Chloride Limits for Short Term Operation - SS347 & SS316 8

[PS6788]

I'm trying to determine the maximum chloride content we can accept in one of our hydroprocessing pilot units before pitting/crevice corrosion and CSCC becomes a significant issue. By significant I mean that we could see issues develop in the span of a few months of operation or expect to replace a significant amount of tubing and fittings in the unit after a run. Does anyone have some relevant experience to share?

We typically run for about a month or two before going into a turnaround. The metallurgy of the unit is predominately SS347 and SS316. The areas I'm mainly concerned about are the wash water injection point which is at elevated temperatures (~250F) and dead ends. Wash water is injected in large excess so there is no chance of drying out. Heat tracing maintains temperature above water dew point in the hotter parts of the unit. The wash water will have some dissolved oxygen. The H2S partial pressure could be as high as 25psi.

 

[EdStainless]

Any Chlorides? Even at trace levels (sub ppm)?
What is the minimum pH?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[PS6788]

Hey EdStainless,

We have some experience operating at about 10 wtppm chlorides on a oil feed basis for about a month. That works out to about 50 mg/l in the sour water. At that level, we saw minimal corrosion or cracking issues in the unit. I'm trying to get some historical pH values from the analytical department.

 

[bimr]

Maybe this link will help for stress cracking:

Link

 

There is no absolute threshold for maximum safe Cl- level. Factors like material condition, geometry, pH, and the presence of other ionic species are all involved. And of course temperature, where the rule of thumb is 140°F or over.

SCC often means 'Short Term Operation'. It is not a phenomenon that you can meet halfway.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[EdStainless]

Both pitting and SCC are real risks.
For pitting I would usually like to see Cl- <100ppm at 80F and <10ppm at 180F at a pH of 7-8.
Welds and crevices will start corroding at lower levels than this, and pH is very critical.
Lower the pH 2 points and these values go down by 10x, raise the pH 2 points and they go up by 10x.
With SCC there is no safe Cl- level.
In an autoclave I have cracked 316L at 10ppb Cl-.
High pH helps, but there are limits.
If you have dead legs, I would figure out how to either eliminate them or inspect them very frequently.
You may get 20 runs without a problem, or you may get failures on run 2.
When you turn around rinse the unit out with clean low Cl- water, <1ppm Cl-.
If the unit sits, water will evaporate in places and concentrate. This would make things worse.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[PS6788]

Thank you for your input EdStainless and ironic metallurgist!

I've been reading up on the issue extensively and I'm coming up with the same very "complicated" answer. I was hoping to get an idea of how fast these issues propagate once they start. I can deal with corrosion and SCC as long as I'm guaranteed to make it 2 months and it doesn't occur extensively in the unit.

I'm still trying to get some pH data. I'm expecting the sour water to be mostly NH4SH (~2wt%), HCl, some dissolved H2S and trace oxygen (from water injection). I don't have a feel for what the pH should be.

I'm also looking into using Hastelloy in the future which I understand to be very resistant to pitting and SCC.

 

[shvet]

Does anyone have some relevant experience to share?

For general service see para. 12 API-581-2-2016.
For particular services see para. 4.5.1+5.2 API 571-2011.
For h/p cooler salts washing service see para. 6.6 API 932-B-2014.
For more information see references cited in those.
Also NACE IP34105 and NACE IP34101 may be useful.

H/p processes corrosion is well researched and it is hard to invent something new other than have already been documented in widelly accepted industry standards, practices, papers and so on.
H/p processes are widely used in downstream industry. What is a reason you develop process yourself instead of to contact an experienced person?

We typically run for about a month or two before going into a turnaround.

Into a "turnaround" or into a "inspection/check/test of metallurgy and integrity of all instrumentation, piping and equipment" in "wet" service? Which kind of inspection/test/check? There is a big difference.
It looks like you manipulate words and are planning maintenance and cleaning only. If this is true run this kind of can last for years in practice and there is no significant difference between the pilot unit and a regular refinery.

For info

Many Licensors I have met use deaerated demi-water for salts washing in h/p processes. I have met a case when not-deaerated demi-water was used for periodic salts washing and this fact reduced service life of coolers from 20 to 5 years (and this was a very experienced Licensor). In that case downstream CS pipe was replaced with inconel.

Note that even sporadic contact with >50 ppmw is prohibited as per para. 6.1.t.7 art. 5.1 ASME PCC-2-2015. Many companies use this 50 ppmw limit in internal instructions. As per my experience some companies use CS (with excess thicknes to compensate uniform corrosion or decreased service life) or SCC-resistant alloys in contact with ~5 ppmw in case of elevated temperature or uncontrolled pH/contaminants. There is a common opinion (not proven) that small details particularly with stagnant flow are more susceptible to SCC and therefore require alloys. Elevated temperature, pH and contaminants like oxygen, carbonates and sulfides decrease this limit.

Note that there is a mechanism of corrosion that in crevices/cavities content of ions can be much much higher than in a neighboring flowing fluid (no time for proofs/references, sorry).

 

[EdStainless]

There are many other alloys before you move to a "C" type.
You need to read up on the experiences of others.
Try searching NACE papers from the past.
As mentioned above Cl- limits more like 5-10ppm are what I have usually heard.
Are you using demineralized and deaerated water?
If so then I would start my search with 2205 and 6%Mo austenitic SS.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

I agree with EdS that one of the elite C alloys may not be necessary here.

First step is always surveying the industry to find out the current state of the art, and of course NACE is your go-to source.


"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[PS6788]

Hello shvet,

Thank you for your feedback. I think the scale and purpose is important here, the pilot unit feed rates are on the order of 200cc/hr. Our lines are generally ~1/2" tubing, vessels are constructed from piping elements. We pilot a large range of feedstocks with highly variable properties for basic research purposes, sometimes in support of licensor process development. We may pilot high chloride feedstocks once for a month and not again for 5 - 10 years. It isn't practical to build a unit that covers every possible corrosion issue that might arise. It also isn't practical to tell a client they need to wait a year while we modify the unit. Its easier to understand what our limits are with existing equipment and work inside of them.

We do destructive inspection on a regular schedule, especially in sensitive areas. I'm recommending even higher frequency for these high chloride feedstocks, even after every run.

Hey EdStainless,

You have a good point, especially because nickel is so expensive right now. I'll investigate your recommendations further.

We use deionized municipal water in our processes. It would be very difficult to deaerate water on site, especially with our current setup.

 

... especially because nickel is so expensive right now.

Over my career I have never known a time when there wasn't a nickel 'surcharge'. Never a chromium surcharge, never a molybdenum surcharge, never a niobium surcharge, only a nickel surcharge.

Possibly something to do with one company having a near monopoly for so long?

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[EdStainless]

We were using Ni surcharges back in 1994.
Ni prices first blew up in 1987, went from $2 to $8 and we thought that the world was ending.
Mo had done it in 1978, but I wasn't working with Mo then.
Cr prices stepped up in 1975, and again in 1989.
The late 1970's also saw Nb prices double, never to drop.

When price upsets were infrequent mills would adjust the base value that they used and in many cases this made surcharges a non-issue.
We tried to follow this pattern but in 2000 we began to quote SS with a 'surcharge at time of shipment' clause.
And this covered all elements.
For some years that amount was zero, but since then there have frequently been price swings that have invoked surcharges.

https://pubs.usgs.gov/sir/2012/5188/sir2012-5188.pdf

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[shvet]

It isn't practical to build a unit that covers every possible corrosion issue that might arise.

It is a disputable point. Many engineers would not agree with you and there are strong arguments to act opposite.
Small/cheap/inconsiderable does not mean not risky. It is enough a neglible amount to kill/injure and a little amount to exposure incredible territory. Especially talking about H2S and H2. I hope you have sufficient safety systems and insurance policy.

Small but deadly

https://youtu.be/ALBWxGik64A

Stress corrosion cracking + high pressure

https://youtu.be/uo7H_ILs1qc

Possibly something to do with one company having a near monopoly for so long?

What's the problem - everyone is welcomed ))
Just build one more giant mine and mill on the edge of nowhere. Route one more endless railroad to the that place. Build a town, seaport and airport on permafrost in tundra.
Every surcharge in present has its price in past and that price was a hundred of thousands of deads. 1/4 of unwilling residents and 1/3 of those during the most severe years. I am talking about Norilsk.

 

[EdStainless]

Ni was never a monopoly in modern times.
Inco (Canada), Falconbridge (Cuba), Russia, Finland, South Africa were all sources in the 1950's.
Today you add Australia, Philippines, Indonesia, Brazil, and a few others.
What was the constraint was the marketing system.
At one time all Ni (like Co) went through brokers and there was no 'market price'.
That was the hard part.
Ni was $2/lb +/- for decades.
At least when they mine Ni they are after the Ni.
Just about every metal with higher value/lower volume that Ni is a by product.
Co being the best example. When Cu prices drop mines cut back, this reduces the byproduct Co production and causes spikes in Co prices.
They can't justify running the mine at a loss just to make Co.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[PS6788]

Hey shvet,

I have a strong background in process engineering and licensing, specializing in heavy oil processes. In my experience with engineers and metallurgists, I think most would agree with my assessment. Selecting appropriate metallurgy can be very challenging even in a well defined process. In my case, our process is not well defined and changes almost every run. I also know that chlorides are a reoccurring problem in every refinery worldwide. I promise you that they are generally not using specialized super alloys for the majority of their at-risk equipment, especially at low chloride concentrations or in areas that would only see chlorides in upset conditions. These situations require proper diligence, risk assessment and regular inspection. I would not want to be the engineer that recommends "gold-plating" the unit with noble metallurgy "just-in-case". From what I understand, when SCC initiates the situation is monitored and the pipe or equipment is replaced when the issue becomes unsafe.

CSB is a valuable tool but it isn't the only authority. Often experts do disagree with their findings. In the second video, they operated those reactors for years without serious incident. The main issue was that they didn't do inspections at all and ignored warnings on the condition of the reactors. Its also concerning that they didn't follow ASME code when building the reactors, which may have played a role. However, wrong metallurgy isn't a good argument. If I know CS would be adequate for 10 years of service and I only needed 5 years of service, why would I upgrade to a material that costs 10x more? Further, I could replace that equipment 10x over before I cover the cost of the upgraded equipment (~50 years service life in my example). This all plays a role in the risk assessment aspect of things. Obviously, if I'm taking on significantly more risk by using the lower metallurgy, it may justify the additional expense.

 

[EdStainless]

SCC can move from initiation to failure in seconds, there is often zero warning.
Taking 20% of the material in a project and raising its cost 10X will usually only add a few % to the total project cost.
The key is 'picking your fights', upgrade in the most critical areas.
And if your organization isn't looking at the Net Present Value of various Total Cost of Ownership scenarios then you are making very shortsighted decisions.
In many cases the usage of a higher alloy will allow for less stringent chemical treatment (such as of cooling water), fewer restrictions on operating conditions, and less frequent inspection.
All of these can save a lot of pain and money.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

The wild card in all of this is 'opportunity crude'. Accountants and not engineers make these calls, balancing profit against increased degradation rates of equipment running dirtier crude. (The 'opportunity' refers to job opportunities for corrosion engineers.)

While it may be true that no 'new' refineries were built for a long time after the 1970s, every refinery has replaced units, added processes like low-sulphur units, and upgraded metallurgy. Equipment originally constructed for plentiful clean crude in the 1960s has had to climb what I call the 'alloy ladder' to cope with increasingly hostile corrosion environments.


"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[PS6788]

Hey EdStainless,

But the rate of SCC would have to be based on relatively severity? If every time you have a minor upset in the unit you have a significant failures due to SCC then you would be forced to cascade metallurgy upgrades to significant parts of the whole unit. Even a loss of cooling or water overfill could result in a significant failure. NACE MR0175 suggests low SCC risk for austenitic SS if certain conditions are maintained, 50mg/l chlorides is permissible at high temperatures if H2S partial pressure is below 1,000 kPa.

For us, metal prices comprise a much more significant percentage of overall cost. Upgrading 20% of the metallurgy at 10x cost will almost double the investment cost. In addition, the only way we can maintain temperature in tubing is with heat tracing. If we lose a section of tracing, I could expect to condense water within a few minutes. If I can expect SCC failure that quickly, it tells me that we need to upgrade all of the hot sections of the unit at least to be sure to eliminate all risk of SCC. The NPV in that scenario doesn't look very good for us.

Hey ironic metallurgist,

A lot of what is driving this is alternative feedstocks, blends with spent lubes and plastics derived py-oil and alternative fuel sources such as vegetable oils. Its the wave of the future...

 

So that's where my used motor oil goes...

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."