Making Every Cement Plant in the Group Compute Metal Loss the Same Way
A corrosion rate module computes a short-term rate from the last two thickness readings and a long-term rate from the original baseline, then applies the more conservative of the two. In cement and lime the standardisation problem is that little of the equipment is code-governed, so each plant invented its own baseline, its own t-min and its own year.
Cement and lime sites carry a mixed asset base: kiln shells, preheater cyclones and risers, tertiary air ducts, clinker cooler casings, baghouse and ESP housings, raw meal and coal handling, plus a small population of genuine ASME Section VIII vessels — air receivers, ammonia storage for SNCR, fuel oil day tanks. Only that last group has a code-derived required thickness. Everything else is thinned by abrasion from clinker dust, by acid dew-point attack where SO3 and moisture condense below roughly 260 to 300 degrees F, and by alkali chloride condensation in the 700 to 1,500 degrees F preheater range. Rates from those mechanisms are not linear. A long-term average taken across a refractory campaign mixes protected and unprotected service and understates the live rate. The module therefore has to keep short-term and long-term rates separate, expose which one governed, and hold the same divisor, baseline and unit rule at every plant in the group.
Source: Method basis: API 510 and API 570 corrosion-rate practice, in which short-term and long-term rates are determined separately and the more conservative governs remaining life; API 579-1/ASME FFS-1 Part 4 for general metal loss averaging and thickness profiling; API 571 for the definitions of dew-point corrosion, erosion and erosion-corrosion, and sulfidation; ASME Section VIII Division 1 UG-27 for required thickness on the pressure equipment that is genuinely code-governed. Regulatory context: 40 CFR Part 63 Subpart LLL, the NESHAP for the Portland cement manufacturing industry, and OSHA 29 CFR 1910.119 Appendix A for anhydrous ammonia threshold quantities where SNCR reagent is stored anhydrous.
| Calculation input | How it varied plant to plant | Effect on the group roll-up | Standardised rule |
|---|---|---|---|
| Elapsed time in the denominator | Some sites divided by calendar years, others by kiln operating years derived from run factor | A plant running at 0.85 run factor reported rates roughly 15 percent lower than an identical plant using calendar years | Calendar years drive the reported rate; operating hours are stored alongside as a second field and never substituted silently |
| Baseline thickness | Nominal from the purchase specification at three sites, first measured survey at four, post-repair plate thickness at one | Long-term rates were not comparable at all, and mill over-tolerance at one site read as negative corrosion | Baseline is the first measured survey where one exists, tagged measured or nominal, with the nominal path flagged on every report |
| Required thickness on non-code equipment | Anything from a structural calculation to a flat 50 percent of nominal to whatever the last contractor wrote | Remaining life could not be ranked across the group because the finish line was in a different place at every plant | Engineering-set required thickness per equipment class, authored once, versioned, and referenced rather than retyped |
| Which rate governs | Five sites used the long-term rate only, two used the last interval only, one used the greater of the two | The group's fastest-degrading equipment was invisible because its short-term rate was never computed | Both rates computed at every location; the more conservative governs and the report names which one it was |
| Units and rounding | Millimetres per year, mils per year, and one site reporting cumulative loss with no rate at all | Sorting the group register by rate produced meaningless ordering | One stored unit with display conversion; rounding applied at presentation only, never written back to stored values |
| Location identity | Point numbering restarted whenever the inspection contractor changed | Two consecutive surveys of the same riser duct appeared as unrelated points, so no short-term rate could exist | Permanent condition monitoring location identifiers owned by the plant, with contractor point numbers retained as aliases |
Cement and lime have no external referee for this calculation
In a refinery, the corrosion rate argument is already settled. API 510 and API 570 tell the owner-user how to derive short-term and long-term rates, which one governs, and how remaining life follows. A jurisdiction may adopt those documents, an insurer will ask about them, and an authorised inspector signs a number that has a defined provenance. Disagreement between two units of the same company is a compliance finding, not a matter of taste.
A cement or lime plant has almost none of that scaffolding. The pressure equipment that genuinely falls under ASME Section VIII is a small minority of the asset list: instrument air receivers, an ammonia storage vessel if SNCR runs on anhydrous reagent, a fuel oil day tank, perhaps a compressor aftercooler. The equipment that actually determines whether the kiln runs — preheater cyclones and their dip tubes, riser ducts, the tertiary air duct, kiln inlet and outlet seals, clinker cooler casings, baghouse hoppers, ESP shells — is not code equipment at all. Nobody outside the company is going to tell you what its required thickness is.
That absence is precisely why a group of eight plants ends up with eight methods. Each site hired a different inspection contractor, inherited a different spreadsheet, and made a defensible local choice about baseline and divisor. None of those choices was wrong on its own terms. Together they make the group register unusable, because a reported rate of 0.22 mm per year at one plant is not the same measurement as 0.22 mm per year at another. Standardisation here cannot be imported from a code. It has to be built into the system that stores the readings.
What the equipment is actually losing, and why that changes the arithmetic
Three mechanisms dominate, and none of them behaves like the uniform aqueous corrosion the two-rate method was originally shaped around. The first is abrasion and erosion from raw meal, clinker dust and coal, concentrated wherever gas changes direction: cyclone inlet scrolls, dip tube ends, duct elbows, cooler grate side walls. Wear is geometric. It accelerates as a scallop forms and the local gas velocity rises, then can slow again once the profile stabilises. A straight-line rate is a convenient fiction over any long window.
The second is acid dew-point corrosion, described in API 571, occurring where SO3 and moisture condense on cold metal, typically below roughly 260 to 300 degrees F depending on sulfur trioxide concentration. This finds cold spots: the shell under external insulation gaps, the bottom of a bag filter hopper, the fan casing after a long stop, ductwork on the raw mill circuit during mill-off operation. Its rate is not a property of the equipment. It is a property of how many hours that surface spent below the dew point, which means it tracks operating pattern, not calendar time, and it can differ by an order of magnitude between two identical hoppers.
The third is alkali and chloride condensation in the upper preheater, in the approximate 700 to 1,500 degrees F band, where potassium and sodium chlorides and sulfates condense on process surfaces and attack metal beneath build-up. Plants with a chloride bypass see a different distribution than plants without one. All three mechanisms share one property that matters for the module: they are episodic. That is the technical case for keeping the short-term rate as a first-class number rather than a diagnostic afterthought.
Calendar years, operating years and the run factor trap
A cement plant naturally thinks in kiln operating hours. Production, refractory life, fan wear and maintenance planning are all indexed to run time, and a plant at 0.85 run factor genuinely puts fewer hours on its ductwork than one at 0.95. It is entirely reasonable that some sites divided metal loss by operating years rather than calendar years. It is also the single change that most distorts a group roll-up, because the divisor difference is invisible in the output. The number arrives as millimetres per year either way.
The arithmetic is not subtle. Two plants lose 1.7 mm of shell over five calendar years. One divides by five and reports 0.34 mm per year. The other applies a 0.85 run factor, divides by 4.25 operating years, and reports 0.40 mm per year. That is a seventeen percent divergence created entirely by bookkeeping, on equipment that behaved identically. Rank the group register by rate and the second plant now sits above the first, and the corporate integrity engineer starts asking the wrong plant the wrong question.
The rule that survives contact with reality is to report on calendar years and store operating hours as a second, always-available field. Calendar time is what an inspection interval and a shutdown window are actually measured in, so remaining life expressed in calendar years is directly actionable. And for dew-point corrosion the calendar basis is arguably more physical anyway, because acid attack continues after a stop as surfaces cool through the dew point with wet dust in contact. What must never happen is silent substitution: no report should be able to switch divisors without saying so on its face.
Why the long-term rate lies after a refractory campaign
The two-rate method exists because the long-term rate is stable and the short-term rate is responsive, and you want both. The long-term rate, taken from the original baseline to the current reading, filters measurement noise across many years. The short-term rate, taken across the most recent interval, catches a change in service. Governing on whichever is more conservative is a deliberately asymmetric rule: it accepts a false alarm more readily than a missed acceleration.
In a preheater and kiln that asymmetry earns its keep, because the metal spends part of its life protected and part of it exposed. While refractory is intact, shell loss is close to zero. As the lining thins, spalls or loses anchorage, the shell sees process gas and dust directly and metal loss can jump by an order of magnitude within a single campaign. A long-term rate computed across two campaigns averages a near-zero regime with an aggressive one and returns a comfortable middle number that describes no period of the equipment's actual life.
Practically, this means the module needs an event on the equipment record — a reline, a plate replacement, a coating renewal — that opens a new short-term window without truncating the long-term series. Both rates stay live, both are shown, and the governing rate is whichever is more conservative on that day. When a plant's short-term rate detaches sharply from its long-term rate, that gap is itself the signal worth escalating, and it is exactly the signal a single-rate site is structurally unable to produce.
The required thickness problem on equipment no code covers
A corrosion rate on its own schedules nothing. Remaining life is the distance from the current thickness to the required thickness, divided by the governing rate, and on a Section VIII vessel the required thickness comes out of UG-27 with a defined joint efficiency, allowable stress and corrosion allowance. On a cyclone cone, a riser duct or a cooler casing there is no such formula, because the failure mode is not internal pressure. It is buckling, distortion under dust load, loss of a support attachment, or a hot gas leak that damages structure around it.
What plants actually did in the absence of a formula is the interesting part, and it is where the group divergence hurts most. One site had a structural engineer calculate a minimum shell thickness per section and reused it. Four sites applied a flat percentage of nominal — usually 50 percent, occasionally 60 — with no stated basis. Three used whatever number appeared in the inspection contractor's template, which changed when the contractor changed. Rates were then compared across the group as if the finish line were in the same place, and it was not.
The workable answer is not to pretend a code applies. It is to make the required thickness an authored, owned, versioned engineering value attached to an equipment class, set once, reviewed on a defined cycle, and referenced by every survey rather than retyped into every report. Then a remaining life of four years at plant A and four years at plant B mean the same thing, and the group can rank them honestly. The value can still be conservative or crude. What it cannot be is different at every site and undocumented at all of them.
What a group roll-up has to reconcile before it means anything
Corporate reporting for a multi-site cement group usually wants three things: which equipment across the estate is degrading fastest, which shutdowns need scope added, and whether the group's total metal loss exposure is rising or falling. Every one of those questions is a comparison, and every comparison fails if the underlying calculation differs. Before any dashboard is worth looking at, six things have to be identical: baseline definition, divisor, governing-rate rule, required thickness basis, stored unit, and location identity.
Location identity is the one most often overlooked and the most damaging. When an inspection contractor changes, point numbering typically restarts from one. The plant now has two surveys of the same riser duct elbow filed under unrelated identifiers, which means the system cannot compute a short-term rate for that point at all — not incorrectly, but not at all. The equipment then appears healthy because its only rate is a long-term average from a baseline it can still see. Permanent, plant-owned condition monitoring location identifiers, with contractor point numbers kept as searchable aliases, close that hole.
Units and rounding are less dramatic but equally corrosive to trust. Storing millimetres at one site and mils at another and converting on the way into a report is fine. Rounding stored values, rather than displayed ones, is not: two rounding steps on a small number over a five-year interval can shift a rate by several percent, which is enough to reorder a ranked list. Round at the point of display and never write the rounded value back.
How to evaluate a corrosion rate module for a multi-plant group
The evaluation question that separates real systems from templated ones is simple to ask and hard to fake: show me the same survey data, and show me the report telling me which rate governed and why. If the output gives a single number with no indication whether it came from the short-term or long-term path, the software has not standardised anything — it has hidden the divergence one layer deeper. Ask to see a location where the short-term rate exceeds the long-term rate and confirm the governing selection is visible on the printed deliverable, not just in a settings screen.
Second, test the migration. Bring a real spreadsheet from your worst site, including the plant that reported cumulative loss with no rate and the plant whose point numbering restarted twice. Ask how nominal-baseline records are distinguished from measured-baseline records after import, and whether that distinction survives onto reports. A system that quietly promotes a purchase-specification nominal into a measured baseline will generate negative corrosion rates from mill over-tolerance and then hide them.
Third, test the roll-up. Load two plants with deliberately different divisors and confirm the group view either reconciles them or refuses to combine them. Silent combination is the failure mode that costs money, because it produces a confident, wrong ranking. Finally, ask what happens to non-code equipment: where the required thickness lives, who can change it, and whether the change is versioned. If the answer is a free-text field on the survey form, you are buying the same problem with a nicer interface. For a scoped walkthrough against your own registers, contact info@atlantisndt.com — demonstrations and consultations are arranged on request.
Why does a cement group get different corrosion rates from identical equipment at two plants?
Because almost nothing in the calculation is fixed by an external code. The plants differ on baseline thickness, on whether the divisor is calendar or operating time, on which rate governs, and on what required thickness means for a cyclone cone. Each choice is defensible alone. Combined, they can move a reported rate by a factor of two on genuinely identical equipment, which is why the numbers refuse to roll up.
Should the divisor be calendar years or kiln operating years?
Report on calendar years, because that is what an inspection interval and a shutdown plan are measured in, and store operating hours as a parallel field. Operating-year rates are useful for comparing a plant at 0.78 run factor against one at 0.93, but they understate the calendar exposure of dew-point corrosion, which continues during a stop as the duct cools through the acid dew point. Never mix the two in one register.
How do you set a required thickness on a preheater cyclone that no code covers?
An engineer sets it, once, for the equipment class, and the number is versioned rather than retyped per survey. The basis is usually structural: buckling of the cone under its own load plus dust burden, or the shell stress a kiln section sees in ovality. What matters for standardisation is less the exact number than that all eight plants reference the same authored value rather than each contractor inventing one.
Does a refractory reline reset the corrosion rate?
It resets the short-term window, not the long-term record. A shell section protected by intact refractory loses almost nothing, then loses metal quickly once the lining thins or spalls. A long-term rate spanning two campaigns averages those regimes and reads low. The module should let a lining renewal open a new short-term window while preserving the full long-term series, and show both, so the campaign-driven behaviour is visible rather than smoothed away.
What happens to the rate when the reading comes from a different point than last time?
You get a positioning artifact rather than metal loss. On abraded surfaces in clinker handling and cooler casings, thickness varies sharply over a few inches, so a probe placed 40 mm from the previous spot can swing the apparent short-term rate wildly. Permanent location identity, a photograph or sketch, and a physical mark are the fix. Without re-registration, the reading is an observation, not a point on a trend.
Is API 510, 570 or 653 inspector certification training part of this offer?
No. This module is inspection management software for calculating and standardising corrosion rates across a multi-plant group. Atlantis NDT does deliver NDT method training to ASNT SNT-TC-1A and ISO 9712 across UT, RT, MT, PT, ET, VT, PAUT and TOFD, ASNT Level III consulting, and report validation, but API inspector certification programmes are separate and are not sold here. Ask for a demo or a scoped quote at info@atlantisndt.com.
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.