It is a great pleasure to be here this morning and to share the platform with distinguished old friends and colleagues in recalling some of the early experiences and impact of the discovery of nuclear fission. When I was invited to participate in this session it was suggested I might speak on the effects on industry and regulatory agencies in Canada of the development of the CANDU line of reactors. When I set about composing what I would say, however, that scope seemed rather too narrow, too parochial, and perhaps even too esoteric for this occasion. Furthermore, having read and enjoyed Bertrand Goldschmidt's "Pionniers de l'Atome" and heard Alvin Weinberg ruminate in his usual interesting way on energy and its uses at the American National Academy of Engineering last year, and being very aware of the majestic sweep of views we could expect from Les Cook and of Geoff Hanna's broad knowledge of modern physics, the scope seemed rather too focused. Finally, the notion didn't accord with my idea of the fitness of things. It didn't seem appropriate, assembled here as we are, to celebrate the fiftieth anniversary of the greatest thing in the field of energy since cooked meat, to devote a lot of time to the contemplation of heliarc welding and the Atomic Energy Control Board. So I have chosen a rather broader subject - "The Development of Nuclear Power".
I was thinking of calling it "The Second Stage". As Bertrand Goldschmidt has told us, the potential to produce useful energy was recognized from the very outset. As we've heard, however, the Second World War intervened and the fission reaction was first applied in the manufacture of bombs. This was the First Stage, the stage of development of the use of nuclear fission which uncovered many of the fundamentals relative to the process and demonstrated the feasibility of building and operating nuclear fission reactions. To develop any use of fission it was necessary to have comprehensive knowledge of the fission products and their properties in order for a chain reaction to be sustainable.
Not long after the war was over people turned their attention back to the constructive use of the energy available through the fission process. For the past thirty-five or forty years we have been creating a nuclear power industry, its related infrastructure and the building of the first generation of nuclear power plants. This I am regarding as the Second Stage - the development and widespread application of the first, basically simple, power reactors.
These first generation plants employ a simple fuel cycle, that is, their fuel is made from fresh uranium and the used fuel goes to storage. There are, admittedly, fuel reprocessing plants, notably in Europe, that separate the unused uranium and plutonium from the waste products in the used fuel and return the recovered plutonium to the power reactors. This recycled fuel, however, contributes very little additional energy in these first generation reactors, does not materially affect their economics, and its use is not impelled by any worldwide scarcity of fresh uranium. There are other motives for recycling fuel in current power reactors, not the least of which is preparation for the future. The time will come when it will be economic, even to the point of necessity, to extract much more of the energy available in uranium than we are doing today.
One of the nice things about nuclear fission - and there are many - is that the energy available in the fuel is not lost through inefficient initial use. When coal was used at 2 or 3 percent efficiency in the early steam engines, all of the useful constituents went up in smoke or out in the ashes and were gone forever. Not so with uranium. The part we don't use in today's comparatively inefficient reactors remains available for use in more efficient reactors in the future. In my chronology, the employment of advanced fuel cycles is the Third Stage in the development of nuclear fission reactors. Although experimentation with such reactors and fuel cycles began in the very early days, and has led to the construction and operation of large prototype fast breeder power reactors in France and the USSR, widespread application - what I think of as the Third Stage - is decades away. Although the high efficiency reactors and associated fuel cycles have been significant, even important parts of many countries' nuclear programs, they are not part of my subject today. I am going to confine myself to what I called the Second Stage - the development and general application of the essentially simple fuel cycle power reactors that provide virtually all of our nuclear power today.
I am going to do this under the following topics:
After the war, goaded by then Captain Rickover, the US Navy, in conjunction with the USAEC, commenced development of nuclear power reactors for naval ship propulsion. For this application, economics was not a prime consideration: compactness, endurance, reliability were. The pressurized light water reactor was a very wise choice. By 1952 the first land-based prototype for submarine propulsion was nearing completion in Idaho. In the middle of that year a 75,000 hp unit for aircraft-carrier propulsion, called CVR, was authorized. In May 1953 this was canceled. Admiral Rickover, as he then was, was not the kind of person to be deterred from developing a reactor for a carrier by a little thing like the lack of a carrier. Congress voted the USAEC $7,000,000 for power reactor development and in the middle of the year (1953) the USAEC authorized the construction of the 60 MW(e) PWR, the project to be run by Admiral Rickover and the nuclear plant to be built by the former CVR contractor, Westinghouse. If I had to pick the single decision that had the most far-reaching effects on the course of the development to date of nuclear power, a leading candidate would be the decision by the US Congress, in 1953, not to proceed with a nuclear aircraft-carrier.
There was continuing interest in the heavy-water reactor because of its simple fuel and the independence this offered with respect to fuel supply. During the period when the American and Soviet light-water reactors were flooding the world, or at least that part of it that could use nuclear power, Canada sold CANDU stations to India and Pakistan.
By the late sixties the American domestic nuclear power program was burgeoning. The country's two major steam generating equipment suppliers had entered the field as suppliers of large components, notably the reactor vessels, to GE and Westinghouse, and as suppliers of PWR nuclear steam supply systems in their own right. With the great success of the light water reactors at home and abroad, the variety of other types that Uncle Sam had experimented with gradually passed out of the picture.
For someone unfamiliar with the times this might seem simply a matter of industrial momentum or expediency. Certainly there was everywhere an urgency to get on with the peaceful application of this new, wonderful resource; but there was real conviction in the choices.
For the British and the French the high temperature available from a gas-cooled reactor was key. I remember this coming up when I was participating in a small round table session in the Hochschule in Lausanne after the 1964 World Energy Conference Sectional Meeting. It was raised by Sir Christopher Hinton, former head of the UKAEA's engineering establishment at Risley, which designed the first production and power reactors, and in 1964 Chairman of the CEGB. He asked: "We thought high temperature was extremely important. Were we wrong?" I think it was Ken Davis, previously with the USAEC, then with Bechtel, and who was later to be Assistant Secretary of Energy in the first Reagan Administration, who said: "No. You weren't wrong, but there is a lot more than thermal efficiency that goes into the success of a nuclear power system."
In this country, the attractions of a reactor system that could be built largely with our own resources, did not require enrichment for its fuel cycle nor reprocessing to make it be economic (as was then thought to be necessary for light water reactors) were the main determinants in our choice. They weighed every bit as heavily as did the experience with the heavy-water reactor.
The simplest nuclear steam supply system is that employing a light-water reactor. The moderating properties of light water are such that there is no room in the reactor for a separate coolant. Fortunately light water itself is an excellent coolant. The result is a common moderator and coolant system using the heat transfer and transport medium we know the best - good old hot water. Drawbacks are the need for large pressure vessels and enriched fuel. In the United States, with manufacturing plants capable of producing all the components required for light-water reactors, including the large thick-walled vessels, and with uranium enrichment facilities in place, a consensus rapidly grew that the light-water reactor was the logical choice for power generation.
In 1986 the No. 2 reactor at Chernobyl, one of the line of boiling-water cooled, graphite reactors in the USSR, suffered its disastrous accident. The USSR has decided to phase out that line and to rely entirely on their PWR line for electric power production.
The heavy-water reactor of the type developed in Canada now remains the only active alternative to the light-water reactor. There are three reasons for this: it uses uranium oxide, water and zirconium alloys like the light-water reactors; it is the variant of water reactor which offers simple, economic fueling; and we did a good job of it.
Besides meeting its own domestic requirements Canada has provided units and related technology to India, Pakistan, Argentina, Korea, and Romania. It is fifth in the world, only slightly behind France and Germany, in power reactor exports. Romania is in the process of building five of these units. India has built four with her own resources and has six more under construction.
I was in India recently and visited some of their facilities. The first unit in their fourth nuclear power plant at Narora on the Ganges, a 235 MW unit of the CANDU type, was just starting up. There is another two-unit plant of this kind in an advanced stage of construction and three more beginning construction. I saw the excavation for the pair that will be located beside the Rajasthan Atomic Power Station which we helped build. Besides these 235 MW units, India is commencing to build several plants with 500 MW units of the same type, to their own design. It seems strange, in walking around their new units which look so familiar, to see important equipment, such as the primary heat transport pumps from West Germany, that has come from third countries. Altogether the country plans to have 10,000 MW installed, in 200 and 500 MW units, by the year 2000. The attraction for countries with these plants is the simple fuel cycle which they can manage themselves. For India there is the added attraction that the heavy-water power reactor promises to be the right vehicle for the thorium/U-233 fuel cycle in the future, because India has large thorium deposits but apparently limited uranium resources.
This arrangement was well suited to the circumstances in Canada. Canadian companies were only a fraction the size of their counterparts in the United States and Europe and the main suppliers of boilers and electrical equipment were subsidiaries of foreign companies, mainly, at that time, American and British. By putting the NSSS engineering with the utilities, it assured the ability to maintain a Canadian program. It also meant that the relatively small Canadian manufacturers were being asked to supply components, such as steam generators, reactor vessels, pumps, etc., against well-developed specifications rather than a whole nuclear steam supply system against a necessarily much vaguer specification. In time, Ontario Hydro has assumed more of the NSSS engineering for its own power plants. This is consistent with its original preference but creates a dilemma as to how best to serve other customers for the CANDU system.
In the United States, with the largest electrical power market in the world, a large number of electric utilities, and a strong belief in free enterprise, there are four large, competing NSSS suppliers and the nuclear steam supply system engineering is performed by the suppliers. In Japan, with the third largest electrical market in the world, several privately owned utilities and with the same belief in private enterprise, the situation is similar but with only two or three suppliers. In Germany, too, the electricity supply environment is similar to that in the United States but, with an even smaller market, there is now only one supplier. However with none of the utilities shopping for nuclear power plants right now, other possible suppliers in the wings, and pan-Europe of 1992, things will probably change. In short, in all these countries, NSSS engineering rests with the suppliers, and the buyers rely on commercial competition to keep things in hand.
Britain and France have, as yet, smaller domestic electricity markets and these are served by publicly-owned national utilities. (This is being changed in Britain but nuclear electricity will remain a national undertaking.) Like Germany, they each now have a single NSSS supplier. The system engineering rests with these suppliers. However, the respective governments have taken major positions in these companies so that control of system engineering cannot be said to rest in the suppliers to the extent that it does in the first three countries but is divided, through interlocking ownership, between buyer and supplier.
Canada has the fourth largest domestic electricity market, between those of Japan and the Federal Republic of Germany, and is the world's largest exporter of electricity, exporting about 10% as much as it uses at home. Most of this is supplied by publicly- owned provincial utilities. Because of hydro and coal resources in other parts of the country the market for nuclear electricity is almost entirely in Ontario at the present time. Unlike the other countries, Canada has no NSSS supplier for its domestic market. (AECL acts in this capacity for export projects.) Instead, the NSSS engineering is performed by an engineering organization (AECL's CANDU Operations) engaged by the buyer, the utility. Control of system engineering rests with the buyer.
So we have the whole spectrum of system engineering control in the organizational arrangements in the various countries, from equipment-supplier control, through a sort of hybrid, to buyer-control.
Clearly, in this paper I am not going to deal with this whole ball of wax. It is not a simple matter to talk about any one aspect, say plant design, without invoking many others for they are all inextricably bound together, except at the very technical level. I will therefore simply say a few things based on experience and personal observations to give some of the flavour of the task of developing nuclear power.
There was another key characteristic of the engineering organizations. The initial professional staff was an amalgam of scientists and specialist engineers, such as reactor, process and control engineers, drawn from the laboratories and other establishments already engaged in the development of nuclear energy, and of equipment and power plant design engineers drawn from the owner's organization or assembled by him. From that point on the organization had a life of its own and acquired whatever personnel it needed for completeness and expansion. Although the number of scientists and specialists was small in relation to the number of engineers in the more common disciplines, it was a new experience for most of the engineers to be working in an environment where the application of fresh scientific knowledge was an integral part of the design process. For the scientists, too, it was a new experience working in an engineering milieu although most of them who came to the Canadian organizations at least tended to be those with an interest in being closer to practical applications. One of the incidental consequences of the introduction of scientists into an engineering office was a heightened skepticism of the basis for designs of products available on the market.
The initial nuclear power plant operating staffs were built the same way, with a mixture of those who had research or production reactor operating experience and those who had power plant operating experience.
For instance, the Americans, who had started with the submarine program and were a few years ahead of us in their nuclear power program, were using stainless steel in the primary heat transport and other reactor water systems. Purity of water is essential to avoid fouling of fuel, other heat transfer surfaces and to prevent the buildup and circulation of corrosion products which would become radioactive and hamper maintenance. Our metallurgist at CGE on the NPD project considered that this did not prevent the use of carbon steel in these circuits and that the risk of stress corrosion cracking with stainless steel was of greater concern. He discussed his opinion with his peers at Chalk River who agreed with him. None of us to whom they reported had any reason not to accept the recommendation and carbon steel was adopted. It was successful and has been used in CANDU systems and their foreign variants ever since. BWR's did run into some stress corrosion problems in particular locations in their stainless circuits but these have long since been resolved. Today the choice of material for these circuits can be almost automatic but, in the beginning, it was a major decision which, if wrong, could, as we were all very aware, jeopardize the whole project and maybe the program.
Let me give you an example concerning a construction method. For strength, corrosion resistance and, above all, neutron economy the pressure tubes in CANDU reactors are madeof zirconium alloys. There are special fittings at each end, made of a high alloy steel, which connect to the external circuits and through which the fuel enters and leaves the reactor. The requirement was to effect a satisfactory joint between the zirconium alloy tube and the steel end fitting. To give some idea of the possible scope, a few years later, for the EL-4 gas-cooled heavy water reactor in France, a compound joint consisting of a screwed joint, a brazed joint, and an electron beam weldment was adopted for a similar purpose. The higher gas temperatures were a factor but it serves to illustrate the scope available at that time. For NPD, the head of reactor design proposed using a rolled joint. Rolled joints, in which the ends of tubes are expanded into other pieces with a mandril and rollers, had been used for a hundred years or more to connect boiler tubes to boiler drums and the tubes of other heat exchangers to their tube plates, although by the 1950's the method was being superseded by welding and other processes. When it was proposed to me it seemed a bit old-fashioned but it was simple, it seemed that it might work, and there were no competing suggestions with any more promise. It was adopted. Of course it was necessary to mount a considerable development program to prove it and there were lessons concerning the detail that had to be learnt down the years in operation. A few years after we had adopted the rolled joint, when we were engaged in a mutual development program with the Americans, Dale Babcock of DuPont who had a responsible role in the engineering of the first production reactors at Hanford asked me how we had come to choose it. I said: "lack of imagination" and his comment was that in engineering, choice of the simplest was very often the right choice.
For nuclear power it was felt that more of the first and less of the second was required to obtain the necessary quality. A system was instituted which entailed inspecting and approving the supplier's quality control organization, staff qualifications, material control, inspection procedures and facilities. This was followed by periodic audits of all of these and of their performance, plus periodic customer inspections of the work in process and of course a final inspection. In Canada we called this system quality surveillance. In time this term was superseded by the term quality assurance which the Americans applied to a similar system and which has now been universally adopted.
In-core components were a whole new ball game since neutron capture characteristics and behaviour in strong neutron, gamma and other radiation fields were of paramount importance. The fuel required a whole new science and laboratories, like those at Chalk River, devoting a major part of their programs to understanding its behavior. For the manufacturers it was a brand new business whose character was entirely different from that of the other product-lines for nuclear power. Being a consumable it required mass production, with very high quality, on a considerable scale. The demand for it, unlike that for the other products was not related to the rate of installation of nuclear power plants, but integrated with the number installed. I remember remarking to colleagues in CGE when, very early in the NPD project, it had been suggested that fuel was not within CGE's scope of supply, a suggestion that came to nought, that making nuclear fuel is tantamount to having a royalty on electricity.
In nuclear power plant operations it was necessary to develop new operating procedures with all the characteristics of other thermal generating stations plus radiation and a new class of control. New levels of staff selection and training were introduced, along with new organizations. In Canada, for instance, the traditional trades employed on station maintenance were replaced by new categories of employees, maintainers with multi-trade capability to avoid the unnecessary radiation exposure required by the traditional organizations.
Because of the hazard from the radioactivity produced, safety with respect to this new risk was a serious concern of all those engaged in any phase of these blossoming nuclear power programs. In addition, in most countries right at the outset, and in all countries eventually, separate national regulatory bodies were established to oversee the conduct of all these phases. Here, also, new policies and practices had to be devised. Different tacks were taken in different places. In the United States the Nuclear Regulatory Commission produced mandatory detailed guidelines and these were adopted by some of the countries that acquired American reactors. In other countries only mandatory principles were established and the practices and results of the operating organization, whether it was a mine, engineering or plant operating organization, were reviewed to see that they complied with the guidelines and met the safety objectives. The distinction in effect between the two approaches appears to be diminishing with the passage of time. There is, however, a more fundamental distinction between American practice and that in other countries. Approval of plans is, in the United States, very much more subject to a process of public argument and judgment than is the case elsewhere, where more reliance is placed on the professional judgment of a regulatory authority. Again, this distinction may tend to get blurred as environmental assessments, based on the adversary process, become more common.
From before the first atomic bomb was dropped there was always the concern of how to extend the benefits of nuclear fission without spreading the capability to produce bombs. Early reactors given or sold to foreign countries, such as the CIR research reactor given to India by Canada, were exported under agreements stipulating that they were for peaceful purposes only. After the International Atomic Energy Agency established a system of international safeguards, it was mandatory that international trade in nuclear materials and facilities invoke these safeguards. In 1968 the Non-Proliferation Treaty received sufficient national ratifications to go into effect. This was an agreement whereby the signatory countries, acknowledged to have weapons, undertook to assist the other countries even with peaceful explosions if these seemed efficacious; and the non-nuclear weapon states foreswore acquiring such weapons so long as they remained signatories. Today Canada requires that before non-nuclear weapons states can receive nuclear materials and equipment they be a signatory of NPT, accept IAEA safeguards and enter into a bilateral agreement containing further restrictions, particularly on retransfer of materials and technology. Many recipient countries find these restrictions denigrating so that export negotiations are greatly complicated by this factor. This is further exacerbated when the international safeguards regimes are evolving, as they were in the sixties and seventies, and when national requirements are also changing. Safeguards against diver- sion of nuclear items to military purposes is an extremely important factor in extending nuclear power from one country to another, and important regimes, national and international, have been established to control it.
In the early stages of development of every phase of this new energy sector, creation of new organizations coupled with innovation in design/ manufacturing, construction, operation, R&D, and regulation were associated with the central task of getting the job done. It would be natural to think that once the organizations and the innovations were in place that the later jobs would get done much more expeditiously and efficiently. This didn't happen. Many later projects have taken up to twice as long to complete as the pioneer projects. There are several reasons for this, some of them internal to the sector itself, others social in origin.
In the established operating (as distinct from regulatory) organizations there are many more knowledgeable people who need to be consulted and there is immeasurably more information, much of it derived from experience with the earlier projects, that needs to be taken into account. In the beginning/ although there was a more collegial atmosphere with broader discussions, there were perhaps only a dozen people who had an important part in all the most important decisions. The regulatory organizations, too, and their practices have been enormously elaborated since the early days and the interplay between them and the operating organizations is much more time-consuming.
But the external factors are at least as great as the internal. In the fifties and sixties the world was rebounding from the depression and World War II. An inadequate, outdated world infrastructure was being replaced. Housing was being built at an unprecedented rate, the automobile population was expanding rapidly everywhere and our modem highway systems were being built; jet fleets were replacing the old propeller-driven aircraft and modern airports were built to handle the exploding traffic and most of the baggage; communications were being transformed by solid-state physics and satellites; health care was improving dramatically and becoming more universal; and all these things required more energy. Not only was there the crying need for all these facilities after the Depression and the War, there was a wealth of pent up scientific knowledge waiting to be exploited and above all, a fertile social climate. After the constraints of the Depression and the War there was an enthusiasm for seeing things happen. Politicians gained points by announcing new projects and eagerly sought opportunities to do so. A residue of the same experiences was a considerable acceptance of authority and a general respect for science. There was a third unrecognized attitude that distinguished those times from these: for a generation that had seen mighty armies assembled, equipped, and established overseas, navies built, air armadas launched, major cities in several parts of the world destroyed and rebuilt all in the space of a dozen years, it was unimaginable that it could take long to build a power plant.
The dynamic building era petered out in the seventies. The new infrastructure was largely in place, in the OECD countries at least. An increasing proportion of hu- man activity swung away from building capital facilities to trade in consumables and wealth, and with a new toy/ the computer, to expand the game. The demand for energy leveled off. The balance of social attitude changed from enthusiasm for what had been termed progress to growing concern for society and the environment, expressed most strongly in negative reactions to large projects. The leveling off of demand for electricity found several power plants in the pipeline that now would not be needed for a long time. Some of these were canceled; some were deliberately delayed. The reduced requirement for new plants allowed authorities to be extremely responsive to the reaction against them and even the utilities to be more tolerant of delay so that many project schedules were considerably protracted. The first wave in nuclear power plant construction subsided, but great things had been achieved.
Figure 1: Number of nuclear power units entering service world wide since 1954.
As can be seen from Fig. 2 the net orders for nuclear units worldwide reached a plateau at about 500 units in 1979, about 400 of which are now in service. There will be relatively slow growth in nuclear power during the nineties. But, in case it should be thought that this is peculiar to nuclear power, Fig. 3 shows cumulative thermalelectric generation additions in North America from 1988 to 1996. As can be seen the expected additions of nuclear and fossil fueled units are the same over that period in the United States although, admittedly the trend favors the latter at the end of the period. In Canada, the four large units at Darlington considerably exceed other thermal generating capacity additions during the period.
Figure 2: The net orders for nuclear power units world wide.
Figure 3: Predicted cumulative thermal-electric generation additions in North America from 1988 to 1996.
Nuclear power plants are now producing about a sixth of the world's electricity, more than the world's total electricity supply when the programs began. Nineteen industrialized countries have nuclear power plants and seven developing countries, although 80% of the capacity in the latter group is in three countries - Taiwan, Korea, and India.
In addition to providing a large amount of electricity, nuclear power has done this safely. Were it not for the Chernobyl accident I should have said remarkably safely. As a direct result of that accident, however, 31 people were killed and more than 200 hospitalized. On the unproven assumption that fatalities are directly proportional to exposure at any level, several thousand more people will have their lives shortened. Nuclear power has had such a good record, however, that even this accident does not remove it from the category of safe industries. A good measure of the social value of a thing is the amount people pay for it. Even taking the estimated delayed fatalities from the Chernobyl accident into account, nuclear power is as safe, on a comparable value basis, as manufacturing or the service industries, 10 times as safe as primary industries. On the same basis, it is as safe as air travel, 10 times as safe as the automobile or life around the home.
As a matter of interest, the amount of energy extractable in existing power plants from Canada's annual uranium exports is twice as great as the energy available in all our coal, oil, gas, and electricity exports combined, and more than three times as great as our net exports in these commodities. Another interesting relationship is that, in the 80,000 tonnes already exported, Canada has transferred resources that have more potential energy, at 50% utilization possible in future systems, than all the energy thought to be available from all the coal, oil, gas, and even the tar sands in this country.
At this time, too, adequate enrichment facilities exist in the United States, the USSR, France, Britain, and in a joint British-German-Dutch plant to supply world needs for some time to come. South Africa has also built an enrichment plant. The large fuel reprocessing plants in Britain and France are serving the international market as well as meeting domestic requirements. Canada is self-sufficient in heavy water, India has several plants in operation and is well on the way to self-sufficiency, and Argentina and Romania have some capacity to serve their domestic needs.
Over the past 35 years a whole new economic sector has been created based on the understanding of nuclear fission. The most obvious results of this are the physical manifestations - the power plants, the mines, the R&D establishments, the fuel cycle facilities. These, however, have a relatively short life and their benefit to mankind limited and transitory. The great result of the past 35 years of worldwide endeavor, by a million or more people in all walks of life, is that the way to put to use one of the world's greatest energy resources is now firmly within human ken.
The burning question is when do we pass through the swamps and reach the ranges ahead. We are fast approaching the time when the industrialized world will need new power plants. In some areas, however, painful experiences in obtaining permission to operate nuclear power plants are current or still fresh in people's minds. So is the economic consequence of Three Mile Island. Real interest rates are, in many areas today, twice what they were when we began to climb that first range of hills. Things are cyclical, however, and human attitudes, which are at the root of these things, can change surprisingly quickly.
Although I can't say just when we can expect to reach the next high ground, I think we can look forward to another cycle of what I referred to in the beginning as the Second Stage - the development and widespread application of basically simple power reactors. There are ample uranium resources for a big program with simple fuel cycles, and the more efficient reactors, the breeders, are not yet ready for general application.
[Canadian Nuclear Society home page]