High-level Risk Safety at Sellafield NS 20 Jun 98 13
MOST of Sellafield could be closed down if it continues to fail to make its high-level radioactive waste safer. Attempts to store the waste in glass at Britain's main reprocessing plant are being thwarted because vital equipment is corroding and failing, faster than expected, according to documents leaked from the UK Atomic Energy Authority (AEA). High-level waste, which comes from the reprocessing of spent fuel from nuclear power stations and military reactors, is the longest-lived, hottest and most dangerous of all the radioactive debris produced by the industry. About 1000 cubic metres of it is stored as a liquid in 21 constantlv cooled tanks at the Cumbrian site. Two vears ago, a study warned that an accident in the tanks could lorce the Irish -ox,ernment to evacuate half a million people from Dublin (This Week, 15 June 1996, p 6), and a scientific report highlighting the risk of an accident is due to be published next week by antinuclear local authorities. Because of the dangers of keeping the waste in the cooling tanks, the government's Nuclear Installations Inspectorate urged British Nuclear Fuels (BNFL), which runs the Sellafield plant, to make the waste more containable by solidifying it into glass blocks for permanent storage "as soon as practicable". A crucial part of this process is the vitrification plant that BNFL has been operating at Sellafield since 1990. But internal memoranda from the AEA, whose waste is treated by BNFL, reveal that the vitrification plant suffers from "continued poor performance". Crucibles in which class is melted have averaged only 900 hours of use, instead of the 2500 hours expected. This is due, says the AEA, "to a combination of fatigue, corrosion and creep". Replacin,the defective crucibles is also problematic. To remove them from the plant then, must be cut into pieces, but according to the AEA the saws bought by BNFL for the task are "ineffective". BNFL is now suing the French company that supplied the saws. The leaked AEA memos warn that failure to sofidify enough high-level waste could prompt the nuclear inspectorate to prevent anv more from being separated by withdrawing permission for reprocessing spent fuel. This would shut down most of Sellafield Gordon Thompson from the Institute for Resources and Security Studies in Massachusetts, the author of the upcoming local authorities' report into the consequences of an accident in Sellafield's waste tanks, points out that delays in vitrification will increase the amount of waste stored in the crucibles. "This will exacerbate the danger," he says. "Vitrification is dramatically safer." Rob Edwards
Meltdown Contamination in Europe's Nuclear reprocessing
NS 13 Jun 98 13
RADIOACTIVE contamination up to 25 times the intemational safety limit has been found on the outside of spent nuclear fuel flasks travelling between Britain and mainland Europe. Scandals about badly contaminated flasks in Cermany and France have already paralysed the nuclear reprocessing business in Europe. Last month all rail shipments of spent fuel from German, Swiss and French nuclear power stations to the reprocessing plants at La Hague in Normandy and Sellafield in Cumbria were suspended indefinitely. The British findings hei-hten the crisis for the industry. In addition, the British govemment announced last week that Europe's only other reprocessing plant, at Dounreav in Scotland, woldd close after 2006 because it is not economic. "These are very big nails in the coffin of reprocessing," says Mycle Schneider of the World Information Service on Energy, an antinuclear group in Paris. British Nuclear Fuels (BNFL) had already admitted in May that since 1995, six flasks out of 312 at its Dunkirk rail terminal had been contaminated with "typically between 10 and 20 becquerels per square centimetre". The safety limit recommended by the International Atomic Energy Agency is 4 Bq/CM2. Last week, BNFL told New Scientist that one flask that left Germany in March this year had contamination of 100 Bq/cml by the time it reached Dunkirk. In addition, over the past three years inspectors had found eight flasks at Barrow docks that had exceeded the safety limit. The worst had contamination of 60 Bq/cml. BNFL insists that the contamination does not pose a risk to health, and says it is caused because the flasks "sweat" in transit. While they are loaded at the reprocessing plant, the paint and metal on their outsides absorb radioactivity that is difficult to remove by cleaning. Isotopes then seep from the surface during transport. Last month, the German environment minister Angela Merkel banned the transport of spent fuel after discovering that 11 out of 55 flasks sent by rail to France in 1997 were contaminated bv up to 13 400 Bq/cm2. She said that the failure of nuclear companies to report breaches of the safety limit to her ministry over the past ten years was "completely unacceptable". The ban is likely to remain until federal elections in September. The French rail company, SNCF, also suspended all spent fuel shipments last month after the French regulatory authorities discovered that a quarter of the flasks and 35 per cent of the rail wagons that arrived at the Valognes rail terminal near La Hague last year were contaminated beyond the safety limit. As in Germany, nuclear companies had failed to report breaches to the govemment for the past decade. Breaches of the safety limit in Britain have also not been formally reported to the Department of Environment, Transport and the Regions, which oversees nuclear shipments. "We do not inform the regulatory authorities because there is no requirement for us to do so," says a BNFL spokesman. Rob Edwards
Making Waves Comprehensive Test Ban Treaty and Seismic Testing
NS 13 Jun 98 18
BEFORE dawn on 11 May, a seismograph in a research institute outside Washington DC recorded a disturbance deep in the Earth on the other side of the world. Over 'the next feiv minutes, dozens of other seismographs all over the planet recorded the same event and transmitted their data automatically to the institute-the prototype International Data Center (pIDC). A computer analysed the signals and gave its interpretation: an "event" of magnitude 5.0 on the Richter scale under Rajasthan in India. Later that morning, seismologists at the pIDC studied the signals and recognised the event as a nuclear test. The job of the institute is to test the technology for detecting nuclear bomb tests around the world. From next year its successor, the real international Data Center in Vienna, will be charged with policing the Comprehensive Test Ban Treaty (CTBT), which outlaws all nuclear explosions. Whether it can do so adequately has become the subject of a fierce debate. For it emerged that the seismic monitoring network under trial by the plDC failed to diagnose India's tests accurately.. In particular, it failed to detect the second explosion, which New Delhi claims yielded between 0.2 and 0.6 kilotons. The reliability of the monitoring system is vital to the success of the CTBT. The treaty is opposed by some in the US who say that testing is necessary to maintain a nuclear deterrent. They claim that the flaws in the monitoring system prove the CTBT cannot work. Such accusations are "already damaging prospects for [US] ratification of the CTBT", says Christopher Paine of the Natural Resources Defense Council, an American pressure group that supports the treaty. The CTBT has been signed by 149 countries since it was agreed in 1996, but to come into force it must be ratified by all 44 nations in the UN Conference on Disarmament that have nuclear reactors. So far only 13, including two "bomb" states, Britain and France, have ratified it. US ratification, as well as being essential for the treaty to come into force, is crucial to encouraging others to ratify. The CTBT calls for four separate global networks to listen for nuclear explosions.
None is yet complete (see Table), but the seismic network-which includes stations that automatically, transmit data as well as some that can Ie used as necessary-is the most developed. In addition, there will be hvdroacoustic monitors to listen for undersea tests, atmospheric sensors to detect the radioactive particles and rare gases, such as xenon, that are released by nuclear explosions, and infrasound receivers to listen for near-surface explosions. All are being tested by the FIDC. The technology at the centre of the current dispute is seismic. Frank Galley, former US assistant secretary of defence under Ronald Reagan and now head of the Center for Security Policy an anti-disarmament think-tank in Washington DC, argues that because the seismic network failed to pick up all of India's tests, "nuclear testing can be conducted in ways that will be unverifiable, if not undetectable". Although India said it exploded 60 kilotons in its first test, the seismic stations recorded only 25 kilotons. However, Roger Clark, a seismologist at the University of Leeds, found that when data from 125 stations-closer to the number required bv the treaty-are taken into account, the estimate is nearer to 60. Moreover, the tests could have had their seismic signals muffled, possiblbly "decoupling" the devices-suspending them within caverns in the ground or burying them in sand. In theory, 10 kilotons of explosive force can be completely hidden in this way. Clarke notes that there could be peculiarities in Rajasthan's -geology that weakened the signals. That problem will diminish with experience of monitoring earthquakes, he says. "The more stations we have feeding seismic data into the system, the more details like that we will understand."
Seismologists point out that the network detected both of Pakistan's tests. But Galley's team maintains that the waveforms that allow seismologists to distinguish earthquakes from explosions are not always clear. "If the Indians hadn't announced their tests, CTBT supporters would have claimed the seismic events that were recorded were natural." Nonsense, says Clark. The waveforms of the explosions in India's first test and those in both of Pakistan's were clearly caused by explosions. Fil Filipkowski, spokesman for the pIDC, says the system performed "admirably".
But these three events were over 4 on the Richter scale. Smaller events are not always so clear. Frode Ringdal, head of Norway's seismic detector and a leading developer of the global network, admits that seismic data alone will not be enough to tell whether small or muffled events are natural or not. "That is why we need different kinds of monitoring," says Suzanna van Moyland of the Verification Technology Information Centre in London. T'he Clinton administration says verification of low-level nuclear tests relies on being able to spot preparations at a test site. Satellite images of drilling or other suspicious activities should provide independent backup for seismic data. This is why there was an outcry when it emerged that India had successfully hidden preparations for its tests from American SPY satellites. Yet the CTBT does not formally include verification by satellite. The scientists who designed the four monitoring networks over the past two decades concluded that these should be sufficient. And they appear to have been vindicated by a seismic event off the Russian Arctic island of Novaya Zemlya, previously a Soviet nuclear test site, in August last year.
The magnitude of the event was 3.5. This, says Ringdal, was "below the size where we can so far reliably tell an explosion from a natural event". A nuclear device of 0.05 kilotons-or a decoupled, larger onecould have given such a signature. And US spy satellites saw suspicious activity at the old test site. The US asked Russia to explain.
Russia denied testing, but the US has not formally abandoned its suspicions. However, the evidence seems to support the Russians' denial. The seismic network located the epicentre of the event to within 10 kilometres of a point 50 kilometres southeast of Novaya Zemlya, 400 metres under the Kara Sea. American spy satellites did not detect any evidence of drilling there, or of tunnelling from the land. Moreover, the seismographs picked up aftershocks, which follow earthquakes but not bombs. In addition, two other monitoring networks picked up nothing from the blast. There were no radionuclides sensed at any atmospheric stations-although there are not yet many of these in place. But the really conclusive evidence, says Paul Richards of the Lamont-Doherty Earth Observatory at Columbia University in New York, is that no hydroacoustic listening posts picked up the boom of a seabed explosion. Gafffiey's team argues that the seismic ambiguity proves CTBT monitoring does not work. But Richards says that when other data were included, the system worked "magnificently". 'A verification system based on so many kinds of information can be powerful," says van Moyland
Moreover, for the CTBT to work, the network need not provide absolute proof. The US Department of Energy concedes that "in many evasion scenarios ... the monitoring systems will not provide conclusive evidence" of a nuclear test. In these cases, treaty members can demand on-site inspections. The DoE is testing technologies for this purpose, such as measuring gases in soil. To its supporters, then, the Indian and Pakistani tests showed that the CTBT can be verified. But they point out that governments will only ever know for sure if they put their faith in the scientists and ratify the treaty. The full network, they stress, will be much greater than the sum of its parts. Debora MacKenzie