Corrosion Rates You Can Defend After the Bridge File Moves Systems
The module computes a short-term rate from the two most recent thickness measurements and a long-term rate from the as-built baseline, then governs on whichever is more conservative. On bridges the migration problem dominates: decades of records describe section loss as a percentage in narrative text, so the baseline has to be rebuilt from shop drawings before any rate is defensible.
A bridge inventory is not a vessel file. Steel superstructures lose section locally — girder webs and bottom flanges under leaking expansion joints, bearing seats, diaphragm connections, the inside faces of built-up members where debris packs — while the rest of the member is untouched. Legacy records rarely hold a thickness. They hold National Bridge Inventory condition ratings, element-level condition states from the AASHTO Manual for Bridge Element Inspection, and sentences like moderate section loss, up to fifteen percent of web at Girder 3 north bearing. Turning that into a long-term rate requires an as-built thickness that only the shop drawings carry, and it requires the software to record that the baseline was reconstructed rather than measured. Short-term rates from two routine cycles twenty-four months apart sit close to ultrasonic repeatability, so the migration must preserve enough metadata to tell real loss from scatter.
Source: Method basis: short-term and long-term rate practice as codified in API 510 and API 570, applied here to structural section loss; 23 CFR Part 650 Subpart C, the National Bridge Inspection Standards, together with the Specifications for the National Bridge Inventory for inspection frequency and inventory data items; the AASHTO Manual for Bridge Element Inspection for element condition states and defect 1000, corrosion; the AASHTO Manual for Bridge Evaluation for load rating and posting with measured section loss; ASTM A709 for structural steel grades including uncoated weathering steel, with FHWA Technical Advisory T 5140.22 on its use; ISO 9223 for atmospheric corrosivity categories; ASTM C876 for half-cell potential and ASTM C1152 and C1218 for acid-soluble and water-soluble chloride in hardened concrete.
| Legacy record type | What it actually contains | Rate it can support | Migration rule |
|---|---|---|---|
| NBI condition rating, 0 to 9, superstructure item | A single ordinal judgement covering an entire superstructure | None | Import as history and context only; never derive a thickness or a rate from an ordinal rating |
| Element-level condition states with quantities | Square feet or linear feet of an element in condition states 1 through 4, with defect 1000 corrosion | A trend in affected quantity, not a metal loss rate | Keep as a parallel trend that can corroborate a measured rate but never substitutes for one |
| Narrative section loss as a percentage of a named component | An inspector's visual or caliper estimate against an assumed original section | A bounded long-term rate, only once the as-built thickness is recovered | Compute only with a drawing-sourced baseline, mark the rate estimated, and exclude it from governing until a measurement exists |
| Spot ultrasonic readings in a consultant's appendix | Discrete thicknesses, frequently without a relocatable position | A short-term rate if and only if the point can be re-registered | Migrate with the sketch or photograph; if the point cannot be relocated, import as a one-off observation, not a monitoring location |
| Paint and coating condition records | Coating breakdown percentage and last overcoat date | No rate, but it changes the rate you should expect | Link to the member so a coating renewal opens a new short-term window instead of corrupting the long-term series |
| Half-cell potential and chloride sampling on concrete | Potentials in millivolts against a reference cell, chloride content by depth | Corrosion likelihood, plus a rebar rate only from polarisation resistance | Store on its own scale and never merge into the steel thickness rate column |
What actually breaks when a bridge file changes systems
Migrations of bridge inspection data fail in a specific and predictable way. The inventory items move cleanly, because the Specifications for the National Bridge Inventory define them tightly and both systems speak that vocabulary. Structure identifiers, spans, materials, year built, inspection dates, condition ratings: all of it transfers. Then the project stalls on the part that matters most for corrosion, which is the measured and described condition of individual members, because that data was never structured in the first place.
The old file holds it as prose. A 1997 report says the bottom flange of Girder 2 exhibits moderate section loss with pack rust at the diaphragm connection. A 2009 report says up to twenty percent loss of web at the north end of Girder 3, measured with calipers where accessible. A 2018 report attaches a consultant's ultrasonic appendix with fourteen thicknesses and a hand sketch. None of those three records is in the same units, at the same location, or to the same definition, and the new system wants a baseline, a current thickness and a date.
The temptation at this point is to force the conversion — assume a nominal thickness, apply the percentage, generate a rate, and get the migration signed off. That produces a database full of numbers with no stated provenance, which is worse than an empty field, because the next engineer trusts it. The discipline that makes a bridge migration succeed is separating what was measured from what was inferred, and carrying that distinction forward permanently rather than only in the migration log.
Reconstructing a baseline from the drawings, not from the inspection record
A long-term corrosion rate is loss divided by elapsed time, and loss is baseline minus current. On a pressure vessel the baseline is easy: nominal thickness from the specification, or the first measured survey. On a fifty-year-old steel bridge the baseline was never measured at all, because nobody performed a thickness survey on a new structure in 1968. The only defensible source is the fabrication record — shop drawings that dimension the web and flange plates, and mill certifications if they survive in the agency archive.
This is why the migration order matters. Drawing recovery should precede data import, not follow it, because the baseline determines whether a rate can exist. Where drawings are unavailable, the substitute is a measurement in an undamaged region of the same member: the mid-span web of a girder whose loss is concentrated at the bearing end is, in effect, the as-built thickness minus a small uniform atmospheric loss. That is a far better estimate than a nominal from a table, and it is measurable today rather than dependent on an archive.
The requirement on the software is that the baseline carries its source as a permanent attribute — drawing, mill certificate, measured undamaged region, or assumed nominal — and that the source appears anywhere the derived rate appears. A long-term rate built on an assumed nominal has an uncertainty band roughly the width of the plate tolerance, which on thin material can exceed several years of actual corrosion. The number should not be forbidden. It should be labelled, and it should not be allowed to govern silently.
Percent section loss is not a thickness, and the conversion invents a number
The most common legacy record in a bridge file is a percentage. Inspectors have recorded section loss that way for decades because it is what a visual assessment supports and what the reporting form asked for. Converting it into a rate requires multiplying by an original thickness, and that multiplication is where fabricated precision enters the database. Fifteen percent loss on a three-eighths inch web is 0.056 inches. Fifteen percent on a half-inch web is 0.075 inches. Same record, same words, a thirty-four percent difference in the loss and therefore in the rate.
There is a second, subtler problem. An inspector writing fifteen percent is not usually estimating a thickness reduction over a uniform area. They are describing an impression that includes pitting depth, area affected, and the visual weight of scale and pack rust. Two inspectors on the same girder end routinely differ by a full condition state. Treating that as a measurement with two significant figures is the error, and it compounds when the resulting rate is then extrapolated over a projection horizon.
The workable handling is threefold. Compute the rate only when a drawing-sourced baseline exists. Store it as an estimate with an explicit uncertainty band rather than a point value. And exclude estimated long-term rates from the governing selection until at least one real measurement exists at that location, so that a narrative from 1997 cannot outvote an ultrasonic reading from last month. The estimate still earns its place — it tells you where to send the next crew — but it does not silently drive a posting decision.
Two-year cycles against ultrasonic measurement noise
The National Bridge Inspection Standards under 23 CFR Part 650 Subpart C set routine inspection at a twenty-four month interval, with extended intervals available for structures that qualify and shorter intervals for those that need them. That regulatory cadence sets the shortest short-term window most bridges will ever have. It is a much longer interval than a refinery turnaround cycle in wall-clock terms, but it is a much shorter one in terms of metal lost, because atmospheric corrosion on painted structural steel is slow.
Run the numbers on the boundary case. A painted girder in an ISO 9223 category C3 environment might lose on the order of a thousandth of an inch per year of steel where the coating is sound. Over twenty-four months that is two thousandths. Manual contact ultrasonic thickness measurement on a field surface — pitted, previously coated, possibly with residual mill scale — carries a practical repeatability that can approach or exceed that figure once probe placement, couplant, surface preparation and operator technique are included. The short-term rate is then dominated by measurement variance rather than corrosion.
This does not make the short-term rate useless. It makes an unqualified short-term rate misleading. The module should attach a confidence treatment to any rate computed over an interval where the expected loss is within measurement uncertainty, should require repeat readings at high-consequence locations rather than a single value, and should not let a noise-driven short-term rate escalate a member into a repair programme on its own. Conversely, when the short-term rate rises well clear of the noise floor — at a leaking joint, at a bearing seat holding chloride-laden debris — that is a real signal and the governing rule should let it win immediately.
Weathering steel is where the two-rate method earns its keep
Uncoated weathering steel, ASTM A709 Grade 50W and its predecessors, was specified across a large slice of the North American inventory on the premise that a stable oxide patina forms and arrests further loss. Where the design intent holds — good drainage, regular wetting and drying, no chloride accumulation, adequate clearance over water — it largely does, and long-term loss curves flatten. FHWA Technical Advisory T 5140.22 exists precisely because the intent does not hold everywhere.
The failure locations are consistent and they are exactly where legacy records are thinnest: beneath leaking expansion joints where salt-laden runoff sheets across the girder end, at bearing seats where debris holds moisture against the metal for weeks, on the inside faces of box sections with blocked drainage, and on fascia girders taking direct spray. In those conditions the patina never stabilises, loss remains roughly linear, and it can exceed the design assumption substantially over decades.
This is the structural argument for keeping both rates on weathering steel rather than trusting the long-term average. A long-term rate spanning forty years on a girder that behaved as designed for thirty and then lost its joint seal is dominated by the benign period, and it will comfortably understate current loss. The short-term rate is the only number that sees the change. Governing on the more conservative of the two is not a bureaucratic convention here; it is the difference between finding a girder end in time and finding it during a load rating.
Concrete and rebar need a different rate on a different scale
Most bridge inventories are mixed. The same migration that carries steel thickness records also carries deck, pier and abutment data where the corroding metal is embedded reinforcement, and the measurement techniques share no units with ultrasonic thickness. Half-cell potential to ASTM C876 gives a potential in millivolts against a copper-copper sulfate reference and describes the probability that corrosion is active, not a rate. Chloride content by ASTM C1152 or C1218 gives concentration by depth and describes whether the threshold for depassivation has been reached.
A rate for embedded steel comes from a different measurement entirely: linear polarisation resistance, giving a corrosion current density in microamps per square centimetre, which converts to a section loss rate in microns per year. Broadly, low current densities indicate passive or low-rate behaviour and values above roughly one microamp per square centimetre indicate active corrosion, though site correction and temperature sensitivity make single readings weak evidence. Cathodic protection systems on atmospherically exposed reinforced concrete, assessed against depolarisation criteria, change the interpretation again.
The migration rule is separation, not integration. These scales must live on their own fields with their own units and their own interpretation guidance, never merged into a thickness-derived rate column because both are described as corrosion. The single most damaging thing a migration can do to a mixed inventory is create one rate field that different subsystems populate with incompatible quantities. What the two families should share is the asset hierarchy and the projection framework, so a deck and a girder on the same structure can be planned together even though their numbers were produced by entirely different physics.
Evaluating the module while you are mid-migration
The evaluation should be run on your own worst legacy file, not on demonstration data. Bring the structure with three inspection consultants across four decades, mixed narrative and ultrasonic records, a weathering steel superstructure and a deck with half-cell history. Ask the vendor to import it and then ask three questions of the result. Which rates are measured and which are estimated, and is that visible on the printed output rather than only in an admin view? Which locations produced no rate at all, and why? And what happened to the records that would not convert?
That third question is the one that exposes a weak migration. A system that silently drops what it cannot parse leaves an inspector in 2031 with no idea that pack rust at a diaphragm connection was documented in 1997. The correct behaviour is that unconvertible narrative is retained, attached to the member, full-text searchable, and surfaced next to the numbers for the same location. The corrosion rate is the calculated layer; the narrative is the evidence layer, and losing the second to tidy the first is a bad trade.
Then test the governing logic against a weathering steel girder end where the short-term rate exceeds the long-term rate, and confirm that the more conservative rate governs, that the report names which one it was, and that an estimated long-term rate built on an assumed nominal cannot outrank a measured short-term rate. Finally, confirm the projection output speaks the language your engineers use — remaining section feeding a rating factor, not a pressure-equipment remaining life. For a migration assessment against a sample of your own bridge files, contact info@atlantisndt.com; demonstrations and scoped consultations are arranged on request.
Can a corrosion rate be computed from an NBI condition rating?
No. A superstructure condition rating is one ordinal number describing an entire structure, and it responds to deck joints, bearings, coating condition and impact damage as much as to metal loss. Two bridges rated 5 can differ by an order of magnitude in remaining section. Ratings are valuable history and they belong in the migrated record, but any system that derives millimetres per year from a 0 to 9 scale is manufacturing precision that was never in the source.
How do you recover a baseline thickness for a bridge built in 1968?
From the shop drawings and the mill certifications if they survive, not from the inspection file. Plate and rolled-shape thicknesses are dimensioned on the fabrication drawings, and the steel grade tells you which ASTM specification governed the plate tolerance. Where drawings are lost, an undamaged region of the same member measured ultrasonically gives a far better baseline than an assumed nominal. Whichever source is used, the record must carry it, because the long-term rate inherits its uncertainty.
Is a twenty-four month interval long enough to compute a short-term rate?
Sometimes, and the system should say when it is not. The National Bridge Inspection Standards set routine inspection at twenty-four months, with extended intervals permitted for qualifying structures. Over two years, atmospheric loss on a painted girder may be well under the repeatability of a manual contact thickness measurement on a rough, previously coated surface. A short-term rate computed from that interval is mostly noise, and the module should present a confidence band rather than a bare number.
Does uncoated weathering steel corrode at a decaying rate or a constant one?
Both, depending on where on the bridge you are standing. Where the patina forms and wet-dry cycling is normal, loss decays toward a low asymptote. Where chlorides accumulate and the surface stays wet — under leaking joints, on the inside of box sections, at bearing seats packed with debris — the patina never stabilises and loss continues roughly linearly. That is precisely why the long-term rate alone is unsafe on weathering steel and the short-term rate has to be kept.
Where does the corrosion rate feed into a load rating?
Into the section properties, not into a remaining-life clock. Under the AASHTO Manual for Bridge Evaluation, measured section loss reduces the area, section modulus and shear capacity used in the rating, which drives the inventory and operating rating factors and any posting decision. The useful output for a bridge owner is therefore a projected date at which a member's remaining section drops the rating factor below one, not the pressure-equipment style remaining life.
What should happen to legacy records the migration cannot convert?
They must be kept, searchable and attached to the structure, not discarded because they do not fit a field. A narrative describing pack rust at a built-up member connection carries information no thickness field can hold, and it is the reason a future inspector looks in the right place. The migration failure to guard against is a system that only accepts what it can compute with, quietly leaving forty years of context in an archive nobody opens.
Built for any business that runs on operations
Most companies do not fail at their craft. They lose time, margin and goodwill in the gaps between the tools they use to run the place — a quoting spreadsheet that does not talk to the job sheet, a job sheet that does not reach accounts, and a compliance folder nobody can search when a client asks. Atlantis closes those gaps by putting the whole operation on one platform, so information is entered once and everything downstream stays in step.
What you can run on it
- Sales and CRM — leads, quotes, follow-ups and the pipeline that tells you what next month looks like.
- Projects and job costing — plan the work, track the hours and materials against it, and see the margin while the job is still live rather than at final account.
- Field and service teams — dispatch, schedules, mobile capture that works with no signal, and sign-off from site.
- Inventory and purchasing — stock, suppliers, reorder points and goods receipt, joined to the jobs that consume them.
- People — records, qualifications and licences with renewal reminders, timesheets, leave and payroll.
- Quality and documents — procedures and forms under revision control, with the audit trail an inspection or accreditation body actually asks for.
- Accounts — invoicing, expenses, multi-currency and the reporting your accountant stops chasing you for.
Affordable, accessible, fully customizable — and we mean each word
Affordable because the whole suite is included rather than sold to you a module at a time, and because implementation is done by people who have run operations rather than by a chain of subcontractors. Accessible because it runs in a browser and on a phone, works for a small team on day one, and does not need a specialist on staff to keep it alive. Fully customizable because your process is the thing that makes you competitive — the software should bend to it, not the other way round.
Industries we configure for
Service businesses and contractors, manufacturing and fabrication, trading and distribution, laboratories and testing houses, engineering consultancies, construction and facilities, and asset owners across energy, marine, aerospace and infrastructure. Inspection and testing is where we started, and it remains the sector we go deepest in — but the platform underneath is general-purpose, and most of what it does has nothing to do with inspection at all.
What happens when you get in touch
A short conversation, not a sales sequence. We ask how the business runs today and where it hurts, show you the platform doing that work, and send a written quote shaped to your region, your team size and the scope you actually need. No obligation, nothing to install first, and no pressure to decide on the call. Reach out and tell us what you are trying to fix.
Related: business management platform · inspection management software · choosing the right category of software · modules · by industry · asset integrity platform. Book a free consultation.