blast logotype

2023 December

Notes • Quotation Sources • Supporting Calculation

The first samples of this number were distributed to the public at LosCon, at the end of November 2023. Based on comments from recipients of preliminary versions, revisions were made until a final version was reached in May of 2024. Hopefully future numbers will progress more rapidly.

The format of this number, in which most of the space is occupied by the one long piece, The Atom in the World Energy Picture, is not intended to be typical. Future numbers may incorporate a greater number of contributions from different hands, and of more modest length.


To request a printed copy (with color cover) by mail, write to :

Post Office Box 1035
Fort Worth
Texas 76101 USA

Be sure your address is legible. If requesting more than one copy, please enclose one dollar per copy, to assist in defraying the cost of printing and postage.

Alternatively, you may send one dollar per desired copy via PayPal, being sure to include your address, a request for the print version of blast №1, and the number of copies desired, in the Notes field provided.


Page 3 — At least one person…
For Frédéric Joliot–Curie’s actions immediately following the first measurement of neutron multiplication, see The Atomic Complex by Bertrand Goldschmidt, published in English translation by the American Nuclear Society in 1982. Goldschmidt, much like Glenn Seaborg, was a scientist first — he was in the room with Halban and Kowarski that day in 1939 — who participated in the Manhattan Project on behalf of Free France, and later came to serve as a senior administrator.
Page 7 — Not even a hundred times the consumption of 1955…
The United Nations Statistical Yearbook for 1962 gives the total world production of (commercial) energy for 1955 as equivalent to 3298 million tonnes of coal. 300 000 million tonnes would be 91 times this.
Page 7 — The nuclear fuel that could be scavenged from ash heaps…
Oak Ridge National Laboratory was set on to the task of evaluating whether the leaching of uranium and thorium from coal waste was a plausible source of material for clandestine nuclear weapons production. This work was summarized in ORNL Review magazine in 1993. Pilot projects have been reported, but as yet, uranium is not being produced from coal ash on a commercial basis.
One conclusion was that a coal–fired kilowatt–hour, as a rule, involves the release of a great deal more radioactive material into the environment than the nuclear kWh. (See also JP McBride et al, Radiological Impact of Airborne Effluents of Coal and Nuclear Plants, Science, 8 December 1978.) This exposes the absurdity of objecting to nuclear power as a source of radioactive pollution, while continuing to burn coal, as Germany has.
Page 9 — So long as an evil seems necessary…
We will have more to say concerning the triumph of the masters of steam engines over the masters of slaves in the American Civil War in a future number. For the present, we may refer to the famous “Am I Not A Man And A Brother?” tokens, which circulated in Britain at the end of the 18th century, among the great mass of copper Conder tokens which made up for the lack of official small change. Such tokens were produced primarily in Birmingham, and the most notable producer was the firm of Boulton and Watt — yes, that Watt. Arguably the first steam–powered manufacturing establishment in the world, Boulton and Watt achieved such technical perfection with tokens that they were ultimately chosen to supply the regal copper coinage of 1797, which rapidly drove the Conders out of circulation.
Page 11 — The very same people…
One of the claims repeatedly made was that electricity has a low Second Law efficiency. Considering how difficult it typically is for the layman to come to grips with the Second Law of Thermodynamics, one might be forgiven for thinking that this line of argument was meant more to obscure than to illuminate. As a result, it comes as a surprise that the utter irrelevancy of it can be proven with a simple table–top experiment. If an LED is connected to a thermoelectric generator, which in turn is heated by a burning candle, even though the efficiency of converting heat to electricity is no more than about 2%, the LED will be much brighter than the flame. How much of the energy released is made available for human use is by far the more meaningful quantity.
Page 12 — Meanwhile in Germany…
On the effectiveness of the Energiewende, see for example Is Germany cleaner than France today?, which uses publicly–available information and internationally–accepted values for the carbon intensity of different electricity sources to compare the electricity supplies of the two countries. The difference is rarely less than ten–to–one in favor of France.
For the telling absurdity of demolishing a wind installation to expand the lignite pits which feed the Garzweiler power station, see this news story. The town of Lützerath is one among many. The forest of Hambach is a particularly ignoble instance of ecological vandalism in the name of the environmentalism which rejects fission and accepts coal.
Page 13 — The whole history of central–station power…
What is regarded as the first steam–electric central station, Thomas Edison’s Pearl Street Station, served only lighting loads when it entered service in September of 1882, but this was not typical. There was already a desire for transmission of power from central stations to operate machinery, if only to avoid the cost and complications of maintaining boilers. Beginning in 1871, power from steam central stations was distributed hydraulically in London, and operations ended only in 1977. (The pipes are now used for communications cables.) Even compressed–air transmission was employed in Paris, despite the remarkable disadvantage that the air had to be re–heated for most uses, typically by burning gas or coal.
For Martin Couney, the opposition he faced in a period when the death of premature babies was often called a wholesome weeding–out of the unfit, and the Coney Island sideshow, see this book review.
Pages 14—15 (Closed Nuclear Fuel Cycle)
This diagram owes a great deal to Bernard Spinrad (Oregon State University), Plutonium–Fueled Fast Reactors as Secure Fuel Cycles, in Nuclear Energy and Alternatives, and in particular to his diagram on page 137. As illustrated by Spinrad, thorium bundles suitable for use in the converter reactor are pre–irradiated in the blanket of the fast breeder to load them with fissile uranium–233. After discharge from the breeder, they are transferred to the converter ; and after discharge from the converter, they are reprocessed, and the 233U blended with depleted uranium, the resulting denatured uranium being fed to a separate converter reactor. The fuel from this converter is then reprocessed, and the plutonium used as make–up for the breeder. (By implication, residual 233U is returned to the uranium–fueled converter.) By his estimate, with all reactors normalized to the same unit annual electrical output, eight breeders can support ten thorium converters, which in turn support nineteen uranium converters. This is a support ratio of greater than 3·5, assuming quite a high breeding gain of 1·6 in the breeder, and a high conversion ratio of 0·94 in the thorium converter, but a conversion ratio in the uranium converter of only 0·7, comparable to present practice.
The present scheme differs principally in assuming a single type of converter, loaded alternately with bundles fabricated from plutonium–enriched uranium and from plain thorium, along the lines of what Bennett Lewis of AECL termed the Valubreeder cycle. Thus all Th—233U conversion occurs in the slow–neutron reactor, where it is most advantageous. This principle has been demonstrated in India, by the use of thorium bundles as absorbers, to hold down the excess reactivity of a full load of fresh normal–uranium fuel, when a new or refurbished reactor is started up. As an alternative, the fabrication of thorium fuel enriched with plutonium is unattractive, because it would require a chemical process to separate all three elements, U, Pu, and Th, from each other and from the fission product wastes. A further alternative is the use of bundles with enriched uranium in the outer rings and plain thorium in the inner rings, but this gives up flexibility, as well as the power and neutron multiplication from fast–neutron fissions in 238U, in this region where the flux of slow neutrons is least. The one advantage would seem to be a reduced production of 234U.
CANDU with normal–uranium fuel typically achieves about one fission per initial fissile atom in the fuel, or a burn–up of about 7·5 MWd/kg (although close to 10 MWd/kg can be obtained, at the cost of increased wear on the fueling machines, and an increased chance of fuel cladding failure). This is no better than for LWRs, although still impressive, as the discharged fuel contains not less than 5 g residual fissile per kg initial uranium. With a very slight enrichment, however, FIFA increases rapidly, as demonstrated in the pressure–vessel heavy–water reactor, Atucha I, in Argentina. Fuel lifetime increased from 6·1 MWd/kg with normal uranium to above 11 MWd/kg with uranium containing 9 g 235U per kg, a gain of about 45% in FIFA.
No attempt is made to lower fuel fabrication costs by using the same assemblies in breeder blanket and converter, because difficulties with the metallurgy of zirconium have yet to be overcome. (Loading discharged converter uranium fuel bundles into the breeder blanket appears likely to improve the economics of reprocessing, not only by increasing the proportion of Pu in the feed, but by regenerating 235U in the second chance fast–neutron n,2n reaction on 236U.) The value for irradiation of thorium is taken from AECL–3081, as confirmed by later sources, and the fissile loading and burn–up for the driver fuel is inferred from this source and from Veeder (1978). EBR–II fuel intended as a prototype for the IFR design (from which PRISM is derived) exceeded 200 MWd/kg burn–up in tests. Use of a U–Pu–Zr alloy fuel metallurgically identical to that planned for the full–size reactor, in the small core of the EBR, required the uranium to be highly enriched. As a result, FIFA was considerably less than 1.
There is, strictly speaking, no need for the complete separation of plutonium from uranium in this scheme, and this may be an advantage technically or politically. The only pure fissile material, even within the safeguarded fuel–cycle center, is then 233U. Pu–bearing fuel discharged from the converter can be partially separated, giving a 80:20 U:Pu mixture suitable for breeder core fuel, excess uranium, and fission–product wastes. The discharged breeder core fuel, after purification from fission products, can then be blended with blanket fuel, converter excess uranium, and 233U in suitable proportions to make up new converter fuel. The resulting small leakage of 233U from the converter to the fast breeder is of little consequence.
Page 15 — OECD
This international body began as OEEC, the Organization for European Economic Cooperation, the body created to implement the Marshall Plan of reconstruction in Western Europe. Alongside the complementary military alliance, NATO, OEEC functioned to define the US sphere of influence in Europe, as CMEA (also known as COMECON) and the Warsaw Pact defined the Soviet sphere.
Page 16 — Having asked the question for the world as a whole…
To properly consider the problem of a transition to nuclear energy would require comprehensive information on topics such as the effect of energy costs on overall economic conditions, and of overall economic conditions on energy demand ; extent of uranium resources, and the variation of cost of production from those resources with annual output ; cost of power from converters and breeders, at different uranium prices ; cost of capital equipment for electrifying various industries ; and the possible competition for capital and manpower between power–system construction and electrification of end uses. In fact, only scant information is available on most of these topics, and its reliability is questionable, despite the best efforts of statisticians, economists, and engineers among others. Some of it may not even be possible to obtain. Therefore, if we are to answer a question, we must ask a simpler one.
Page 18 — The important thing is…
The breeder reactor has proven a conundrum for economists. Partly this is because of the strong influence of technical details such as the cost of fuel refabrication, or the loss of fissile material in reprocessing. It is also difficult to evaluate the capital cost of breeders, when few have been built. Even with two recent nuclear power projects so similar as Hinkley Point C and Barakah, both using pressurized–water reactors (the most common type) of about 1500 MW electrical output each, the spread in cost per watt installed is more than a factor of five. A survey of the literature, however, reveals a total lack of consensus on methods. The price of plutonium, for instance, has variously been set based on an equivalent quantity of enriched uranium, which depends on the design of the reactor it is to be used in, as well as on the costs of uranium and enrichment ; at the full cost of reprocessing the converter fuel it is recovered from ; or even at zero, where reprocessing is considered a waste–management cost, to be paid by the operator of the reactor discharging the fuel. Consequently, it is possible to find a study supporting pretty much any position.
It does not help that breeders of different designs produce different average numbers of new fissile atoms per fissile atom consumed (up to a maximum of about 2·1, which would be very difficult to obtain in practice), different quantities of net fissile material per unit of energy output, and different annual quantities of net fissile material per tonne of fissile investment in the core. The concept of the support ratio can help greatly in disentangling the confusion. Converters of different designs, operated in a closed fuel cycle, require different amounts of make–up fissile material per unit of output. If we require a certain amount of energy from the overall nuclear installation, and the deficit on the converter side is to be balanced by the surplus on the breeder side, then we see two main possibilities.
With breeders which produce only a modest fissile surplus per unit of energy (typical of designs using ceramic fuel, such as the French Superphénix), and converters with a large make–up requirement (such as the PWR), it would take more than 1 W of breeder capacity to feed each watt of converter capacity. In this case, few converters will probably be built after the breeders begin to be built on a large scale, and the surplus fissile from the breeders will mostly be dedicated to starting up further breeders. For breeders which produce more fissile (typical of metal–fuel designs such as EBR–II and its derivatives) and converters which require less (such as CANDU), each watt of breeder capacity can support considerably more than 1 W of converter capacity. This leads to a mixed system of breeders and converters.
Page 22 — The Pickering accident of 1983…
On 1 August 1983, while operating under full load, Reactor 2 at Pickering Nuclear Generating Station, near Toronto, Ontario, suffered the rupture of a pressure tube. A sudden failure of the primary pressure boundary is considered perhaps the most severe kind of accident that could occur to a power reactor — in a pressure–vessel water reactor such as PWR, it would be absolutely catastrophic, and must be prevented at all costs. Operators brought the stricken reactor to a safe, depressurized condition without needing to invoke the emergency systems, and without any radiation exposures to personnel or radioactive releases to the environment beyond the very small ones of ordinary operation. The reactor was subsequently repaired and returned to service, although this required more than two years, because a detailed examination found that further failures were likely. To prevent this, all the pressure tubes of Pickering reactors 1 and 2 were replaced with an improved design, fabricated from a different zirconium alloy, already adopted for later CANDU reactors.
Page 22 — Not only winter heat, but also summer cooling…
Extracting low–temperature heat from the bottom of the thermodynamic cycle at a power station, where the conversion efficiency is least, is a well–established way of economizing on fuel. A particular advantage is that it makes available to domestic users heat from fuels such as lignite or heavy oil, which are cheap but require complex and expensive equipment. Nuclear energy reaches an extreme in this regard, and the use of fission for combined–heat–and–power operation goes back at least to the R3 Ågesta experimental station near Stockholm, Sweden, which operated 1964—74. (The very first instance of nuclear space heating appears to have occurred in the autumn of 1951, at Harwell in England. See Utilization of Waste Heat from the British Experimental Pile, Nucleonics, 1952 March.) At Beznau in Switzerland, as a typical case, approximately eight kilowatt–hours of heat are supplied for each kWh of electric production lost. By pumping hot water, at above atmospheric pressure, transmission distances of tens of kilometers become practicable. For the advantages of this system over steam, see Piping Problems Simplified with High Temperature Hot Water Distribution, Architectural Record, 1953 February.
US Patent 1,781,541, granted 11 November 1930 to Albert Einstein and Leo Szilard, and titled simply Refrigeration, describes the absorption chiller, marketed by the Servel Corporation as The Miracle of Ice from Heat. It is really an invention worthy of those two great lateral thinkers : a purely thermodynamic machine, composed of a sealed loop of tubing, actuated by heat, and with circulating fluids as its only moving parts. Absorption chillers today are typically either very large units, or very small ones, such as for travel trailer refrigerators. In the former case, they are typically fired by pipeline gas, sometimes by oil or even coal — whatever fuel is used by the building boiler. In the latter, they are often fitted for operation by bottle gas when the vehicle is parked, and electric heat when in motion.
For actuation by low–temperature heat, such as from a district heating system or solar collectors, ordinary water under sub–atmospheric pressure is used as the working fluid, with an admixture of lithium bromide, an extremely hygroscopic salt. This system is an established one ; see Air Conditioning Works with Design, Architectural Record, 1950 January. As a refrigerant, water is not only as cheap as could be desired, but non–toxic, and utterly incapable of forming an explosive mixture with air. The prototypical refrigerant for compression–type chillers, anhydrous ammonia, is both poisonous and explosive. The chlorinated fluorocarbons were invented to solve these problems, but have been banned as injurious to the ozone layer of the atmosphere, which protects us from the harmful rays of the Sun. HFCs, brought in as substitutes for CFCs, themselves must now go, because of their exceptional global warming potential. And what is the new favored working fluid for small installations, typical of retail stores, homes, and vehicles? Propane.
Page 24 — The Hanford site in eastern Washington…
Gatlinburg II : An Acceptable Future Nuclear Energy System (MW Firebaugh and MJ Ohanian, editors) is the summary Proceedings of a 1979 conference held by the Institute for Energy Analysis and the Oak Ridge Associated Universities, and consists (unusually, but usefully) primarily of the discussions rather than the papers presented. Section II of this volume, Siting, emphasizes the power park and fuel cycle center concepts. Several sites suitable for supporting 10 or even 20 GW of generating capacity are identified. Work relating to the Hanford site specifically is reported by Ronald K Robinson.


Supporting Calculation

Approach to the nuclear–energy economy
Quantity123456789 10111213141516171819 2021222324252627
Assumed starting stock : spent LWR fuel equivalent to 80 000 t initial uranium. Composition 95% U, of which, 1% 235U ; 4% fission products ; 1% total plutonium.
Enrichment tails sufficient to produce 76 000 t at 1% 235U by stripping to 0·1% 235U.
FBR initial loading : 3·63 t total Pu per gigawatt electrical capacity
FBR annual Pu surplus : 184 kg (fissile)/GWe a (Tripplett et al, 2010)
CANDU burn–up with uranium fuel of 1% 235U content is taken as 20 MWd/kg, with thermal efficiency of 30%, requiring 60 833 kg/GWea of fuel.
Plutonium content of discharged fuel is taken as 0·62%, or 377 kg/GWea. (Croff and Bjerke, 1980)
For convenience in calculation, U–fed and Pu–fed HWRs are accounted for separately. In practice, we would expect to start up a reactor with a load of recycled LWR uranium, approximately 1% 235U, interloaded with plain thorium to hold down the excess reactivity. As the core approaches equilibrium, with reactivity of Th bundles rising by ingrowth of 233U, and reactivity of U bundles falling by depletion of 235U and accumulation of fission products, U–Pu bundles would be fed in. The plutonium yield from the initial fuel load would necessarily be somewhat less than shown, owing to the loss of neutrons to thorium.
Pu from reprocessing of LWR fuel is dedicated to FBR cores only.
Pu surplus from FBRs is added to recovery from uranium–fueled HWRs, in that year, and the Pu requirements for HWR fuel in that year deducted. The remaining Pu is passed along to the fabrication of FBR cores in the following year.
Plutonium recovered from Pu–fueled HWRs is taken as recycled into HWR fuel, and thus does not appear in the material balance.
Uranium from LWR fuel [t]—3800— —0—
U from enrichment tails [t]—3800— —0—
U to (from) stockpile [t]59263445(228)(5670)(3472) —0—
Plutonium from LWR fuel [t]—40— —0—
Pu from HWR fuel [t] 17·242·780·5136114 —78·2——0—
Pu from FBR surplus [t] 2·05·09·516·125·834·4 41·147·854·661·568·575·5 82·589·796·9104111119126134 140141142143145146148
Pu passed along to FBR cores [t] 19·347·890·015313092·0 93·394·595·897·098·399·6 10110210410510610810911025·8 26·026·326·526·827·027·3
Net Pu to HWR fuel [t] —0—10·120·526·031·5 37·142·748·354·059·8 65·671·577·483·489·4 95·5101113115116117118119120

FBR capacity constructed [GWe] 11·016·324·235·853·046·7 36·336·737·037·437·738·1 38·438·839·239·539·940·3 40·641·030·47·17·27·2 7·37·47·4
Total FBR capacity [GWe] 11·027·351·587·3140187223 260297334372410449487527566606 646687728758765772780787794802
HWR capacity constructed [GWe] 27·540·860·489·513311790·9 91·792·693·494·395·296·1 97·097·998·999·7101102103 76·017·717·718·118·218·4 18·6
Total HWR capacity [GWe] 27·568·3129218351467560650742 8369301025112112181316141515151615 171718201896191319311949196719862004
HWR capacity fed with U [GWe] 27·568·3129218182 —125——0—
HWR capacity fed with Pu [GWe] —0— 1693424335256187118059009961093 119112901389149015921695 1896191319311949196719862004

Electric Generation [GW] 38·595·61803054916547819101040 117013021435157017061843198121212262 240425472654267927042729275527802806

To simplify calculation, a 100% annual load factor has been used. In other words, each gigawatt of reactor capacity is taken as generating 8760 gigawatt–hours of electricity a year. This is not possible in practice, and as a result, in any real system, the investment of fuel in reactor cores, per gigawatt–year of generation, will be greater. As a further simplification, only the annual fuel use of the converter reactors has been considered. Of course, a reactor cannot be started up without a full load of fuel, and if a fuel bundle stays in the reactor for three to five years, then a corresponding amount is required in the beginning. This can be offset to an extent, for instance, by part–loading with plain thorium or depleted uranium, and by fuel shuffling, but it is a complicating effect which must be accounted for in framing a realistic model. In this instance, it is partly represented by the uranium which is initially stockpiled.

The construction rates shown, peaking at about 150 GW/a, may appear astonishing, far beyond anything so far achieved. The vital point is, however, that for the kind of universal electrification now being advocated, something will have to be built at this kind of rate. The products of well–established heavy industries, which can be installed on a limited number of modestly–sized sites, and do not require the wholesale reconstruction of power networks, have a strong claim to consideration. Indeed, wind and solar would require installation of a much greater (perhaps 5×) nameplate capacity to generate the same number of annual units, only to require replacing at 12—20 year intervals. Electrochemical storage systems, often promoted in conjunction with wind and solar, are even worse. Not only would they have to be installed on a scale hitherto altogether unheard–of, but they can scarcely be expected to last more than 3—5 years in heavy service. It is difficult to see how anything can be achieved with such inadequate means.

Clearly, the weak point in this calculation is the support ratio of 2·5. If this cannot be justified, all else falls. Certainly 2·5 is easier to obtain than the 3·6 of Spinrad (vide supra). Hans Bethe (Relative Merits of Alternative Fuel Cycles, in Nuclear Energy and Alternatives at page 175), relying on work from Argonne National Laboratory, quotes a support ratio of 2·75 for the combination of oxide breeder and thorium–fueled HWR, which is heartening.

A calculation based on the CANDU performance given by Croff and Bjerke is not promising : an input of 608 kg 235U is offset by an output of 377 kg total plutonium. (Considerably more Pu is produced, and its consumption in situ contributes to the excellent overall fuel economy.) Because of the large slow–neutron capture cross–section of 239Pu, less of it than of 235U is required for criticality. Even assuming the discharged Pu to be 1:1 equivalent to 235U, however, a deficit of at least 231 kg must be made up, which is greater than the Pu surplus per GWea of S–PRISM in the high gain configuration. Adding in the residual 235U, about 46 kg, only brings us to a support ratio of unity. Even adding in the 236U, about 89 kg, which would correspond to 240Pu in a fully plutonium–fueled system, does not get us to anything like the 60 kg/GWea make–up fissile assumed.

The use of thorium as fertile material helps to economize on fissile material, because the 232Th nucleus has a capture cross–section for slow neutrons more than 2·5× that of 238U. (See figure, page 6.) This makes it a more effective competitor against neutron–absorbing fission products. The fissile species produced, 233U, has a smaller fission cross–section than either 235U or 239Pu, so that more of it is required for criticality, but its tendency to absorb a slow neutron and not fission is particularly small, making for better overall economy. In other words, the required fissile inventory is greater, but the make–up feed rate is less.

As metal or oxide, thorium also changes its properties more slowly under neutron bombardment than uranium does, meaning that thorium fuel bundles can be left in the reactor longer, saving on reprocessing and refabrication costs. It has been suggested that a fast breeder could be built to accept fuel bundles fabricated from plain thorium in its outer blanket, which would be discharged at a certain fissile content, and then loaded directly into a converter. Set against this is a much smaller probability of fast–neutron fission compared to 238U, which contributes substantially to the breeding gain of a metal–fuel breeder.

A word about the treatment of 240Pu, produced by about a quarter of all slow–neutron absorptions in 239Pu : although this isotope is not fissile — that is, it does not fission with neutrons of all energies, and thus cannot sustain a chain reaction by itself — it has a much larger fast–neutron fission cross–section, and a lower neutron–energy threshold for fast fission, than nuclei such as 238U. For this reason, depending on reactor design, it is typically assigned an equivalent worth, gram for gram, in the fast reactor of about 0·3× 239Pu, taken as a reference fuel. It has been stated that the equivalent worth, for use in a fast–neutron reactor, of reactor–grade Pu (containing about 30% 240Pu) is not much impaired by being recycled once through a slow–neutron reactor. In a plutonium–fueled converter, 240Pu absorbs a neutron, producing 241Pu, which has an even higher fission cross–section and neutron yield than 239Pu. The resulting disappearance of an absorber and appearance of a fissile nucleus helps to maintain reactivity despite the consumption of 239Pu, and for this reason, plutonium with a high content of 240Pu has been termed phoenix fuel.

blast Main Page