Piece of the Sun : The Quest for Fusion Energy (9781468310412) Page 14
Meanwhile something happened that would give Hirsch’s plan new urgency. On 6th October, 1973, Egypt and Syria launched a surprise attack against Israel. Starting on the Jewish holy day of Yom Kippur, the attackers made rapid advances into the Golan Heights and the Sinai Peninsula, although after a week Israeli forces started to push the Arab armies back. The conflict didn’t remain a Middle Eastern affair for long. On 9th October the Soviet Union started to supply both Egypt and Syria by air and sea. A few days later, the United States, in part because of the Soviet move and also fearing Israel might resort to nuclear weapons, began an airlift of supplies to Israel. What became known as the Yom Kippur War lasted little more than two weeks, but its effects reverberated around the world for much longer.
Arab members of OPEC, the Organisation of Petroleum Exporting Countries, were furious that the US aided Israel during the conflict. On 17th October, with the war still raging, they announced an oil embargo against countries they considered to be supporting Israel. The effect was dramatic: the price of oil quadrupled by the beginning of 1974, forcing the United States to fix prices and bring in fuel rationing. The prospect of fuel shortages and rising prices at the pump was a shock to the American psyche. All of a sudden the huge gas-guzzling cars of the 1960s seemed recklessly wasteful and US and Japanese carmakers rushed to get more fuel-efficient models onto the market. In response, President Richard Nixon launched Project Independence, a national commitment to energy conservation and the development of alternative sources of energy. That meant more money for fusion, lots more. In 1973 the federal budget for magnetic confinement fusion was $39.7 million; the following year it was boosted to $57.4 million, and that was more than doubled to $118.2 million in 1975. And the increases continued: by the end of the decade magnetic fusion was receiving more than $350 million annually.
Reading the political runes in 1973, Hirsch was keen to get his plan moving but wanted to make one significant change – he wanted the scientific feasibility experiment to use deuterium-tritium fuel. The fusion labs had previously assumed that they would use simple deuterium plasmas so that they wouldn’t have to deal with radioactive tritium, radioactive plasma vessels or the added complications of alpha-particle heating. They wanted a nice clean experiment in which they could get a deuterium plasma into a state in which, if it had been D-T, they would get the required energy output – a situation known as ‘equivalent break-even.’ For Hirsch, that wasn’t enough. He suspected that many of the scientists at the fusion labs were just too comfortable working on plasma physics experiments, but he wanted them to get down to the nitty-gritty of solving the engineering issues that a real fusion reactor would face. And he knew that a real burning D-T plasma would be PR gold. The White House, Congress and the public would never understand the significance of equivalent break-even but if a reactor could generate real power – light a lightbulb – using an artificial sun, that would get onto the evening news and every front page.
Not all fusion scientists were against moving quickly to a D-T reactor. Oak Ridge was not afraid of radioactivity. The lab had been set up during the wartime Manhattan Project and had pioneered the separation of fissile isotopes of uranium and plutonium to use in atomic weapons. Since then it had branched out into many fields of technology, some of which involved handling radioactive materials. Oak Ridge’s tokamak, Ormak, was performing well and researchers there saw a D-T reactor as a natural next step. In fact, they offered Hirsch more than he had asked for: they proposed a machine that would reach not just break-even but ‘ignition,’ a state where the heat from alpha particles produced in the reactions is so vigorous that it is enough to keep the reactor running without the help of external heat sources – a self-sustaining plasma. To reach ignition would require very powerful magnets made from superconductors, another area in which Oak Ridge already had expertise.
Seeing all the government money that was being thrown at new energy sources in the winter of 1973-74, Hirsch wanted to speed up his fusion development plan. Instead of building a deuterium-only feasibility experiment by 1980 followed by a D-T reactor by 1987, he proposed that they should move straight to a D-T reactor, starting construction in 1976 and finishing in 1979. Such a timetable would require a much steeper increase in funding as a D-T machine, at $100 million, would be twice the cost of a feasibility experiment.
None of the fusion labs liked this accelerated plan. Princeton didn’t want to get involved in D-T burning yet and the new plan would eliminate the deuterium-only feasibility experiment they had hoped to build next. Oak Ridge, although enthusiastic about D-T, thought the timetable was too short. And Los Alamos and Livermore, which were planning new pinches and mirror machines, feared that a big D-T tokamak would consume all the fusion budget and squeeze them out entirely.
An issue that would play a key role in the move to larger tokamaks was plasma heating – how to get the temperature in the reactor up to the level necessary for fusion. Early tokamaks simply relied on ohmic heating, where the resistance of the plasma to the flow of current heats it up. Using ohmic heating alone, these machines were able to get to temperatures of tens of millions of °C, but fusion would need ten times that much. Theory predicted that as the temperature in the plasma got higher, ohmic heating would get less efficient, so another way of heating the plasma was needed. US researchers were pinning their hopes on neutral particle beams. These were being developed at Oak Ridge and the Berkeley National Laboratory as a way of injecting fuel into mirror machines, but tokamak researchers realised that they might work as plasma heating systems.
Neutral beam systems start out with a bunch of hydrogen, deuterium or tritium ions and use electric fields to accelerate them to high speed. If those ions were fired straight into a tokamak they would be deflected because its magnetic field exerts a strong force on moving charged particles. So the ion beam must first be fed through a thin gas where the ions can grab some electrons, neutralise, and move on through the magnetic field undisturbed. Once in the tokamak’s plasma, the beam gets ionised again by collisions with the plasma ions but because the beam particles are moving so fast when they do collide they send the plasma ions zinging off at high speed, thereby heating up the plasma.
During 1973 a race developed to see who would be first to demonstrate neutral beam heating in a tokamak. The team running the CLEO tokamak at Culham, using a variation on Oak Ridge’s beam system, were first to inject a beam but their measurements didn’t show any temperature rise above the ohmic heating. Princeton’s ATC, using the Berkeley beam injector, came in next and managed to get a modest rise in temperature. Oak Ridge lost the race, but got the best heating results in Ormak. These first efforts had low beam power, typically 80 kilowatts, and temperature gains were small, around 15% above ohmic heating. But within a year ATC was doing better, boosting the ion temperature in its plasma from around 2 million °C to more than 3 million °C. The signs were promising that neutral beam heating would be able to take tokamaks up to reactor-level temperatures. That would really be put to the test in an upcoming machine, the Princeton Large Torus (PLT), which had begun construction in 1972. Designed to be the first tokamak to carry more than a million amps of plasma current, PLT would have a 2-megawatt beam heating system to boost the ion temperature above 50 million °C.
The sensible thing would have been to wait and see how well the PLT worked before embarking on a larger and more ambitious reactor, but Hirsch didn’t want to wait. In December 1973 he called together the laboratory heads and other leading fusion scientists to discuss plans for the D-T reactor. The Princeton researchers were highly critical of Oak Ridge’s proposal for a reactor that could reach ignition. Such a machine would require a huge leap in temperature from what was then possible and would cost, they had independently calculated, four times the allotted $100 million budget. Then the question of timetable came up. Oak Ridge’s head of fusion Herman Postma was asked if his design would be ready to begin construction in 1976, Hirsch’s preferred start date. The reactor’s am
bitious design and superconducting magnets would take some time to get right, so Postma said that he didn’t know. Hirsch was furious.
After a break for lunch, the head of the Princeton lab, Harold Furth, got up and made a surprising proposal. He sketched out a machine that he and a few colleagues had first proposed nearly three years earlier. Since it was difficult to get plasma temperature up to reaction levels, they had reasoned that you could build a tokamak that was only capable of reaching a relatively modest temperature and fill it with a plasma made of just tritium. Then with a powerful neutral beam system they would fire deuterium into the plasma. While there would be no reactions in the bulk of the plasma, at the place where the deuterium beam hits the tritium plasma the energy of the collisions would be enough to cause a reasonable number of fusion reactions. Furth referred to this setup as a ‘wet wood burner’: wet wood won’t burn on its own, but it will if you fire a blowtorch at it.
Such a reactor would never work as a commercial power-producing plant because it could only achieve modest gain (energy out/energy in). Reactor designers had always assumed that beam heating systems would only be used to get the plasma up to burning temperature and then the heat from alpha particles would sustain the reaction. It was never the idea for beams to be an integral part of the reactor. But Furth suggested that this would be a quick and relatively cheap way to get to break-even.
Hirsch gave the two labs six months to come up with more detailed proposals. When those plans were revealed in July 1974, Hirsch had a difficult choice. On one hand was a bold, technologically inventive machine that had the potential to get all the way to ignition in one step, although it came from a lab that was relatively new to the fusion game. The alternative was a much more conservative choice. Princeton’s wet wood burner was not dissimilar to the Princeton Large Tokamak that the lab was currently building. It would not go a long way towards demonstrating how a fusion power reactor would work but Princeton had nearly twenty-five years’ experience of building these things and if Hirsch simply wanted a demonstration of feasibility, this was more likely to give it to him. He was aware that European labs were working together to design a large reactor and Russia had ambitious plans too, so he could not afford to delay – as always, American prestige was at stake. So he played safe and gave Princeton the nod to begin work on the Tokamak Fusion Test Reactor (TFTR) with an estimated total cost of $228 million.
Like the JET design team across the Atlantic, the designers of TFTR didn’t have a lot of information to work with. But unlike Rebut’s daring design for JET, the Princeton team opted to keep it simple. No D-shaped plasma for them; they stuck with the tried-and-tested circular design. It was a trademark of the Princeton lab to keep their devices as simple as possible – the simpler they are the faster they can be built and the easier it is to interpret the results.
In 1975 the PLT began operation and produced some impressive results using neutral beam heating, raising the ion temperature to 60 million °C. Theorists had predicted that a beam impacting with the plasma would cause instabilities, but these failed to materialize. Altogether the Princeton researchers were happy with the results, but there was one thing that caused some concern: confinement time got worse the more beam heating was applied. Although this cast a small dark cloud over the future of TFTR, in the rush to finish its design and get construction started there was little time to consider the issue.
Ground was broken for the new machine in October 1977 and it was scheduled to start operating in the summer of 1982. Much of the work was parcelled out to commercial contractors, a significant fraction was built by other government labs, and the rest was done in-house by Princeton staff. Like almost any science project of this size, there were numerous technical headaches along the way. New buildings had to be built with thick concrete shielding to protect people from the neutron flux when D-T reactions were taking place. The power supply system, involving the usual giant flywheel, proved unreliable and had to be virtually rebuilt, which bankrupted the contractor involved. For the first time, computers were bought to help analyse results from the reactor but, being a new technology, it took some years to get them working properly. As the summer of 1982 passed and moved into autumn the diagnostics systems for monitoring the reactor were nowhere near finished. TFTR project director Don Grove was determined for TFTR to get its first plasma before the end of the year, and that meant before Christmas. In desperation Princeton staff took over the installation of diagnostics from the contractor on 12th December and worked around the clock to get it in place.
By 23rd December they had installed the very minimum set of diagnostic instruments. The cabling that connected these instruments, via a tunnel, to the nearby control room had not been installed so they set up a temporary control room in the reactor building. A thousand and one things had to connected, checked, rechecked and tested. The Princeton researchers had never built such a large and complex machine before and everything was new and unfamiliar. The sky darkened and the team worked on into the evening. Grove decreed that whether they finished or not they would stop working at 2 a.m. The clock ticked past midnight into the early hours of Christmas Eve. They were very close, but not there yet. No one admits to knowing how it happened, but the clock on the control room wall mysteriously stopped at around 1.55 a.m. Since it was not officially 2 a.m. yet, the team kept working.
Staff celebrate first plasma in the Tokamak Fusion Test Reactor at Princeton on Christmas Eve 1982. Note the clock, stalled at 1.55 a.m.
(Courtesy of Princeton Plasma Physics Laboratory)
About an hour later they attempted their first shot. There was a flash in the machine and it was done. Grove ceremoniously handed a computer tape to Furth containing measurement of the first plasma current and Furth handed over a crate of champagne. The giant machine was duly christened and everyone went home for Christmas. TFTR would not produce another plasma till March as the team had to finish installing all the things that had been left out in the rush to meet the deadline.
Despite the head start that the JET team originally had in designing their machine, the delay over deciding its location put them firmly in second place. The ground breaking ceremony for TFTR in October 1977 was only days after the end of the hijack drama at Mogadishu. Construction of JET didn’t begin until 1979 so the Culham team were two years behind their US rivals, but they soon made up lost ground thanks to the steely determination of Rebut. The Frenchman did not, however, get the job of JET director. Euratom passed him over and gave the job to Hans-Otto Wüster, a German nuclear physicist. Wüster had made a name for himself as deputy director general of CERN during the construction of its Super Proton Synchrotron. Although not a plasma physicist, Wüster had an easygoing style that allowed him to talk with equal ease to construction workers and theoretical physicists – very different from the blunt Rebut. But underneath the charm he was an adroit politician, something that proved very useful in keeping JET on track. Rebut, however, was knocked sideways by the decision and considered resigning. But the job of technical director still allowed him to supervise the construction of the design he fought so hard to bring to life, and Wüster gave him complete freedom in the construction.
Palumbo also did his bit to set JET off on the right track. Although 80% of JET’s funding came from Euratom – with another 10% from the UK and the remaining 10% split between the other associations – he insisted that JET be set up legally under European law as a Joint Undertaking, in other words an autonomous organisation with its own staff of physicists, theorists and engineers, and at arm’s length from interference by Euratom and the national labs. Even at Culham, where JET was sited, it remained separate from the national laboratory – it had its own buildings and its own staff. Culham researchers worried that JET would suck all the vitality out of their own lab and tended to view the JET researchers as a superior bunch who kept themselves to themselves. There was one issue that caused more than cool relations: pay. The JET undertaking instituted a system of secondment i
n which researchers from Euratom association labs would come to JET and work there for a while. During these sojourns they enjoyed the generous rates of pay typical of people working for international organisations. But the Culham staff seconded to JET continued to get the local salary rate which was less than half what their overseas colleagues were getting. This disparity caused huge resentment among British employees at JET, forcing them eventually to take the matter to court.
Just as in the building of TFTR, there were numerous technical hurdles to overcome in JET’s construction, but Rebut ruled with an iron hand and allowed very few changes to the design. On 25th June, 1983, almost exactly six months after TFTR produced its first plasma, JET fired up for the first time. ‘First light, a bit of current,’ the operator wrote in his log book. It was only a bit, just 17 kiloamps, but the race between the two giant machines was on. Rebut had made a bet with his opposite numbers in Princeton that, even though JET started later, it would achieve a plasma current of 1MA first. The loser would have to pay for a dinner at the winner’s lab and bring the wine. JET duly passed the milestone first, in October, and so the two teams dined together, at Culham, drinking Californian wine.