Hydrogen Embrittlement: The One Thing Cathodic Protection Makes Worse

A note: We are publishing a bit out of order. The next intended issue was discussing MIC and is the one of those topics where every answer begins with "it depends," and I want to be sure ours works. It will come. In the meantime, a subscriber asked for this topic, and it moved to the front of the line.

For a long time I told our technicians the same thing. In stations with stainless steel piping, keep your polarized potentials less negative than -1.1 volts. Hydrogen embrittlement. The stainless cannot take it.

I do not remember where I picked that up. Somewhere in a class, or from somebody I worked for early on, or out of a specification I read once and absorbed without noticing. That happens in this trade. You carry a rule long enough and it stops feeling like something you were told and starts feeling like something you know.

Then a technician asked me why.

Not in a challenging way, he wanted to understand it, which is exactly what you want from a technician. I got about a sentence and a half in before I realized I could describe the rule but I could not explain it. We also recently had a subscriber ask about it.

So I went back through the CP manuals, the course material, and the standards. I found out two things. The number was fine. The reason I gave for it was wrong.

Hydrogen damage is counted among the forms of corrosion, and it belongs there. But it is the only one on the list that leaves the wall alone.

Everything else we work with takes metal away. External corrosion, pitting, MIC, DC interference - different mechanisms, same ending. Iron leaves the pipe, and eventually there is not enough wall left. Our whole toolkit is built around measuring how much is gone and how fast it is going.

Hydrogen damage does not work that way. Nothing dissolves and nothing leaves. Measure a wall thickness reading on an embrittled joint and you get the same number you would have gotten the day it went in the ground. What changes is inside the steel. Hydrogen gets in, and the steel stops being willing to bend before it breaks.

That is also why it fails the way it does. Metal loss can lead to a leak, usually with some warning. This gives you a crack.

Cathodic protection forces the steel surface to act as a cathode, and something has to consume the electrons we are pushing. At normal levels that is mostly oxygen reduction. As the polarized potential drives more negative, the second reaction takes over:

2 H₂O + 2 e⁻ → H₂ + 2 OH⁻

The hydrogen produced there does not start out as gas. It starts as single atoms on the steel surface, and that distinction is the whole story. Molecular hydrogen (the two-atom gas) is too large to get into solid steel. It has to come apart first. The single atom is the smallest thing in the periodic table and can move through the crystal lattice without much trouble at all.

Cathodic protection is not the only source. Atomic hydrogen forms at cathodic sites during ordinary corrosion too, with no CP involved. What CP does is raise the rate.

Once that atom is on the steel surface, it has options. It may combine with oxygen and become water. It may react with something else in the electrolyte. It may find another hydrogen atom, pair up, and leave as gas. Or it may dissolve into the steel.

It is a race, not a switch. Some fraction finds its way into the metal even on a normally protected pipeline, and that is ordinary. Steel carries a certain amount of hydrogen without consequence.

Two things tilt the race.

Polarization. Below the hydrogen evolution potential the cathodic reaction is mostly oxygen reduction and atomic hydrogen production is slow. The AMPP CP 2 manual puts hydrogen evolution in neutral environments at -1.044 V CSE. Past that, water reduction takes over and the atoms arrive faster. The more crowded the surface gets, the more of them find a way into the lattice before they find a partner.

Anything that blocks the pairing step. Certain chemical species get in the way for the recombination of atomic hydrogen into gas. Block that exit and the same potential drives considerably more hydrogen into the steel. Sulfides are the ones to know about, which in soil means anaerobic ground and the conditions that come with it.

So there is no cliff. There is a ramp, with a knee at the hydrogen evolution potential. And the number that matters is the polarized potential, the instant-off. An ON reading carries the IR drop with it and cannot be read as a polarized potential at all.

There are two of these, and they want opposite things from the steel. One wants hard steel. The other wants dirty steel.

Embrittlement is the hard-steel version, and it is the one the potential limits are written for. Hydrogen collects where the steel is already under the most strain, and the steel loses its give. Nothing looks different. The pipe just stops tolerating a load it used to carry, and cracks instead of bending.

Three things have to be true at once: hydrogen getting in, steel that is hard or high strength, and static tensile stress. And the stress is mostly not operating pressure. Residual stress is what matters, and residual stress from welding runs close to the yield strength of the material.

Blistering is the dirty-steel version, and it runs almost backwards. It shows up mostly in low-strength steel, and what it needs is not hardness but flaws. Inclusions and laminations left inside the wall by how the steel was made. Hydrogen finds one, turns back into gas inside it, and cannot get out. Pressure builds until the steel bulges or splits. A bubble forming inside the wall rather than on it.

Old steel can be caught by both. Newer steel is better off against both.

Corrosion Basics: An Introduction tabulates what cathodic polarization does to each of the environmental cracking mechanisms. It suppresses stress corrosion cracking. It accelerates hydrogen-induced cracking. AMPP CP 3 says the same thing in one sentence: except for hydrogen-induced cracking and fatigue cracking, cathodic protection can be effective in mitigating all of these forms of corrosion if the structure is buried or immersed. The two sources differ on fatigue. On hydrogen they agree.

There is one more here that catches people, because it’s opposite. Everything else in our industry gets worse when it gets hot. Corrosion rates climb with temperature. The validity of the CP criteria themselves falls off as temperature rises. So the instinct is that heat is the enemy.

This one runs the other way. Susceptibility peaks near room temperature and below, and many steels are most susceptible below 0 °C. Warm helps. Cold hurts. Which puts the highest-susceptibility window over winter work, when cold weather is already handing you enough to think about that this is nowhere near the list.

R/W gets reused. A new line goes in beside an old one, they get made common, and one CP system protects both. That is ordinary practice and there is nothing wrong with it.

Now put a 1950s line next to a line built last year, with the same rectifier, test stations side by side.

The new pipe is clean modern steel: controlled chemistry, low sulfur and phosphorus, consistent microstructure, modern welding procedures.

The old pipe is not. Peabody notes that older steels generally contain higher levels of impurities such as sulfur and phosphorus, and that the microstructures associated with hard spots and welds are typically more susceptible to hydrogen damage than the wrought base metal around them.

And notice it catches the old line both ways. The hard spots and weld microstructures are the embrittlement risk. The inclusions that come with dirtier steel are the blistering risk. The new line beside it has neither.

Same instant-off reading on both. Two different risk pictures.

The potential is a property of the CP system. The susceptibility is a property of the pipe.

Now the stainless.

The AMPP CP 2 manual lists the materials that atomic hydrogen generated by cathodic protection can damage: steel, particularly high-strength steel; martensitic stainless steel; high-strength aluminum alloys, particularly the 7000 series; prestressed concrete; and titanium, which forms hydrides.

Martensitic stainless steel — the hardenable kind. Not stainless generally.

Stainless comes in families. The material most of us picture at a station. 304 and 316, the tubing and small-bore piping is austenitic: soft, ductile, and it cannot be hardened by heat treatment. Martensitic grades, the 400 series types like 410 and 420, can be hardened much like carbon steel, and as strength goes up ductility comes down. Corrosion Basics gives their applications as valve parts, ball bearings, and surgical instruments, and adds that these steels are not normally used in process equipment such as tanks and pipelines.

That last part relocates the whole problem. If martensitic stainless is at your station, it is not the pipe. It is in the components.

Which points at the real driver, and it is not a material family at all. Hardness is a major contributor to hydrogen embrittlement — harder, stronger materials are more susceptible than softer, weaker ones. It is why the tightest CP limits in the standards are the ones written for the hardest steels. So the better suspects are valve internals, bolting, and anything hardened or heavily cold worked. And it may not be the stainless at all. It may be a hard spot in the carbon steel header ten feet away.

The second thing I had backwards took longer to untangle, because the standard changed underneath it.

SP0169-2013 gave stainless steel piping a polarized potential criterion of -450 mV CSE or more negative in neutral or alkaline conditions, well short of the -850 mV we use on carbon steel, because the material is already passive. The 2024 revision dropped that number. It now lists two criteria for stainless: documented empirical evidence of effectiveness, or a minimum of 100 mV of cathodic polarization. It also adds a caveat worth knowing: if a protective oxide film cannot form, or if chloride ions are present, a polarized potential of -850 mV CSE should be considered.

So the number moved and then went away. What did not change is that no edition of the standard has ever asked stainless steel piping to go anywhere near -1.1 V. And on the upper limit the standard declines to give one at all — for ferritic, martensitic and duplex stainless steels, and for austenitic grades in acidic conditions, it says the safe potential limit shall be determined by testing.

Which means a conservative house rule is a reasonable answer to a real gap. It just was not the rule I thought it was. I would keep the caution where it belongs: on station hardware and fittings of unknown metallurgy, treated as suspect until somebody confirms what they are.

This is the part I would have found most useful ten years ago. There is no single upper limit for cathodic protection. The numbers you hear quoted come from different places and answer different questions.

-1,044 mV CSE. From AMPP CP 2. Not a criterion at all. It is where hydrogen starts being generated in a neutral environment. Electrochemistry, not policy.

-1.05 to -1.1 V CSE. From Peabody, described as the general consensus in the industry for avoiding overprotection generally (coating damage and hydrogen damage together) then restated specifically for hard spots and welds in older steels. This is where my rule came from. I had it attached to the wrong metal, but the number is straight out of the reference.

-1.20 V CSE. The current consensus, in ISO 15589-1 and in recent AUCSC course material, using nearly the same sentence Peabody uses, and tightening further as steel strength increases. Some operators plot a -1.20 V line directly on close interval survey data so overprotection is visible at a glance.

And -1.20 V turns up in more than one role. In AMPP CP 2 it appears as a pipeline coating limit, as the ISO 15589-1 limiting potential for pipelines, and as the alkali limit for aluminum. Same number, three jobs. If you were handed -1.2 V and told it was about hydrogen, that is one of the things it does — not the only one.

None of these are wrong. They answer different questions, and the older, tighter numbers reflect a more conservative reading of the same physics. Standing on unknown metallurgy, the conservative end is a defensible place to be.

Writing these articles and building training material has taught me more in the last couple of years than I expected, and this one made the point better than most.

I carried that -1.1 V rule for years. It was a good rule, and the crews were better off following it than not. But I could not explain it, and the explanation I would have given (that stainless steel is what is at risk) was wrong. It was never really about the stainless. It was about hydrogen, hardness and stress, and the stainless was along for the ride.

Nobody in this trade knows everything, and I have never claimed to. What is worth defending is the habit of going to find out when somebody asks. A technician asked me why. A subscriber asked us to write about it. Between the two of them I ended up understanding something I had been repeating for a decade.

If you are carrying a number you cannot source, go find out where it came from. It might be the right number. Mine was. Kind of.

You will still know more when you are done.

Roberts Corrosion Services, LLC

Established in 2011, Roberts Corrosion Services, LLC delivers comprehensive, turn-key cathodic protection and corrosion control solutions nationwide. Our end-to-end expertise encompasses design and inspection, installation and repair, surveys and remedial work. We provide drilling services for deep anode installations and a full laboratory for analysis of samples and corrosion coupons, as well as custom CP Rectifier manufacturing.

While our initial focus was on the Appalachian Basin area, we complete field work all over the US. We are a licensed contractor in many states and can complete a wide range of services.

Our biggest strength is in our flexibility for our clients. Solutions and Results.

Let us know how we can help.

Our New Online Training Site

Newsletter Site

Website

LinkedIn

Location: 39.251882, -81.047440

(304) 869-4007