Decommissioning inspection is not construction inspection done later. The assumptions that make new-build NDT straightforward — known material, known history, clean access, a drawing that matches what is in front of you — are exactly the assumptions that fail on a plant being taken apart.
We work across the West Cumbrian nuclear estate from our Lillyhall facility, and the questions we get asked on decommissioning scopes are consistently different from the ones we get asked on fabrication. This is what those differences are, and what they mean for scoping a job.
You are inspecting something nobody designed to be inspected
A weld in a new pressure vessel was made to a qualified procedure, by a qualified welder, in a position chosen partly so it could be examined. A weld in a 1960s plant was made to whatever was normal at the time, and the surrounding structure was built around it afterwards and gone through decades of potentially undocumented changes.
The practical consequences arrive early. Surface condition is poor — paint, lagging, corrosion product, sometimes decades of it. Access is whatever the plant left behind. Original material certification may not exist, or may not match what is actually there. Weld caps were never dressed because nobody expected an ultrasonic probe to need to sit on them.
This changes the order of work. On fabrication scopes the inspection is the job. On decommissioning scopes, surface preparation and access provision are frequently the larger part, and pretending otherwise produces a quote that bears no relation to the eventual cost. We would rather survey first and price second.
Material identification comes before method selection
You cannot select an ultrasonic technique without knowing the material — velocity, attenuation and grain structure all change the answer, and austenitic and duplex materials behave very differently from ferritic ones under ultrasound. On legacy plant, "it should be 316" is not a basis for a procedure.
Positive material identification therefore often runs ahead of everything else on a decommissioning scope. Handheld XRF confirms the alloy on site quickly, which is usually sufficient to select the inspection technique and to make waste-routing decisions.
One limitation is worth stating plainly, because it catches people out: XRF measures elemental composition but does not measure carbon. It cannot distinguish 304 from 304L, or 316 from 316L. Where the carbon grade genuinely matters — and in sensitisation and stress-corrosion contexts it often does — optical emission spectroscopy or laboratory analysis is required. Any report that claims an L grade from XRF alone is overstating what the instrument did.
Dose is a design constraint, not a footnote
On a conventional site, inspection method selection weighs capability against cost. On an active nuclear site it also weighs capability against dose uptake, and that reorders the priorities.
Techniques that reduce time-at-component become attractive even where they cost more. Encoded phased array that captures a weld in one pass and defers analysis to an office beats a manual technique that keeps a technician in the area for an hour. Remote visual inspection through an access port beats opening a system. Digital radiography with a faster panel beats film, because exposure time is time in a controlled area.
It also puts a premium on getting it right first time. A re-shoot on a fabrication job is an inconvenience. A re-shoot in an active area is a second dose to a person, and it will be questioned. This is where a documented technique, a rehearsed sequence and a technician who has seen the geometry before earn their cost.
Radiography inside constrained space
Site radiography on a decommissioning scope has to solve the controlled area problem in a building that was not laid out to accommodate one. Adjacent occupied spaces, floors above and below, and other contractors working to their own programme all constrain where a boundary can go.
Three things usually resolve it. Source selection — selenium-75 in place of iridium-192 where the thickness allows, because the lower energy shrinks the boundary. Timing — out-of-hours exposures when surrounding areas are clear. And relocation — bringing the component to a shielded facility rather than taking the source to the component.
That last one is underused. Our Cumbria site runs two radiation compounds capable of taking components up to eight metres, which means a surprising amount of pipework, valve bodies and structural sections can be shot under controlled conditions rather than in situ. If a component can be removed anyway as part of the decommissioning sequence, radiographing it after removal is almost always cheaper, faster and lower-dose than radiographing it in place.
Chain of custody is part of the deliverable
Decommissioning generates decisions with long tails: this material is free-release, that one is not; this section met its acceptance criteria, that one was cut out. Those decisions get audited, sometimes years later, by people who were not there.
So the inspection record has to be more than a pass/fail. It needs to say what was examined, where exactly on the component, by whom, to which certification level, against which procedure and which acceptance criteria, with which equipment, calibrated when. If it is radiography, the images need to be retrievable. If it is encoded ultrasonics, the data file needs to survive.
This is where paper reporting quietly fails. A PDF emailed to a project manager who has since moved on is not a chain of custody. Reporting into a system where results are timestamped, attributable and retrievable is worth more on this kind of work than on almost any other, and it is why our results are recorded digitally and available while the technician is still on site rather than written up days later.
Sequencing is where programmes go wrong
The technical work on a decommissioning scope is rarely what causes the delay. Co-ordination is.
Inspection on an active site competes for the same constrained resources as everything else: access permits, scaffolding, isolations, controlled area availability, and the attention of whoever can authorise a boundary. An exposure that takes ninety seconds may need a four-hour window in which the surrounding area is clear, and that window has to be booked against a programme written before anyone thought about it.
Three things reliably shorten the calendar. Book the inspection window at the same time as the scaffold rather than after it — the scaffold usually determines when the component is reachable, and inspection is usually remembered last. Get the acceptance criteria signed before mobilisation, because a query at the component costs a return visit. And identify early which components are coming out anyway, because those should be inspected after removal in a controlled facility rather than in situ.
That third point bears repeating, because it is consistently the largest single saving available and it is almost always settled by habit rather than analysis. If a spool is being removed next month, radiographing it in place this month is paying twice — once in dose, once in the controlled area you had to build around it.
What we get asked to do most
Across West Cumbrian decommissioning scopes, the recurring package is fairly consistent:
PMI for material confirmation and waste routing, usually first.
Ultrasonic thickness survey on pipework and vessels to establish remaining wall before lifting or cutting decisions.
Radiography of welds and internals where the geometry defeats ultrasonics
Surface methods — magnetic particle on ferritic material, penetrant on austenitic — for crack detection on lifting points and structural connections.
Remote visual inspection where opening a system is the expensive option.
Ferroxyl testing where free iron contamination on stainless matters, which on nuclear work it frequently does.
Rarely all of them, rarely in the order the programme first assumed.
Scoping one of these properly
The scopes that run smoothly share a few features. Somebody has walked the job with the inspection contractor before the price was fixed. Surface preparation and access are in the scope rather than assumed. Material has been confirmed rather than inferred from a drawing. The acceptance criteria have been agreed in writing before anyone mobilises — because "is this acceptable" is a much harder question to answer standing next to an indication than it is a week earlier at a desk.
And someone has been honest about whether the component can be moved. It is the single biggest lever on cost and dose, and it is usually decided by habit rather than analysis.
We have been working this estate for over twenty-five years, from a facility on its doorstep in Workington. If you have a decommissioning scope that needs inspection input before it is priced, that walk-round is worth more than a quotation. Our case studies show the shape of the work we take on.
