Failed Gate Valves / Aggressive Deterioration and Damage 3

[myname_earl]

Background: Two 30-year-old, 20", 304L, triple offset, butterfly valves are replaced in October 2021. They are replaced with 20" carbon steel gate valves in preparation for the ten-year internal inspections of stripper reboilers (These valves isolate the reboilers from the stripper column). In Fall 2022, reboilers are still able to be isolated for yearly cleaning but fail to close in Spring 2023 for internal inspections.
[ul]
[li]SDS: 90% H2O, 6% ammonium sulfate, 2% acrylonitrile, 2% organics and cyanides.[/li]
[li]Operating Temp: 240F[/li]
[li]Operating Pressure: 13 psig[/li]
[li]pH should be controlled to 7.8[/li]
[li]Designed to be 10% vapor (90%) liquid[/li]
[li][/li]
[/ul]

Other items to note:
[ul]
[li]all piping is 304L,[/li]
[li]tubes on the reboilers are 304L,[/li]
[li]shell of reboiler is CS,[/li]
[li]50# steam is on shell side,[/li]
[li]Soda Ash and Sulfuric acid are used in this process stream as well.[/li]
[li]"Liquid in" valves for the reboilers were also replaced in 2021, very few or no signs of degradation[/li]
[li] Flashing is a constant or common occurrence[/li]
[/ul]


The piping shows no signs of similar degradation, the SS valves pulled in 2021 also did not show signs of degradation similar to the CS valves installed.

Really I am curious as to what may cause this type of damage, and so quickly. Looking into API 571 I see very similar damage compared to Amine Corrosion or CO2 Corrosion. Asking process engineering, it's not likely that these could cause damage as the scenario would be hard to develop. From what I see, it looks like mesa-type corrosion, tunneling I compare to ash bore beetles, and straight-lined damage I could only compare to damage seen in 600# steam services from cavitation.

Determining the root cause of this failure seems to have me on a roadblock so any guidance would be appreciated.

https://files.engineering.com/getfi...-a038a7fc074a&file=Valve_failure_pictures.pdf

Mechanical Integrity Specialist (Year 1)

 

[MJCronin]

I am guessing here, of course ... But, I do not believe that this is a corrosion issue, but an application error

I also believe that your premium SS Butterfly valves did a better job at full closure than your low-ball CS gate valves.

(Large diameter gate valves of any material are notorious for failure to close .... Was the decision to change to a cheaper valve made by an MBA ???

I believe that your observed failure in a short period is commonly called "wire draw"

Anybody else ?

MJCronin
Sr. Process Engineer

 

[myname_earl]

You are absolutely right. The gates were used as this job was a late notice adder to our turnaround. I don't know why CS was installed rather than SS.

The gate decision was decided by either planning or my predecessor for reliability.

I am to investigate the failure of the valves and the two questions I think are important or interesting for future use is

How/Why were the CS gate valves installed rather than a replace-in-kind?
Why is carbon steel a poor choice in this service?

I know my immediate response would be to replace with SS for starters, and make it a butterfly if maintenance allows it (lead times are typically what hinders these decisions)


I am still interested in the damage mechanism and finding that answer will better guide my response or answer questions that others may have.



Mechanical Integrity Specialist (Year 1)

 

[EdStainless]

Are these valves only ever used full open or full closed?
While this does look a lot like amine/CO2 damage in CS it also looks a lot like erosion corrosion from poor flow dynamics.
Waiting until you are going into an outage to find out that some parts have 6 month lead times is a sign of people not doing their jobs.
Nailing down sourcing is part of the pre-planning process.
Yes, first back to SS.
Second, a more robust valve style is in order. Especially since you use these for isolation.
This becomes a safety issue.

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

 

[myname_earl]

Correct, they are not used for throttling flow, just for positive isolation.

Would poor flow dynamics still be evident in the SS piping?

I dug into the work order to initiate the valve change and it was almost exactly 6 months prior to the outage... Somehow it was missed and later picked up and considered a rush.

Unfortunately when investigations occur, gathering evidence from personnel involved is pretty cumbersome. Of course, who would want to dig their own "grave"?
It is hard to portray that there isn't going to be a fall guy and some of it just came down to an awful pipe spec. The spec for some reason allows planners to choose between CS and SS and offers no guidance on when to use which. If I were a planner, given a bi-material spec by the reliability engineer, on a rush job that shouldn't have been, I would surely choose the component that takes the shortest time to arrive.

There is bad planning, advising, and documentation that makes this event very gray. Diagnosing the damage mechanism is the fun part of the investigation and something that can be completed in confidence... I'll be herding cats to solve the "How did these valves get installed?" question...

What are your thoughts on plug valves in this environment? Operations does not want to go back to butterflies as they were hard to close due to cleanliness.


Mechanical Integrity Specialist (Year 1)

 

[MJCronin]

Knife gate valves are the big "more robust" brother to the generic ANSI/ASME gates that you show in your picture. Knife Gates have been in use for decades in positive shut-off service. Knife Gates are commonly specified in more demanding shut-off services ... Perhaps the confused MBA who specified the defective CS valves was unaware of these ?

The biggest difference between these two types of gate valves is that gate valves are built to ANSI dimensional standards while knife gate valves adhere to TAPPI (Pulp and Paper industry) standards. Therefore, dimensionally the gate valve is wider, flanged, and ANSI pressure rated. It also has API leak tightness standards must be met.

I would begin to talk to a valve Engineering Consultant and informally ask for help from a company like DEZURIK

I also suggest that you explore an "all stainless" blade and trim, or a hard material specified trim for this service

DeZurik has been around forever ... But there are others.... EMERSON, CRANE and FERGUSON come to mind ....

(NO, project planners ... These will NOT be cheap !!! .... You will get what you pay for !!)

https://www.dezurik.com/products/product-type/knife-gate-valves/knife-gate-valves/42/

MJCronin
Sr. Process Engineer

 

[myname_earl]

Thank you, I'll consult with our Ferguson rep.

Planners aren't to blame; there are several discoveries of how this could happen and planner ego or skimping isn't the root cause... It doesn't help, but it is certainly not to blame.

Thank you for the guidance as well EdStainless.

Mechanical Integrity Specialist (Year 1)

 

[EdStainless]

MJC, Thanks, Knife gate valve was the term that I couldn't recall.

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

 

[myname_earl]

We're also looking at installing these for positive isolation and limiting exposure that may be of interest:

https://www.youtube.com/watch?v=ZIQCh9bZFM8

Mechanical Integrity Specialist (Year 1)

 

[TugboatEng]

You have 30 years of good experience with the triple offset butterfly? I would start there. If you need longer life upgrade the materials.