IC TRACK · INTRODUCTION TO INTERNAL CORROSION

Oxygen Corrosion in Pipeline Systems

Oxygen corrosion — why oxygen is an intrusion rather than an ingredient, differential aeration, and the field signatures that flag it.

Basics & Theory ~10 minutes PDH/CEC eligible

They’ve cut a section out of the gathering line to tie in a new block valve. The line’s isolated, blown down, and drained — the old spool is out on the ground and the two open ends sit waiting for fit-up. You put a light on the inside of the pipe, and along the bottom of the bore, that 6-o’clock stretch where water drops out and lies, it’s crusted with rust-orange mounds. Knock one off and there’s a pit under it, going down into good steel. This is sweet gas — mostly methane, a little CO₂, no H₂S to speak of — and it came out of the ground carrying no oxygen at all, the way formation gas always does. And yet the bottom of this pipe is pitting from the inside, and the iron showing up in the samples keeps climbing.

Orange rust is the tell. CO₂ leaves a gray-green carbonate; H₂S leaves a black sulfide. Bright orange-brown rust is the fingerprint of oxygen — and oxygen is the one aggressive gas on this whole track that has no business being in the pipe at all. Pipeline gas is held to a tight oxygen spec for exactly that reason. So the question this cut-out puts in front of you isn’t only “how bad is it.” It’s “where is the air getting in?” — because on a line that should carry none, oxygen corrosion is first and foremost the story of a leak.

This module takes oxygen end to end: what it does to iron and why it speeds corrosion up, how the numbers we chase split across two very different scales, the one mechanism that explains most of the damage it does, where it sneaks into a system, what it leaves behind, how you measure it without fooling yourself, and how operators keep it out. By the end, that pitted spool on the ground will point you straight at the question worth asking.

The odd gas out

Oxygen isn’t an ingredient — it’s an intrusion

Start by placing oxygen against the two gases the last two modules covered. CO₂ and H₂S come up out of the ground with the gas. They’re part of the stream — ingredients you expect, measure on the gas analysis, and manage. Oxygen is different in kind. Gas straight from the formation carries essentially no oxygen, and pipeline-quality gas is held to a tight oxygen spec on top of that. When oxygen shows up inside the line, it almost never rode in with the product. Something let it in.

That single fact reshapes how you hunt it. With CO₂ or H₂S, the gas is a given and the work is managing its effect. With oxygen, the gas being present is the finding — a system reading measurable oxygen is telling you a seal, a blanket, or a purge has failed somewhere upstream. Most oxygen in a pipeline system is there because something mechanical is failing, not because the chemistry of the gas changed. Hold that thought — we’ll come back to it, because it’s the reframe that separates a tech who chases oxygen from one who chases the leak that’s feeding it.

CO₂ and H₂S are ingredients that come up with the gas. Oxygen is an intrusion — formation gas holds almost none, and pipeline gas is specced to keep it that way. When you measure oxygen inside the line, the reading itself is the alarm: air is getting in somewhere it shouldn’t.

What oxygen does to iron

You already know the reaction at the center of every module in this set. Iron in water gives up metal one atom at a time, and each atom that leaves sheds two loose bits of charge behind — electrons:

The reaction the whole set is built on
Fe → Fe²⁺ + 2 e⁻
Iron leaves the wall as a dissolved ion and drops two electrons at the surface. This is the metal-loss event — the thing every aggressive fluid is just a different way of feeding.

Here’s the part that matters for oxygen. Those shed electrons can’t just pile up — something has to take them away, or the reaction chokes on its own leftovers and slows down. In water with nothing hungry for them, the takers are limited, so the whole thing runs slowly and tends to stall. Oxygen is the hungriest electron-taker a pipeline is likely to meet. Dissolved oxygen grabs those electrons eagerly and combines with water:

Why oxygen speeds everything up
O₂ + 2 H₂O + 4 e⁻ → 4 OH⁻
Oxygen consumes the electrons iron sheds — fast. With the electrons cleared out of the way, the iron-dissolution reaction is free to keep running hard.

That’s the whole story of why oxygen is so aggressive, and it’s worth saying in plain field terms: oxygen doesn’t change what corrosion is — it clears the traffic jam so the same corrosion runs faster. Same iron loss, same reaction the sweet and sour modules taught, just more of it, because oxygen is greedy for exactly what the dissolving iron is trying to get rid of. It’s a strong oxidizing agent, and a strong oxidizing agent is precisely a substance that’s hungry for electrons.

And it doesn’t stop at speeding up general loss. When oxygen is in the water, it tends to drive corrosion into pits — and when it rides along with CO₂ or H₂S already in the stream, it can push pitting rates up by roughly ten times over what those gases would do alone. Oxygen doesn’t just add to the problem. It multiplies the problem that’s already there.

Every iron atom that dissolves sheds two electrons that have to go somewhere. Oxygen is the hungriest taker in the pipe — it clears those electrons fast, so the same iron-loss reaction keeps running hard. Same chemistry, more of it. And it steers the damage toward pitting.

How small a number we’re chasing

Oxygen gets measured in two different places, on two different scales, and mixing them up is the fastest way to talk past each other. Keep them in separate lanes:

In the gas — oxygen shows up on the gas analysis, in parts per million (ppm), the same report that gives you CO₂ and H₂S, read off a gas chromatograph or a dedicated oxygen analyzer. This is the number that lands on your desk.
In the water — down where the corrosion actually happens, at the low points where water drops out, oxygen is measured as dissolved oxygen, in parts per billion (ppb), with a portable meter or an ampoule kit.

Same element, two phases, two yardsticks — and you can’t line the two numbers up against each other. Oxygen-by-volume in the gas and dissolved oxygen in the water aren’t the same measurement, so “50 on the gas report” and “10 in the water” don’t compare head to head. Take them one at a time.

On the gas report. Gas straight out of the Marcellus or Utica should read essentially no oxygen — it came out of the ground carrying none. So any oxygen on the analysis is telling you air got into the system somewhere. In our area the working rule is to flag anything over about 50 ppm and chase the cause, and to hold oxygen as low as practicable the rest of the time. That number isn’t “how much oxygen is safe” — it’s a practical line that says “stop and investigate this one” instead of note-and-watch.

Oxygen on a gas report is contamination, full stop. Native gas carries none, so a real reading means air is getting in — an integrity flag worth chasing to root cause, not a shrug. And at high enough levels it stops being only a corrosion problem: oxygen and hydrocarbon together in a pipe is a flammability and process-safety concern in its own right. Either way, oxygen on the analysis is never something to wave off.

In the water. Here the numbers get very small, because this is where oxygen does its work. Open surface water — a pond, a pit, an open tank — holds roughly 4 to 12 ppm of dissolved oxygen just from sitting in contact with air. A sealed system should hold almost none. The old rule of thumb is that dissolved oxygen above about 50 ppb — that’s 0.05 ppm — is enough to be considered corrosive, and systems built around water, like injection and disposal, are held tighter still, down toward a 10-ppb target. Those are parts per billion, a thousand times finer than the ppm on your gas report — which is exactly why the two numbers live in separate lanes. (And cold water holds more dissolved oxygen than warm, so the same system can run higher in January than it does in July.)

Two places, two units, one enemy. The gas analysis (ppm) is your early warning that air is in the system; the water at the low points (ppb dissolved) is where it bites. On a line that should carry none, either reading is the alarm — and wet plus oxygen is the worst combination there is.

Same wall, different oxygen: the mechanism to really get

If you take one mechanism out of this module, take this one. It explains more oxygen damage than any other single idea, and it’s the root of a whole family of problems you’ll meet by other names later in the track.

Picture two patches on the same stretch of pipe wall. One sits out in the flowing bulk stream, washed with whatever the fluid carries. The other sits under a deposit — a mound of settled solids at the bottom of the line, a patch of scale, the tight gap under a gasket edge, the crevice at a poorly made fit-up. That sheltered patch is starved of oxygen, because oxygen has to work its way through the deposit to reach it and gets used up on the way.

Now the two patches are living in different worlds, and that difference does the damage. Out in the open, oxygen is plentiful — it mops up electrons easily, and the steel there stays relatively protected. Under the deposit, oxygen is scarce — so that starved patch becomes the spot that keeps giving up iron. The electrons it sheds travel through the metal to the oxygen-rich areas where oxygen is waiting to take them. The result is the part that surprises people the first time: the metal dissolves fastest exactly where the oxygen is lowest — down in the sheltered pit under the deposit — even though the oxygen driving it all is coming from the water nearby.

When one patch of wall sees less oxygen than its neighbors — under a deposit, in a crevice, at a gasket edge — the oxygen-starved patch is where iron dissolves fastest. The rich areas are fed by oxygen and stay protected; the starved area quietly eats a pit. Oxygen from the open water, damage under the deposit.

This uneven-oxygen setup has a name — a differential aeration or oxygen-concentration situation — but the name matters less than the picture. Anything that changes how much oxygen reaches one part of the wall versus another can set it up: settled sand and solids, scale, biofilm, a gasket lip, a lap in a fit-up, the shielded bottom of the pipe where water and debris collect. That’s why one plain mechanism shows up wearing so many different faces:

Under-deposit attack — pits growing beneath mounds of solids or scale at the low points.
Crevice corrosion — attack in the tight gap of a fit-up, a threaded joint, a lap.
Gasket-edge and shielded-spot attack — wherever geometry hides a patch of wall from the flow.

One more field detail worth carrying: which patch is doing the dissolving isn’t fixed forever. If the deposits shift, the flow changes, or a spot gets cleaned, the roles can move. The lesson isn’t to memorize which spot loses — it’s to recognize that uneven oxygen across a surface concentrates the attack, and that the sheltered, low-flow, under-deposit places are where oxygen quietly does its worst.

Where this goes next. IC-005 gives you the mechanism — uneven oxygen drives localized attack. The morphology and metallurgy of pitting itself, and the deep treatment of under-deposit corrosion as its own phenomenon, live in later modules in the track. Your Foundation job is to see a deposit or a crevice on an oxygen-bearing line and know what it’s setting up underneath.

Follow the air

Where oxygen gets in

Back to the reframe from the top: on a system that should carry none, measurable oxygen means something is letting air in. Ingress is almost always mechanical, and it very often happens where the system runs below atmospheric pressure and actively pulls air in through any weakness. The usual suspects on gas systems:

Compressor and vapor-recovery (VRU) seals — the suction side runs under vacuum, so a worn seal or a leaking VRU draws air straight into the stream. The classic one.
Valves and fittings opened during maintenance — anything broken into for a repair, or a tie-in or new line put in service without a proper purge, can leave air behind in the system.
Frac fluid and methanol — flowback fluids and injected methanol can carry oxygen in with them if they weren’t handled to keep air out.
Tanks and vessels breathing — a tank without a working gas blanket breathes air on every level change; let a nitrogen or gas blanket lapse and the vessel does the same.
Low-pressure gathering under vacuum — a gathering line pulled below atmospheric pressure pulls air in through any leak it has.

Notice what almost every item on that list has in common: it’s a piece of hardware that’s worn, opened, or left unfinished — not a change in the chemistry of the gas. That’s the gold nugget of this whole module, and it’s the reframe a good tech carries into any oxygen investigation:

Most oxygen ingress is mechanical, not chemical. Solving an oxygen problem starts with finding the leak — the worn compressor seal, the lapsed blanket, the line that went back in service without a clean purge. Chemical treatment comes after that, as the polish, never as the first answer.

Put it on a real system to make it concrete. Walk a low-pressure gathering system from the wells forward: gas comes off the wellheads, into a compressor station where the suction side runs under vacuum (a worn seal or a VRU leak here pulls air straight in), past valves and fittings that were opened during the last maintenance turn (any one left un-purged is an air path), and out into the gathering line, where water drops out and collects at the low points — exactly where the oxygen that got in will do its work. If an oxygen reading climbs on the analysis, you don’t reach for a chemical — you walk that system and find the station where the air is getting in. That walk is the heart of oxygen work.

What it leaves behind

Oxygen has a look, and learning to read it saves you a lab trip. Its corrosion products are iron oxides — the same family as ordinary rust — and they run orange to red-brown, and they’re bulky: they take up far more room than the metal they came from, so they pile into loose mounds rather than tight films. Those mounds are tubercles — the crusty orange-brown knobs you find wherever water sits in a line, very often with a pit hiding underneath.

The overall pattern usually reads as two things at once. There’s broad general thinning where oxygen-bearing water washes the wall — more uniform-looking than the localized bite of CO₂ or H₂S in a lot of cases. And laid on top of that, the localized attack from the differential-aeration mechanism: deep pits concentrated under deposits and tubercles, often with a relatively clean rim of metal around a deposit and the real damage down at the deposit’s edge. That combination — orange bulky product, general loss, and hot-spot pits under the mounds, right down the 6-o’clock of the line where water lies — is the field signature that says oxygen before any number comes back.

Reading the color. Gray-green or white carbonate scale points at CO₂. Black, sooty sulfide points at H₂S. Bright orange-to-brown, bulky, mounded rust points at oxygen. It isn’t proof on its own — but on a line that’s supposed to carry none, orange tubercles are a strong reason to go looking for where air is getting in.

Measuring it without fooling yourself

Because oxygen where there should be none is the whole finding, the first question about any oxygen reading is whether it’s real — and that’s a discipline on both the gas side and the water side.

On a gas sample, the reading is only as good as the bottle. A sample is supposed to come in clean — techs are trained to purge and fill without letting air in, and done right, that’s exactly what happens. The backstop for the rare bottle that does catch air is a quick check the lab runs: compare the oxygen against the nitrogen. Air is about one part oxygen to four parts nitrogen, so if the two show up tracking that ratio, the bottle most likely picked up air during sampling — the reading gets set aside and the location resampled. If the oxygen is there without air’s matching nitrogen, it’s real oxygen in the line. A contaminated bottle means a false reading, a wasted trip, and a resample — which is exactly why pulling it clean the first time is what matters. The ratio check is the safety net, not the plan.

On the water side, dissolved oxygen is read with a portable meter or a colorimetric ampoule kit, working down into the low ppb. It’s the same trap in a different form: draw the sample carelessly, let it splash or stand open, and you’ve measured the air instead of the line. So calibrate the meter the day you use it — a meter off by a few ppb is off by the whole range that matters — keep the sample off the air, and read it right at the point.

The first question about any oxygen number is whether it’s real. On gas, the lab checks oxygen against nitrogen to catch a bottle that caught air; on water, you keep the sample sealed from air and calibrate on the day. Then the rule the whole track runs on: one reading is a note, two that agree are a signal.

Keeping oxygen out, in priority order

Operators control oxygen in two broad moves, and the order is the whole point — cheapest and most effective first.

First, keep it out. On a gas system this is nearly the whole game, and it’s all mechanical: seals and packing kept in good order, compressor and VRU maintenance, gas or nitrogen blankets that are actually maintained on tanks and vessels, and lines properly purged before they go back in service after a tie-in or repair. Fix the ingress and most oxygen problems simply stop, with no chemistry at all. This is where the mechanical-ingress reframe pays off.

Second, take out what still gets in. Where oxygen can’t be kept out entirely, it’s removed. An oxygen scavenger — a chemical dosed to react with and consume the remaining oxygen — is the common tool, used on both gas and water systems; the usual families are the sulfites (often with a catalyst to react fast enough at line temperatures) and hydrazine (a clean reaction, but a probable carcinogen whose use is increasingly restricted). On high-volume water-handling systems — produced-water and injection is the classic case — the workhorse is mechanical vacuum deaeration: pull the water under vacuum and the dissolved oxygen comes out of it, no reagent needed, with a scavenger to polish the last few ppb. You won’t run a deaerator on a gas line, but you’ll meet these on the water side of the business, and it’s the same oxygen either way.

The cheapest oxygen is the oxygen that never gets in. On a gas system, prevention — sealing, maintenance, blanketing, clean purges — is nearly the whole program. Removal, by scavenger or (on water systems) by deaeration, is what handles the rest. Chemistry is the polish, not the plan.

Whose call the chemistry is. Which scavenger, what catalyst, what dose rate, how it fits with the other chemicals on a system — that’s mitigation design, done by the integrity engineer or corrosion specialist and taught in the mitigation modules higher in the track. Your Foundation job is to know that prevention comes first, and to recognize that a scavenger running hard on a system with a bad seal is treating a symptom while the leak keeps feeding it.

When oxygen meets sour: a trap worth naming

One flag before we put it together, because it catches programs off guard. If oxygen gets into a system that also carries H₂S — air leaking into a sour line, or a scavenger program added without accounting for the sour side — the oxygen and the H₂S can react part-way and throw elemental sulfur, which is itself corrosive. The nasty result is that a half-treated sour system can end up more aggressive than one left fully sour or kept fully sweet. It’s not a mechanism to work through here — just a tripwire to recognize: mixing an oxygen intrusion into a sour line is not a simple “two problems added together,” and it deserves a specialist’s eye.

There’s a flip side worth carrying too, and it points at the next module. Oxygen and certain bacteria have opposite tastes: the microbes that do some of the worst internal damage thrive precisely where oxygen is absent. Controlling oxygen doesn’t just slow oxygen corrosion — it changes the whole biological picture inside the pipe, for better and for worse. That link between oxygen and the bugs is exactly where this set goes next.

From a suspicion to a plan

Drop the whole module into the working chain you already know — indirect data, the repeatability gate, direct data, mitigation designed upstream. Here’s how it runs when oxygen is the suspect.

Indirect data. It starts on the gas analysis: the oxygen column against that 50-ppm flag, read alongside the rest of the report — water content and dew point, CO₂, H₂S, and any liquids — because wet plus oxygen is the worst combination there is. Then the field signs: orange tubercles and pits in a cut-out, bulky rust on a pulled coupon, the operating story of a recently commissioned line or a compressor seal that’s been weeping. And the move from the hook works in reverse too — find the orange pitting first, and the next step is to pull the recent gas analyses and read the oxygen column to see what it’s been doing. Read together, these say “look at oxygen.”

The repeatability gate. Same rule as always, and it bites hardest here because oxygen readings are so easy to spoil: one aggressive read is a note, two consecutive reads that agree are a signal worth acting on. Confirm before you escalate.

Direct data — the number you hand up. When the signal clears the gate, coupons go into the spots this module flagged — low points, drips, dead legs, downstream of a suspect compressor — and they come back wearing that orange rust. The coupon turns exposure into a hard number: weigh the metal it lost, and you get a corrosion rate. That rate is the module’s one working calculation, and it’s the language the integrity engineer speaks:

The one calculation in this module
corrosion rate (mpy) = (3.45 × 10⁶ × W) ÷ (A × T × D)
W = mass the coupon lost (grams) · A = coupon surface area (cm²) · T = time in service (hours) · D = steel density, about 7.87 g/cm³. “mpy” is mils per year — thousandths of an inch — the way most North American coupon reports read. Swap the constant to 8.76 × 10⁴ and the same formula gives millimeters per year (mm/yr) for metric programs.

Worked example — is this line holding?

Setup. A carbon-steel coupon rides a gathering-line drip for 45 days. Pulled and cleaned, it has lost 1.0 g; its exposed area is about 40 cm². The program flags anything over 5 mpy for a closer look.

Step 1 — time in hours. 45 days × 24 = 1,080 hours.

Step 2 — put it in the formula. rate = (3,450,000 × 1.0) ÷ (40 × 1,080 × 7.87) = 3,450,000 ÷ 339,984 ≈ 10 mpy (about 0.26 mm/yr).

Read. 10 mpy against a 5-mpy flag — twice over. This line isn’t holding; the number says investigate. That’s not the tech’s fix to design — it’s the tech’s job to hand the specialist a clean, repeatable rate that says the program needs tightening.

Mitigation, designed upstream, executed by you. The rate crossing the threshold sends the problem up the chain. The response comes back to the two moves from earlier — find and fix the ingress first, then remove what still gets in — and which lever, at what setting, is designed by the integrity engineer or corrosion specialist. The tech runs it in the field and keeps the coupons and the oxygen reads coming, because follow-up data is what proves the fix is holding. On oxygen especially, the most valuable thing a tech ever contributes isn’t a chemical — it’s the observation that closed the leak.

Reading the rate. There’s no single pass/fail number that fits every line — the corrosion-rate categories in the industry monitoring standard are guidelines, and each system sets thresholds that fit its own metal, pressure, and risk. The smart move is often an investigation threshold rather than a hard action limit: a rate that says “something changed, go find out why,” which is exactly the tech-to-specialist handoff this set is built around.

Back to the spool on the ground

So — the cut-out spool, orange tubercles down the 6-o’clock with pits under them, on a sweet gathering line that’s supposed to carry no oxygen at all. The easy read was “some rust, throw a scavenger at it.” Here’s the read this module builds instead.

That orange is oxygen’s fingerprint, and oxygen on a sweet line means air is getting in somewhere — probably a mechanical leak, and a compressor seal running under vacuum is the first place to look. The pits under the tubercles are differential aeration doing its work: the metal dissolving fastest right where the deposits starve it of oxygen. The move isn’t to dose a chemical and hope. It’s to pull the recent gas analyses and read the oxygen column against that 50-ppm flag; confirm any live reading with a clean, bracketed sample you trust; walk the system from the compressor forward hunting the way in; drop coupons at the low points and let them come back with a rate; and hand that rate up against the line’s threshold. Find the leak, and the scavenger you might have reached for first turns out to be the last thing the line needs — or a small polish on top of a fix that actually holds.

Same spool, different eyes. The other guy saw rust. You see a gas that doesn’t belong in the pipe, a way in that hardware failure opened, and a plan that starts with the leak instead of the symptom.

Three gases down — next the pipe fights back with something alive

This is module 5 of 8 in the Corrosion Mechanisms in Pipeline Fluids set. You now carry oxygen end to end: the intrusion-not-ingredient frame, the electron-clearing reason it speeds corrosion up, the two-scale numbers and why you never compare them directly, differential aeration as the mechanism behind under-deposit and crevice attack, the mechanical ways air gets in, the orange-tubercle signature, disciplined measurement with the oxygen-to-nitrogen check and the repeatability gate, the prevent-first order of control, and the mass-loss corrosion rate you hand up the chain.

The last three modules covered gases — CO₂, H₂S, and now oxygen. The next one turns to a threat that isn’t a gas at all: the microbes living in the water, doing their worst exactly where oxygen is kept out. Same four ingredients, same one reaction — this time the culprit is alive.

Key takeaways

Oxygen is an intrusion, not an ingredient. CO₂ and H₂S ride in with the gas; formation gas holds almost none and pipeline gas is specced to keep it out. Measuring oxygen inside the line is itself the alarm — air is getting in.
Why it speeds corrosion up: iron dissolving sheds electrons (Fe → Fe²⁺ + 2 e⁻); oxygen is the hungriest taker of those electrons (O₂ + 2 H₂O + 4 e⁻ → 4 OH⁻), so it clears the traffic jam and the same iron-loss reaction runs harder. It also steers damage toward pitting — up to ~10× worse alongside CO₂ or H₂S.
Two places, two units, one enemy. In the gas, oxygen reads in ppm on the gas analysis — native gas ≈ 0, so flag anything over ~50 ppm and chase the cause. In the water, dissolved oxygen reads in ppb (surface water 4–12 ppm; >~50 ppb corrosive; ~10-ppb injection target). Don’t compare the two numbers directly — and wet + oxygen is the worst combination.
Differential aeration is the mechanism to master. Where a deposit, crevice, or gasket starves one patch of oxygen, that starved patch dissolves fastest — fed by the oxygen in the water nearby. One mechanism behind under-deposit attack, crevice corrosion, and shielded-spot pitting.
Most ingress is mechanical. Compressor/VRU seals under vacuum, valves opened in maintenance, un-purged tie-ins, frac fluid/methanol, breathing tanks, lapsed blankets. Find the leak first; chemistry comes after.
The signature is orange. Bulky orange-to-brown rust and tubercles with pits underneath — general thinning plus hot-spot pits under deposits, down where water lies.
Measure it with discipline. On gas, the lab checks oxygen against nitrogen to catch a bottle that caught air (resample); on water, calibrate on the day and keep the sample off the air. One read is a note, two that agree are a signal.
Control, prevention first: keep it out (sealing, maintenance, blanketing, clean purges) → then remove what gets in (scavenger, or vacuum deaeration on water systems). Chemistry is the polish, not the plan.
The one calculation: corrosion rate (mpy) = (3.45×10⁶ × W) ÷ (A × T × D) from a weighed coupon — the number you hand up against the line’s investigation threshold. (Swap the constant to 8.76×10⁴ for mm/yr.)
Flag: oxygen leaking into a sour line can throw corrosive elemental sulfur — a half-treated sour system can be worse than fully sour or fully sweet.

References & further reading

Internal Corrosion Field Guide — the canonical field reference for this track; operating-parameter effects, where corrosion occurs, liquid analysis, and coupon-based corrosion-rate measurement in working-tech terms.
Field Guide to Internal Corrosion Mitigation and Monitoring for Pipelines — companion reference connecting oxygen chemistry to monitoring and to deaeration and scavenger practice.
Internal Corrosion: Sweet and Sour — Field Notes from RCS; the aggressive-gas context this oxygen module sits alongside.
NACE SP0106 — control of internal corrosion in steel pipelines and piping systems; the canonical internal-corrosion practice standard.
API RP 1160 — managing system integrity for hazardous liquid pipelines, including water-handling and oxygen context.
49 CFR Part 192 — federal gas-pipeline regulation; internal corrosion control requirements.
Corrosion Basics: An Introduction — foundational text on corrosion by water and oxygen and the forms corrosion takes.

Apply — three field-grounded problems

Turn a coupon into a rate, score the risk, plan the handoff

Three short problems, all built from the oxygen read you just finished, all kept at the level you’d use on the job. There’s exactly one calculation — a coupon’s mass loss turned into a corrosion rate — and you run it once. The other two are reading and judgment: in the second you score the differential-aeration risk on a real segment and rank the fixes; in the third you take a sweet gas line wearing oxygen’s fingerprint and carry it from signal to handoff. Read each setup, think it through, then click each step to compare with our working.

How to use this lesson. Read the setup. Think before you click. Each step reveals what we’d say.

Problem 1 · Turn a coupon into a rate

Weigh the loss, run the number, make the call

Setup. A carbon-steel weight-loss coupon rode a gathering-line drip — a low point where water collects — for one quarter. It’s back from the lab, cleaned and weighed:

Exposure90 days

Mass before24.60 g

Mass after (cleaned)24.25 g

Exposed area32 cm²

Steel density7.87 g/cm³

Program flag5 mpy

What’s the general corrosion rate in mpy — and does this line pass, or does it get a closer look?

Step 1 — find the mass lost and the time in hours

Mass lost is just before minus after: 24.60 − 24.25 = 0.35 g. Time in service goes into hours for the formula: 90 days × 24 = 2,160 hours.

Step 2 — run the corrosion-rate formula

Same formula from the read, in mils per year:

rate (mpy) = (3.45 × 10⁶ × W) ÷ (A × T × D)
= (3,450,000 × 0.35) ÷ (32 × 2,160 × 7.87)
= 1,207,500 ÷ 543,974
≈ 2.2 mpy (about 0.056 mm/yr)

W is the 0.35 g lost, A the 32 cm² of exposed coupon, T the 2,160 hours, D the 7.87 g/cm³ density of steel. (Swap the constant to 8.76 × 10⁴ if you want the metric answer.)

Step 3 — make the call

2.2 mpy against a 5-mpy flag — the line’s holding on the general rate. But don’t stop there, because this is oxygen. A weight-loss coupon gives you the uniform rate, spread evenly over the whole coupon. Oxygen’s real danger is localized — deep pits under deposits — and a low general number can hide a nasty pit that ate a small area fast. So the read is: general loss is under the flag, log it and trend it, but check the coupon for pitting under any deposit before you call the line clean. A passing rate is still data you hand up — the value is the trend over time, not the single number.

A coupon turns exposure into a rate you can hand up — but on oxygen, the uniform rate is only half the story. Always look for the pit a general number can hide.

Problem 2 · Score the risk

Where does the metal go, and what do you fix first?

Setup. A survey on a gathering segment turns up a picture worth reading:

Servicesweet gas, wet

Gas-analysis O₂above the 50-ppm flag

Low pointssediment / solids piled up

Upper crownclean, flow-swept

Wall loss (inspection)worst at 5–7 o’clock, under deposits

Three questions: which areas of the inside wall are oxygen-rich and which are oxygen-starved? Where does the metal dissolve fastest, and why? And what three actions would you rank to cut the loss — in what order?

Step 1 — rich vs. starved

The oxygen-rich areas are the ones the flow washes: the upper crown and any wall the moving stream keeps swept. The oxygen-starved areas are down under the sediment piles at the 5–7 o’clock low points — oxygen has to work through the deposit to reach that wall, and it gets used up on the way. Same pipe, two very different worlds on the surface.

Step 2 — where the metal goes, and why

The metal dissolves fastest under the deposits — the oxygen-starved patches — which is exactly where the inspection found the worst wall loss. That’s differential aeration: the starved patch keeps giving up iron, and the electrons it sheds travel through the metal to the oxygen-rich areas where the oxygen is waiting to take them. The oxygen driving it comes from the water in the open flow, but the damage lands under the deposit. The clean upper crown isn’t lucky — it’s protected precisely because it’s washed with oxygen.

Step 3 — rank the three fixes

In order:

1. Find and fix the oxygen ingress. The gas analysis is over the flag — oxygen is getting into a line that should carry none. Walk the system for the mechanical way in (a compressor seal under vacuum, a valve left open on the last maintenance, a lapsed blanket). The cheapest oxygen is the oxygen that never gets in.
2. Take away the shelter. Clean the deposits out — pigging and water/low-point management — so there are no shielded patches for the starved cells to form under.
3. Then chemistry, if it’s still needed. A scavenger is the polish once the leak is closed and the line is clean — and that dose is the specialist’s call.

The trap is starting at #3. Dosing a scavenger while air keeps pouring in through a bad seal is treating the symptom while the cause keeps feeding it.

The pits are under the deposits, fed by oxygen from the open flow. The first fix is the leak, then the deposits — the chemical is last, not first.

Problem 3 · Signal to handoff

A sweet line, orange rust, and the move

Setup. A gathering line that’s read sweet for years — CO₂, no H₂S — gets a section cut out to tie in a valve. The bore shows orange tubercles down the bottom with pits underneath. You pull the recent gas analyses:

Two quarters ago< 10 ppm

Last quarter15 ppm

This month70 ppm

Iron countsclimbing

H₂Snone

What do you conclude, and what’s the move — before anyone reaches for a treatment?

Step 1 — read the signal

Orange tubercles with pits, on a sweet line, with the gas-analysis oxygen climbing from a trace to 70 ppm — over the 50-ppm flag — and iron counts rising with it. That’s oxygen ingress: air is getting into a line that came out of the ground carrying none. The color, the pits under the deposits, and the trend on the analysis all point the same way. It’s contamination, not a change in the gas.

Step 2 — clear the gate before you act

One reading is a note; two that agree are a signal. Before this goes anywhere, make sure the 70 ppm is real — that it isn’t a bottle that caught air during sampling. The lab check is the oxygen-to-nitrogen ratio: if the oxygen tracks air’s nitrogen signature, resample; if it’s oxygen without that nitrogen, it’s real line oxygen. Confirm with a clean, repeated read. Here the rising trend across quarters and the physical pitting back each other up — but you still confirm before you escalate.

Step 3 — the handoff

This isn’t the tech’s to fix by dosing a drum of something. The move is a clean handoff:

Flag it up the chain. A sweet line taking on oxygen is a reclassification — the integrity engineer or corrosion specialist needs to know so the materials and inspection questions get reopened.
Keep the data coming. Bracketed gas-analysis oxygen, coupons in the low points for a rate, water checks where they fit.
Walk the system for the leak. The oxygen has a mechanical way in — a compressor seal under vacuum, a valve opened on the last turnaround, a tank breathing. Finding it is the fix that actually holds.

The specialist designs the response; you deliver the clean data and the found leak that make it possible.

Oxygen on a sweet line is contamination with a mechanical cause. Confirm it’s real, hand the reclassification up, and go find the way in — the treatment is someone else’s call, downstream of your leak.

That’s the oxygen toolkit, used the way the job uses it. You turned a coupon’s mass loss into a rate and made the call, scored a differential-aeration segment and ranked the leak above the chemical, and carried a sweet line from orange rust to a clean handoff — confirming the number was real before it moved.

Next in the set, the pipe fights back with something alive: the microbes that thrive exactly where oxygen is shut out. Same four ingredients, same one reaction — this time the culprit is a living thing.