No-one has ever tackled a waste disposal job like the one facing the Dounreay clean up team, says Trevor Jones, the senior project manager for waste retrieval at the shaft and a nearby silo.
The heterogeneous nature of the waste in both shaft and silo, and the way it is jumbled together, makes the task especially difficult, he says. Large blocks of concrete, sealed drums, small flasks with liquid in them, and lengths of scaffolding pole are all surrounded by a 'brown gooey' sludge of metal turnings and dust - including some 10kg each of plutonium and uranium in the shaft - corrosion products and organics. There are even plastic bags, timber and other light debris floating on the surface of the water. A wide range of techniques will be needed to handle the very different items.
'Most schemes handle similar items like fuel rods, ' Jones says, 'but here there is a mix of things that might well have jammed together.' For example the scaffold poles, used to poke down the waste heap and then, being radioactive themselves, thrown into the shaft, might get jammed against the rough rock walls or caught in the wire mesh which lines most of the lower part of the shaft.
Or they may simply be locked together.
To devise a scheme the team has scoured the world, from the US to Chernobyl, and visited places like Vandellos in Spain 'the only scheme remotely like this'. But even there the challenge was only a silo handling graphite.
Emptying the silo is an immense challenge, but pales in comparison to that posed by the shaft. 'There is no scope there to move the waste around in order to pick out items in the best order, ' says Jones.
Despite considering radical options, like carving the entire shaft out of the rock and jacking it up, or ground freezing the area, the solution agreed is dry sorting of the material working with remote handlers from a sealed cell above the shaft. The same principle will be applied to the concrete-lined silo.
Detail design will continue for the next couple of years or more. Construction is likely to take four years with a two year commissioning period. 'We think it will then be eight years of retrieval work and another four for decommissioning, ' says Jones.
Both the shaft and silo pit will have handling buildings constructed above them, each with a heavily lined hot cell immediately over the openings. No-one can go into these. Other 'cells' around the hot room will allow maintenance and control with some human access, depending on what is in them at the time.
For the silo, mast mounted remote handlers will be used to sort and pick the waste from inside, lifting it out through the hatches in the silo roof into 2m 3 rail mounted skips which will take it to another reprocessing building alongside.
Liquid from both waste facilities, both water and sludge, will be brought out with a long reel hose with different pump attachments at the end. The liquid has to go, both for its own treatment and to allow visibility for the handlers.
The shaft is more difficult, because unlike the silo it was not purpose built. The silo is relatively shallow while there is 60m depth of waste in the shaft. 'And the silo had a limit on the waste size because the openings were just 700m square.' Not only that, there is an unknown amount of waste in the stub tunnel at the bottom.
The shaft building will be bigger and more complex.
The hot cell will house a spider type grabber and also a platform-mounted remote handler that will be able to move and manipulate waste, possibly even cutting it if very large or jammed pieces are found. 'We may have to do some rock bolting on the way, ' says Jones. Grab and platform may be lowered into the shaft one above the other.
A cell above will have winches and machinery, and other cells alongside will be used for maintenance work on the equipment. Inside the hot room trays will store awkward items known to be at the top of the shaft, to await treatment.
Once in the skips the waste will be lidded and passed along a concrete lined tunnel to the treatment plant. Experiences from the Chernobyl plant have given the team a concept for an open sorting area where waste is first tipped on to vibrating screens to remove string and plastic, and then washed.
Remote mini-excavator type arms will pick off any 'specials' - particularly radioactive or awkward objects.
Waste then goes into toothed shredders of the type now current in scrap yards.
'The modern ones can tear apart an engine block, ' says Jones, 'and this helps solve a key problem - how to open the sea disposal drums that were put in the shaft.' As with much of the thinking the aim is to find commercial equipment that can be used - perhaps with modification - rather than 'going down the R&D route' of purpose made handlers. Keeping the project budget within bounds is part of the challenge.
Shredded waste will be separated into three types - ferrous, aluminium and organics - by automated sorters and passed into temporary storage bins.
'[Waste storage company] Nirex specifies exactly the proportions of each you are allowed in waste, ' says Jones, 'so we will pick and mix the amounts when filling the final disposal drums.' Waste is 'supercompacted' and embedded in concrete in 500 litre approved drums while homogenised sludge goes into 200 litre drums.
These will then be stored on site until a national repository is available. But that is another issue.