WHY "THE DREAM OF UNLIMITED CHEAP FUSION POWER" IS A LOAD OF HORSESHIT ----------------------Warren D. Smith Oct 2000------------------------ There are a class of "scientists" - for whom I have little but contempt - who view their job, not as the unveiling of natural mysteries, but rather as the production of deceptive propaganda in an effort to delude gullible politicians into giving them hundreds of millions of dollars in research funding for projects that will never bear fruit. Unfortunately fusion power research comes pretty close to matching this description -- and the stakes are literally the future of civilization. The public has been perpetually fed a stream of lies and intentionally deceptive statements by the big-money fusion research community. I'm not talking here about "cold fusion" - essentially everybody agrees THAT was a bogus hoax - I'm talking about the mainstream plasma fusion community. LIE #1: "We achieved energy breakeven - we're producing more energy than we input! We've succeeded!" ACTUALLY: In the cases I've seen, such statements did NOT measure the amount of energy coming in the door from the utility company, and also did NOT measure the amount of energy their fusion project generated, and then notice E_out>E_in. Instead, they measured certain "energy input" measures like "kinetic energy of neutral beams we inject into the reaction chamber" which are far, far, smaller (due to huge and ignored inefficiencies in the conversion of utility-supplied power into neutral beam kinetic energy). And the energy "output" does not actually exist in any useful form, and instead always is something like "number of fusing nuclei, times energy theoretically produced from each such nuclear event" - in fact if they tried to convert this to a useful form of energy, which they didn't, again large inefficiency factors would come into the calculation. (If you check the fine print, the terminology for all this is "scientific breakeven" or "breakeven plasma conditions", NOT genuine energy breakeven.) The total effect of ignoring these real-world inefficiencies is probably a factor of 100. In short, their fusion reactors are a great way to LOSE energy. [Example: JET (Joint European Torus) in 1997 claimed to have "neared breakeven" with fusion power/input power = 60%. This was claimed to be a world record. Their fusion power was 13 MegaWatts peak, also a claimed world record. I'd thought PPPL TFTR had claimed actually reaching breakeven, but I cannot find such claims on their web pages so perhaps they've now abandoned them (?) - all I now see is a claim they'd achieved (fusion energy)/(input energy) = .33 in a run producing a record 10.7 MW of fusion power for a fraction of a second. Again this .33 figure seems to depend on conveniently bogus definitions of both input and output energy, as I just described.] LIE #2: "Don't worry about the fact we are going to run it with tritium now - it is perfectly safe! We have many layers of protection to prevent any leak! That is impossible!" ACTUALLY: PPPL (Princeton Plasma Physics Lab), after numerous public assurances of the type above, managed to leak and lose 100% of their tritium into suburban Princeton / Plainsboro NJ. (These clowns are so incompetent, they cannot even keep their tritium bottled?) Tritium is an extremely hazardous radioactive gas (beta emitter, halflife 12.3 yr; note that tritium is about a billion times more radioactive than uranium, plus uranium has the advantage of being solid hence not leaking.) They then told the media not to worry - probably all that tritium rose into the upper atmosphere, never to be seen again. Well, maybe. But in fact, maybe some stuck around near the ceiling for awhile? And tritium in air will react to form radioactive water, either quickly (if you light a match, or expose to catalysts) or slowly (which is why the atmosphere contains no free hydrogen). The Giant Organs of Propaganda at PPPL neglected to mention THOSE facts. Were there suitable catalysts around? They did not tell us. And the true outrage was that they were even using tritium in the first place. Namely, they'd been running for years with deuterium or H1 only, which, while producing far less fusion energy, was thought entirely adequate for research purposes since it was comparatively easy to extrapolate how DT would have performed. (Deuterium and H1 are essentially non-hazardous.) In a move purely motivated by Public Relations, in an effort to be able to claim they'd achieved "breakeven" (see lie #1), and knowing their funding was likely to get reduced really soon, PPPL switched over to tritium. The T runs caused their entire setup to become too radioactive to be accessible by humans for many months - which they knew was going to happen. In other words they intentionally sabotaged their whole scientific effort, creating an expensive permanent radioactive hazard, purely so they could try to generate deceptive propaganda. LIE #3: (direct quote from PPPL web page!) "The major fuel, deuterium, may be readily extracted from ordinary water, which is available to all nations. The surface waters of the earth contain more than 10 million million tons of deuterium, an essentially inexhaustible supply." ACTUALLY: Fusioneers love to spout this nonsense to the media. But the only kind of fusion power that is anywhere near close to feasibility, is DT fusion, not DD fusion which is far far more difficult due to a >100 times smaller cross section. [R.F.Post: Fusion power, Scient. Amer. 197,6 (Dec 1957) 73-84 has a cross section vs energy graph showing DD/DT cross section ratio exceeds 100 for D energies throughout the range 10-100 kev.] And golly. The Earth does not contain any natural tritium at all. Oops. So where is that boundless multimillion-year energy reserve now? So their plan would actually be to create the tritium by atomic transmutation of lithium via neutron irradiation. Lithium is a rare element - a heck of a lot rarer than deuterium in seawater, and not much more common that uranium and thorium: Li's crustal abundance is 17 parts per million by weight [CRC handbook], versus Uranium's 1.8 and Thorium's 6. Furthermore only the rare Li-6 isotope (7.6 atomic percent; the rest of natural Li is Li-7) can be feasibly used to make Tritium. Thus, the total mass of Lithium-6 in the Earth's crust is actually about 6 times LESS than the total mass of Thorium and Uranium! (Note: ALL isotopes of Th and U are useable as fuel if breeder reactors - already commercially available and running in France - are employed. I admit about 4 times less energy is released per kg of fissioning U or Pu nuclei than per kg of Li6 nuclei, in principle. But even so, since 6>4, fission is still winning this comparison. Plus, fusion reactors are going to be a lot less efficient than fission reactors, making fusion look worse still.) Making it even worse: Lithium has other important, non-fusion uses which would compete for the limited supply: batteries, lubricants, alloys, reagents, glass/ceramics, drugs. As far as I know there are only 2 US lithium producers - and atomic transmutation is a very expensive process, and tritium is extremely hazardous. Finally estimates of the total economically accessible Earth fusion energy reserves, as lithium, seem indeed actually to be SMALLER than the total economically accessible Earth FISSION energy resources as thorium and uranium, although that comparison seems to vary with time. Remember how fusion was going to be that wonderfully cheap and abundant energy source, a vast improvement on fission? Baloney! LIE #4: Don't worry about the fact lithium is rare. We can just extract it from seawater and the energy used in the extraction will be far less than the fusion energy it will later generate. I.e. here is a direct quote from "The President's Committee of Advisors on Science and Technology (PCAST)" (report dated 11 July 1995, chaired by Honorable John A. Young, former president and CEO of Hewlett Packard co.; given in full on the PPPL web page): "The cost of the raw fuel for fusion -- lithium and deuterium extracted from seawater -- would be a very small fraction of the total cost of the electricity produced." ACTUALLY: It is correct that Lithium IS a lot more common in seawater (.18 ppm by weight, says CRC handbook), than either Uranium (.003) or Thorium (.00005). That is the good news. The bad news is: The total amount of any of these elements in seawater is a tiny fraction of the total amount in crustal rocks, and the concentration, even of lithium, in seawater is almost 100 times lower than its concentration in average rocks (which in turn have vastly smaller Li concentrations than do ore deposits). But the hope of fusioneers, of course, is that despite this low concentration, it will be so easy just to pump seawater through some kind of lithium extractor, that we'll be awash is easily obtained lithium, with no need for any mining, and everything will be great. BUT NO! Assume (probably optimistically) that the cost of extracting the lithium from seawater is the same as the cost of boiling that water. (Actually, desalination methods are known to remove salts from water for less than the cost of boiling, but considering we also must seperate the lithium from all the other kinds of salts, present in far higher concentrations, and we must isotope-seperate the lithium-6 from lithium-7, my present assumption still seems optimistic.) Well, the fusion energy we can get from the .18 milligrams of Lithium (and hence 11.8 micrograms Lithium-6) in a liter of seawater (assuming Deuterium is free and ignoring the Li isotopic enrichment costs) is (assuming 100% fusion efficiency) 4.2 megajoules. Meanwhile the energy needed to boil that liter is 2.3 megajoules. So if fusion power generation is <=53% efficient (and note, the best steam turbine power plants are only 40% efficient, so it is absolutely certain fusion plants will be <40% efficient) then in fact, it will cost MORE energy to extract the lithium from seawater, than you can get from fusion. More, not less. They lied again and notice they are not just lying to the general public, but also to the US president and congress in a report by "the President's highest level, private sector advisory group for science and technology." So we now clearly see that what matters is not how much lithium is in all the world's oceans, but rather how much lithium is economically extractible from the world, which is a tremendously smaller amount. We've just seen zero ocean lithium is economically extractible. If you want to be economically reasonable you have to go with lithium mining (the method actually used by the present day economy) and as we've already seen (cf. lie #3) it seems plausible that less energy is extractible from Lithium mining using hypothetical future fusion technology than from Uranium and Thorium mining using present-day breeder-fission technology. LIE #5: "Oh, don't worry about the need to produce tritium from lithium. We'll just surround the fusion reactor with an Li blanket which will absorb neutrons and emit tritium into the reactor. Trivial." ACTUALLY: No such blanket has ever been tested and I don't think it is going to be very trivial. It would probably have to be several meters thick, for example, and might need to be molten. It might cause vacuum contamination problems. The thickness is a severe problem because the thicker you are, the further away your magnetic coils (on the other side) have to be, causing the magnetic fields to become weaker or the magnets to drink more power. (Similarly if steam turbines were ever actually to be hooked up to fusion reactors, their water/steam pipes and insulation for them, would also be thick, reducing efficiency further.) Also, note that a lithium blanket this thick covering even ONE fusion reactor, would all by itself consume a substantial fraction of the entire world's annual production of Lithium, and if this blanket were made of isotopically pure Lithium-6, it would probably exceed the world's annual production of Li-6 (and certainly *far* exceed the world's present production of isotopically pure Li-6). The hope here is to combine the reactions Li6 + n --> He4 + T + 4.7MeV T + D --> He4 + n + 17.6MeV to get the net reaction Li6 + D --> He4 + He4 + 22.3MeV. (This is as contrasted with fission reactors in which U or Pu turns into smaller atoms plus neutrons with about 200 MeV released per fissioning nucleus.) But actually, a considerable fraction of the neutrons will necessarily be lost, hence NOT all the needed tritium could be supplied by this means. Hence, a considerable fraction (perhaps 20%?) of the needed tritium would have to be produced in a FISSION reactor of a type featuring a lot of excess neutrons flying around (and thus also of a type inherently suitable for "breeding" plutonium...). Hence the bullshit about how fusion could hope to "replace" or "supplant" fission reactors, eliminating fission wastes, is indeed bullshit, and also especially bullshit is this quote from the PPPL web-page: "The world-wide availability of [lithium] would thus eliminate international tensions... [currently arising from some nations attempt to control Plutonium and Enriched Uranium supplies and nuclear technology]." Since in fact, at present not enough tritium is available from all commercial sources combined to operate even one fusion plant - and the only source of enough tritium is, in fact, (if there is one) top secret US and Russian plants which produce tritium for their H-bombs. So you'd have to negotiate with US and Russian militaries to run your fusion plant, as matters stand! This is hardly a tension "reduction" due to "wider availability"! (Also of course, the fusion plant's neutrons could be used to breed plutonium, anyway - yet another reason these claims are nonsense, and flatly contradicting this quote from the PPPL web page: "Another significant advantage is that the materials and by-products of fusion are not suitable for use in the production of nuclear weapons." In fact by breeding plutonium with my fusion reactor, and siphoning off some of the Deuterium, Tritium, and Lithium-6, it would be easy to get all the ingredients of an H-bomb.) LIE #6: "Fusion will be clean, cheap energy, not generating lots of nasty radioactive waste, and with no worries about nasty meltdowns, hence there will be no need for expensive safety precautions like with fission power." ACTUALLY: It won't be clean. It won't be cheap. DT fusion releases "fast neutrons." (Fission plants use much slower and more confinable neutrons, and release far fewer neutrons per megawatt generated.) Fast neutrons turn everything they touch radioactive or busted. The entire enormous fusion vacuum vessel, confinement field coils, cladding, expensive sensors, expensive windows, and all, will turn radioactive and will have to be completely replaced and disposed of on time scales of, e.g., 1 year. Fusion plants will be extremely expensive and complicated: As far as I know, even the research-scale fusion places so far already are more expensive than any actual power plant - and you've seen nothing yet when it comes to cost. The present cost (year 2000) of power plants is about 1-2 dollars per watt of generating capacity. The JET power 13 MW record I mentioned before was achieved at a plant cost of billions, thus the plant cost was over 100 times larger per megawatt than commercial power plants. (The ITER international reactor planned to replace JET is estimated at 3.2 $billion plus 0.2 $billion in decommissioning costs, with construction taking 10 years. It is hoped to be able to generate about 40 MW more pseudo-power out than power input.) And that is totally ignoring the facts this 13MW was only produced in a short burst timespan surrounded by a far larger downtime (multiplying the "100" by at least another 100), it was produced at the cost of inputting far more power than 13 MW so negative power actually was generated, zero actual electrical power was generated, and fusion plants would have to be replaced at enormous rates. Thus, really, it would seem the capital costs of fusion plants per watt of generating capacity will be over 10000 times more expensive than conventional power plants, and the maintenance and design costs will also be enormous. Wow! THIS is supposed to be the Great Hope? The prospect of rebuilding such an incredibly complicated and expensive plant every year or few, is an extremely economically unappealing one, and what to do with all that radioactive waste - i.e., the old plant? And I've already discussed the hazards of tritium. (Being a gas, tritium is in many ways more dangerous than the usual fission products, which generally are solids.) LIE #7: "We're real close. Just give us 10 more years." ACTUALLY: After 50 years of such claims, it is time to shut up. The claim they are "close" is like a stone age man claiming: "I'm finally getting close to finding out how to burn rocks so they'll produce more energy than I use to ignite them. I admit I haven't managed this feat yet. But in no time I'll have gigawatt power plants with steam turbines, etc, based on burning rocks." Ha. Now in fact, the only way fusion ever will be clean and economical (if ever) is going to be if it is based on some fusion reaction NOT generating neutrons, so that the extremely expensive fusion plant does NOT have to be rebuilt every couple of years. But the known neutron-free reactions (boron is the lowest element commonly available on earth participating in one: H1 + B11 --> C12 + 16.0MeV) are far harder to achieve than DT fusion - 5 times larger coulomb repulsion. DT is especially easy since D & T each have charge +1 only, hence have the minimum possible repulsive barrier that we need to overcome to cause fusion, and also T is extremely unstable leading to a high fusion energy and a high fusion cross section. DD has a >100 times smaller cross section. That means far larger confinement times are required with DD - and all the while heat energy has to be input, increasing the debit (loss) side of the energy ledger while meanwhile the production side of the ledger is made smaller since DD is a less energetic reaction. So forget "breakeven" (which was a lie anyway) - you'll be lucky to get 10% of your energy investment back out. Known *neutronless* reactions like the boron reaction involve far larger charge-repulsion barriers to overcome than DD, as well as far smaller cross sections than DT, requiring far larger temperatures AND confinement times, and hence far larger losses. So all of these seem far more difficult than DD fusion, which itself is regarded by all as unfeasible - only DT is anywhere close to breakeven in their wildest dreams. As far as I know, no research lab is even looking at neutronless fusion reactions, despite the fact that everybody knows they are the only hope, ever, for economically attractive fusion power. FOOD FOR THOUGHT: Make your cyclotron beam T ions into a D target. (Or D,D, or D,T.) Say the total kinetic energy of the incoming ion is E_in. Say the total energy produced by DT fusion events during the ion-target interaction (on average), is E_out. Experimental fact: No matter what energy you set your cyclotron to, E_out << E_in. Perhaps this well known experimental fact should make you worry that maybe, just maybe, fusion power is not going to be easy to achieve? This does not actually prove it is impossible (Indeed: the sun works, due to the fact the sun loses energy at an extremely slow rate due to hundreds of thousands of miles of insulation; albeit the sun has an extremely slow rate of fusion far too slow for any economical use on Earth, i.e. one must wait billions of years per hydrogen atom for it to fuse in the sun. Earth based fusion reactors would have to cause fusion at 10^16 times higher rates.) but it sure is food for thought. Constrast the utter simplicity of fission. MORE FOOD FOR THOUGHT: Studies from the USA "super" program (attempting to develop the H-bomb) concluded that deuterium "burning" following "ignition" by an A-bomb, would NOT occur. Of course, the USA ultimately did develop H-bombs, including one based on DD fusion, but only by using the A-bomb not merely to ignite, but also to compress and confine, the fusion fuel. Again, this factoid does not prove fusion is impossible, but does indicate its difficulty. STILL MORE FOOD FOR THOUGHT: "Plasma instabilities" are a well known bugaboo that has bedeviled them for decades. But only in the last 5 years have they brilliantly come up with the notion of trying to use feedback control to fight back against such instabilities. Duh. I mean, they are actually so incompetent that such control was not incorporated into fusion reactor designs until just now. AND STILL MORE FOOD FOR THOUGHT: The total amount of Lithium produced in the world in the year 2000, was 13 thousand tonnes, according to US Geological Survey mineral yearbook & commodity summaries. The per capita world consumption of power is 2.1 kwatts, according to the United Nations. Energy used by humans in 2000: 2.1*kilowatt/person *6*10^9*people = 4.0*10^20 Joules Fusion energy available from Lithium produced in world in 2000: 13*kilo*tonne * .0658 KgLi6/KgLi * 3.58*10^14 Joule/kgLi6 = 3.1*10^20 Joules So: This amount of lithium was (even assuming all the lithium-6 in it was totally used to generate fusion power, and even assuming we somehow attained 100% efficiency in obtaining that power from the lithium-6) not sufficient to generate that much energy - it fell short by 23%. Oops. Meanwhile, the total amount of uranium mined in 2000 (about 61 thousand tonnes) WAS over 12 times more than sufficient (if used in breeder reactors) to power the world that year, assuming 100% efficiency, and assuming (more realistically) 30% efficiency, it still was nearly 4 times more than sufficient. (Fusion plants would be a lot less than 30% efficient, I think.) WHAT ABOUT USING H-BOMBS TO GENERATE ENERGY? Quote from "Scientific American": The Chinese have proposed harnessing the energy of underground thermonuclear explosions for generating electricity. The explosions would supposedly take place in a subterranean cavity lined with massive steel tubes, which would conduct steam to turbines on the surface. "It's possible that this kind of controllable nuclear electric station will become the main energy supplier around the world in 30 to 50 years," wrote He Zuoxiu. I'm going to argue that this insane-sounding Chinese plan is actually the most promising way known to generate fusion energy - but it probably still will not be economically useful. Re the quest to use minimal fission, and maximal fusion, in bombs: It presumably would be possible to use subcritical masses of fission fuels. The critical mass of Pu239 or U233 as a sphere surrounded by air is about 16 kg, but under compression by a length-factor of X, presumably the critical mass decreases by about a factor of X. Also surrounding the mass with substances other than air decreases the critical mass. (The presence of light element "moderators" also can decrease critical mass, but would slow down neutrons while speeding up thermal and hence sonic velocities - thus perhaps preventing the fission chain reaction from proceeding to near-completion before bomb disassembly, hence probably not a good idea.) Extremely high compression might be achieveable by assembling the critical mass with the aid of magnetic "railguns" or "high velocity light gas guns" instead of chemical explosives. (Recently railguns and light-gas-guns have been used to shoot a bullet with a bullet to get 2X-speed collisions.) The combination of these ideas might enable making a supercritical mass, and a fission bomb initiator for an H-bomb, with as little as 500 grams of Pu239 or U233. Although such a bomb might not be useful as a weapon, it might be quite useful for our purposes. Observe that H-bombs CAN detonate pure deuterium (indeed the "Mike" shot, the first H-bomb test, used liquid deuterium) and thus WOULD be capable of tapping the enormous ocean deuterium supply without need for tritium or lithium. The problem would be to generate most of the bomb's energy by DD fusion, but very little of its energy using fission (since the Earth's supply of fission fuels is far smaller than the Earth's supply of deuterium). The radioactive residues would hopefully be confined underground and would not be a problem. If the bomb were detonated inside a "salt dome," perhaps a mass of molten salt, serving as a heat-storage reservoir, would be formed which could then be tapped by technology similar to today's "geothermal." The 10 December 1961 "Gnome" test detonated a 3.1 kiloton nuke in a salt bed formation 360 meters underground near Carlsbad, New Mexico (32 degrees 15'46.7"N / 103 degrees 51'57.3"W) An announced purpose of Gnome was "to study the possibility of converting the heat... to steam for the production of electric power." (It was also thought that salt, since water-soluble, might enable easy recovery of radioisotopes produced by the blast.) According to press accounts: Gnome created a permanent cavity 50 meters in diameter and 24 meters high which was entered by humans 5 months later, at which time it had a temperature of about 140 Fahrenheit. A few weeks after the explosion it was still more than 1000F inside. The cavity is still highly radioactive. (It was also wondered whether such cavities could be useful for, e.g., storing oil. However it is cheap to create such cavities by solution mining without need of explosions...) The ground jumped about 2 meters directly above the explosion, settling about 0.6 meters above its initial altitude. Any power plant, heat transfer pipes, etc, would have to withstand such stresses. Gnome also caused an unintended release to the atmosphere of radioactive steam - mostly during the first 30 minutes, but continuing for another day. An estimated 2400 tons of rock were melted and mixed with an infalling mass of 13000 tons of unmelted rock salt. Presumably the cavity volume would be multiplied by 330 for a 1 megaton instead of 3.1 kiloton blast, i.e., the cavity diameter would be multiplied by about 7. Much of the energy in each explosion would necessarily be wasted by heat leakage into surrounding rocks. You would need a heat-transfer/collector area comparable to the area of the cavity walls to avoid great innefficiency, and it, and the power plant, would have to survive the stresses of the explosions. So: the Gnome test makes it seem pretty difficult to use underground nuclear explosions to generate useful energy, but I'm not sure it is impossible. Indeed, this whole idea seems MORE (not less) promising, than every other proposal for generating fusion power. Namely: (a) it is feasible with today's technology, (b) it really would work with DD fusion and thus really be capable of tapping an enormous energy supply, (c) while it may or may not ever be economically feasible - it is closer to being economical than all other proposals. Note that humanity's present power consumption (2.1 kW/person, 6*10^9 people) is equivalent to exploding a 1 megaton H-bomb (assuming garnering its energy with 100% efficiency; note 1 weapons megaton = 10^12 calories) every 5.5 minutes. Incidentally, that "Scientific American" piece continued: That scenario is highly unlikely, according to Richard L. Garwin of the IBM Thomas J. Watson Research Center, who participated in a U.S. study of this power-generation concept in the 1970s. The investigation, he says, showed that a minimum of two detonations a day, or more than 700 a year, would be required to keep such a generator running. The costs of this technology, Garwin says, would exceed those for conventional nuclear reactors, which are already hard-pressed to compete economically with hydropower, fossil fuels and other energy sources. Garwin contends that nonnuclear methods are also cheaper, more effective and less damaging to the environment than PNEs (Peaceful Nuclear Explosions) for applications such as excavation and oil-well stimulation. SUMMARY: In short, all big-money fusion research is known to be looking in directions that can never yield an economically feasible fusion plant, EVEN if every hope they have, actually does pan out as successfully as they could hope for. Furthermore, the genius designers and builders of these huge projects sometimes look more like idiots, and that group of scientists and administrators is absolutely relentless in lying to the general public and the federal government.