Deadlegs in API 570: Why They Corrode and How They Are Inspected

Short answer: A deadleg is any part of a piping system that normally sees little or no flow, such as a blanked branch, a line behind a closed block valve, a spare pump connection, a level bridle or a drain. API 570 treats deadlegs as a special-concern area because stagnant water, solids and concentrated corrosive species can thin them faster than the active line. The code expects them to be identified, monitored at the stagnant end and the tie-in, examined with methods able to find localised loss, and removed where they are not needed.

This guide explains how API 570 frames deadlegs, why they corrode differently from the line they hang off, how inspection plans place condition monitoring locations (CMLs) on them, which NDE methods work at which diameters, and the mistakes that let a deadleg leak between turnarounds. The deadleg text described here follows the API 570 4th edition (2016). The 5th edition was published in February 2024; secondary summaries report the deadleg provisions carried forward, but you should confirm wording and any clause numbering against your licensed copy of the current edition (API 570 announcement, api.org). Nothing here reproduces code text; it is a plain-language explanation.

What API 570 counts as a deadleg

API 570 defines deadlegs by flow, not by shape. The test is whether the component normally has little or no significant flow, and whether it is still connected to the process. If both are true, it is a deadleg, however short it is and whatever it was originally designed for.

The 4th edition gives a long list of examples, and it is worth walking a unit with that list in hand because many deadlegs are not obvious on a P&ID:

• Blanked branches left from a project that never came back, or from equipment that was removed.
• Lines with normally closed block valves, including lines that are opened only during start-up, shutdown or regeneration.
• Lines with one end blanked and pressurised dummy support legs, where a pipe stub used as a support is open to the process.
• Stagnant control-valve bypass piping, which is closed for years at a time.
• Spare pump piping, where the standby pump suction and discharge sit full but idle.
• Level bridles and instrument stand-pipes.
• Relief valve inlet and outlet header piping, which only flows when the valve lifts.
• Pump trim bypass lines, high-point vents, sample points, drains and bleeders, and instrument connections.
• Piping that is no longer in service but is still connected to live process, the classic "abandoned in place" line.

The last item is the one most plants underestimate. A line taken out of service by closing a valve is not isolated in the corrosion sense. It still holds process fluid, it still breathes moisture through the valve seat over time, and it often loses its owner when the unit is reorganised. API 570's position is that such piping either gets inspected as a deadleg or gets physically disconnected.

Why deadlegs corrode faster than the active line

The short version is that the process chemistry inside a deadleg is not the process chemistry of the main line. Without flow, the fluid separates, settles and concentrates, and corrosion inhibitors or neutralisers dosed into the main stream do not reach it.

API 570 names the drivers in general terms: contaminated water, solids, temperature differences and concentration of corrosive species. In refinery and petrochemical practice those show up as:

• Water drop-out. In hydrocarbon lines, free water settles to the low point of a stagnant leg. The bottom of a horizontal deadleg becomes a water-wetted surface even when the main line is dry.
• Under-deposit corrosion. Scale, corrosion product, salts and sand settle in the stagnant section. The metal under the deposit becomes a separate electrochemical cell and pits.
• Concentration of corrosive species. API 570 specifically mentions ammonium salts, organic acids, H2S and acidic deposits. In overhead systems, chloride and ammonium salts deposit where temperature drops in a stagnant branch.
• Temperature differences. A deadleg off a hot line cools along its length. Somewhere along that gradient the stream crosses its dew point, and condensation corrosion starts. Older API 570 text also noted that high points in hot systems can corrode from convective currents set up in the stagnant leg.
• Microbiologically influenced corrosion. Water-bearing deadlegs in cooling water, firewater and some hydrocarbon services are classic sites for MIC, because bacteria thrive where flow is absent.

These mechanisms are described in more detail in API RP 571 (see our API 571 damage mechanisms guide). The practical point is that a deadleg's corrosion rate cannot be borrowed from the parent circuit. A line that loses a few thousandths of an inch a year in the flowing section can lose much more at the stagnant end of a bridle or bypass.

API 570 also points out that deadlegs on primary piping deserve extra weight, because they generally cannot be valved off from the process. A leak at a deadleg on a main header is a leak of the full unit inventory.

What API 570 expects the inspection plan to do

API 570 does not set a separate deadleg interval. Instead it requires the owner-user's inspection plan to recognise deadlegs as an area of special concern and to monitor them in a way that will actually catch localised loss. The 4th edition text sets out several expectations that a defensible programme follows.

Monitor both ends

The stagnant end and the connection to the active line both need thickness data. The stagnant end collects water and deposits. The tie-in sees turbulence, temperature cycling and the boundary between flowing and stagnant fluid. A deadleg with readings only at one end is not covered.

Track potentially corrosive deadlegs as separate circuits

API 570 says deadlegs with CMLs that are potentially corrosive should be tracked separately from the mainline circuit. The reason is arithmetic. If the deadleg readings are averaged into the parent circuit, the corrosion rate of the deadleg disappears into the parent's rate, and the remaining-life calculation for the deadleg is wrong. Separate circuits give each deadleg its own long-term and short-term rate (our guide to long-term and short-term corrosion rates explains the calculation). Where several deadlegs share the same anticipated damage mechanism and rate, they may be grouped, but not with the active line.

Cover all four quadrants on horizontal legs that may not be liquid full

For horizontal deadlegs that may hold a liquid level, the 4th edition calls for examination points in all four quadrants. The bottom of the pipe sees water and deposits; the top sees vapour-space and condensation corrosion; the liquid line sees interface attack. A single reading on the top of the pipe, where access is easiest, can miss the worst metal entirely.

Ask a corrosion specialist where the CMLs go

API 570 recommends consulting a corrosion specialist on deadleg CML placement, especially above and below liquid interfaces. This is the code acknowledging that deadleg damage is location-specific and that the specialist's knowledge of the process chemistry is what puts the CML on the right spot.

Consider removal

Finally, API 570 recommends considering removal of deadlegs that serve no process purpose and are potentially corrosive. Every deadleg removed is one fewer circuit to inspect for the life of the unit. Removal is a management-of-change item for the owner, but the inspection programme is usually the first place where the case is made.

NDE methods for deadlegs, and what each one can and cannot find

Spot UT thickness readings are the default CML method on most piping, and they are the wrong default for many deadlegs. A single transducer reading measures the metal under the probe. Deadleg damage is often localised pitting, under-deposit attack or a groove along the water line, which a spot reading can miss by a few millimetres. API 570's 4th edition text points towards methods that cover area or profile, and names several of them.

The pattern most programmes settle on is: locate the problem area (thermography, guided wave screening or profile RT), quantify it (scanning UT or PAUT, RT measurement), then trend it at fixed CMLs. For small-diameter deadlegs, profile radiography is usually the most efficient single method, because it shows both the wall and what is sitting inside the pipe. Where fouling or under-deposit corrosion is suspected, the 4th edition specifically points to profile RT.

For more on choosing methods by damage type, see our NDE method selection by damage mechanism guide.

Worked example: a stagnant control-valve bypass

Consider a carbon-steel control valve station on a wet hydrocarbon line in a crude unit. The bypass around the control valve has two block valves and a globe valve, and it has been closed for years. The main line circuit has a steady, low corrosion rate from spot UT trending.

Applying API 570 thinking, the inspector and corrosion specialist ask four questions. First, is the bypass a deadleg? Yes: it normally sees no flow and is connected to the process. Second, what does the stagnant fluid do? With the line horizontal and the bypass at the same elevation, free water can settle on the bottom of the bypass section between the block valves and the tee. Third, where are the CMLs? Points at the tee on the main line, at the bottom quadrant of the horizontal bypass section, and on the low side upstream of each block valve, with top and side quadrants covered at least once to confirm the bottom is the worst position. Fourth, which method? The bypass is small bore, so profile RT is shot at the tees and the low section, and any area showing loss is proved up with UT.

The result in a case like this is usually one of three outcomes. The bypass is in reasonable condition and is set up as its own circuit with its own rate. It shows active under-deposit pitting, and the inspector raises a recommendation for replacement and for a flushing or operating practice that keeps it from sitting stagnant. Or the owner decides the bypass is not needed and removes it under management of change. All three are owner and inspector decisions. The NDE provider's job is to give them data that represents the worst metal, not the most accessible metal.

Common mistakes with deadleg inspection

• No register. Deadlegs are not identified as a population, so they are only found when they leak. A walk-down against the API 570 example list, cross-checked with P&IDs and isometrics, is the starting point.
• Averaging into the parent circuit. Deadleg CMLs are kept in the mainline circuit, so their rate is diluted and their remaining life is overstated.
• Top-of-pipe readings only. Readings are taken where access is easiest rather than where the water and deposits sit.
• Ignoring the tie-in. The stagnant end is measured, but the junction with the active line is not.
• Spot UT on pitting damage. Single readings are trended for years on a deadleg where the damage is pitting between points.
• Abandoned lines without an owner. Lines taken out of service but left connected drop out of every circuit list.
• Forgetting CUI. Insulated deadlegs are often cold or cycling, which makes them prime corrosion-under-insulation locations as well as internal-corrosion locations. See CUI inspection.
• Not re-checking after a process change. A change in crude slate, water content or wash-water practice changes deadleg chemistry. API 570 expects inspection plans to be reviewed when operating conditions change.

Deadlegs, small bore and injection points: the three related special-concern areas

API 570 lists deadlegs alongside other areas of special concern, and three of them overlap so often that programmes manage them together. Injection points concentrate damage over a short length where a chemical or wash water enters the stream; see our injection point inspection guide. Small-bore and auxiliary piping, including vents, drains and instrument connections, overlaps heavily with the deadleg list and adds vibration fatigue at socket welds near rotating equipment; see small-bore and auxiliary piping. And deadlegs that are insulated are also CUI locations.

The practical consequence is that a single walk-down can populate all three registers. The same small-bore drain may be a deadleg (stagnant), small bore (vibration and mechanical damage) and a CUI location (insulated low point). Recording it once with all three attributes keeps the programme from inspecting it three times or, more often, not at all.

Deadleg circuits also take a piping class like any other circuit. The class sets the maximum thickness and external visual intervals; see API 570 piping service classes. Many owners assign deadlegs the same class as the parent line and then shorten the interval based on the deadleg's own measured rate.

Documentation the inspector expects

When a deadleg programme is audited, by the owner's own inspection group, a PSM auditor or a jurisdiction, the inspector of record expects to see a trail from identification to decision. A complete record usually includes:

• A deadleg register: location, line number, parent circuit, service, reason it is stagnant, and whether removal has been evaluated.
• Circuit and CML definitions on an isometric, with deadlegs as separate circuits where potentially corrosive.
• The basis for CML placement, including the corrosion specialist's input on liquid interfaces.
• NDE reports that state the method, procedure, technician certification level, coverage and the actual values, not only "acceptable". A structured piping inspection record format keeps these fields consistent.
• Corrosion-rate and remaining-life calculations per deadleg circuit.
• Inspection recommendations, with due dates and closure evidence, including any removal recommendations.

In US facilities covered by OSHA's process safety management standard, piping inspections and tests must follow recognised and generally accepted good engineering practice and be documented (29 CFR 1910.119(j)). API 570 is widely used as that practice for process piping, which is why deadleg coverage often comes up in PSM audits. In Canada, provincial pressure-equipment regulators set the in-service framework; in Alberta, for example, owners operate under ABSA-accepted integrity management programmes. Confirm the requirements that apply to your jurisdiction.

How Atlantis supports deadleg programmes

Atlantis NDT performs the NDE that a deadleg programme depends on: profile radiography of small-bore deadlegs by licensed radiography crews, scanning UT and corrosion mapping on larger legs and low-point quadrants, phased array UT at tie-ins and branch welds, spot UT trending at defined CMLs, and guided wave screening of long inaccessible runs. Work is done by ASNT-certified technicians under ASNT Level III oversight, and results are delivered to the owner's API 570 authorized piping inspector, who remains inspector of record and sets the circuits, intervals and recommendations. See piping circuit and CML inspection for how we structure CML data. Request a deadleg NDE quote; we respond within 24 hours.

Frequently asked questions

What is a deadleg in piping?

A deadleg is a section of piping that normally has little or no flow but is still connected to the process. Examples include blanked branches, lines behind closed block valves, control-valve bypasses, spare pump piping, level bridles, relief valve headers, vents, drains and sample points.

Why are deadlegs a special concern in API 570?

Because stagnant fluid lets water, solids and corrosive species settle and concentrate, and inhibitors in the main stream do not reach them. Deadlegs can therefore corrode faster than the active line, and API 570 asks inspection plans to treat them as an area of special concern.

Where should CMLs be placed on a deadleg?

API 570 expects thickness monitoring at both the stagnant end and the connection to the active line. On horizontal deadlegs that may not be liquid full, the 4th edition calls for points in all four quadrants, and it recommends asking a corrosion specialist about placement above and below liquid interfaces.

Should deadlegs be a separate circuit from the main line?

Potentially corrosive deadlegs with CMLs should be tracked separately from the mainline, so their corrosion rate and remaining life are calculated on their own data rather than averaged into the parent circuit.

What NDE method is best for small-bore deadlegs?

Profile radiography is usually the most efficient choice for vents, drains and instrument connections, because it shows both wall loss and internal deposits. Areas of concern are often proved up with UT.

Is spot UT enough for deadleg inspection?

Often not. Deadleg damage tends to be localised pitting or under-deposit attack that a spot reading can miss. Scanning UT, corrosion mapping or profile RT give area coverage; spot UT is then used to trend fixed points.

Does API 570 require deadlegs to be removed?

API 570 recommends considering removal of non-essential deadlegs that are potentially corrosive. It is a recommendation for the owner to evaluate under management of change, not an absolute requirement.

What inspection interval applies to a deadleg?

There is no separate deadleg interval. The circuit takes its piping class maximum intervals, and the measured corrosion rate and remaining life can shorten them. The owner-user and authorized inspector set the interval.

Can abandoned piping still be a deadleg?

Yes. Piping that is no longer used but remains connected to the process is a deadleg under the API 570 definition, and it should be inspected as one or physically disconnected.

Where can I find the current API 570 deadleg text?

In the current API 570 edition, purchased from API. The 5th edition was published in February 2024. Free copies circulating online are usually old editions and should not be used to set requirements.

Planning a deadleg survey for a turnaround or a PSM audit? Talk to Atlantis NDT about deadleg NDE, or read our API 570 code requirements overview. For the whole piping programme, get a quote within 24 hours.