Repairing P91 headers on HRSGs

[athomas236]

We have a problem with a HRSG header that is designed to ASME section I (but the HRSG is not ASME stamped). The header is from A691 grade 91 and was damaged during final manufacture.

The damage has resulted in both the innner and outer surfaces of the pipe being pushed outwards by a maximum of 10mm over a distance of 15% of the outside surface, This damage extends approx 75mm from the weld preparation.

Instead of reheating and rerolling the affected end of the header, the manufacturer build up the inside surface of the header with weld overlay by more than 10mm, then machined the inside and outside surfaces so they comply with ASME I tolerances.

The header is nominally 600mm OD by 31mm thick and is designed for 34 bar gauge and 565C.

Details of the weld procedure for the overly, subsequent PWHT and effect of repair on header life have been requested and have not yet been received.

My immediate concerns are:

1. That the material which is in the form of a longitudinally welded pipe is not acceptable to ASME I.

2. The repair procedure which did not involve rerolling does not comply with ASME I.

3. The header life will be significantly reduced below the 25years stated in the contract.

4. The HRSG cannot be said to have been designed, manufactured and tested unless it is stamped with the ASME stamp.

In the longer term, the weld overlay will result in a heat affected zone that cannt be replica tested in the future to assess creep damage.

athomas236

Does any one have any comments or sugestions on this

 

[metengr]

Item 1;
The material furnished to SA 691 Specification, which is for Carbon and Alloy Steel, Electric Fusion Welded Pipe for High Pressure Service is not approved for ASME B&PV code Section I application. You can certainly use the formulas and follow the heat treatment requirements in Section I.

Item 2;
Repair to components that are in fabrication are agreed upon between you and the fabricator. According to SA 691 Section 10.2, the manufacturer can perform weld repairs on areas where defects have been removed provided the depth does not exceed 1/3 of the thickness.

The SA 691 specification contains supplemental requirements for hardness testing, metallography and NDT. I would suggest you review this specification and discuss the repair(s) performed by the manufacturer.

At this point, until you have more information I would be hesitant to provide more advice regarding weld repairs to the cast cylinder. Grade 91 is sensitive to weld filler metal and post weld heat treatment. I would ask the fabricator to provide you with actual material certs for the weld filler metal and preheat and post weld heat treatment charts for review. Based on this data, you can decide if supplemental NDT and other testing is required.

 

[davefitz]

The longitudinally welded P91 pipe is vulnerable to type IV creep failure . The weld intercritical HAZ has 1 mm wide "soft zones" on both sides of the weld line that are overaged and have approximately 60% of the creep strength of the parent metal at the typical operating temp of 1050 F. This soft zone can be located by conducting a micro-hardness traverse.

Except for conducting a N+T, there is not much you can do about it except either apply a 0.6 creep stregth reduction factor when designing the header or install creep growth monitors ( etch a triangle of dots along the weld line)and check the growth rate annually.

If the header has not yet been installed, I would see if there is a furnace that can N+T the header witin trucking distance.

 

[metengr]

athomas236;
After giving your post some thought, why are you even considering a HRSG that is not designed AND fabricated to ASME Boiler and Pressure Vessel Code? I would presume this is being installed at a location that does not have any boiler or pressure vessel code requirements. Is the HRSG being built to some other standard?

I believe you are accepting way to much risk with this fabrication!

 

[athomas236]

Gentlemen,

Thank you both for your prompt responses.

The situation is that I have been involved in this project for only two days, the plant is already in operation and has been for several months. I am not employed by the contractor or owner but I am trying to offer technical support to a colleague who is technical adviser to the bodies that are financing the project.

The contractor is saying the the repair is in accordance with ASME II and therefore acceptable and will not result in a reduced header life.

The owner is saying the the repair is not in accordance with ASME Section I which they say is the governing code and will result in a header life rduced to some 50,000hours.

The financing bodies want advice to the actual situation and we are still in the very early stages of our investigations.

Now to your specific points:

metengr

1. At its maximum the weld overlay is 280mm wide by 10mm deep by 75mm long. I am not sure yet how such an extensive weld could qualify as a repair especially as I understand the overlay does not cross over the longitudinal weld and covers only base metal.

2. It is, in my experience, quite common to have boiler plant nominally designed and fabricated to ASME I but without the stamp. Fabrication/erection etc are inspected but not by ASME authorised inspectors. Part of the difficulty is the availability and cost of such authorised inspectors in various parts of the world.

davefitz

You make reference to the 0.6 creep strength reduction factor as you have on other posts on this topic. Having looked through ASME I, I have been unable to identify the location of this requirement. Can you assist with a reference please.

athomas236

 

[metengr]

If the header was fabricated to SA 691, the base metal can be weld repaired per this specification. The only limitation of the repair is depth not area. If the header is 31mm in thickness, by specification the depth or thickness of the weld repair can be 31/3 = 10.33 mm.

davefitz brought up some interesting information regarding normalize and temper heat treatment that should have been performed versus the standard post weld heat treatment. Unfortunately, these are items that should have been addressed at the time the contract was awarded.

The susceptibility to Type IV creep damage stems from sub-critical post weld heat treatment of Grade P91 material, and the level of service stresses. For circumferential welds, the level of stress is typically 50% of the hoop stress -however if the header is not properly supported or contains elevated bending stresses, this will result in Type IV creep/fatigue cracking in the fine grained regions of the base metal heat affected zones. Having said this, your only option at this point is to monitor creep deformation of the header, and to conduct periodic inspections using NDT and replication. It is impossible to predict the impact on actual creep life without some type of isostress creep testing of actual header material, and knowledge of service stresses. At this point, I would institute a rigorous in-service inspection program to monitor the header.

 

[davefitz]

athomas:

I will loook for the references for a 60% weld creep strength. ORNL papers by V Sikka ( 1980's), papers presented by IHI (Japan), and in the line tech publication <

http://www.ommi.co.uk>

would be where I would start.

The exact value of creep strength in the HAZ is a function of the weld history and how much "reheat aging" has occurred during the welding process. Interpass temp, welding amps, number of passes of welds (number of reheat cycles) , weld grrove geometry .. pressure part thickness all play a role in degrading the HAZ properties.

To overcome this aging, the Japanese had pioneered the use of very narrow groove welding , but this has not caught on over in the US. Some people incorrectly atribute the narrow groove weld process as being aimed at reducing weld time and weld electrode consumption, but IHI papers circa 1989 indicate it was rooted in the need to minimize weld reheat aging in advanced creep resistant alloys.

I'll let you know what I find, but as indicated above, it is an inexact estimate unless you know exactly the welding process details ( ie reheat history) .

 

[davefitz]

OK, to not delay a response:

see <

http://www.ommi.co.uk>

vol 1 issues 3 1st paper by I Shibli- see his fifure 3 . As per his references 1 and 6, the creep strenght reduction across a "typical" weld is 20-30% , and Ref 6 ( Allen) indicates a 40% reduction is common.


I also recall that section I had a 0.6 weld strength factor back when they still listed the allowable stresses ( 1989?)in section I. I can no longer find this note in section II.

 

[athomas236]

Gentlemen,

Once again thank you for your advice.

I suspect that the header was not made in accordance with A691 since this does not cover Gr91 material although it does cover a Gr9 material. What I suspect happened was that the header was made from rolled and welded A387 Gr91 plate with the allowable stresses for the plate used to determine the minimum header thickness. Confirmation of this is currently being requested.

I also suspect that unless ASME Sections I or II require a weld reduction factor of 0.6 or so or such a value is included in the allowable stresses, the supplier will not have included it. Again, confirmation of this is being requested.

I have looked at the Shibli paper and it contains some useful information as advised. Didn't realise the guy used to work at Mitsui Babcock. Their offices are about 20 miles from ours. We know each other well. In fact they are our subcontractor on a boiler remnant life study we are doing.

athomas236

 

[davefitz]

Athomas:
It is a small world. And Mitsui Babcock seems to be a major player in the industry that may build large coal fired boilers in the near future. Very impressive recent record.

 

[metengr]

athomas236;
SA 691 - 2004 does indeed cover Grade 91 material. The plate spec referenced in this Standard Specification is SA 387 Grade 91 plate in Table 1.

 

[athomas236]

metengr,

You are correct that Table 1 of the specification SA691 makes reference to SA387 plate specification. My understanding is that this reference is to Gr 9 not Gr 91.

In any case use of SA691 is not permitted by ASME Section 1.

Regards,

athomas236

 

[metengr]

This reference does not treat Grade 9 and Grade 91 as the same - two completely different animals. The reason why I had mentioned this in the above post is that this specification does allow certain weld repairs to Grade 91 material. The SA 691 Specification was intended to provide guidance for full penetration, electric fusion welded pipe in high pressure applications for application under ASME Section III. SA 691 references rolled and welded plate with specific heat treatment and inspection requirements.

 

[athomas236]

metengr,

Thank you for the clarification.

What I have done is to pose questions to the HRSG supplier which (a) seek to clarify the original design and (b) clarify the method of repair.

For (a), the questions centre on the actual specification of the header as installed (because some documents say SA335 gr 91 others say SA691 gr91) and the weld/heat treatment procedures.

For (b), the questions relate to appropriate code requirements for the repair and the weld/heat treatment procedures.

I expect some response by end of Jan 05

Merry Christmas to you and davefitz.

athomas236

 

[athomas236]

Gentlemen,

Further to the above, I have just been advised that the boiler supplier has agreed to replace the complete header.

athomas236

 

[davefitz]

athomas236:
Good deal ! It looks like you hit the jackpot. If only we could get the same response from our HRSG vendors.

 

[CiccioBaliccio]

We just started a project for a Combined Cycle here in Italy and there's a big discussion going on about P91 appropriate Thermal Pre Heating and P.W.H.T.
Can someone indicate me where I can read specifically about this matter?

 

[davefitz]

see thread378-125940 for references to other threads on this subject.

There are 2 large conferences on this subject in Europe this September ( Hamburg 7-9 Sept and London 12-15 Sept) - see the notice at the online technical journal OMMI (

http://www.ommi.uk.com).

A good primer and basis for planning the welding is the free, small book published by Mannesman- Vallourec " the P91 book" - one should interpret all of their recommendations as requirements.

 

[athomas236]

davefitz

Gives good advice, there is also a good article in the magazine Combined Cycle Quaterly.

We have some projects with major international companies and ther still believe the ASME codes will ensure good high temperature creep strength