H₂S (Sour) Corrosion: Mechanisms and Effects

You’ve sampled this gathering line a dozen times. It’s always come back sweet — carbon dioxide and water, a straightforward CO₂ line you know how to read. Today the gas analysis has a new row on it: hydrogen sulfide, 180 ppm. Small number, way down the list, a rounding error next to the methane. The operator glances at it. “Little bit of sour now. We’ll keep an eye on it, same as we watch the CO₂.”

Here’s the honest answer this module is built around: you can’t treat sour like “the CO₂ line, plus a little extra.” That 180 ppm doesn’t just add some corrosion to the tally. It switches on a second kind of damage — one that can crack a pipe or blister its wall with the thickness gauge still reading nearly full. And there’s a threshold, smaller than almost anyone guesses, where that line quietly changes categories: it becomes sour service, and the rules for what steel is even allowed to carry it change with it.

The last module took you through the first aggressive gas: CO₂ is acid attack — wet walls, carbonic acid made fresh around the clock, iron leaving one atom at a time. Sour is that same acid attack plus a hydrogen problem inside the steel itself. This module takes H₂S end to end: what it does in the water, the one number that flips a line into sour service, the three ways it damages steel, what the damage looks like, where it hides, and what you collect when you suspect it. By the end, that 180 ppm will read very differently than it did to the operator.

What “sour” actually means

One gas, two separate threats

First, the name. Gas that carries hydrogen sulfide is sour — the old-timers named it off the rotten-egg smell H₂S gives off. Gas without it is sweet. That’s the whole vocabulary — and the difference between the two isn’t a matter of degree. Sweet gas has one corrosion story to tell. Sour gas has two.

Because H₂S does two jobs at once, and they’re separate problems. Hold onto this frame, because everything else in the module hangs off it:

• Job one — it drives metal loss. H₂S dissolved in water is an acid, and like any acid on the wall it eats iron. This part rhymes with sweet corrosion: acid, iron leaving, a scale building up. If this were all H₂S did, sour would just be another flavor of acid attack.
• Job two — it pushes hydrogen into the steel. This is the one CO₂ never does. H₂S drives loose hydrogen into the metal instead of letting it all bubble harmlessly away. Once hydrogen is inside the steel, it can crack it and blister it from within — damage that owes nothing to how much wall the acid removed.

A sweet line has one failure story: it thins until it can’t hold pressure. A sour line has two, running at the same time and on different clocks. That second story — hydrogen inside the steel — is what makes sour dangerous out of all proportion to how much metal it actually removes. A tech trained only on wall loss will read a sour line wrong, because the worst thing happening to it may not be showing up on the thickness gauge at all.

The chemistry, in field terms

You don’t need to balance any of this. You need the picture — enough to see why sour behaves the way it does. Start where any acid gas does: dry gas touches nothing. H₂S in a dry line rides along harmlessly. The trouble needs liquid water on the wall, and then H₂S dissolves into it and comes apart in two steps:

The H⁺ does what any acid’s hydrogen ion does — it attacks the iron, and the wall gives up metal:

So far, that’s ordinary acid attack. Here’s the fork in the road — and it’s the single most important idea in the module.

Look at that hydrogen forming at the surface. Normally, two hydrogen atoms find each other, pair up into hydrogen gas (H₂), and bubble off the wall — gone, harmless, exactly as the equation shows. H₂S interferes with that pairing-up. With H₂S present, a share of those hydrogen atoms are kept from combining into gas and leaving. They stay on the surface as single atoms — and a single hydrogen atom is the smallest thing there is. Small enough to slip between the iron atoms and soak into the steel itself.

That’s the whole mechanism, and it’s worth saying plainly because it ties the module together: H₂S doesn’t make the hydrogen — the acid attack does. H₂S gets in the way of its exit, so more of the hydrogen goes into the metal instead of leaving as a bubble. Every one of the cracking and blistering problems ahead traces back to this one idea.

The sulfide left over from that first step, meanwhile, meets the dissolved iron and builds a scale:

The working number: H₂S partial pressure

How much sour trouble a line can have starts with a familiar move: the raw content — a bare percentage, or a bare ppm — isn’t the number that matters. The number that matters is the partial pressure, the share of the total line pressure that H₂S is carrying. Same formula shape as CO₂:

But here is where sour parts ways with sweet, and the difference is dramatic. Sweet corrosion doesn’t get serious until partial pressures of a few psia — its moderate zone starts around 7 psia. Sour service trips at a number about a hundred times smaller:

Why a measured threshold, and not a simple yes/no? Because how much hydrogen works its way into the steel scales with how much H₂S is present — more H₂S means more charging, not a switch that flips on. The 0.05 psia line is just where that rising risk becomes enough to change the rules for the materials.

Now sit with how small that trip point is. Seven pounds of push before sweet corrosion gets your attention; five hundredths of a pound before a line is officially sour. That gap is the reason a trace of H₂S nobody would blink at can flip a line into a different category. Run the arithmetic once, properly:

Get this calculation into your hands until it’s automatic, exactly like the CO₂ one. It’s ten seconds with a phone, and it’s the difference between “there’s a little H₂S on the report” and “this line just crossed a line that changes what it’s allowed to be made of.”

Three ways sour damages steel

This is the heart of the module. H₂S produces three distinct kinds of damage, and a tech should leave able to tell them apart — because they fail differently, they’re found differently, and only one of them shows up on a routine wall-thickness check. Two of the three come straight out of that charged-in hydrogen from the chemistry section.

1. Uniform corrosion — the metal loss

This is the acid-attack side, the part that behaves like ordinary corrosion: general thinning of the wall wherever sour water sits against steel, usually under and around the black iron sulfide scale. It’s the most familiar of the three because it works the way metal loss always works — metal leaves, the wall gets thinner, a thickness gauge can track it. If uniform loss were the only thing H₂S did, sour would be a manageable problem. It isn’t, because of the next two.

2. Hydrogen-induced cracking (HIC) — no stress required

Follow the charged-in hydrogen. Single hydrogen atoms soak into the steel and wander until they reach an internal flaw — a tiny non-metallic inclusion, a seam, a void left from how the steel was made. There, in that little pocket, the atoms finally do what they couldn’t do on the surface: they pair back up into hydrogen gas. But now they’re trapped inside the metal, and gas takes more room than loose atoms. Pressure builds in the pocket — enormous pressure — and the steel has to give:

• Near the surface, the trapped gas lifts the metal into a blister — a visible bulge on the wall, like a bubble under paint, with sound steel stretched over a gas pocket.
• Deeper in, pockets at different depths crack toward each other and link up in a stair-step pattern that runs along the wall — stepwise cracking, the classic HIC signature, most common in older, lower-strength steels.

The thing to burn in: HIC needs no outside stress at all. Nobody has to be pulling on the pipe. The hydrogen does the whole job from the inside, on its own schedule. A line can be sitting at steady pressure, minding its business, growing internal cracks the entire time.

3. Sulfide stress cracking (SSC) — stress plus hard steel

The third mode needs two extra ingredients on top of the charged-in hydrogen: tensile stress and hard steel. When hydrogen soaks into steel that’s both hard and under a pulling stress, it makes that steel brittle, and brittle steel under tension cracks — suddenly, and at stresses well below what the same steel could carry in a sweet line. This is the fast, scary one: not years of thinning, but a crack that can run when the conditions line up.

Two field facts carry the weight here. First, the stress doesn’t have to come from operating pressure — leftover stress locked into a weld during welding is more than enough. Second, susceptibility tracks hardness: harder, higher-strength steel is the danger zone for SSC, and softer steel is far more forgiving. The sour-service materials standard sets the hardness limits, and they govern a lot of the material and welding decisions for a sour line. Which is why welds — hard zones plus locked-in stress, sitting right next to each other — are the classic place SSC shows up.

The black scale is not your friend

Every sour line builds that iron sulfide scale from the chemistry chain — a black, sometimes sooty or smeared layer on the steel. It’s tempting to read it the way you’d read a tight carbonate scale on a CO₂ line: scale on the wall, maybe it’s protecting the metal underneath. Don’t make that read on a sour line.

Iron sulfide scale is a poor protector and, worse, an active troublemaker. It tends to be porous and loosely stuck — it flakes and sloughs off in patches instead of forming a tight, even jacket. And where it comes off unevenly, it does something a protective scale doesn’t: the patches that stay behind and the bare steel next to them set up small, stubborn cells that concentrate attack into pits right at the edges of the scale. So the presence of black scale isn’t evidence a sour line is coping. Often it’s the opposite — a sign the line is generating sulfide freely, with localized loss hiding under and between the flakes.

The damage you can’t gauge

Here is the sentence that makes experienced hands respect sour service: the wall thickness can read almost full while the steel is seriously damaged.

Think back to the three modes. Uniform corrosion thins the wall — a thickness gauge catches it. But HIC and SSC don’t remove much metal at all. A stepwise crack running through the middle of the wall, or a brittle SSC crack at a weld, can be advanced and dangerous with the outside surface looking clean and a straight thickness reading coming back near nominal. The workhorse tool of internal-corrosion monitoring — measure the wall, watch it thin — is nearly blind to the most dangerous thing sour does.

That’s not a reason to distrust thickness readings; it’s a reason to know their limit. Finding hydrogen damage takes methods built for it — a crack-detection in-line inspection run (a “smart pig” sent down the line), or specialized ultrasonic at a dig — both built to see cracking through the wall rather than just clock its thickness. Worth knowing: the ordinary metal-loss pig shares the thickness gauge’s blind spot, so catching cracks takes a tool built to look for cracks. The details of those methods belong to the assessment modules later in the track. What this module plants is the instinct: on a sour line, a good thickness number is not a clean bill of health. The line can be losing its integrity in a way that gauge was never built to see.

Sour is a moving target

One more idea before we put it all together, and it’s the one that catches programs off guard. A line’s sour classification isn’t fixed at commissioning — a line that started sweet can turn sour over its life.

The usual reason is biological. Certain microbes living in the water inside a system produce H₂S as a byproduct of how they feed — a process the industry calls souring. A gathering system that tested clean of H₂S at startup can, years later, carry enough microbe-generated H₂S to cross the sour-service threshold from the inside, without a single change in the gas coming out of the ground. The mechanism of those microbes — how they live, what feeds them, how they’re controlled — is a whole module of its own further along in this set. Here it earns one job: name it as the reason sour classification is about a line’s current service, not its birth certificate.

The practical takeaway is a habit. A line being sweet on the day it went in the ground doesn’t keep it sweet. Sour service is a current condition, tested for on a schedule — and in our part of the world it’s a live question, because shale-play systems have a long track record of starting sweet and trending sour as the field matures. Mike has watched exactly that happen on Appalachian Basin lines. What was a sweet line on your paperwork last year is a question you re-ask with every sample.

Where sour lives

Put all of it on a map of a real system, and sour corrosion concentrates in the same kinds of places every internal threat does — wherever water and the acid gas collect and sit. Sour gas reservoirs and the gathering lines off them, obviously; but also separators and scrubbers, amine treating units, and the quiet spots throughout a system: low points where water lays, dead legs with no flow to sweep them, and stagnant zones under deposits where the souring microbes shelter. If you’re hunting sour trouble, you start where the water stops moving.

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 what it looks like when H₂S is the suspect.

Indirect data. The gas analysis gives you the H₂S content; the operating pressure turns it into partial pressure; the 0.05 psia threshold tells you whether the line is sour service. The water fills in the rest: pH trending down, dissolved iron climbing, sulfide showing up in the analysis. The field signs are strong on a sour line — a rotten-egg smell at a sample point, black sulfide scale on a pulled coupon or a pig’s catch. And because souring is biological, bacteria checks on the water belong in the picture too, since they speak to whether a line’s sour side is likely to grow.

The repeatability gate. Same rule as always, and worth repeating because it protects everyone’s money and credibility: one aggressive sample is a note in the file; two consecutive reads telling the same story is a signal worth acting on. Sampling has too many ways to mislead for a single result to move a program.

Direct data — with one sour-specific catch. When the signal clears the gate, coupons go into the spots this module flagged — low points, dead legs, downstream of separators — and they come back telling you the uniform corrosion rate and whether it’s pitting, often wearing that black sulfide scale. But remember the damage you can’t gauge: coupons and thickness readings track the loss side well and the cracking side poorly. Catching HIC and SSC takes the crack-detection methods from earlier — a smart-pig run built for cracks, or ultrasonic at a dig — and choosing and running them is assessment work designed above your tier. The instinct that matters here is knowing the ordinary tools don’t tell the whole sour story.

Prediction and materials — whose work that is. There are models and standards that take the numbers you collect and turn them into corrosion-rate estimates and material requirements for sour service. Running them, and deciding what steel and what welds a sour line is allowed to have, is the integrity engineer’s and corrosion specialist’s job, taught in depth far up the track. Your stake is the same as always, and worth taking pride in: the model and the material decision are only as good as the field data feeding them — and the tech who delivers clean, repeatable numbers and catches the moment a line crosses into sour service is the one who makes those upstream calls possible.

Mitigation, designed upstream, executed by you. When the data confirms a real sour problem, the response comes in familiar families: chemical — corrosion inhibitors, and H₂S scavengers, products dosed to pull the H₂S out of the stream before it can do its two jobs; operational — drying, cleaning pigging, keeping water from collecting where it sours; and material — qualified sour-service steels and controlled-hardness welds for the parts that have to live in it. Which lever, which product, what dose is designed by the integrity engineer or corrosion specialist and taught in the mitigation modules much later in the track. The tech runs the program in the field and keeps the samples coming, because follow-up data is what proves the fix is holding. The chain loops; it never really ends.

Back to the analysis

So — the report on the hood of the truck. H₂S: 180 ppm. The operator wants to watch it like the CO₂. The line runs at 850 psig.

Ten seconds: 850 + 14.7 ≈ 865 psia. P_H₂S = 0.000180 × 865 ≈ 0.16 psia. Against a 0.05 psia threshold, that’s more than three times over — this line is sour service. “A little bit of sour” was never the read; the read is that the line just changed categories.

And now the module fills in what that means. It’s not simply “more corrosion to watch.” It’s a line that can now crack and blister from the inside, where the thickness gauge can’t see it; whose black scale is a warning, not a comfort; whose welds are the first place to worry about brittle cracking; and whose classification might keep climbing if microbes are souring it further. The next move isn’t “keep an eye on it.” It’s to flag the sour-service classification so the materials and inspection questions get asked by the people upstream who own them — and to start the collection that backs it up: water chemistry with sulfide and pH read, bacteria checked, the gas re-confirmed, and the same sampling again next cycle, because one read is a note and two are a signal.

That’s the difference one module makes. The operator saw a small number low on the list and reached for the sweet-line playbook. You saw a line cross into sour service at a sixth of a pound of push — and you know the real threat may never show up on the gauge he trusts. Same report, different eyes, different plan.