Going underground

Imagine building a structure that has to last 20 times as long as Egypt’s pyramids. It has to be deep underground, hollowed out of the bedrock. It must withstand the crushing weight of glaciers and the rise and fall of civilisations. And it must be impregnable because it will be filled with millions of tonnes […]

Imagine building a structure that has to last 20 times as long as Egypt’s pyramids. It has to be deep underground, hollowed out of the bedrock. It must withstand the crushing weight of glaciers and the rise and fall of civilisations. And it must be impregnable because it will be filled with millions of tonnes of some of the most lethal substances in existence.

The structure is the proposed deep repository for the UK’s intermediate- and high-level nuclear waste (ILW and HLW). Although the description above sounds like apocalyptic science fiction, it is now being planned in an unprepossessing office building just outside Oxford.

It is almost certain that the UK’s nuclear waste will be stored underground. After two years of deliberations, the Committee on Radioactive Waste Management recommended disposing of ILW and HLW in a repository excavated deep in the bedrock. Its report is with the government, which means engineering, technological and sociological tasks are just beginning.

In other countries with nuclear waste stockpiles, deep geological disposal is already the method of choice. In the US, work has begun on the Yucca Mountain repository in Nevada. In Europe, studies are well advanced for repositories in Finland, Sweden, France and Spain. The UK lags behind, but with official approval on the horizon, the planning can begin.

The task of designing the repository will fall to NIREX. Based in Harwell, NIREX has been researching nuclear waste disposal options since 1982. ’Our mission is to advise on safe, environmentally sound, publicly acceptable options for long-term management of radioactive materials,’ said Bruce McKirdy, NIREX’s director of science and technology.

The problem is the safe disposal of the UK’s wastes, spent nuclear fuel and separated uranium and plutonium from fuel reprocessing. The UK has more waste, and of a wider variety, than most other countries, as it was a pioneer of nuclear research. ’We’ve got so much exotic intermediate level waste from reprocessing at Sellafield, from weapons programmes, from the very early reactor research, that we have quite an interesting cocktail of wastes,’ McKirdy said. ’The legacy ponds and silos at Sellafield are crammed with all kinds of things. Some of them, people don’t have an idea of what’s in there.’

NIREX estimates that by 2012, when the current fleet of nuclear reactors have been decommissioned, there will be 478,000 sq m of higher-activity wastes, of which ILW will comprise 350,000 sq m.

In other countries with nuclear programmes ILW tends to be a small fraction of the total waste; in Finland it is disposed of in shallow repositories beneath nuclear power stations. But in the UK, ILW makes up the majority of the waste inventory. Although it accounts for less than five per cent of the total radioactivity of the UK waste, it contains long-lived, very toxic and unstable materials and must be stored for tens of thousands of years before its radioactivity has subsided to safe levels. Also, the ILW is a mixture of unstable substances. As well as being radioactive, it will break down and rot, producing various radioactive gases.

The HLW is more radioactive than the ILW; it produces heat and will be hazardous for more than 100,000 years. But it is a physically and chemically stable form of waste and well understood.

The UK therefore needs two types of repository, said McKirdy: one for the ILW and another for the HLW and spent fuel. These could be part of a single complex, which would reduce the total cost considerably, but could also be separate.

The NIREX plan is known as a generic concept: this means it will serve as the basis of the structure, with adjustment to take into account the type of rock at the host site. The location is yet to be decided, so the generic concept allows planning and research to begin soon.

McKirdy said that the concept for both the ILW and HLW repositories is underpinned by two design philosophies: they use multiple barriers to prevent the radioactivity from reaching the surface and allow for phased disposal of the waste. ’Phased geological disposal is a series of reversible steps rather than going straight into it in one go, sealing it up and walking away,’ he said. ’It allows you decision points, the option to keep it open and take everything out if that’s what’s decided.’

The ILW repository will be 300-1,000m underground, deep enough to be safe if the top layers of rock are scraped away by glaciers, said engineering manager Brendan Breen. ’You immobilise the waste, encapsulate it in cement and put it in stainless steel drums or boxes — that’s the first barrier. The next stage is geological isolation. In terms of security, it’s better to have half a kilometre of hard rock in the way than to have it in a shed.’

The waste vaults will be 16m x 16m in cross-section, and 300m long. The waste will be emplaced or removed by tried-and-tested technology. ’We have a design philosophy to base everything on proven technology and, if possible, not go for novel techniques,’ McKirdy said. ’So we’ll know that, having got the stuff in, we can take it out again with the same system.’

The emplacement systems will be based on crane technology and much of the machinery will be located outside the radioactive zones. This will ensure it can be maintained and replaced without using remote handling techniques.

Once the ILW is emplaced, the vaults are likely to remain open possibly for as long as 300 years to allow for monitoring but eventually they will be sealed by backfilling the vaults with an alkaline cement. This, McKirdy said, reduces the solubility of some of the radionucleides in any groundwater within the rock.

’The backfill includes a lot of sorbing surfaces,’ he said, ’and it’s quite porous, which ensures there’s an escape route for any gases produced by the waste.’ The radioactivity will not stay entirely within the containers. ’Something like 95 per cent of the radioactivity never gets out of the drum. About 99 per cent never gets out of this near-field; one per cent gets into the surface but very little of that gets back to the geosphere because it gets sorbed on to rock or cement surfaces.’ Even the backfill is a reversible process, he said. ’The cement isn’t a strong as structural cement; it’s quite soft. You can cut it with a water jet but it would be a major task.’

The strategy for HLW and spent fuel is different. The system, developed by Sweden’s nuclear waste management body SKV and now being co-developed with Finland, also uses a multiple barrier concept. The waste is placed inside a cast-iron container which is sealed inside a canister made from five cm-thick copper. It is then placed in a hole drilled into the bedrock and packed around with bentonite, a clay that swells on contact with water. This seals the canisters in tight and allows nothing to escape. Like the ILW, the UK’s HLW repository will be 300-1,000m below ground.

The SKV system was designed for conditions in Scandinavia, where the bedrock is fractured all the way to the surface. Most of the UK bedrock lies beneath sedimentary layers that are not fractured, McKirdy said; these act as an extra barrier to prevent radioactivity from reaching the surface. The engineers developing Sweden and Finland’s repositories must assume water will percolate through the rocks and their containment systems need high integrity.

Timo Äikäs, vice-president in charge of engineering for Posiva, Finland’s HLW repository programme, said copper was chosen because metallic copper is abundant in Sweden and Finland’s bedrocks, proving the metal is suitable. ’We’ve done several years of of lab work and long-term corrosion testing and we’ve determined that 1.5 cm thickness of metal is enough for a corrosion barrier, but we’re still going with five cm.’

The conditions in Scandinavia are likely to include ice-ages over the 100,000 year period during which the waste is hazardous and this has also been researched, Äikäs said. ’We conducted pressure-testing of the canisters, where we pressurised the full-scale canister in conditions that are similar to those under an ice-sheet. We’re confident that the canisters will not rupture, even if the rocks around them were to crack.’

For UK conditions, the canisters are over-engineered. ’These are currently £70,000-£80,000 apiece,’ McKirdy said. ’We have this huge safety margin we don’t need, but we get the SKV design for a few tens of thousands a year in royalty fees. For that, we’re tapped into tens or hundreds of millions worth of R&D that’s already been done.’

Both ILW and HLW repositories in the generic concept have common features: the vaults for the waste; shafts and access routes for the waste and staff; and underground roadways that double as ventilation. The ILW repository also has two types of vault — one for waste that does not require extra shielding and can be emplaced using forklift trucks, and another where shielding used for transportation is removed from the waste before emplacement using remote-controlled handling equipment.

The concept will be modified to the type of rock at the host site — and siting the repository is possibly the most difficult part of the project. In the 1980s a British Geological Survey of UK geology concluded that 30 per cent of the UK had suitable rocks to support a repository. ’For geological isolation, you make sure that your repository is in an area where there is no water flow, physically very stable, at an appropriate depth to ensure you get very long return times before you get any activity back to the surface,’ saoid McKirdy.

Later in the decade NIREX did preliminary studies at Sellafield, one suitable site, and applied for permission to excavate an underground laboratory, at the depth of the proposed repository. This was refused after a long, acrimonious public inquiry. This time, NIREX proposes to use the method which is proving successful elsewhere in Europe — identify all the regions with suitable geology; where transportation links for the waste already exist or could be constructed; and where land is available, and seek a volunteer community in one area.

’You can’t have national policy foisted onto a community that doesn’t want it,’ said McKirdy. ’Somebody has to stick their hand up and say “yes, we’ll have it”. And there will probably have to be something in it for them other than the jobs that go with the facility.’ The community would have a say in aspects such as the design of the surface facilities and the level of retrievability of the waste.

The suitable geologies in the UK cover a wide variety of rock types, however. Some are granites, others hard rocks or clays. David Savage, chief geologist to UK environmental risk consultancy Quintessa, said each has pros and cons. ’The rest of Europe is moving very much towards clays, if the country has them as a choice, as they are a more favourable host rock than what we’d call fractured hard rocks,’ he said. One reason for this is because water diffuses through clay, leaving mineral traces behind it.

’That gives you the idea that the main mode of transport through these clays has been diffusion for millions of years and gives you great confidence that it’s going to be the main transportation method through the future. You can’t do that with fractured hard rock, because there’s no preserved history of water flow.’

However, clays have a drawback. Retrievability is easier if the waste is stored in large vaults but these are much easier to build in hard rocks, Savage said. ’In clay, they’d have to make much smaller diameter tunnels. That would need a wider area for the repository, which might impose constraints in terms of land ownership.’ Clays and hard rocks are both available in the UK — clays mostly around southern England and the hard rocks in the north and Scotland.

McKirdy and Breen say much work is still required. ’The key to characterising the site is getting information on and fully understanding the the groundwater regime in the region and what’s forcing it to be the way it is,’ McKirdy said. ’You have to understand the speed the water is going and have confidence that what you’re doing won’t perturb the system.’

Co-location of the HLW and ILW repositories will also require a great deal of research. The two would have to be separated by a reasonable distance underground, to ensure that even the small amounts of water flowing through the alkaline cement ILW near-field would never intersect with the HLW store while the waste was still hazardous. ’The alkaline plume would need to be kept away from the HLW, because it could have an adverse effect on the bentonite and vitrified waste,’ McKirdy said.

The cost saving would be considerable. The estimated costs for an ILW repository is £6.1bn, said McKirdy, and for a separate HLW repository £5bn. Locating them together would cost about £10.1bn — a £1bn saving.

NIREX thinks site characterisation could not begin until next year at the earliest and is unlikely to be complete before 2020. Construction is likely to take another 20 years, with emplacement of waste beginning no earlier than 2040. Different techniques would be used to excavate the vaults depending on the rock at the site, Breen said, but standard tunnelling techniques will be used.

’This isn’t like any other construction project,’ said McKirdy. ’You have to continually check the site back against the concept, making sure that nothing you find would have an adverse effect on the models, constantly checking back against the geology and chemistry.’

In Finland, meanwhile, work is about 20 years ahead of the UK. A site has been selected, at Olkiluoto in the southwest of the country, which was one of six volunteer sites investigated by Posiva. Construction of a rock characterisation facility, 420m below the surface, is well under way. ’We should reach that level by 2009,’ said Timo Äikäs. ’If everything goes well, and the site meets all the requirements of safety, then it will become one of the accesses for the repository.’ Rock characterisation facilities are also being built in France and Sweden.

Filling the sites could take 60 years. but once sealed, they are designed to be be entirely passive. The local community could request continued monitoring, said McKirdy, but the concept is designed so that this should be unnecessary. ’You can walk away from the thing. The whole system must be such that it doesn’t require any more intervention.’

It is from this point that the calculations of the engineers, geologists and chemists will face their sternest tests. How can they be certain that the repository will remain safe?

’The Geological Society was asked that,’ said McKirdy. ’They said that, for the earth sciences, hundreds of thousands of years isn’t a particularly long timescale. Rates of change are very, very slow — it would be hundreds of millions of years before you could expect to see things happening.’

The NIREX research takes account of societal change — studies are determining the best ways of recording vital information on the repository that will survive any political upheaval.

The Egyptian Pyramids are 5,000 years old and the neolithic monuments of the UK some thousand years older — and they were produced by a civilisation of which we know barely anything. Although the reason for repository is controversial, and no signs should remain on the surface, it could yet become one of the engineering wonders of the world.