Every refinery turnaround (TAR) budget has a line for inspection labor, a line for NDT services, a line for scaffolding and access, and — if the planning team is honest with itself — a contingency line that exists specifically because scope always grows once the unit is actually opened up. That contingency line is usually the biggest tell in the whole budget. It's not there because inspectors are bad at their jobs; it's there because a huge amount of pre-TAR scope development still depends on what an engineer can physically see and remember from a site walk-down months before the unit shuts down, cross-referenced against drawings and inspection reports that may or may not reflect the asset's current condition.
\nRefinery downtime is widely cited across industry literature as running into the millions of dollars per day in lost throughput for a major processing unit — a well-known industry figure, not something specific to any one operator or vendor. That number alone explains why TAR planning is so risk-averse and so scaffolding-heavy: when a single day of schedule slip costs more than an entire additional access package, the rational move is almost always to over-scaffold, over-scope, and build in schedule float, because the alternative — running short and extending the outage — costs far more than the padding did. Scaffolding and temporary access are consistently among the largest non-labor line items in a TAR budget for exactly this reason; getting eyes and hands onto every piece of equipment that might need inspection or repair requires physical access to be built first, torn down after, and rebuilt again if scope discovered mid-turnaround requires reaching somewhere the original access plan didn't cover.
\n\nPre-turnaround scope development traditionally runs through a sequence of physical site visits: an engineering walk-down months out to identify candidate equipment and access requirements, a follow-up inspection-focused visit to review specific findings against RBI-driven inspection due dates, and often a final walk-down closer to shutdown to confirm nothing has changed. Each of those visits requires travel, site access coordination, PPE and safety orientation, and — critically — the availability of the specific engineers and inspectors who understand that unit's history, which is rarely a large pool of people at any single site.
\nThe breakage shows up in two predictable places. First, scope gets missed not because anyone was careless, but because a walk-down engineer simply can't hold the entire condition history of a complex unit in their head while standing in front of one vessel — the corrosion trend on a nozzle they're looking at might be perfectly fine in isolation, but if they don't have the full circuit's RBI-driven risk ranking in front of them at that moment, a genuinely high-priority item can get deprioritized simply because it wasn't visually dramatic. Second, and more expensively, scope grows once the unit is physically open: insulation comes off and reveals corrosion under insulation nobody scoped for; a vessel internal inspection reveals damage that wasn't visible from the manway; a piping circuit that looked fine on paper shows thinning that wasn't caught because the last UT survey's CML didn't happen to land on the affected area. Every one of those discoveries mid-TAR triggers a change order — additional scaffolding, additional inspection time, additional repair scope — all under schedule pressure that makes each change order more expensive than it would have been if planned in advance.
\n\nWhy "more site visits" doesn't actually fix this: the instinctive response to missed scope is to add another walk-down cycle — but that just adds more travel cost and more scheduling friction for the same underlying problem: the engineer doing the walk-down still only has whatever they can see and remember in the moment, cross-referenced against documents that are hard to correlate spatially with what they're standing in front of. Adding visits without changing how the asset's condition data is organized and presented mostly just adds cost without proportionally reducing missed scope.
\n\nThis is where a digital twin's value proposition for TAR planning is genuinely different from just "better documentation." A digital twin platform that holds a persistent 3D model of the unit — piping, vessels, tankage — with inspection and corrosion data layered directly onto it lets an engineer do a meaningful chunk of scope development work remotely, before ever needing to physically be on site, and lets the physical site visits that do happen be far more targeted.
\nReducing redundant site visits: when RBI-driven inspection due dates, corrosion trends, and prior finding locations are all visible on the same model an engineer can navigate from a desk, a large share of initial scope screening — which equipment is even a candidate for this TAR window, based on inspection interval and condition trend — can happen without a site visit at all. The site visits that remain get reserved for what actually requires physical presence: confirming access logistics, visually verifying a specific finding that needs eyes on it, or scoping a repair that genuinely can't be assessed remotely. That's not zero site visits — it's fewer, and each one is better targeted because the engineer already knows exactly what they're going to look at and why before they arrive.
\nTighter pre-TAR scope definition: scope-growth change orders during an open turnaround are disproportionately expensive compared to the same scope identified during planning, because mid-TAR discovery competes with schedule pressure, access that has to be modified on the fly, and inspection resources already fully allocated to the planned scope. A model that surfaces the full corrosion history and RBI risk ranking for every circuit and vessel in the unit — not just the ones an engineer happened to flag on a walk-down — catches more of the genuinely high-priority items during planning, before the schedule and budget are locked. It doesn't eliminate mid-TAR discovery entirely; some damage mechanisms (like CUI hidden under intact-looking insulation) are inherently hard to catch without physical removal. But it substantially reduces the share of scope growth driven by information that existed and simply wasn't connected to the right context at planning time.
\nFaster RBI-driven planning cycles: an RBI program per API 580/581 is only as useful for TAR planning as the speed at which its output can be translated into an actual scope list tied to physical locations and access requirements. When RBI risk rankings live in a separate system from the 3D asset model, someone has to manually cross-reference which physical equipment and circuits correspond to which RBI-flagged items — a translation step that takes real engineering time and is a common source of scope-list errors. When the RBI output is visualized directly against the model, that translation step effectively disappears: the flagged equipment is already spatially located, already shows its access requirements, and already shows what surrounding equipment or piping might be affected by the same access package.
\n\nIt's worth being concrete about the scale of spend this space typically involves, because it frames why even incremental reductions in redundant site visits and scope-growth change orders matter financially. Enterprise asset management and turnaround management platforms from major vendors — SAP EAM/S/4HANA-based turnaround management modules, IBM Maximo, and specialized TAR planning software — are typically licensed at the enterprise level with implementation projects that can run from the low hundreds of thousands to several million dollars depending on scope, plus ongoing licensing and maintenance costs, according to industry-published estimates and analyst commentary on EAM implementation costs. Scaffolding alone on a major unit turnaround is commonly cited in TAR industry literature as running into the millions of dollars for a single event, and engineering walk-down and scope-development labor — spread across multiple site visits by senior engineers and inspectors over the months leading up to a TAR — is a recurring, substantial cost center in its own right even before any tools or repairs are purchased.
\nFor a unit turnaround where scaffolding and access already run into seven figures and even one avoided major scope-growth change order can run into six figures once schedule impact is factored in, the case for better data organization tends to make itself once the underlying costs are laid out this plainly. Pairing that visualization layer with an ERP system built for inspection and turnaround scheduling closes the loop between what the model shows and what actually gets put on the TAR work list.
\n\nWhat Atlantis actually offers here: Atlantis doesn't publish a flat price for the digital twin platform, because implementation scope varies enormously by asset count, data sources being integrated, and how much historical inspection data needs to be migrated into the model — the same reason enterprise EAM and TAR platforms are quoted per-project rather than off a price list. What Atlantis offers instead is a tailored quote built around a specific unit or site's actual scope: contact the team for a walkthrough of what integrating your piping, vessel, and tankage inspection data into a persistent 3D model would look like for your next TAR cycle, and get a quote built for that scope rather than a generic number that doesn't reflect your asset count or data complexity.
\n\nThere's a second layer of savings that TAR planners tend to undercount because it doesn't show up as a single line item: the efficiency of the NDE crews once they're actually on site with the unit open. A well-known inefficiency in turnaround execution is scaffold sections getting climbed multiple times by different disciplines — a UT crew accessing a platform for shell readings on Tuesday, a PT crew accessing the same platform for weld inspection on Thursday, a visual inspector climbing it again on Friday — because the work packages were sequenced by discipline rather than by physical location. Each of those repeat climbs costs real inspection hours that could have gone toward actual measurement and evaluation instead of access logistics, and on a large unit with dozens of access packages, that inefficiency compounds fast.
\nA model that shows exactly which CMLs, welds, and vessel internals sit within a given physical access zone lets a TAR planner batch inspection tasks by location rather than by discipline silo — UT, PT, and visual work that shares a scaffold section gets scheduled together, so the access is built once, used by every discipline that needs it, and torn down once. This is a scheduling change more than a technology change, but it only becomes practical when the spatial relationship between CMLs, welds, and access packages is visible and queryable rather than buried across separate discipline-specific work order lists.
\nField crews working from the model instead of a paper packet: the other piece is what happens once a technician is actually standing in front of the equipment. A UT technician working from a printed CML location sketch on a vessel with 40-plus nozzles and multiple historical finding locations spends real time just confirming they're measuring the correct point before they even start collecting data — and a wrong-location reading, if it goes undetected, corrupts the corrosion-rate trend for that CML going forward, as covered above. A technician working from mobile access to the model, with the exact CML location, prior reading history, and any known findings pulled up against the physical equipment in front of them, spends less time on location confirmation and starts actual data collection faster, with less risk of a mislocated reading making it into the record uncaught.
\nTying access sequencing to RBI output: this connects back to RBI-driven planning directly: when the RBI risk ranking, the corrosion trend data, and the physical access zones are all part of the same model, the TAR schedule itself can be built around physical zones rather than discipline handoffs — inspect everything reachable from access package 3, then move to package 4, rather than "all UT work across the whole unit" followed by "all PT work across the whole unit" with the same scaffolding being occupied and vacated multiple times by different crews. That sequencing change is a scheduling and coordination discipline as much as a software feature, but it depends on having the spatial data organized well enough to plan it — which is exactly the gap a digital twin closes.
\n\nIt's worth being direct about the limits here, because overselling this undermines the case for the parts that are genuinely true. A digital twin doesn't replace the physical inspection itself — UT thickness readings, MFL floor scans, and visual inspections still have to be performed by qualified personnel, and the model is only as good as the data fed into it. It doesn't eliminate mid-TAR discovery of damage mechanisms that are inherently hard to detect without physical access, like corrosion under insulation that hasn't yet caused visible insulation jacket damage. And it doesn't substitute for the judgment of experienced inspectors and engineers in deciding what a given finding actually means for fitness for service — it makes that judgment easier to apply consistently and remotely, but the judgment itself still has to come from qualified people, whether that's an in-house team or outside ASNT Level III consulting support brought in to review scope and findings ahead of a major TAR.
\nWhat it does solve is the specific, expensive problem of scope development and condition review depending almost entirely on what a person can hold in their head during a physical walk-down, disconnected from the asset's actual inspection history until someone manually pulls the relevant reports. For a refinery running multiple TARs a year across a large asset base, closing that gap is where the real planning-cycle savings show up — not in any single dramatic number, but in the steady reduction of redundant travel, missed scope, and mid-TAR change orders across every turnaround the plant runs.
\n\n\nAtlantis NDT pairs field expertise with software: NDT inspection management software — Atlantis ERP, a digital twin platform for asset integrity, and NDT reporting software. Build your team with NDT training & certification (ASNT SNT-TC-1A) and ASNT certification pathways, or bring in ASNT Level III consulting. Affordable, accessible, fully customizable — book a free consultation.
\n