation assessment of the technology to determine whether SILEX "would (not) be inimical to the common defense and security" of the United States.

As opposed to past techniques which exploited the slight difference in mass between the two isotopes, SILEX uses a laser, tuned to a particular excitation of U²³⁵, to differentiate it from the U²³⁸. More than 20 countries have dabbled in laser enrichment over the past two decades, including South Korea and Iran, without much success. SILEX was developed by the Australian company Silex Systems, and is now being commercialized exclusively by GE Hitachi. GE Hitachi has applied for a license from the NRC to operate a full-scale commercial SILEX plant in North Carolina. Currently, the U.S. is the only country in which GE-Hitachi has applied for a license. While they have not yet made a decision whether to commercialize the technology, the license is a necessary step in their development path.

There is clear private benefit to laser enrichment: If GE-Hitachi successfully commercializes SILEX, it stands to make hundreds of millions of dollars a year. A more challenging question is whether there is a net public benefit. To evaluate the public benefit, we consider three separate issues: technical, economic, and legal.

Technical Issues

The proliferation risks of an enrichment technology can increase as the technology becomes more efficient. In general, if the technology occupies a very small space, its construction may no longer be observable through satellite surveillance. And, if it operates on very low power, it may no longer require an observable dedicated power source or have a detectable heat signature. In other words, an extremely efficient enrichment facility could be below the detection limit. The answers to just a few technical questions can begin to establish the proliferation risk of an enrichment technology:

• Can the technology be used to fabricate weapons grade fissile material?
• Are all the components of the technology "dual use" (i.e., do they have other non-military applications.)
• Is the technology undetectable in its construction and operation?

SILEX details are proprietary and undisclosed; however, based on publically available information, it is clear that the answers to some – perhaps all - of these questions are "yes" [2]. For example, the company has stated that SILEX occupies a space 75% smaller and has substantially lower energy requirements than a centrifuge facility, which would make it extremely difficult to detect [3]. Centrifuge technology is already challenging to detect, as evidenced by the recently declared facility in Qum, Iran [4] and a facility 75% smaller only increases the detection challenge.

Clearly, then, SILEX raises proliferation concerns. Indeed, it was in part this very issue that compelled the members of the APS POPA Report on Nuclear Weapons Downsizing to recommend that NRC should address non-proliferation threats in the licensing process [5]. Others have raised similar concerns [6].

Economic Issues

Could the public risks be outweighed by public economic benefits? Certainly, there is a potential for consumer savings in making enrichment more efficient. Today, ten percent of the cost of electricity from nuclear power is attributed to the cost of fuel, and of that, roughly 50% is due to enrichment costs. The Energy Information Administration estimates average fuel costs from nuclear power in the United States at 0.45 cents per kilowatt-hour [7]. In 2007, 806 million MWh of electricity was generated in the United States from nuclear power, yielding a value to enrichment services in the U.S. of approximately $1.8 billion dollars.

In 2006, a SILEX executive stated that it anticipated the technology to be anywhere from 1.6 to 16 times more efficient than first-generation gas centrifuges [1]. Since details are undisclosed, the efficiency claims are impossible to verify. But assuming a continuation of historical trends in enrichment efficiency (which follows Moore's law, as many technologies do [8]), a fair assumption is the doubling of today's best efficiency by 2020. In that case, if laser enrichment lowers enrichment costs by half and all the savings are passed on to consumers, then the average household would save approximately 66 cents per month [9].

Over time, the savings could be greater than 66 cents/household/month. For example, household savings from cheaper enrichment would rise if demand for nuclear power increases. Doubling nuclear generation in the United States by 2025 (an ambitious growth scenario for the nuclear industry) could double the value of enrichment savings to $1.32 per household a month. In addition, a change in the relative prices of enrichment services (lower) and natural uranium (higher) will increase the demand for SWUs in the production of fuel. If the price of the former halves and the price of the latter doubles, we calculate, based on a cost-optimization of formula for enrichment processes from the Massachusetts Institute of Technology [10], that demand for SWUs will increase by 40% for the same level of electricity production from nuclear power. These factors would therefore increase savings by an additional 53 cents per household a month.

The above calculation gives laser enrichment technology every possible consumer advantage, namely

• All cost savings are passed along to the consumer.
• Nuclear power doubles over the next 15 years and thereby increases demand for uranium.
• The process is twice as efficient as the best available technology tomorrow – and not just today. In fact, centrifuge technology may improve as well, so that the relative advantage of SILEX may be much lower than we estimate here.
• 100% market penetration of the technology within 15 years.

All of these assumptions are far too generous to the technology, the last one particularly so. Indeed, SILEX's own documents indicate that they anticipate, at best, achieving only 25% market penetration in 15 years. In that case, conceding the other four assumptions for the sake of argument, the average household could expect a savings on their electricity bill due to this innovation to be at most 45 cents per month. That maximum savings is too meager to compel a decision to pursue laser enrichment based on public benefit to the consumer.

Legal Issues

We raised the preceding issues in an article in Nature magazine [11]. Representatives of GE-Hitachi did not challenge our conclusions; instead, they informed reporters that proliferation issues are not the business of NRC. As evidence, they point to a 2005 NRC judgment in which the NRC declined to consider proliferation issues for an enrichment facility [12]. Therefore, GE-Hitachi claims, a jurisdictional precedent was established that NRC should not consider proliferation issues in the case of SILEX.

GE-Hitachi has an overly narrow reading of the 2005 decision. Two Acts govern the jurisdiction of the NRC: the Atomic Energy Act (AEA) and the National Environmental Policy Act (NEPA). Under the AEA, the NRC must determine whether a technology "would (not) be inimical to the common defense and security" of the United States. Under NEPA, the NRC must consider the environmental impacts on the United States. The 2005 decision did not refer to any AEA authority relating to non-proliferation issues, but instead relied exclusively on NEPA and therefore cannot serve as a comprehensive precedent. In fact, at this time, there is no precedent that would require or disallow non-proliferation assessments as part of a licensing process.

Nevertheless, the non-proliferation assessment that we propose is within the jurisdiction of the NRC based on the interpretation of the phrase "inimical to the defense and security" of the United States in the AEA. Smaller, less detectable technologies for the production of nuclear materials are potentially proliferation game-changers. Such technologies would be far more difficult to detect within the existing International Atomic Energy Agency (IAEA) Safeguards inspection regime and therefore would indeed be "inimical" to the security of the United States. The fact that SILEX is a departure from previous technologies and therefore requires close scrutiny by the NRC was acknowledged by NRC Chairman Gregory Jaczko in a recent interview:

"It's a very new technology, or a novel technology. It's not similar to the kinds of enrichment facilities we've licensed in the past. So, I certainly think there may be some things we need to take a look at and make sure we've got the right approach to ensuring that kind of protection of the technology and the material. [13]"

Additional Considerations

There have been three common responses to our analysis. First, an argument has been made that by developing laser enrichment technology in the United States, US entities can ensure that the technology is adequately safeguarded against proliferation. History does not instill confidence in this approach. Previous enrichment technologies - the calutron, gas centrifuge and advanced centrifuges - have all created proliferation risks over the past 50 years despite efforts to withhold the information. By having the NRC carry out a non-proliferation assessment to determine, among other things, whether the facility design and core technical discoveries can be secured against theft, there would at least be a possibility of reducing the risk of repeating the type of breech that led to A.Q. Kahn taking centrifuge designs from the Netherlands to Pakistan.

A second response is that "the genie is out of the bottle" and if the United States doesn't develop it, some other country – one potentially hostile to our interest – will develop it. Actually, the genie is not "out of the bottle." Historically, proliferation to rogue states doesn't occur until a technology has progressed along the learning curve to the point it has been industrially proven or has achieved production scale. The technology may be ultimately commercialized, but we believe that in this case any delay would be useful: first, in each year that the technology exists, costs exceed benefits, and second, a thorough review of risks today, together with delayed commercialization, might incentivize the manufacturers to include safeguards in the package.

A final response is that the non-proliferation assessments we propose might stymie the US nuclear industry. In fact, we believe the assessments would have the opposite effect. Careful consideration of proliferation risks is in the best interests of the expansion of nuclear power, a view that was the subject of an APS report [14].

References

[1] http://www.silex.com.au/public/uploads/announce/House%20of%20Reps%20Presentation%20090206.pdf

[2] http://www.fas.org/sgp/othergov/doe/lanl/docs4/silex.pdf

[3] http://www.forbes.com/global/2009/1116/outfront-nuclear-power-uranium-ge-riches-in-enrichment.html

[4] http://www.time.com/time/world/article/0,8599,1929088,00.html

[5] "Technical Steps to Support Nuclear Arsenal Downsizing," a Report by the APS Panel on Public Affairs, 2010.

[6] Charles D. Ferguson, Laser Enrichment: Separation Anxiety, Bulletin of Atomic Scientists, March/April 2005; James M. Acton, "Nuclear Power, Disarmament and Technological Restraint", Survival, August/September, 2009.

[7] Electric Power Annual, EIA, 2007.

[8] Gordon E. Moore, Cramming more components onto integrated circuits, Electronics, Volume 38, Number 8, April 19, 1965.

[9] These numbers are all for 2007. That year, according to the Census Bureau, there were about 113.6 million households in the United States. Note these savings are relative to average enrichment costs in the US today, which includes the very-high cost gaseous diffusion plants. Were these plants – at $160/SWU – simply replaced by modern centrifuge plants – at a cost of $80/SWU – US households would also save. In the subsequent calculations we compare savings between laser enrichment and a modern centrifuge plant.

[10] MIT, The Future of Nuclear Power, (2003), Appendix 5.

[11] "Stop Laser Enrichment", F. Slakey & L. Cohen, Nature, vol 464, no. 7285, pp. 32-33 (4 March 2010).

[12] http://www.docstoc.com/docs/7265175/2005-RAS-COMMISSIONERS

[13] Laser Nuclear Technology Might Pose Security Risk, by Richard Harris, April 12, 2010.

[14] "Nuclear Power and Proliferation Resistance: Securing Benefits, Limiting Risk", a report by the APS Panel on Public Affairs, 2005

Francis Slakey is the Upjohn Lecturer on Physics and Public Policy at Georgetown University and the APS Associate Director of Public Affairs. Linda R. Cohen is a Professor of Economics and Law and Associate Dean of Research at the University of California, Irvine.

---

These contributions haven not been peer-refereed. They represent solely the view(s) of the author(s) and not necessarily the views of APS.