Main steam pipe failure ( crack) 3

[mjsamman]

What could be the cause of the attached picture and video.
This high pressure and temperature pipe of the main steam line to turbine.
Press. 170 bar
Temp. 540 DEGREE C.

 

[weldstan]

There are a few possibilities. It appears that crack initiation was at the geometry change in the reducer and that a major weld repair was done possibly indicating that multiple PWHTs were performed. There appears to be creep deformation of the matching pipe diameter portion of the reducer. With the stringer bead welding technique, I assume the material is P91. It's possible that the reducer was improperly heat treated and multiple PWHT's adversely affected the creep ductility of the alloy. Only a metallurgical failure analysis will determine root cause.

 

[MJCronin]

Let us begin the long series of questions !!!

IMHO, that steam piping failure is massive and indicates a defective manufacturer and a callous disregard for human life

First, 540C does not necessarily mean that the MOC is P91 .... At that design temperature, it could be seamed P11 or P22

But of course, we have very few details and 540C may be an operating OR a design temperature

We do have a video however !!!!

Tell us the piping material of construction and age of your piping system. IS this a relatively new piping system ?

Was the system subject to repeated shock ?

Was the piping system repeatedly weld repaired in this area ?

Is this your first job since graduation from the University ?

Just a wildass geezer guess here, but was the reducer sourced from China ? Do you have any idea where your sources of material came from ?

.... You always get what you pay for, you know .....

MJCronin
Sr. Process Engineer

 

[TugboatEng]

That looks loud. This crack opened explosively. That can be fatigue related. The SS Norway proved that with its superheater header failure. But, I won't speculate any further. You're lucky this didn't kill anybody as far as we know.

 

[blacksmith37]

Does/did the reducer meet specification ? How was it verified ? ( My vote would be defective reducer material).

 

[GD2]

mjsamman,
Don't forget the operational issues with a typical steam line. Check for upsets encountered through its history. Condensate is an nuisance and operational risk when it is not removed. Most often steam traps functions are ignored and drip legs may not be designed at appropriate locations. Most often it leads to water hammer that causes piping damage and even rupture.
I would assume there must have had a high local stress point in the HAZ in the weld joint or an inherent metallurgy defect that may have initiated the crack. A kind of water hammer was enough to give the longitudinal thrust to open up the crack.
Another situation I would explore is if the facility is in a cold ambient environment where the steam had been charged suddenly to full pressure ignoring the initial warm up procedure, if there is one stated in the operation manual.


GDD
Canada

 

[TugboatEng]

Looking at the reducer, the the crack mostly follows the radius where the reducer transitions from straight to flare. It's a fairly sharp radius and it appears to be the result of a roller being used to upset the pipe for the flare. I don't think this is normal practice?

 

[hacksaw]

In the video, the main pipe appears to be of differing outside diameter than your main run. See attached screen shot.

Interestingly, the crack has insulation in it, and one the bottom side of the pipe, evidence of condensate leakage over time, and rusting.

Is this a steam vent from a safety valve?

 

[davefitz]

Not enough information from the post and video. You will need to send the failed section to a metallurgist for a full analysis and then have a qualified stress analyst review the reactions at the reducer for all real startup and shutdown conditions, including any on-site observations regarding piping hitting bumpers or other interferences. Such contact of bumpers causes an exponential increase in load at some piping locations, and this is further magnified at a reducer

The location at the smaller pipe section implies (a) piping reactions are greatly magnified at such a location and (b) the distance from the weld implies the PWHT heating coils may have overheated that section. See papers by Superheat that shows that an excessive distance between the induction heating coils and the monitoring thermocouples can lead to an overheat above the first critical temp under the coils. Trace back the field QC documents to determine the location of the PWHT coils and T/C's if the metallurgical analysis proves the 1st critical temp was exceeded.

Other field issues that are rarely considered but can have a great impact on pipe reactions include : forgetting to remove spring hangar pins following turbine rebuild ( TVA Bull run==> 12" extra growth) and differential settlement between boiler and steam turbine ( texas clay with insufficient piling depth==> 12" extra diff settlement after 10 yrs)

"Normally" such a crack might be expected to occur at the weldment between the 1.5" thk P91 main steam pipe and the 3"-6" thk F22 turbine stop valve due to thermal stress and a creep fatigue failure of the weldment, but this particular crack has other causes. " may you live in interesting times".



"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[davefitz]

see papers and presentations by Bill Newell ( Superheat) re: P91 PWHT and AWS guidlelines to see the effect of exceeding the first critical due to a fast preheat or fast PWHT with the TC's too far from the heating pads. OOps. " it was then that he knew he had effed up"

https://cdn.ymaws.com/

http://www.vma.org/resource/resmgr/2013_technical_seminar/d_1035am_bill_newell_revised.pdf

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[MJCronin]

davefitz, et al ....

We will never know if indeed this is a P91 weld failure because,.... Najeeb has lost interest in this topic and refuses to post any more information...

Or perhaps he has gotten what he wants from the West and could care less about anyone else.



MJCronin
Sr. Process Engineer

 

[mjsamman]

Dear All,
thank you all for great speculations and investigation.
the pipe now is repaired temporarily and has to be replaced for the nearest time.(attached video)

I'd like to add some operation data:

- this line is located before entering the HP turbine
- before a week from the failure, previous crack was observed from the weld joint and welded and heat treated.
- during the start up of the unit (after the first crack joint repair), another failure was reported by operation which located down stream of the previous crack (not from the welding joint).

the pipe is P91 and 26 mm thick.
It could be some condensation was existing in the line but the operation won't say the truth.
Note: this failure was observed in many units of the same location

regards,

 

[weldstan]

You stated, "this failure was observd in many ubiits of the same location." Were all the units exhibiting similar failures made during the same time frame. If so what was/wee the year(s) built? During the first EPRI P91 seminar, it was quite evident that P91 piping and bends were misfabricated and field welding and heat treating were improperly performed. This still seemed to be true during the second EPRI seminar.

 

[MJCronin]

Weldstan, davefitz.....

Based on what you have learned from this thread, and based on your expertise and knowledge of P91 welding skills ...

How long do you think that this Main Steam weld repair will last before failure ?

and ..... exactly type of failure would you expect ???

Best Regards

Michael J Cronin

MJCronin
Sr. Process Engineer

 

[weldstan]

He indicated that there were other failures in this pipe system not associated with a weld. If that's the case, The pipe and fittings (ells and reducers) are due for more cracking incidents and the repaired reducer may not last a year and I believe it will be of shorter duration. This system appears to be at the end stages of its life. I believe there was creep damage to the reducer prior to the repair. It could fail during the next startup.

 

[MJCronin]

My Geezer guess ---- Nine to fourteen months .... until the next rupture---

It would certainly help if we had some idea of the age of the system .... but that is all "top secret"

People may die in the next main steam piping explosion ... It has happened before

https://www.technicalsafetybc.ca/al...ngulfs-boiler-building-high-temperature-steam

"mjsamman" will, of course, be blamed for the next failure ...


MJCronin
Sr. Process Engineer

 

[GD2]

mjsamman,
It's the way, it works in an operating plant. If something gets blown off, the maintenance to fix it ASAP. Operation Manager is the King of the plant and you wont too much of investigation data from them. This is the regular modus operendi- temporarily fix it, run it until next major plant turnaround.

The Maintenance Manager can set up a Root Cause Analysis which will take months to conclude the findings.

If you sit on the Maintenance or Engineering side, listen to the King. If you want to go against the management plan, you will be the bad actor.

I have by this interactive forum, you have gained good knowledge of what may go wrong.

GDD
Canada

 

[LittleInch]

MSJAMMAN,

Take note of the above and make sure any recommendations you have about replacement of the items or concerns you express are written down in an e-mail or report and then TAKE A COPY.

Then when the plant finally dies, explodes and kills or seriously injures your fellow workers, at least you might not be thrown in Jail. The upper management will NEVER be thrown in jail...

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[davefitz]

The post has value in that it shows how not to operate a modern power plant that uses P91. The number of deaths and injuries that have occurred since the plant has been built should be determined so as to act as a cautionary tale for other plants.

If they have no metallurgical analysis of failures and no review of the condition of the pipelines then one cannot predict when the next accident will happen- one can just avoid being hired at that plant or at any plant operated by that company.

P91 was approved for use around 1986 and the ASME did not finalize the correct method of fabricating it unitl around 2005, and those rules only apply strictly to units designed and built after 2005. There are a lot of pipelines built in that 20 yr period that are questionable, and I am familiar with some major problems that occurred due to incorrect fabrications and installation procedures. This could suggest that there may be a few ticking timebombs out there.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick