The Nuclear Decommissioning Authority (NDA) has been making extensive use of drones as it works to clean up the toxic legacy of the Sellafield site and other decommissioned nuclear sites — a task it earlier described as “the largest, most important environmental restoration programme in Europe”.
The NDA’s role includes cleaning up over 60 years of bad practice and poor storage choices for nuclear waste at 17 locations across the UK. In an R&D report this week, the agency detailed its growing use of technology such as drones, machine learning and 3D imaging technology to help it tackle the challenge.
(Photos leaked from Sellafield in 2014 showed cracked concrete tanks holding radioactive water, seagulls bathing on the water, a mess of discarded items on elevated walkways, and weeds growing around the tanks)
The NDA’s efforts to clean the UK’s nuclear legacy will take over a 100 years and cost the tax payer more than £100 billion. Last year the authority invested approximately £90 million into R&D in a bid to help drive those costs down.
Some of the more difficult parts of the decommissioning lie ahead; such as the removal of nuclear waste from open ponds at the controversial Sellafield site, as the NDA’s notes in its Draft Business Plan for 2020 to 2023: “After years of painstaking preparations on the 60-year-old facilities and complex engineering solutions, work will begin in 2020 to retrieve waste from the 2 silos and bulk sludge from the site’s 2 ponds.”
The UK currently has 15 operational nuclear facilities that generate roughly 21 percent of the country’s electricity. However, nearly half of these plants are expected to be retired by the year 2025, creating more work for the NDA.
Nuclear Decommissioning Authority Technology
The R&D report, published this week, particularly drone use on sites, as the technology, as carrying capacity and battery life continue to improve.
Due to drones’ versatile nature, the NDA has deployed them across its sites in order to complete tasks such as the inspection of tall chimneys, pipelines, and roofs. The drones have also been instrumental in collecting high quality imagery and sensors readings from areas containing radioactivity and have even been used to find previously unknown areas of radioactivity.
“The information has eliminated working at heights and radiation dose to operators. Compared to traditional manual methods, there have been savings in both time and money” the authority said.
Machine Learning
The NDA has also co-funded several projects to increase the level of technology and automation used on site with the aim of chipping away at the daunting clean-up costs. One such project, co-funded by Sellafield, aims to create a machine learning model that can be used to glean greater insight from the health, safety and environmental data gather on site in order to predict safety issues.
In the last 15 years Sellafield has produced over 100,000 safety reports and observations. Currently all of this data is stored within Sellafield’s internal document management systems and is inspected and processed manually.
The facility is in the middle of a Proof of Concept project in order to better understand how machine learning could be used to interrogate all the safety reports that can be made available as comma separated variable (csv) files.
Two key benefits of such a system, as Sellafield note, are “improved analysis of safety reports and prediction of issues.” and “reduced time spent reviewing data, enabling skilled Health, Safety and Environmental professionals to focus on value added mitigation activities.”
3D Imaging
At the Sellafield site there is a significant amount of spent nuclear fuel held in different storage conditions. One is spent oxide fuels that have been placed in interim storage while a decision is made on whether to declare them as waste that will be buried underground.
The NDA wants to understand how these wet fuels will react over time and as such are investigating how 3D imaging technology could be used to monitor the waste’s condition. This will help spot material changes such as corrosion, cracking, voids or cladding stress/failure.
The NDA notes that: “Methods considered included X-ray, neutron, gamma, ultrasound, muon and atom probe technologies. The capabilities of differing techniques were assessed, to understand how they might support R&D work on spent AGR fuels, whether they were readily available in the UK and suitability for a nuclear environment. With some techniques, it is possible to virtually “walk” inside the material, find interesting sections of the image and inspect it more closely by zooming in.”
Read More: Robotics Helping with Sellafield Clean Up “Headache”
