Outline for An Acceptable Nuclear Future

by ALVIN M. WEINBERG

Nuclear energy is in trouble. Despite the reassur- account that I believe nuclear energy will never be ances offered by its proponents, a substantial, and accepted to the unquestioning degree that electricity has possibly growing fraction of the public is uneasy about been accepted. the course we are following. 'To be sure, the nuclear The third possibility is that we car] devise fixes- moratorium bills in the United States have been technological and institutional-that will make nuclear defeated at the polls by a vote of 2 to I. Nuclear pro- energy acceptable. Can we draft a peace treaty between ponents point to the two-thirds who favor nuclear those who oppose nuclear energy and those who sup- development and congider this a mandate to go ahead. port nuclear energy'? Can we outline an acceptable But I think it is fair to say that a primary energy system nuclear future? that is feared or rejected by 33 percent of the public is Acceptabil~tyand need are conjugates. What is not go~ngto survive. acceptable, and how much risk we are willing to take, Three possibilities emerge. First, that nuclear energy depends on how badly we need, or think we need, will gradually disappear and, except for the bomb, the nuclear energy. This perception of need depends both world will revert to the situation that existed before upon our projection\ of energy demand and upon the December 2, 1942, when the first man-made chain energy situat~onat a given time. To he sure, the rate at reaction was established. One must remember that which nuclear energy was introduced in the United fission itself is rather a fluke. The conditions for estab- States was determined pretty much by competitive lishing a self-sustaining chain reaction might not have market forces rather than by perception of energy been met had the number of neutrons per fission been demand. (And it is notable that in the Soviet Union less than I, or if man had evolved after all of the 235U nuclear power has been introduced at less than one- had disappeared. 'There was nothing pre-ordained about fifth the rate it has been introduced here.) But once fission energy or the discovery of fission in 1938. Had nuclear power has been introduced, once a $75 billion this discovery been dclayed by 50 years, we would industry is in being, the need for continuing the enter- somehow be doing without it. prise is overwhelming. It would be disastrous, at least in The second possibility is that our fears about fission the short run, simply to shut down the nuclear enter- energy may simply subside. As nian acquires more prise, what with oil embargoes and other energy experience with reactors, and more particularly, as the shortages. public acquires familiarity with radiation. it is quite What is at issue, then, is not so much what ik to hr possiblc that the opposition to fission will wane. The done with the nuclear system already in place; it is analogy with the introduction of electricity is close. what is to be donc in the future. If this future is per- There arc still some elderly people around who were ceived to be a high-energy future, then the need for once uncomfortable about electricity; yet these fears nuclear energy and thc acceptable risk are corresponcl- have largely disappeared as electricity has becotne part ingly higher; if it is a low-energy future, the need, and of everyday living. therefore the acceptable risk, is lower. An essential element is missing in this analogy: the bomb. Even though we resolve our concerns about SEMANTIC CONFUSIONS AND CONSENSUrlI- CLIMATES personal safety, or ever1 genetic hazard, the bomb and Let us first dispose of a semantic confusion-thc proliferation make nuclear energy special. It is on this two meanings of the word "acceptable." As the nuclear enterprise now stands, it is acceptable neither to those ent levels. First is growth. Those who are opposed to who oppose nuclear energy nor to those who favor it. growth as a matter of principle are opposed to nuclear It is unacceptable to the former because they consider energy since nuclear energy, insofar as it is unlimited, it somehow too dangerous. It is unacceptable to the gives the technological base for unlimited growth. nuclear industry because they are frustrated by adminis- Related to this concern is centralization and bureau- trative delay and uncertainty, by serious cost overruns, cratization; growth can be managed only by centraliza- and by continual bickering. tion, and centralization is bad. Since nuclear energy is But most of the frustration of the proponents simply the epitome of centralized technology, it evokes fears reflects the basic unacceptability of nuclear energy to among those who long for a decentralized and, one the opponents; many of these frustrations would dis- hopes, a more resilient society. appear if nuclear energy were acceptable to the op- The second concern has to do with proliferation; ponents. The regulatory process is lengthy and court- indeed, this at the moment seems to be the main objec- ridden not because the regulators are poorly organized tion to nuclear energy. There are many who insist that or doing a bad job, but rather because the underlying no nuclear energy system can be devised to be prolifera- technology that is being regulated has not received an tion-proof, and that this alone warrants a rejection of adequate consensus. In the absence of such consensus, nuclear energy. Proliferation is of course not indis- the regulatory process has become subverted; it be- solubly connected with nuclear power. The best one comes an instrument for bringing to focus the profound can hope for is a way of delaying, not stopping, differences in perception between pros and antis, and proliferation. in the process it frustrates nuclear energy. Indeed, the Finally, there are the concerns over the intrinsic process becomes a background for what has been safety of nuclear energy-waste disposal, reactor acci- described as a religious war. Thus, although some of the dents, routine emissions, possible accidents during fixes that I propose will be aimed at getting nuclear transport, toxicity of plutonium, vulnerability of the energy off dead center, I consider these secondary. Unless we can arrive at a system that commands a con- sensus, any fixes that satisfy only the pros will be sub- merged in the overall opposition. A Nuclear Glossary It can be argued that we set an impossible task. z33u -A fissile isotope of uranium How do we know when an adequate consensus has produced from 232Th been achieved? Who, after all, speaks on whose behalf? 235U -A naturally occurring fissile isotope Which antis are to be placated, which pros are to of uranium negotiate? In a democratic representati-~esociety, are 238U -The abundant naturally occurring isotope of uranium we not constrained to use the duly constituted instru- "'Th - Naturally occurring thorium ments of authority-our elected representatives, our 239P~ - Artificially produced plutonium; regulatory bodies, our judicial system? I doubt that this isotope can be used as a anyone has a clear conception as to how to fully legiti- nuclear fuel just as can *"U and mize dissent when it is deep-rooted and widespread. 235u. 240P~ - An artificially produced isotope of Nevertheless, our system has been resilient enough to plutonium establish what I might call "consensual climates" even Breeder -A reactor that produces more on issues that at one time were bitterly divisive. Civil nuclear fuel than it burns rights was deeply divisive, yet it was finally largely Burner -A reactor that burns more nuclear resolved. We have achieved a consensual climate with fuel than it produces EBR-II -A 60-megawatt experimental respect to this issue. It is this consensual climate with breeder reactor respect to nuclear energy that we seek to establish with- MW(e) -The unit of electrical power of a out giving up nuclear energy. nuclear reactor as opposed to thermal power CRITERIA TO BE MET GWe -A billion watts of electric power Quad What really bothers opponents of nuclear energy? - 1 quadrillion British thermal units The opposition is concerned with issues at three differ- An Acceptable Nuclear Future

nuclear system to sabotage and diversion. Most of my always plagued development of nuclear energy. Some proposals will be aimed at remedying the present of the early workers, notably Walter Zinn and Eugene system's deficiencies in this general area. Wigner, hoped to avoid this dilemma by skipping The first concern, growth, we cannot remove by Phase I altogether. It was on this account that most of devising an acceptable nuclear future. Indeed, if one is the original civilian reactor development at both convinced that growth is intrinsically bad, then one has Argonne National Laboratory and Oak Ridge National relatively little incentive to devise such a future, since Laboratory was centered on the breeder. Others, no- any nuclear future makes growth, or at least a shift to tably Bennett Lewis of Canada, disagreed: Nuclear electricity, more feasible. If this belief comes to domi- energy based on "TJ, possibly enhanced with the intro- nate, and we adopt an extremely low-energy, non- duction of advanced converters, was a sufficient goal. electric style of living, then it is doubtful that nuclear "Breeders are not necessary," thundered Bennett Lewis energy can survive in any case. But the remaining two in the 1950s; and even Eugene Wigner, who considered concerns-proliferation and safety in the broad sense the enterprise valid only if the breeder were developed, -I believe can be ameliorated without rejecting conceded that the breeder might develop out of the nuclear energy. technology of burners, rather than developing entirely The nuclear energy system comprises mining, de nouveau. enrichment, fuel fabrication, reactor construction and Phase I is self-limiting. We can estimate its magni- operation, reprocessin~and waste disposal. It is com- tude if we can estimate the reserve of uranium cheap plex and intricate. The larger the system, the greater enough to be used in a burner. We may take for this the are the chances for system breakdown, since there are official figure of 3.5 million tons in the United States, more points that are vulnerable. All of the cpncerns though we recognize that the acceptable cost, both increase as the amount of nuclear energy increases. If economic and environmental, of uranium usable in a the nuclear enterprise were small, and served merely as burner may increase if we are badly strapped for a non- a short-term transition to other, more benign forms of fossil energy source. The uranium required to inventory energy, the concerns would be small and limited. The and fuel a 1000-megawatt Light Water Reactor issues become stark and urgent only when the nuclear (LWR) for 30 years without recycle is about 6,000 system becomes very large, and is regarded as the tons, with full recycle about 4,000 tons. Thus our energy mode that will continue far in the future. assumed uranium reserve will support between 600 and Thus if we are to design an acceptable nuclear 900 large LWRs for their lifetime. The Institute for system, we must first agree on criteria for acceptability, Energy Analysis's most recent "low" estimate of not merely when the system is small but when the nuclear electricity in the year 2000 corresponds to system is large, and the full systems problems emerge. about 300 LWRs, growing to, say, 400 LWRs by 201 0. In theory we must decide, for example, what calculated Phase I then might run its course, say, 50 years later reactor accident probabilities, or how much flow of -by 2060. plutonium, or how big an inventory of wastes in a fully Phase 11, based on breeders, could last irnrneawr- deployed system are acceptable. We cannot, of course, ably longer, since the reserve of low-grade ?"U or state precisely the acceptable thresholds for these ':Vh usable in the breeder is so great. Let us consider, probabilities. We can, however, estimate these prob- somewhat arbitrarily, an ultimate Phase 11 comprising abilities and their implications in a fully deployed 1000 large breeders each operating at ahout 2000 system, and then see what can be done to reduce these MW(e). This system corresponds to 120 quads (q) probabilities. being produced by nuclear energy. An alternative ulti- mate system might be half as large-500 breeders THE TWO PHASES OF NUCLEAR ENERGY corresponding to 60 q. Let us now estimate the risks Nuclear energy will develop in two phases, Phase I implied in systems of this magnitude. and Phase 11. Phase I is based on reactors that burn We do not have figures for probability of meltdown ""U, Phase I1 on breeders that essentially burn 23sU for Liquid Metal Fast Breeder Reactors (LMFBRs)- or ':"Th. We cannot say with assurance when Phase I1 assuming these are what we deploy--comparable to the will displace Phase I, since we do not know how much estimates for LWRs. The latter probability as estimated uranium we have. This is the strategic dilemma that has in the well-known Rasmussen study is .00005 per reactor pcr year, of which only one-fourth grossly nautical engineer estimated that if as many airlines breach aboveground containment. If the LMFBR flew as fly today. we could expect a crash every two accident probabilities are the same as those estimated days! for LWRs, then the expected accident rate for the large 'I'he system will inevitably be self-limiting: It will system is .05 per year; for the small system, .U25 per expand only to a size with which the society is comfort- year-it., we can expect one accident every 20 or 40 able, and this size bill depend primariIy upon the state year$, depending on thc size of the system. of the technology. But no matter the state of the tech- Before asking whether this IS good enough, we nology, we shall have to exercise non-technical inge- must recognire that by the time the ultimate Unitcd nuity to reduce the risks, even though they cannot be States system har been reached, the rest of the world quantified. In particular, what we seek are mechanisms will have also deployed many breeders. Scenarios have that will: been developed at the Institute for Energy Analysis Minimize the likelihood of physical disaster and thc International Institute for Applied Systems Analysis that contemplatc 10,000 large breeders. If the Minimize the consequences of disaster Rasmcssen probabilities given above are taken serious- Ensure institutional responsibility for as long ly, then one estimates an accident, on the average, as the nuclear system requires care every two years. I would suggest the following measures will be re- This calculation illustrates the dilemma. From the quired if Phase 11-which may be very large and last point of view of an operator or utility operating a for a very long tirne-is to be acceptable. single reactor, 1 chance in 20,000 per year is accept- able. On the other hand, from the point of view of the I'hyrical irolation. It seems evident that only a system as a whole, it woulcl seem the probabilities relatively small fraction of our planet ought ever to be ( 1 every 2 years ) are too high. Just as all DC-1 0s are in contact with high-level radioactivity-and the grounded if too many DC-1 0s anywherc in thc world smaller the better. This leads to the idea of committed misbehave, so one would imagine that breeders simply sites surrounded by sparsely populated areas-in short, would not survive if, on the average. one of them melted to a strongly collocated system. One thousand reactors every two years-and this is independent of whether might be accommodated in 5.000 square miles within the meltdowns occurred in thc U.S. or elsewhere. the United States-say, 100 sites each containing 10 The same problems of system vulnerability, as reactors and supporting chemical facilities occupying opposed to individual vulnerability, apply to the other, 40 square miles apiece, and 1,000 square miles for less quantifiable risks. The U.S. system would contain waste disposal. Some of these sites would be in the some 2,500 to 5,000 tons of plutonium (Pu) ; the world oceans; many of them would represent expansions of system possibly 10 times as much. Whatever the risk current nuclear plant sites. of diversion or of contamination when the world inven- The appropriate degrce of collocation is negotiable. tory is, say 100 tons of Pu, these risks are certainly If breeders and their chemical plants are collocated much larger when the inventory is 250 to 500 times (a return to the original concept of the breeder as a larger. Or consider the matter of retired reactors. In a closely coupled reactor and a chemical plant), one world of 10,000 reactors, some 300 would be retired minimizes transport of plutonium; on the other hand, every year, 300 new ones completed to replace them. the optimal size of the chemical plant may not match Is this really credible? the output of the cluster of reactors, and in any case, as When one looks at the matter from this point of Phase I1 gets under way, transport of fuel between the view, one has almost reduced the full-scale deployment, reactors and off-site chemical plants is inevitable. worldwide, of breeders to an absurdity. Yet we cannot A firm national commitment to the principle that say any of this with certainty. SureIy the gradual evolu- the nuclear enterprise is to be confined to as few sites as tion of the technology will reduce the a priori risk possible seems to me the very minimum. A more far- probabilities. The legislated meltdown probability for reaching policy-that all breeders and their supporting LMFBRs of .000001 per reactor year, if achieved, facilities shall be collocated-seems to be indicated, would relieve much of our concern about large acci- though this stronger policy is more open to argument. dents. Some 40 years ago a distinguished Swedish aero- conrinued on page 26

INGiNLLRlNC AND SCIENCE An Acceptable Nuclear Future . . . conrinuedfrornpage7

Strengthened security. In a world Idaho Falls, Savannah River. Each of plants in New England, and govern- beset by terrorism, we are confronted these places, somewhat isolated, some- mental generating entities such as the wlth unsatisfactory alternatives- what self-contained, tends to create a Power Authority of the State of New terrorism with few constraints on cadre, a sense of professionalism and York or the Tennessee Valley Authori- individual rights, or order with consid- of commitment, that is possibly less ty are also examples. erable restraints. Yet I believe we can easy to create in smaller sites that give The nine exlsting Reliability Coun- have order without losing freedom on less scope for a cadre of critical size cils in the U.S. conceivably could serve a large scale. We can cope with terror- to develop. as nuclei for the creation of such gen- ism, sabotage, and diversion by erating consortia. However, our pur- strengthening security at vulnerable Establishment of nuclear generation pose here is not to restructure the spots, without having to impose repres- consortia. When nuclear reactors electrical utility industry in order to sive measures upon the entire society. were first introduced, they were viewed rationalize the distribution of electrici- Thus, physically isolated nuclear sites simply as replacements for convention- ty; it is to create entities equal to the lend themselves to being made secure al boilers. This was a mistaka. The responsibility imposed by the genera- by virtue of their isolation. But this nuclear system is a far more complex tion of nuclear energy. may not be enough. We will probably and demanding enterprise than is a Not the least of these responsibilities have to reimpose on our nuclear plants coal-fired boiler. The responsibility is the handling of a serious accident, the same kind of security we imposed inherent in and the complexity of a even one that does little or no harm to on them during the war. In effect, we nuclear system, particularly a breeder, the public. One can never simply will have to buy order at the expense go much beyond that involved in a abandon a nuclear plant, least of all of freedom, as has been the case fossil-fuel system. one that has suffered a meltdown. Yet throughout man's history, but we con- The breeder system, with its intrin- what assurance do we have now that fine these encroachments on freedom sically closer coupling between the utilities operating reactors are to 5,000 square miles and perhaps chemical reprocessing and reactor robust enough to withstand the stress 200,000 people-the cadre that operation, places demands on the utility imposed by an accident; and if a utility operates the nuclear plants. industry that go even further beyond goes bankrupt, is the nuclear plant that industry's tradition. Indeed, the assured the resources needed to keep Professionalization of the nuclerrr split between chemical reprocessing it from causing damage? Thus, along cadre. In the final analysis, the safety and reactor operation inherent in the with the restructuring I speak of, I and integrity of the nuclear system will ceramic-fueled LMFBR was to a degree would contemplate some means of depend on the caliber of the staff that encouraged by the utility industries' preventing responsibility from lapsing, mans the system. If one concedes that disinclination to go into the chemical in the event an accident causes bank- the nuclear system is special, because a reprocessing business. But it is by no ruptcy. malfunction could cause very great means clear that ceramic-fueled harm, and of a kind that our society LMFBRs will always dominate the Longevity of institutions responsible has not become accustomed to (and nuclear enterprise. As the technology for nucleclr energy. The nuclear may never become fully comfortable is rationalized, other breeder systems system, once started, can hardly be with), then it follows that a higher in which chemical reprocessing and abandoned. Who, for example, is degree of professionalism and dedica- reactor operation are more closely responsible for cleaning up once a tion is required to manage and operate coupled could well come into being. nuclear fuel chemical plant has gone the nuclear system than is required for It would seem therefore that in the out of business? This residual responsi- the conventional power system. The ultimate system the gcnerat~onof bility is not i~niq~leto n~~clenr energy. responsibility borne by the superin- nuclear electricity, as contrasted with Abandoned strip mines continue to tendent of a nuclear power plant is at its distribution, ought to be placed in cause acid drainage. and no one can be least as great as that borne by the pilot the hands of utilities that are specifi- held responsible. The difference per- of a 747. This must be recognized-by cally geared for this job. A siting policy haps is that we know in advance that a the superintendent, by his management, such as we have enunciated would nuclear plant cannot simply be aban- and by the public. lend itself to just such separation. doned; the same knowledge was not so I believe this sense of professionalism Generation would be the responsibility apparent in the case of acid mine is enhanced by the siting scheme that I of the company or consortium or gov- drainage. propose. I draw this inference from ernment entity established specifically We have already recognized that our experience at the existing large to carry out this task. Consortia of this waste disposal, which invnlves some nuclear sites-Hanford, Oak Ridge, sort already operate the nuclear power surveillance over a long time, must be a governmental responsibility-as lish the sites for nuclear energy now, LMFBR were not deployed ~intilthe much as anything because of all our and adopt the principle that only sites year 2000, and the total uranium re- institutions, it is the government which so committed, most of which will be sources were only 1.8 million tons. This is longest-lived. Probably other ele- occupied first by LWRs, will be used relatively leisurely introduction of ments of the nuclear systcm will a130 for the breeder~;and that we shall keep breeders would limit our total nuclear demand such longevity and possibly the number of such sites to a minimum? capacity of 400 GWe until around stronger governmental involvement. We may achieve this policy de facto 2020. By then the system could resume At the very least, in the ultimate simply because it is becoming so hard growth as old breeders fed fissile system we must be assured that what- to license new sites. C. Burwell esti- material into new ones. In the interven- ever entity is responsible for the opera- mates that 80 of the 100 existing sites ing years one would presrimably de- tion of the reactors is likely to remain could be expanded into large centers. pend on additional lower grade ore, viable as long as the reactors contain Would ~tnot be prudent to confine should demand exceed 400 GWe. appreciable amounts of radioactivity. the future nuclear system, including Although most of us in the nuclear the rest of Phase I, essentially to the enterpr~sehave always believed that THE ACCEPTABILITY OF PHASE I AND existing sites, and to adopt this as our the breeder is the essence of nuclear THE TRANSITION TO PHASE I1 national policy? Reprocessing and fab- energy, we could never mount a com- The requirements we lay down for rication facil~tiesshould be collocated pletely compelling argument for intro- an acceptable Phase I1 are stringent: -all seem to agree on this. I would duc~ngit very quickly, except the one physical isolation and collocation, go further and urge that chemical com- based on an extremely rapid and most strengthened security, professionaliza- plexes be collocated wlth ex~sting unlikely rise in energy demand, or a tion of the cadre, establishment of nuclear sites. Though I believe very milch smaller uranium ore reserve generating consortia, and longevity of eventually all s~~chreprocessing com- than we now consider likely. In the the operating entities. These require- plexes ought to be collocated with absence of such a strong demand or ments seem to me necessary if the ulti- breeders, I am prepared to leave this lower reserve our argument for quick mate nuclear energy system is as large question open for the time being. deployment fell back on economics- as 500 or 1,000 breeders in the United Simply do not put new complexes any- we believed breeders would be cheaper States, and ten times that many in the where except in places where nuclear than burners, in which case the market world; and if the system is to be with energy centers already exist or are would force rapid deployment of us for an indefinitely long time. planned. breeders. How much of this is req~~iredto The key question with respect to Eventually this will be the case; but make Phase I acceptable? I would Phase I1 is the rate at which it develops, it is not at all clear when, since neither argue, not very much, because Phase I and the speed with which breeders are the future price of uranium nor the is limited in size and duration. Again, introduced. The original plan for intro- capital cost of the breeder is known. assuming the 3.5 million tons for the duction of the breeder as outlined in To be sure, if breeders instead of raw material in the United States, we the 1962 White Paper of the Atomic burners were now deployed, most of compute one expected meltdown in all Energy Commission (AEC) was predi- the questions concerning separative of Phase I. And it is reasonable to cated upon a much faster growth of work capacity and uranium ore would expect that the incremental improve- electricity than we now consider plau- disappear-the nuclear system would ments in the technology of safety will sible-an electrical demand of 70 q by have fewer links, and therefore would reduce this expected number. The like- 2000 and doubling every 15to 25 be less subject to total failure as a lihood is that Phase I will pass without years thereafter. No wonder the U.S. result of failure of a single link. any serious meltdown. The same argu- drive toward fast deployment of the Though this cannot be looked upon as ment holds with respect to the other breeder seemed so obviously sound. a completely compelling argument for issues. Because Phase I is limited, in- But more recent estimates of our rapid deployment of the breeder, it cremental improvements in security, demand for electricity are much more reinforces this fundamental fact: in professionalization of the cadre, modest-for example, at the Institute Nuclear energy based on breeders is and in ens~iringfinancial responsibility for Energy Analysis we now project much less beset by uncertainties related ought to be sufficient. demand for nuclear electricity of 300 to our energy demand, our ore reserve, On the other hand, it seems to me to 400 GWe (18 to 24 q) by 2000. And or our separative work capacity than is that Phase I ought to develop along A. M. Perry and M. 3. Ohanian at IEA nuclear energy based on burners. paths that smooth the transition to an have estimated that a total nuclear This, in final analysis, is the strongest acceptable Phase 11. Of major import demand of 400 GWe could be achieved argument for fast rather than slow is the siting policy. Ought we not estab- by 2000 even if the first commercial deployment of breeders.

~NGfNEkRlNGAND SCIENCE An Acceptable Nuclear Future . . . continued

PROLIFERATION-THE BOGEYMAN But it is clear that institutional examination of reactor systems with a President Carter's new look at arrangements must be part of any pro- view to hardening them against nuclear energy, particularly his de- llferation-resistant system, and this prol~feration. ferral of recycling in LWRs and his certainly was anticipated in the It must be noted that these proposals deferral of LMFBRs, has been largely Acheson-Lilienthal ideas for an Inter- are aimed more at improving the motivated by his concerns over pro- national Atomic Energy Authority. safety, rather than the proliferation- liferation. It is curious that technolo- I would argue that the policy of col- resistance, of the nuclear system. gists tend to look upon proliferation as located nuclear energy centers espoused Nevertheless collocation should en- a political problem amenable only to here could itself strengthen resistance hance the proliferat~on-resistanceas political solutions, whereas politicians to proliferation even without a full- well as the safety of the nuclear regard the problem as amenable to fledged international author~ty.The system. Resident inspection is more technological fixes. Hence the Presi- point is that such a siting policy, which feasible if collocated energy center dent's call for development of prolifer- confines nuclear energy to relatively sltes are adopted. A natlon bent on ation-resistant reactors. History has few large sltes with a minimum of milking its nuclear power plants of come full circle; 3 1 years ago Dean transport of fissile material, is in the weapons materials would almost surely Acheson and David Lilienthal proposed first place easier to inspect by the have first to expel its resident inspec- a plan for control of nuclear energy instruments of the present International tors. This would be tantamount to that depended heavily on a techno- Atomic Energy Agency; and it lends publicly and explicitly going nuclear, logical fix, the division of nuclear itself to resident inspectors who could in much the same sense that Egypt's activities into dangerous and non- insinuate themselves into the local expulsion of UN observers from the dangerous ones. The technological key situations and detect unauthorized Sinai signaled its intention to go to war. to this distinction was the possibility activities much more readily than could Moreover, a collocated system-for of extracting power from denatured fis- non-resident inspectors. Indeed, I example, a closely coupled breeder and sile materials, is., isotopic mixtures of should think that once the principle of chemical plant (like EBR-I1 or the ZaW, 23W, and 'a'.U, or (as was then resident inspection is adopted, the Molten Salt Reactor)-would inher- somewhat mistakenly believed) "Vu IAEA would have taken an important ently be more proliferation-resistant X1"Pu.Thus nuclear activities could be step toward becoming the sort of inter- and diversion-proof than would a dis- separated into dangerous and non- national entity conceived in the persed system since the fuel need never dangerous ones, the former being Acheson-Lilienthal plan. be fully decontaminated. supervised by an international author- Let me put the pieces of what I Is it likely that suggestions such as ity, the latter being conducted under think is an acceptable nuclear future these will quiet the concerns over national auspices. Though a special together. My basic point is that if we nuclear energy enough to establish a committee set up by General Groves can devise an acceptable long-range "consensual climate" in which the regu- warned that denaturing could not be system-that is, Phase 11-then much latory process can work during Phase foolproof, the Acheson-Lilienthal plan of the opposition to the much smaller I? At the moment I would judge that nevertheless relied on such approaches. and limited Phase I ought to gradually these proposals do not go far enough Today we again seem to be casting subside, without either a dismantling for the antis, and go too far for the about for a technological solution to of Phase I or a rejection of Phase IT. pros. The situation is ripe for imagina- proliferation based on denaturing. Now The basic elements of Phase I1 are tive, constructive thinking. I look on it is certainly true that the world physical isolation (and therefore col- these proposals as tentative first steps- community of reactor developers has located energy centers and resident indeed, as an invitation for those who never considered building in resistance IAEA inspection) ; heavier security; are interested to suggest means of to proliferation as a design constraint. professionalization of the cadre; remedying the ills of the nuclear busi- I suspect that schemes such as Molten longevity of the operating entities; and ness. Consensual climates in a democ- Salt Breeder Reactors (MSBRs) restructuring of the nuclear energy racy are not easy to forge, especially fueled with "W/Z3'3NU mixtures are system, with the establishment of when the issues are bitter and impor- more proliferation-resistant than are consortia for the generation of nuclear tant. Yet, unless we establish such a reactors that use pure ':qPu; and I power. In preparation for the transition climate, we run a danger of losing the believe it is useful for the reactor com- to Phase 11, I would urge that any new nuclear option. I believe the butden munity to examine reactor systems that reactors, whether breeders or LWRs, we would thereby impose on our incorporate technical barriers to pro- be confined essentially to existing sites. descendants is much heavier than the liferation, or at least weaken the link I also urge further strengthening of the one they would have to bear in manag- between power and bombs. nuclear cadre during Phase I, and re- ing an acceptable nuclear future.