Piping inspection where the data quality decides the answer
Atlantis runs piping inspection to API 570: circuits defined by damage mechanism, condition-monitoring locations placed where the metal actually thins, thickness measurement and external visual at the class interval, and CUI, injection-point and dead-leg coverage. Every reading is tied to a numbered CML on an inspection isometric, so corrosion rate and remaining life come out of paired readings rather than guesswork.
Piping is the API program where the data quality decides the answer. A vessel has one nameplate; a unit has hundreds of circuits, thousands of readings, and a corrosion rate that is only as good as the pairing behind it. API 570 groups piping into circuits by expected damage mechanism, then places condition-monitoring locations inside each circuit to characterize how that circuit corrodes. Recommended maximum intervals are five years for Class 1 thickness measurement and external visual, ten years for Class 2 and Class 3 thickness measurement, and three years for injection points — capped at half the remaining life. The failure mode we see most in inherited data sets is not a missed reading. It is a reading recorded against the wrong baseline, which silently corrupts every remaining-life number downstream, and UT thickness report review is where that surfaces first.
Source: API 570, Piping Inspection Code: In-service Inspection, Rating, Repair, and Alteration of Piping Systems, Fifth Edition (February 2024) with Addendum 1 (2025) for current edition status and the recommended maximum interval table; clause text read directly from the Fourth Edition (February 2016) with Addendum 2 (2018) — 5.1.1, 5.6.1–5.6.3, 5.8, 5.10, 6.2.1, 6.3.3, 6.3.4, 7.1–7.4, 7.6, 7.9–7.11. Also 29 CFR 1910.119(j)(4) (OSHA PSM). Verified August 2026.
| Circuit class | Example services | Thickness measurement | External visual |
|---|---|---|---|
| Class 1 | Auto-refrigerating flammables, rapidly vaporizing C2–C4, H2S over 3 % weight in gas, anhydrous HCl, hydrofluoric acid, piping over water or public throughways | 5 years | 5 years |
| Class 2 | On-site hydrocarbons that slowly vaporize, on-site hydrogen, fuel gas and natural gas, on-site strong acids and caustics | 10 years | 5 years |
| Class 3 | On-site hydrocarbons below flash point, off-site distillate and product lines, tank farm piping, off-site acids, caustics and fuel gas | 10 years | 10 years |
| Class 4 | Steam and condensate, air, nitrogen, water, lube and seal oil, ASME B31.3 Category D services | Optional | Optional |
| Injection points | Water, steam, chemical or additive injection circuits | 3 years | By class |
| Soil-to-air interfaces | Buried-to-aboveground transitions | See API 574 | By class |
What piping circuit inspection establishes
A piping inspection program has to produce, for every circuit in the unit, a defensible corrosion rate, a remaining life for the limiting component, a next inspection date, and a repair list that someone owns. Nothing about that is exotic. What makes piping harder than vessels is scale and identity: a refinery unit holds thousands of measurement points, and each one is worth exactly as much as the certainty that this year's reading was taken where last year's was.
We build the program around that certainty. Circuits are defined before points are chosen, points are numbered and drawn before readings are taken, and readings are paired against a measured baseline rather than a specification. Where a program is inherited rather than built, we audit it first — the arithmetic is quick, the data archaeology is not. Independent review of that layer sits under our ASNT Level III consulting services.
Circuitization: dividing the unit by damage mechanism
API 570 requires piping to be broken down into systems and circuits before an inspection plan can be written. Systems are often defined at process flow diagram level; circuits are usually defined at P&ID level. The organizing principle is the expected damage mechanism, not the line number — sections of piping sharing a damage mechanism, a material and a set of operating conditions belong in the same circuit, because that is what lets one set of readings speak for the whole group.
Get this wrong and everything downstream inherits the error. A circuit that spans a spec break, a temperature change or an injection point averages two corrosion behaviours into one number that describes neither. Dead legs may be combined into a single circuit only when their anticipated damage mechanisms and corrosion rates are alike. Where the population is large enough that circuit-level judgement needs a formal risk basis puts likelihood and consequence behind the grouping.
CML selection: which points earn a place in the program
CMLs go where the metal actually thins. API 570 directs attention to elbows and tees, reducers, valve discharge points, dead legs, low points, insulated sections, injection points, soil-to-air interfaces, weld heat-affected zones, and immediately upstream and downstream of material changes such as spec breaks. The allocation must consider service-specific damage mechanisms described in API 571 and API 574 — localized corrosion, stress cracking, CUI and high-temperature hydrogen attack are all named as CML conditions, not just wall thickness.
Density follows risk, not geometry. More CMLs go on circuits with higher leak consequence, higher expected or experienced corrosion rates, higher localized-corrosion potential, more fittings, branches and dead legs, higher CUI potential, greater thickness variability, higher short-to-long-term rate ratios, and greater process variability. Fewer go on long, straight, relatively non-corrosive runs. Where CMLs are substantially reduced or eliminated, a corrosion specialist must be consulted and the reasoning recorded.
CML placement and examination points
A CML is an area; an examination point is a spot within it. Where appropriate, thickness measurements should include each of the four quadrants on pipe and fittings, with particular attention to the inside and outside radius of elbows and tees where flow turns and erosion-corrosion accelerates. As a minimum, the thinnest reading — or an average of several readings within the point — is recorded at each examination point.
Finding the minimum is a scanning task, not a spot-reading task. API 570 states that the minimum thickness at a CML can be located by ultrasonic scanning or profile radiography, with electromagnetic techniques used to identify thin areas that are then measured by UT or RT. A single shot in the middle of a 6-inch square that happens to miss a groove is not wrong arithmetic; it is the wrong measurement, and no downstream calculation can recover from it.
CMLs should be marked on inspection isometrics, and the piping itself may be marked so the same locations get measured every cycle. That physical repeatability is what turns a series of numbers into a trend.
A CML is not a random reading
The practical difference between a CML and a random UT reading is identity across time. A CML has a number, a drawing position, a documented baseline, a history and an owner. A random reading has a value. Two randoms taken a year apart on the same spool produce a difference that looks like corrosion and may simply be two different points on a plate with mill tolerance variation, or a probe placed 40 mm away on the far side of a weld cap.
This is why campaign data collected by a crew that did not draw the isometrics so often fails on arrival. The readings are real, the instrument was calibrated, the technician was certified — and the numbers still cannot be paired against history, so the corrosion rate has to be rebuilt from scratch. We check pairing integrity before we accept any inherited data set, which is the same discipline described in what makes an NDT report defensible.
Corrosion rate from paired readings, and which rate governs
API 570 gives two point-to-point rates. The long-term rate is the initial thickness minus the actual thickness, divided by the years between them, where initial means the first measurement at that CML or the thickness at the start of a new corrosion rate environment. The short-term rate is the previous thickness minus the actual thickness, over the years between those two. Remaining life is actual thickness minus required thickness, divided by the governing corrosion rate.
Both rates are calculated and compared to see which produces the shorter remaining life. The authorized inspector, in consultation with a corrosion specialist, selects the rate that best reflects the current process. When the short-term rate changes significantly from the previously identified rate, the inspector should consult a corrosion specialist to determine the cause rather than simply re-baselining. Required thickness is the greater of the pressure design thickness and the structural minimum thickness — not whichever one is convenient.
Statistical analysis is permitted as an alternative to point-to-point, and must be documented and conservative. It is explicitly not applicable to circuits with significant localized, unpredictable corrosion mechanisms, which rules it out for exactly the circuits where the temptation to smooth the data is strongest.
The nominal-thickness trap
API 570 states plainly that installation inspection should document baseline thickness measurements to be used as initial readings for corrosion rate calculations, in lieu of nominal and minimum design thickness data from specifications, data sheets and drawings. That sentence exists because using nominal as the baseline is the most common corruption in piping data, and the most invisible. Pipe is delivered within a mill tolerance, so a nominal figure can overstate or understate the metal that was actually installed.
The consequence compounds. A long-term rate built on a wrong initial thickness is wrong; remaining life divides a correct margin by an incorrect rate; the next inspection date follows the wrong remaining life; and the MAWP calculation, which uses actual thickness minus twice the estimated corrosion loss before the next inspection, inherits it too. If the calculated rate turns out to disagree with observed loss, API 570 requires the rate used for the next period to be adjusted to agree with the actual rate found.
This is a data governance problem more than an NDT problem, which is why we push baselines, tolerances and provenance into a system of record rather than a spreadsheet. Our fully customized ERP for NDT companies holds circuit and CML registers with the audit trail attached.
Injection points, dead legs, soil-to-air interfaces and CUI
Injection points get their own circuit and their own interval. The recommended upstream limit is a minimum of 12 in. (300 mm) or three pipe diameters upstream of the injection point, whichever is greater. The downstream limit is the second change in flow direction past the injection point, or 25 ft (7.6 m) beyond the first change in flow direction, whichever is less. CMLs go on appropriate fittings within the circuit and at both its upstream and downstream limits, and the preferred methods are radiography and UT scanning or closely spaced UT grids.
Corrosion under insulation is inspected on externally insulated carbon and low-alloy piping operating between 10 °F (−12 °C) and 350 °F (175 °C), and can be conducted as part of the external inspection. Dead legs, small-bore piping, threaded connections and soil-to-air interfaces each have their own extent rules, because each breaks the assumption that a circuit's average behaviour describes its worst point. Where these circuits sit inside an OSHA PSM-covered process, the documentation rules in OSHA PSM mechanical integrity NDT apply to every one of these examinations.
What the piping record has to show
API 570 specifies what to record per circuit: material of construction and piping specification, diameter, operating and design pressures and temperatures, flange rating, process fluids, class where RBI is not used, insulation, heat tracing and PWHT status, whether the circuit is a dead leg, injection point or intermittent service, the corrosion rate and remaining life of at least the limiting examination point, the maximum interval for external inspection, the maximum interval for thickness measurement, any unusual localized corrosion mode needing specialized techniques, and features that could corrode rapidly during an upset.
Inspection isometrics carry all significant components, material and spec breaks, diameter, insulation status, secondary piping, all CMLs with enough information to locate them, orientation and legible detail, circuit numbers, continuation drawings, and pipe support locations. Repair recommendations must be tracked with corrective action and date, priority or target date, the circuit identifier affected, and a list of temporary repairs needing follow-up. A management system for reviewing outstanding recommendations is required, not optional.
That register is an asset. Once circuits, CMLs and histories are structured, they can be surfaced against the plant model so an engineer sees the thinning elbow in place rather than in a row of a spreadsheet — which is what our digital twins platform is built to do. Send us your isometrics and last campaign data through contact and we will scope the work: affordable, accessible, fully customizable, quote on request.
What is a CML in piping inspection?
A condition-monitoring location is a designated, recorded point on a piping circuit where inspection data is collected repeatedly so damage can be trended. It carries a location, a documented baseline thickness and a history, and it may hold several examination points — four quadrants on an elbow, for instance. CMLs replaced the older term thickness-monitoring location and cover cracking, CUI and hydrogen attack as well as wall loss.
How often does process piping need inspecting?
By class, and never longer than half the remaining life. Class 1 takes thickness measurement and external visual at five years. Class 2 takes thickness measurement at ten years and external visual at five. Class 3 takes both at ten years. Class 4 is optional. Injection points run at three years. When remaining life drops below four years, the interval may be the full remaining life to a maximum of two.
How many CMLs does a piping circuit need?
Enough to characterize how that circuit corrodes. API 570 sets no count per foot. More CMLs go where leak consequence is high, corrosion rates are high or localized, the circuit is complex with branches and dead legs, CUI potential is high, or the short-to-long-term rate ratio is high. Fewer go on long, straight, non-corrosive runs. A circuit with perfectly uniform corrosion could in theory be monitored with one.
Why does a wrong nominal thickness invalidate remaining life?
Because the long-term corrosion rate is the difference between an initial thickness and the current thickness, divided by the years between them. Substitute a spec-sheet nominal for a real measured baseline and the numerator is wrong by the mill tolerance plus any as-built difference. That error flows straight into remaining life, into the next inspection date, and into the MAWP calculation that uses actual thickness minus twice the projected loss.
Who is qualified to perform piping inspection?
An API 570 authorized piping inspector owns the inspection and the evaluation. Examiners perform the NDE under approved procedures and do not need inspector certification, though the owner/user can require ASNT or equivalent credentials. Thickness measurements are obtained by the inspector or by examiners working to the inspection plan, and the inspector selects the corrosion rate that best reflects current process conditions, in consultation with a corrosion specialist.
What does a piping inspection company deliver?
Inspection isometrics showing every CML with enough detail to find it again, circuit records carrying material, diameter, design conditions, flange rating, service, class, insulation and special-circuit status, the corrosion rate and remaining life of the limiting examination point, both maximum intervals, and a tracked repair recommendation list with target dates. That package is what a successor inspector can actually re-use.
How to select CMLs and TMLs on a piping circuit: a practical guide
Short answer: select condition monitoring locations (CMLs) circuit by circuit, starting from the damage mechanisms you expect, not from the geometry. Put them where those mechanisms act first: fittings, flow disturbances, dead legs, low points, injection and mix points, soil-to-air interfaces, spec breaks and insulated sections with CUI potential. Add more where the consequence or the corrosion rate is higher, and fewer on long, low-risk straight runs. Then mark each one so the same spot is measured every time. API 570 and API RP 574 set out the criteria. The authorized piping inspector applies them.
The sections above describe how we run piping circuit inspection. This guide goes one level deeper into selection itself. It covers what the terms mean, including the "CML port" and "thickness monitoring location" phrasing people search for. It then covers a step-by-step selection workflow, placement by component, how to choose the measurement technique at each CML, and how to add, move or retire CMLs without breaking the history. Code points are paraphrased from API 570 (current edition: 5th, February 2024, with Addendum 1 in 2025) and API RP 574 (reissued as a 5th edition alongside it). Where we refer to clause content, it was read from the 4th edition of API 570. Check your licensed current copy before quoting it.
CML, TML, examination point, CML port: what the terms mean
Direct answer: a CML is a designated area on a piping circuit where periodic examinations are made to monitor its condition. TML (thickness measurement location) is the older term for the same idea when only thickness was monitored. An examination point is a spot inside a CML where a reading is taken. A "CML port" is usually an inspection plug or opening in insulation that gives access to a CML without stripping the insulation.
CML vs TML. API moved from TML to CML in the API 510 9th edition and API 570 3rd edition era. The change recognized that a monitoring location may track more than wall thickness. It might be a crack-susceptible weld, a CUI-prone area or a hot spot checked for creep or HTHA. Many plants still say "TML" for thickness points, and both terms appear in older data sets. When you merge inherited registers, map TML IDs to CML IDs one to one. Do not renumber them.
Examination point. A CML is an area, and an examination point is a location within it. Industry practice often treats a CML on pipe as a band or circle a few inches across, with one or more examination points. API 570 lets the inspector record either the thinnest reading or an average of several readings at an examination point. Choose one approach per CML and keep it. Switching between minimum and average from one campaign to the next produces corrosion rates that are artefacts.
Thickness monitoring location. This is a common plain-English name for a TML or thickness CML. Some owners also use "thickness monitoring point". The meaning is the same: a fixed, recorded spot where wall thickness is measured repeatedly so loss can be trended.
CML ports. On insulated lines, removing and replacing insulation at every CML each campaign is slow and expensive and can damage the jacketing. Owners therefore fit inspection ports: removable plugs or small openings in the insulation and jacketing at each CML, often with a cap or label. API 570 lists insulation plugs among the ways to identify examination points, along with stencils, isometrics, RFID tags and monitoring buttons. The risk is that a poorly sealed port lets water into the insulation and creates the CUI it was meant to help monitor. Ports need good caps, sealing and periodic checks. Spot readings through a small port can also miss localized attack beside it. That is why CUI programmes use ports alongside, not instead of, insulation removal at susceptible areas. See our page on corrosion under insulation inspection.
CML "in construction". In piping construction and commissioning, CML usually comes up as the baseline. API 570 says installation inspection should record baseline thickness readings at CMLs, to be used as the initial readings for corrosion rates instead of nominal thickness. A CML defined and measured at construction gives you the first point of every future trend. (In general construction, "CML" can also mean other things, such as a consolidated materials list. On a piping isometric it means a monitoring location.)
A step-by-step workflow for selecting CMLs
Direct answer: circuitize first, then assign damage mechanisms, then rate consequence and expected corrosion, then walk the isometric component by component placing CMLs where those mechanisms act. Size and mark the examination points, take baselines, and review the selection after the first two campaigns.
- Define the circuit. Group piping with the same material, process conditions and expected damage. Break the circuit at spec breaks, temperature changes, injection points and dead-leg boundaries.
- Assign credible damage mechanisms. Use API RP 571 and process knowledge: general thinning, localized corrosion, erosion-corrosion, under-deposit corrosion, CUI, high-temperature mechanisms, and cracking. API 570 says CML allocation should consider service-specific damage mechanisms, as described in API 574 and API 571. Our API 571 damage mechanisms guide is a starting point.
- Judge consequence and rate. Higher leak consequence, higher or more variable corrosion rates and higher localized-corrosion potential all call for more CMLs. Low consequence and low, stable rates allow fewer.
- Walk the isometric. Place CMLs at the components where the assigned mechanisms concentrate (see the placement table below). Add representative CMLs on straight runs so general thinning is also tracked.
- Decide coverage at each CML. How many examination points, which quadrants and which technique: spot, grid, scan, mapping or profile RT.
- Mark and identify. Number every CML uniquely on the inspection isometric and, where practical, on the pipe. Record orientation and offset from a permanent feature.
- Baseline. Measure every new CML before or soon after start-up. If the line is already in service, the first reading becomes the initial value for the long-term rate.
- Review. After two or more campaigns, compare rates across CMLs. Add CMLs where readings are scattered or rates are high. Consider reducing where history shows low, stable rates. Record the reasons.
The authorized piping inspector owns this process and consults a corrosion specialist where the code calls for one, for example before significantly reducing or eliminating CMLs.
Where CMLs go, component by component
Direct answer: fittings and flow disturbances get priority because turbulence and impingement concentrate damage. Dead legs, low points and soil-to-air interfaces get CMLs because water and deposits collect there. Spec breaks get CMLs on both sides. Straight runs get representative CMLs, more on longer or more corrosive runs.
| Component or feature | Why it gets a CML | Placement notes |
|---|---|---|
| Elbows and bends | Erosion-corrosion and impingement on the outside radius | Cover outside radius, inside radius and cheeks; four quadrants as appropriate |
| Tees | Flow split and impingement at the branch and opposite wall | Run and branch sides; area opposite the branch |
| Reducers and expanders | Velocity change; turbulence downstream | On the fitting and just downstream |
| Downstream of control valves and orifices | Turbulence, flashing, cavitation | Several diameters downstream, plus the first fitting |
| Dead legs | Stagnant fluid, water settling, deposits | At the low point and near the connection to the active line |
| Low points and drains | Water and solids collect | Bottom of line, at the 6 o'clock position |
| Injection and mix points | Localized attack from poorly mixed fluid | Own circuit; see our injection point guide |
| Soil-to-air interfaces | Moisture and coating breakdown where pipe enters ground or concrete | At and just below the interface; may need excavation |
| Spec breaks | Material change; different corrosion behaviour | Immediately upstream and downstream |
| Insulated sections with CUI potential | Water under insulation | At penetrations, terminations, supports and low points |
| Long straight runs | General thinning trend | Representative spacing; fewer on low-risk runs |
| Small-bore and threaded connections | Thin wall, vibration, threads remove metal | Follow API 570's specific provisions; often covered by VT and RT |
API 570 also lists conditions that justify more CMLs: higher leak consequence, higher or variable corrosion rates, localized-corrosion potential, complexity such as dead legs and injection points, CUI potential, process variability and industry failure history. Conditions for fewer CMLs include low consequence, relatively non-corrosive service and long straight runs. CMLs can be eliminated only in narrow cases, such as very low consequence with non-corrosive history and no expected change. Write down the reasoning whichever way you go.
Choosing the measurement technique at each CML
Direct answer: match the technique to the mechanism. Spot UT is fine for uniform thinning. Localized mechanisms need grid UT, scanning, PAUT corrosion mapping or profile radiography. A CML is only as good as its ability to find the minimum thickness in the area it represents.
| Expected damage | Suitable technique at the CML | Note |
|---|---|---|
| General uniform thinning | Spot UT at fixed examination points | Keep the same points and orientation each time |
| Erosion-corrosion at fittings | Grid UT or scanning over the fitting | Extra attention to the elbow's outside radius |
| Pitting, under-deposit, MIC | PAUT corrosion mapping or close grid UT | Spot readings commonly miss isolated pits |
| Bottom-of-line water attack | Circumferential band scan or mapping | Focus on 4 to 8 o'clock |
| Insulated lines | Profile RT, or UT through ports after screening | Profile RT shows wall and deposits through insulation |
| Long runs, limited access | Guided-wave screening, then UT/RT at indications | Screening locates; UT or RT sizes |
| Cracking mechanisms | MT, PT, shear-wave UT, PAUT or TOFD | Choose by mechanism and geometry |
Thickness measurement should follow a written procedure with calibration on a reference block of the same material, temperature correction where the line is hot, and coating handling. Our article on UT thickness measurement requirements covers the details. For localized mechanisms, see corrosion mapping, and for profile work on insulated lines, radiographic testing.
Adding, moving and retiring CMLs without breaking the history
Direct answer: treat the CML register as a controlled record. Add CMLs when data or process changes justify it. Never move an existing CML silently; retire it and create a new one. Never reuse an ID. Record the reason for every change, especially reductions.
- Adding. Add CMLs when readings at a location scatter widely, when the short-term rate jumps, when a new mechanism is found, or after a process change. A new CML starts its own history from its first reading.
- Moving. If a CML was placed in the wrong spot, retire it and create a new one at the right place. Moving it under the same ID mixes two locations into one trend.
- Retiring after replacement. When a spool or fitting is replaced, retire its CMLs and baseline new ones. The new metal has a new initial thickness.
- Reducing. CML optimization is legitimate, but reductions should be reasoned, recorded and reviewed with a corrosion specialist where the code calls for it.
- Management of change. Feedstock, temperature, velocity or chemical-treatment changes can bring new mechanisms. Link MOC to a CML review.
Our article on building a CML register that survives ten years covers ID rules and contractor handover in more depth.
A worked example, described qualitatively
Consider a carbon-steel naphtha line between two units. It runs about 200 feet, with eight elbows, two tees, a control valve, a dead-leg bypass and a section that goes underground at a road crossing. Part of it is insulated. The inspector sets it up like this:
- Circuits. The main line is one circuit. The bypass dead leg is its own circuit because water can collect there. The road-crossing interface is flagged as a soil-to-air location.
- Mechanisms. General thinning and erosion-corrosion at fittings, under-deposit corrosion in the dead leg, CUI on the insulated section and external corrosion at the soil-to-air interface.
- CMLs. One on each elbow and tee, one downstream of the control valve, representative CMLs on the straight runs, two in the dead leg (low point and junction), CMLs at insulation terminations and supports, and CMLs at both sides of the soil-to-air interface.
- Techniques. Grid UT on fittings, mapping in the dead leg, profile RT on the insulated section with ports at selected CMLs, and VT plus UT at the interface after excavation.
- Review. After the second campaign, two elbows show higher rates. The inspector adds a CML on the next elbow downstream and changes those CMLs from grid to mapping. Straight-run CMLs with stable, low rates are kept but not increased.
The point is that every CML has a reason tied to a mechanism, a technique matched to that mechanism, and a record of why it exists.
Common CML selection mistakes, and the jurisdiction note
Direct answer: the usual mistakes are placing CMLs where access is easy rather than where damage happens, spacing them evenly by length, using spot readings for localized mechanisms, and changing CMLs without records. All of them weaken the corrosion rates the inspection interval depends on.
- CMLs placed on easy-to-reach straight runs while fittings are skipped.
- Even spacing by length, with no link to mechanism.
- No CMLs in dead legs, at low points or at soil-to-air interfaces.
- Insulation ports left unsealed, creating CUI.
- Single spot readings where pitting is the expected mechanism.
- CML IDs reused or renumbered when isometrics are redrawn.
- Nominal thickness used as the initial reading because no baseline was taken.
In the United States, OSHA PSM (29 CFR 1910.119(j)) requires inspection and testing following recognized and generally accepted good engineering practice, and API 570 with API RP 574 is the common reference for process piping. In Canada, provincial regulators such as ABSA in Alberta and TSSA in Ontario oversee pressure piping, with CSA B51 as the base code and owner-user programme requirements in some provinces. Confirm the rules that apply with your jurisdiction.
Refinery piping corrosion: where the inspection effort goes
In refinery process piping, thinning concentrates where the damage mechanisms described in API RP 571 predict it: high-temperature sulfidation in crude and vacuum unit piping, naphthenic acid corrosion in hot crude streams, ammonium chloride and sour water corrosion in overhead and effluent systems, amine corrosion in treating units, and corrosion under insulation. API 570 ranks piping into classes by the consequence of a leak, with the highest-consequence services in Class 1, and singles out injection points and dead legs for extra inspection; API 574 describes the piping inspection practices used to find the damage, thickness measurement at CMLs and radiographic profiling among them.
The circuits, techniques and intervals come from the plant's inspection plan. Atlantis NDT supplies the examination and the records: UT thickness at the CMLs, corrosion mapping services where loss is localised, and profile radiography where insulation stays on. See also ultrasonic inspection services.
Sources: API RP 571, Damage Mechanisms Affecting Fixed Equipment in the Refining Industry; API 570, Piping Inspection Code; API 574, Inspection Practices for Piping System Components.
More CML questions
What are CML ports?
Inspection plugs or openings in pipe insulation at condition monitoring locations. They let technicians take UT readings without stripping insulation. They must be well sealed, because a leaking port lets water in and causes corrosion under insulation.
What is a thickness monitoring location?
A fixed, recorded spot on a pipe where wall thickness is measured repeatedly to trend corrosion. API 570 called it a TML in older editions and now uses the broader term condition monitoring location (CML).
What are condition monitoring locations (CMLs)?
Designated areas on piping or equipment where periodic examinations monitor condition. That usually means thickness, but it can include other damage such as cracking or CUI. Each CML has a unique ID, a marked position and a reading history.
What does CML mean in inspection?
Condition monitoring location. It is the basic unit of a piping inspection plan: readings at the same CML over time give the corrosion rate, remaining life and next inspection date.
What does CML mean in construction?
On piping construction, CML refers to the monitoring locations defined on isometrics and baselined at installation. Those first readings become the initial thickness for future corrosion rates instead of nominal thickness.
Is a TML the same as a CML?
Essentially yes for thickness. TML is the older term; CML is broader and can include monitoring for other damage mechanisms.
Who selects CMLs on a piping circuit?
The owner's API 570 authorized piping inspector, with a corrosion specialist where needed. NDE contractors measure at the CMLs and can flag locations that look under-covered.
Need a CML survey or a review of an inherited CML register? Send us your isometrics for a quote; we reply within one business day. You can also ask about corrosion mapping at high-risk CMLs.
Next step
Atlantis performs the NDE this work needs and reports to your inspector of record. To go further, request an inspection quote. We reply within one business day.