Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here


Weld P91 to P22

Weld P91 to P22

Weld P91 to P22

Does anybody have any experience welding P91 material to P22.  I have heavy wall P22 nozzles that are to be welded to P91 pipe.  Should an interface buttering material be used?  Any other recommendations?  thanks

RE: Weld P91 to P22

What kind of service? Normally, for this type of dissimilar metal weld (DMW) you want to locate it in a low stress region to avoid bending stress in the P91 base metal heat affected zone (Type 4 cracking).

The DMW can be deposited using either 9Cr-1Mo filler metal or 2.25 Cr-1Mo filler metal. No buttering is required because PWHT will be required for Grade 91 and P22. So, you need to carefully control the PWHT temperature to ensure adequate tempering for the Grade 91, and not overshoot the lower critical transformation temperature for the P22 material. I would suggest using a 1400-1425 deg F for PWHT. Monitoring of this PWHT is critical. This material is unlike other Cr-Mo materials.

Also, after welding the Grade 91 make sure you allow this weld to cool to below 200 deg F before PWHT. Use low hydrogen consumables.

RE: Weld P91 to P22

Service if steam -4000 psi 1050 F (fossil power main steam piping).  Question was generic and I did not provide you the exact configuration at the weld.  The weld configuration would be a full penetration full circumferential corner joint reinforced with an appropriately tapered fillet weld.  I would be welding a P91 material to an existing P22 pipe.  This would result in some local secondary bending (Q) stresses at the weld joint location due to transfer of the P22 pipe bending moment loadings into the P91 component through the attachment weld joint at the discontinuity.

Design and analysis of the weld joint configuration would be by finite element methods (to the satisfaction of B31.1 - 2010 edition, paragraph 104.7) to assure that all code primary and secondary stress limits were satisfied.

Provided, Code stress criteria are satisfied and the appropriate precautions are taken regarding welding and PWHT, do you see any Code or long term service issues that need be considered.


RE: Weld P91 to P22

Yes, I do see a real material issue. At 1050 deg F the P22 is marginal material in terms of allowable stresses for creep strength. As a metallurgist I would not recommend welding a creep strength enhanced ferritic steel (much stronger in creep strength), to an existing lower creep strength material using a corner joint design. This mismatch in creep strength will result in local creep damage or accelerated creep damage to an already marginal P22 pipe material.

Why is Grade 91 being considered when the weak link is the existing P22 pipe?

RE: Weld P91 to P22

Actual temperature is 1035 and pressure is 3690 psig. Existing pipe is P22 and was designed per B31.1.  Operating has been about 200,000 hrs.  Pipe remote from welds is sound and in good condition.  However, due to differential creep of the pipe and the pipe tot pipe Girth welds, cracks have been found at the weld joint areas.

As the pipe remote from the weld joints is sound and only the weld joints are the problem we are investigating encapsulation of the weld joints using P91 material encapsulation components.  We would remove any existing cracks and drill through the P22 encapsulated pipe - This would remove the encapsulated portion of the pipe from the pressure boundary.  By equalizing the pressure on both sides of the encapsulated pipe it would totally remove circumferential pressure stress loading.  The encapsulation component would be designed to take all pressure loading and system moment loadings.  As the encapsulated portion of the pipe would remain, some load would still be transferred through it (although all analysis of the piping system with the encapsulation components and the encapsulation component itself would be performed considered that it did not exist).  The low stress levels is the existing encapsulated P22 pipe would not provide for a condition that new cracks would form. - However, if this were a concern - a full 360 cut of the encapsulated pipe could be made to totally remove any loading on the encapsulated pipe.

Any thoughts/concerns/ideas on this would be greatly appreciated.

RE: Weld P91 to P22

I gave this some serious thought, and if it we my steam line I would not use this repair method. Why? If you have creep damage at existing pipe to pipe girth welds, I would remove the entire girth weld and re-weld, PWHT and be done with it.  Especially if the P22 pipe spool is in acceptable condition. I know that at some point all of my main steam line girth welds will fail because of the Type 4 , fine grain cracking along the base metal heat affected zone. However, my current plan is to eventually replace each entire girth weld with new material (as it becomes necessary) and leave the P22 remaining pipe spool.

Using creep strength enhanced ferritic steel to encapsulate creep damaged P22 girth welds sounds costly to me, and more importantly, one is setting up conditions for accelerated Type 4 cracking in the P91 attachment weld fine grained region. I would not allow this ECrepair method on any of my steam lines.

RE: Weld P91 to P22

This is an excellent thread...much to be learned here.

I am sure that this P22 to P91 interface issue has come up many times in the past considering the steam line retrofits that have occured.

Metengr, should there be changes in the ASME code rules regarding the special metalurgy in these kinds of joints ?


RE: Weld P91 to P22

There is a fine line between writing Code to provide safe design and writing an engineering guideline that provides step by step guidance (aka consulting). The Code continues to evolve based on experience and examples of this includes Code Cases related to T23 material and later versions of Grade 91 where post forming has shown to significantly effect creep life. As we get more experience, Code Cases will be revised and eventually code revisions will happen.

At this point, I am unsure what we would change other than if you are working with creep strength enhanced ferritic steels that are enough cautionary statements in the code to raise alarms, and seek technical guidance.

RE: Weld P91 to P22

I agree with the comments by Meteng- the repair proposed by PMCap does not seem to be optimum solution in any pipeline case I can imagine, and it doesn't appear that the Code practice on creep strength enhanced steels is going to change anytime soon.

Given that the Code will not provide an all-encompassing guide for the design and use of these alloys, and not more than  1% of the practicing designers are likely to obtain a graduate degree in metallurgy, it would seem that ther is a need for a central database  of (a) correct design and fabrication procedures for a range of standard piping applications, including typical specifications and (b) an anonymized listing of past errors.

Also of Note- ASME will be holding 2 sessions of a 3-day course on use of these alloys this year.  

RE: Weld P91 to P22

One document that has been recently made public is

EPRI Document 1023199 Guidelines and Specifications for High Reliability Fossil Power Plants, Best Practice Guideline for Manufacturing and Construction of Grade 91 Steel Components.

RE: Weld P91 to P22

Is the referenced EPRI publication available for purchase by non EPRI members?  

RE: Weld P91 to P22

Yes. Go to the EPRI web site and I was informed at the recent Code meeting that the public can download it. It was decided to allow the general public access because of a safety concern.

RE: Weld P91 to P22

Gentlemen - Thanks to all of you who offered your experience on this issue.

I'm still investigating a way to remove/minimize loading on the  existing degraded P22 to P22 girth welds by encapsulation as previous mentioned.  There were concerns noted in making the dissimilar welds and additional concerns regarding cost.  

Regarding cost, I believe the encapsulation method would be less expensive as the need to remove existing pipe (most of which is in good shape) and procure and install new pipe (which would be quite expensive.) would not be required.  Additionally, encapsulation of the welds would save downtime as this method of restoration could be done in significantly less time than a wholesale pipe replacement.

However, I note your concerns and value your opinion on the issues regarding the welding of these 2 dissimilar metals and the potential for Type IV cracking.

I found a recent research paper dealing with using Inconel 625 to weld P91 to P22 which indicates favorable results.


Disregarding cost considerations (which I believe can be managed) would use of Inconel weld wire with appropriate PWHT result in a sound weld and alleviate your concerns?

Would other type material for the encapsulation component be preferable or perhaps some other combination of material and weld type that you may recommend?

Thanks in advance for looking into this issue for me.

RE: Weld P91 to P22

I believe it would be far better to locally remove the creep damaged girth weld and re-weld this joint versus encapsulation. Second, using Inconel filler metal does not remove the concerns related to Type IV cracking because this location is into the fine grained region of the Grade 91 base material.

I really see no advantage to using Grade 91 because of the Type IV cracking concerns. I even thought of pre-fabrication of encapsulated weld joints in the shop and using P22 to P22 girth welds for installation in the field.

During shop fabrication you could weld the DMW by first buttering the P91, PWHT and complete the weld joint to Grade 22 using 2.25% Cr- 1%Mo filler metal. However, this shop cost and staging would exceed conventional weld repair using field machining, prepping and welding P22 to P22 weld joints.

RE: Weld P91 to P22

Metengr – Thanks for all the insight.

You said - I believe it would be far better to locally remove the creep damaged girth weld and re-weld this joint versus encapsulation.

Question:  Done properly, I agree that this would be a sound restoration method.  However, in removing the creep damaged girth weld and adjacent HAZ, there would be a resultant large gap between the remaining pipe sections that cannot be easily welded up without using a filler pup piece requiring 2 girth welds (yes – same as would be required with an encapsulation restoration).  Also, as we are all aware – when you cut out the weld the pipe will most likely want to spring – thus requiring that the pipe be restrained locally to prevent movement and maintain alignment after cutting (no easy task as we are generally talking about large diameter main steam piping that is in the size range of say 20"OD with >3" wall thickness.  

You said - Second, using Inconel filler metal does not remove the concerns related to Type IV cracking because this location is into the fine grained region of the Grade 91 base material.

Question:  Your response seems appropriate.  I assume that you are saying that Type IV cracking in the P91 or P22 HAZ would not be a concern when using P22 type filler metal or that using P22 weld metal would be less of a concern?  

Question:  The report I cited seemed to infer that use of Inconel 625 weld metal and with appropriate PWHT may result in less of a cracking concern than in using P22 weld metal.  Did I read the report correctly or is this just not true?

You said - I really see no advantage to using Grade 91 because of the Type IV cracking concerns. I even thought of pre-fabrication of encapsulated weld joints in the shop and using P22 to P22 girth welds for installation in the field.
During shop fabrication you could weld the DMW by first buttering the P91, PWHT and complete the weld joint to Grade 22 using 2.25% Cr- 1%Mo filler metal. However, this shop cost and staging would exceed conventional weld repair using field machining, prepping and welding P22 to P22 weld joints.

Question:  The shop prep method is a good idea that I have been giving consideration.  I could make the encapsulation component from P22 material (I was thinking that a P91 encapsulation component would be lighter due to the increase strength of P91 material – only that location where the P91 were to be welded to the P22 would I need a greater thickness) – Your thoughts on using P22 for the encapsulation component or perhaps some other suitable material?

Question:  Regarding your comment on cost consideration I believe that having to deal with the gaps resulting from cutting out the bad welds and HAZs (see above), the shop cost and staging + installation would be less or as a minimum a tradeoff (shop work is generally less costly than field work) and using the encapsulation method does not have the associated risk of not being able to maintain pipe alignment in the field.  Again – if the cost considerations can be reconciled – your suggestion of buttering and PWHT in the shop would seem a good option to consider.  Do you see any other concerns that I should be aware of or investigate?

RE: Weld P91 to P22

If encapsulation and conventionl repair are compared and similar in costs, I would use P22 on P22 ecapsulation.

RE: Weld P91 to P22

Metengr -

Thanks so much for sharing your expertise and insight

RE: Weld P91 to P22

The use of inconel 625 weld filler on P91 ( buttered)  has recently been a cause of concern due to its tendency to form an embrittled layer in the P91 parent metal adjacent to the HAZ after 6-10yrs service. If bending moments are applied to that weld, it may fail ; the total weld interface fails at once. It had been used for dissimilar metal welds ( P91 to 304 or 316), and failures of such a DMW in piping in the workspace has developed into a significant concern- for tubing inside the boiler such failures lead to outages and loss of availibility.

Granted, a T22 to T91 weld  is not exposed to the high, alternating  shear stress that a T91 to austenitic tube woud experience, but the high chrome leveles  of the inconel 625 weld filler will yet remove carbon from the P91 parent metal and lead to embrittlement. Other alternative fillers  could be considered.  

RE: Weld P91 to P22

davefitz - thanks to you an al who have responded to this issue - looks like metengr's suggestion to use P22 all around is currently on the top of the list.  

RE: Weld P91 to P22

Excellent thread, gentlemen - thanks.   

RE: Weld P91 to P22

Assuming that your proposed encapsulation approach uses a clam shell design involving 2 longitudinal weld seams on the P91 reinforcement, I cannot see any benefit over similar P22 pipe reinforcement as per metengr. The Type IV creep properties of the the P91 reinforcement weldment are near equivalent to the creep properties of P22 in hoop and the difficulty of volumetrically examining the P91 longitudinal welds make the repair all the more dubious. Like metengr, I would recommend removal of the damaged weld or a pipe segment containing the creep damaged weld and replacing in kind.      

RE: Weld P91 to P22


A typical fossil plant main steam pipe run has quite a number of weld joints.  I was considering constructing one piece components that were attached to existing pipe using full penetration circumferential corner joints with appropriate tapered welds.  

Installation approach would be to remove a small section of existing pipe at as few location as necessary and the slip encapsulation components (quantity as necessary) over existing pipe (noting that encapsulation component has larger ID (root gap) than existing pipe OD).  

Restoration where existing pipe was removed or installation of the encapsulation components would be repaired by installing a new spool piece as suggested by metengr.  

Of course cost is always an issue - However, it is thought that the encapsulation approach could be less costly than wholesale removal of all weld joint areas as each area would require installation of restraints to maintain alignment - not an easy task.  

RE: Weld P91 to P22

While I have used a similar technique to repair leaks in live gas transmission lines (still a clam shell design), I cannot see any real benefit to your approach over replacement of damaged pipe sections. You do have to cut the existing line and remove a section thereof sufficient in size to place the seamless pipe reinforcement piece.   

RE: Weld P91 to P22


You fellows sure make me think - good for the brain cells.

I would not classify my component as a clam shell design - that is if one interprets a clam shell design as a 2-piece wrapper plate that has the longitudinal seams full pen welded together and the ends attached by fillet welds. The component I am considering would be a 1-piece component with an inside diameter larger than the pipe OD, long enough to encapsulate the weld and HAZs, with its ends configured with a hub so that it can be full penetration welded (corner weld) to the pipe.  This end hub / full penetration weld would be appropriately tapered and the weld ultrasonically testing for soundness.  All code stress criteria would be satisfied.

Yes, I agree that cost issues would need to be considered.  IMHO, if one were faced with having to restore a large diameter heavy wall piping system (think main steam line where pipe OD = 20" and pipe wall = >3"), then the economics of my method could govern.  Consider the cost of cutting out and replacing all the bad girth weld areas which would require installing a large number of temporary restraints, perhaps on both sides of each cut area.  For my proposed method, only a few sections of pipe would have to be cut and restrained as several of my components could be slipped onto the pipe at each cut location.

RE: Weld P91 to P22

I assume that you intend to swage the ends of the SA-335 P91 encasement pipe to permit the repair as stated. You do need to determine the costs for making these items and note that the componenets will require N&T afte swaging. If I were the Owner, I would not contemplate such a repair without finite element analyses of the corner groove weld and at the reinforcement at the swaged ends.  

RE: Weld P91 to P22


I agree with you.

Components would be fabricated from SA335GRP91 (Pipe) or ASME SA336 Gr. F91 (Forging) with appropriate axial and through thickness directions examination and material testing - both before and after component construction.  

I have obtained preliminary hardware costing information that appears reasonable. Preliminary estimates for replacement of an entire piping system have been developed.  Cost estimates for installation of encapsulation components and costs for installation of spools - including additional work for design and installation of the required restraints will be developed.  

Affects on schedule to perform each of the pipe restoration approaches will be looked at as time is money.

Piping restorations using the encapsulation method would include detailed finite element analysis to document ASME B31.1 Code compliance of the encapsulation component itself and the effect of installing the new hardware on the piping system Code qualification of record.


RE: Weld P91 to P22

We have made a number of small replacement spools in P91 pipe to repair locations where the original pipe was mis-heat treated. Spools were between about 3.5 feet long to 10 feet long and from 10" to 24" diameter. Average cost to cut out the existing pipe, make the replacement spools in a shop, and weld them into the existing lines was approximately $115,000 each, including all PWHT and NDE.

After you make your localized encapsulation repairs, what guarantee do you have that the remaining creep life of the "nondamaged" locations will be realized?     

RE: Weld P91 to P22


Thanks for the cost information – this will be really helpful to me.  

Based on my experience and hardware and field costs numbers I seen to date, I'm reasonably confident that my encapsulation repair method could be implemented for less than what your costs have been.  

Again, I'm talking here about not a single spool replacement but restoration of a complete system including main steam lines and turbine bypass lines.  My restoration method for this type system restoration, where I'd have to cut out a spool to get access to install 1-piece component, works best through an economy of scale.  Also, for a large number of repairs on a single system, I believe my method could be implemented in significantly less time.

Regarding your question "After you make your localized encapsulation repairs, what guarantee do you have that the remaining creep life of the "nondamaged" locations will be realized?"  Very good question with no direct answer or guarantee.  Remaining creep life for the non-damaged pipe outside the HAZ areas would depend on the length and type of service the system has seen at the time the damages was uncovered in the HAZs requiring repair.  What we know is that creep life in HAZs for P22 material is significantly less than for the non-HAZ areas.  I would employ a team of experienced and qualified engineers and metallurgical "experts" familiar with the creep behavior of P22 material to assess what the remaining creep life of the system would be following restoration.  The economics would have to be there to support a restoration of the damages areas vs. wholesale pipe replacement.  For many systems, I believe that significant life of the non-damaged portions in the pipe would remain and thus warrant restoration of the weld/HAZ areas.     

RE: Weld P91 to P22

As long as you are developing a detailed stress model of the asembly, you might as well check the thermal stresses developed during startup and shutdown for differnt assumed ramp rates- this series of calcs can be used to determine the allowable piping system temperature ramp rate that will not over-stress the fillet welds due to the DT between the inner P22 pipe and outer P91 sleeve- and confirm this is a faster ramp rate than other inherent limits ( STG, headers, etc).

RE: Weld P91 to P22

Also, there may be 2 corrosion related issues with the sleeved design that need to reviewed. First, the annular space acts as a crevice, and corrosion products may tend to open up the pre-existing crack that is being developed at the inner root of the fillet weld.

Second, the corrosion products may actually develop a high shear force between the sleeve and the inner pipe- similar issues have plagued some gas turbine bolts.

The steam conditions you have defined indicate a supercritical unit using "holy water", so most corrosion products are excluded, but even oxidation in steam environment yields corrosion products.  

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close