Brave New Nuclear World

Transcription

1 The Future of Nuclear Energy Conference, The Bulletin of the Atomic Scientists November 1-2, 2006, Chicago, Illinois, USA Brave New Nuclear World The Expansion of Nuclear Power and its Relevance for the Proliferation of Nuclear Weapons Alexander Glaser Program on Science and Global Security Princeton University November 2, 2006 Revision 7, web 1

2 We would not make atomic weapons, at least not to start with, but we would start out and build enormous plants, and we would call them power plants maybe they would produce power; and these plants we would design in such a way that they could be converted with the maximum ease and the minimum time delay to the production of atomic weapons, and we would say, this is just in case somebody twotimes us; and we would stock-pile uranium, we would keep as many of our developments [as] secret as possible, we would locate our plants, not where they would do the most good for the production of power, but where they would do the most good for protection against enemy attack. J. Robert Oppenheimer (on the prospects of a nuclear weapons convention) in the Bulletin of the Atomic Scientists Vol. 1, No. 12, June

3 Scale of a Hypothetical Global Expansion of Nuclear Power 3

4 Capacity Buildup ,000 Slower growth scenario 4,000 Target 2 Installed nuclear capacity [GWe] 3,000 2,000 1,000 Expansion begins Target Year 4

5 Capacity Buildup ,000 Faster growth scenario 4,000 Target 2 Installed nuclear capacity [GWe] 3,000 2,000 1,000 Expansion begins Target Year 5

6 Capacity Buildup Annual nuclear capacity added [GWe/yr] (includes new construction and replacement) Expansion begins A nuclear expansion would have to be extremely aggressive (both in scale and speed) in order to make a significant contribution to climate-change-mitigation efforts by 2050 Hypothetically, 4500 GWe by 2100 would be easier to achieve than 1500 GWe by 2050 because the infrastructure (to build reactors) would already be in place Year 6

7 Global Nuclear Expansion Scenario (1500 GWe in 58 countries, based on 2003 MIT study) More than 10 GWe installed At least 1 GWe installed 7

8 Enrichment Demand and Distribution (for 1500 GWe Global Nuclear Expansion Scenario) 17,650 37,200 20,850 36,600 27,750 12,750 3,150 2,850 5,100 2,850 2,250 24,450 5,400 4,800 10,300 11,050 tswu/yr Total SWU-production in country Combined SWU-demand of countries importing all their enrichment services: 11,850 tswu/yr Global enrichment capacity: 1,500 x 150 tswu/yr (225,000 tswu/yr) 8

9 The Example of the Gas Centrifuge for Uranium Enrichment 9

10 Centrifuges for Uranium Enrichment tails Source: IPFM 2006 Report feed product top scoop baffle Enriched uranium Depleted uranium center post casing rotor bottom scoop electromagnetic motor Source: Presentation by Mohammad Saeidi (AEOI) bottom bearing 10

11 Source: Urenco 11

12 Centrifuges in the Past Centrifuge technology for uranium enrichment has been around for more than 50 years Technology considered too complex to be a proliferation threat R&D classified since 1960 in those countries that were already exploring the technology Focus was on the back-end of the fuel cycle (discouraging reprocessing / separation of plutonium) Until recently, export controls were considered sufficient by those who held the technology Proliferation occurred when export controls were violated 12

13 Why Are Centrifuges Different? 13

14 Crude Breakout Scenario (using an early-generation machine) 60 Source: Urenco Number of machines in stage Total number of machines in cascade: Assumed characteristics of P-2-type machine peripheral velocity = rotor diameter = rotor height = separative power = m/s cm cm SWU/yr Feed = Product = 32.4 kg/d 3.3 kg/d Stage number Performance of reference LEU-cascade of UF6 w/ natural uranium of UF6 w/ 4.4%-enriched uranium 14

15 Crude Breakout Scenario (using an early-generation machine) % Enrichment [wt%] Serial batch operation: 10 cascades feed into 1 additional cascade Simple batch operation: 1 cascade using pre-enriched feed-stock 20 26% Time [hours] (compare to equilibrium time for gaseous diffusion process, which is on the order of months) 15

16 Crude Breakout Scenario (using an early-generation machine) Batch operation in series 11 cascades (10 cascades feeding into 1 additional cascade) about machines or SWU/yr Production of 25 kg HEU in less than 12 days (contained in 37 kg of UF6 and requiring 3900 kg of UF6 LEU feed-stock) PREPARATION FOR BREAKOUT Collection of LEU feed-stock over a period of less than 4 months (110 days) Note: up to 2500 kg of enriched UF6 can be stored in one standard product cylinder (Type 30B) 16

17 Clandestine Options (Undeclared centrifuge facilities are virtually impossible to detect) Detectability (Selected Criteria) Identifiable Structure Thermal Signature Effluents Plutonium Production Reactor Reprocessing Yes Yes No No No (Yes) Uranium Enrichment Calutron/EMIS Gaseous diffusion No Yes Yes Yes Yes Yes Centrifuge No No No 17

18 What Are Our Options? 18

19 Possible Strategies to Limit the Front-End Proliferation Risks of the Nuclear Fuel Cycle STRATEGY Increase the effectiveness of (and the confidence in) safeguards Increase the ability to detect undeclared facilities Contain technology to existing or selected producers Focus on the demand side (i.e. devalue nuclear weapons) TARGET/OBJECTIVE Preclude covert misuse Deter clandestine activities Know-how held by trusted users Motivation 19

20 Containment Strategies Have and have-not approaches Bush Proposal (2004) or other criteria-based proposals Black Box approaches with or without Poison Pills and combined with multinational operation of facilities Export Controls Deter, delay, detect procurement efforts 20

21 Containment Strategies (cont d) PROBLEMS Strategies do not effectively address the unique proliferation concerns of centrifuge technology (breakout and clandestine option) Since the idea of further restrictions on nuclear fuel cycle technologies has been revitalized in 2004 (e.g. supplier-/client-state arrangements), several countries have expressed renewed interest in domestic enrichment In addition to Iran and Brazil, these are Argentina, Australia, Canada, Kazakhstan, South Africa and the Ukraine Economic incentives to forego domestic enrichment (e.g. assurances of supply) are largely irrelevant Even if a country is willing to pay five times the market-price for enrichment services in order to have a domestic uranium enrichment capability, this would only raise the overall cost of electricity by about 10% To what extent are containment strategies durable anyway? Underlying assumption that indigenous R&D efforts are irrelevant/insufficient 21

22 Genealogy of the Gas Centrifuge U.K. The Netherlands Germany Urenco Russia USA France Iran (Iraq) China N. Korea (Libya) Pakistan Japan India Brazil Australia Original centrifuge R&D (pre-commercial, Zippe-connection ) Technology transfer (confirmed or planned) Independent development or unconfirmed foreign assistance Status or achievement unclear Last revision: 09/

23 Timeline of Centrifuge Programs Germany Netherlands U.K. Japan Australia Pakistan India Brazil Iran R&D as part of URENCO/ETC R&D as part of URENCO/ETC R&D as part of URENCO/ETC R&D Machine at least 2-5 SWU/yr Test cascade at least 100 machines Pilot plant and further developments DRAFT version, August by Alexander Glaser, Princeton University (arrows indicate uncertain dates of respective events or milestones) 23

24 Timeline of Centrifuge Programs (cont d) How long does it take to develop centrifuge technology? Length of required R&D-period has not significantly changed over the past decades (It takes about years to go through all phases of the R&D-process) Even important outside assistance does not shorten the R&D-period excessively (possibly up to 50%, e.g. the case of Pakistan) Will more countries be able to successfully develop centrifuge technology? Timeline suggests that countries may begin to pursue a centrifuge program sooner or later, depending on when they feel sufficiently confident to be able to carry out such an R&D project Key technologies that were previously used specifically for centrifuge-component manufacturing are expanding into additional sectors of modern industry and/or require less experience or expertise to be operated (Examples are rotor balancing and flowforming techniques) 24

25 Possible Strategies to Limit the Front-End Proliferation Risks of the Nuclear Fuel Cycle STRATEGY Increase the effectiveness of (and the confidence in) safeguards Increase the ability to detect undeclared facilities Contain technology to existing or selected producers Focus on the demand side (i.e. devalue nuclear weapons) TARGET/OBJECTIVE Preclude covert misuse Deter clandestine activities Know-how held by trusted users Motivation 25

26 Viability of a Nuclear Expansion Conclusion and Outlook Despite current efforts to set-up a system of assurances of supply, incentives to acquire national enrichment capabilities remain high If the incentives to acquire national enrichment capabilities, for either peaceful or military purposes, continue to exist, we can indeed expect successful independent development and deployment of centrifuge technology in more states The only effective approach to reduce these incentives is to increase the countries sense of security Progress in nuclear disarmament has to be a central element of such an agenda To consider a global expansion of nuclear energy before one reestablishes confidence in the future of the nuclear nonproliferation and disarmament regime is an imprudent proposition 26