ASME P91 hardness measurements and acceptable values

athomas236;
I have been heavily involved with evaluating Grade 91 piping in our HRSG units across the US. First off, portable hardness testing for Grade 91 material must be performed using a procedure that qualifies the organization that is doing the field hardness testing on actual Grade 91 pipe material. I have seen many questionable results from inspection personnel that claimed they can perform field hardness testing. However, these so called qualified inspection organizations new little about proper surface preparation specific to Grade 91 pipe material (not removing surface decarburization, etc) and actual use of the portable hardness tester. Ask the inspection organization for their hardness testing procedure, specific to Grade 91 material, if they do not have one I would use someone else. How is the hardness testing going to be performed - in quadrants around the pipe and fitting?

Once there is a qualified hardness testing procedure, hardness data should be evaluated against hardness requirements in ASME SA 335 for Grade 91 material - N&T (Grade 91, 265 HV or 251 BH, maximum). For the low end of hardness for N&T, I normally would use 190 HV for minimum hardness, 180 HV minimum (as verified by field replication of the microstructure). This should be your basis for N&T material.

For subcritically PWHT weld joints in the field, the hardness range should be 190 to 270 HV for base metal HAZ and weld metal. If you have any field hardness data that is out of range, use field replication to evaluate the microstructure. Both replication and hardness testing should be used before any remedial action is performed.

Through extensive hardness testing and microscopy, I would not accept hardness values in the base metal below 190 BHN or equivalent without additional microscopy to verify the appropriate microstructure. About 5% of materials tested between 184 and 189 exhibited an inappropriate microstructure. Hardness below 180 exhibited unacceptable microstructures more than 50% of the time and hardness below 174 nearly 100%.

Microstructures with hardness < 170 BHN are typically ferritic indicative of being annealed.

Weld hardness > than 280 BHN should receive an additional PWHT.

Gentlemen,

Thank you both for your advice.

The hardness test procedure used is summarised below:

1. Personnel shall have a working knowledge of and a min of 10 hours experience of the specific test equipment used.

2. The instrument shall be calibrated in accordance with the manufactureer's specification and shall be calibrated by a certified laboratory every 12 months.

3. The calibration of the instrument shall be checked before each shift using a calibration block of similar compositionm micro stucture and hardness of the parts being tested. The hardness of the calibration block shall be measured by a certified laboratory instrument.

4. The surface to be tested shall be flat and free of scale and oxides, grease, paint etc and shall be prepared by smooting the surface with successively finer abrasives finishing with 22 micron material.

Best regards,

athomas236

athomas236,
As metengr stated one of the keys to hardness testing of P91 is to assure removal of the decarburized layer. Initial removal of approximately 0.060-inch of metal is recommended, assure that min. wall is not breached. Note that the decarburized layer can extend deeper than 0.060" and further removal of the surface layer with retesting should be explored.

stanweld,

Thank you for reminding me of metengr's comment.

Is there a way of detecting that the decarburized layer has been removed?

Best regards,

athomas236

athomas236
If the hardness readings indicate softness (below minimum range) continue to remove material by selective grinding and obtain additional hardness readings. If the readings increase in hardness, you may need to locally grind further until the hardness readings are relatively consistent. This is the only method to determine if decarburization is effecting hardness.

By the way, there is a Utility Users Group that deals with Grade 91 material issues, and they evaluated field hardness testing. This Group experimented with various portable hardness testers and found that there is a difference in testers and results, and this should be taken into account for field hardness testing of Grade 91 material.

metengr,

Thank you again.

I have looked at the contractor's procedure for hardness testing and it makes no mention of removing the decarburized layer only about removing scale etc as in my post of 27 May 08. As a consequence, if the decarburized layer has been removed in the 1000 individual hardness measurements made to date it will have been more by accident than design.

If the decarburized layers have not always been removed then any measured hardess below the minimum acceptable limit could be false low values which could result in the contractor taking un-necessary corrective action. Although the contractor has not yet agreed to make any corrective action for low hardness.

In a similar way, for any measured hardness above the acceptable maximum could also be false low values with the true hardness being above the measured values. This, in theory at least, could result in corrective actions not being made when they are required. This could be more of a worry.

In this instance, the contractor has agreed to PWHT one weld with the highest measured hardness, as a test, this coming Saturday. He has not proposed any relica testing to establish the microstructure either before or after the PWHT.

Are you aware of the approximate difference in hardness caused by not removing the decarburized layer compared with that when the layer has been removed.

Best Regards,

athomas236

athomas236:

Are you also including in your review the other QC items that are used to indicate proper P91 weld and hot-bend fab procedures? The proper hardness range, following proper surface prep, proper hardness meter selection, proper operator training in the end only yields one neccesary but not sufficient indication of correct P91 metallurgy.

Other items that may be included in the review:
-records of certified electrode used
-low hydrogen welding procedures specification + training
-time-temperature histograms of the weld zone pre-heat, interpass, cool-down to 100 C, and PWHT temps for weld zones
-time-temp histograms of hot bent el N+T of entire work piece, including(a) was the entire piece N+T'd following bending ? and (b) was it cooled from 1040 C at a rate faster than -5.0 C/min?

athomas236;
Typically, I have seen hardness readings in the decarb layer as low as 170 VH. Once we removed the decarb material for hardness testing, the hardness increased to 195 HV. No further action was required.

metengr,
Thanks for that information.

davefitz
Shown below is a summary of the weld procedure being used by the contractor.

Best regards,

athomas236

SUMMARY OF WELD PROCEDURE
Welding materials
The filler metals shall be E9015-B9 and E9018-B9 for the Shielded Metal Arc Welding process; ER90S-B9 for the Gas Tungsten Arc process; and EB9 for the Submerged Arc Welding Process. Grade 91 welding materials shall be purchased with the combined nickel and manganese content to be 1.2% or less. Also, the minimum nitrogen content shall be [(0.52 X Al) + 0.032]%.

Weld preparations
Butt joints in tube or piping shall have the root pass welded using the gas tungsten arc welding process and the interior of the component shall be purged with 100% argon gas.

Preheating and inter-pass temperatures
The preheat temperature shall be a minimum of 205°C for the pipe. A preheat of 300°F (150°C) is acceptable for attachment to pipe welds using the GTAW process only. Results of preheat must be documented.

Preheat shall be maintained throughout the entire welding cycle, including tack welding.

The inter-pass temperature shall not exceed 400°C.

The preheat shall be measured and monitored within three inches of each side of the weld groove. The temperatures shall be measured by the use of temperature indication crayons, optical pyrometers, thermocouples with calibrated chart recorders.

Prevention of hydrogen induced and stress corrosion cracking
After all welding is completed an extended hydrogen bake shall be performed using Electric Resistance Elements.

Prior to performing the extended hydrogen bake, the weldment shall be cooled in still air, to below 135°C, but not below 95°C.

The temperature shall be raised to 315 - 400°C and held for three hours minimum before being allowed to cool in still air to a temperature not below 95°C

To minimize the risk of stress corrosion cracking, following the completion of the extended preheat / hydrogen bake operation and prior to the post weld heat treatment, the weld must be maintained absolutely dry. One accepted method to assure dryness is to maintain the weld temperature at 200°F (95°C) minimum. The weld temperature may be reduced to below the specified minimum if another method is employed to assure the weld remains absolutely dry.

Furnace PWHT
Thermocouples shall be placed evenly spaced (~ every 90º) at five circumferential planes along the length of the component.

1. Near mid-length of the component.
2. Near each end of the component.
3. Near ¼ length from each end of the component [This location my be omitted for component length of less than 3 meters]

Each butt weld shall have four thermocouples evenly spaced (~ every 90º) around the circumference of the weld.

Each attachment weld shall have four thermocouples evenly spaced (~ every 90º) around the circumference of the weld.

Post weld heat treatment shall consist of heating to the temperature of 730-774°C for one hour per inch of thickness, 30 minutes minimum, and then cool in still air. To avoid over tempering, the holding time shall be limited to 1.2 hours per inch of thickness with the thickest weld in the component being the governing dimension.

Appropriate calibrated recording devices shall be used to record the time at temperature during all postweld heat treatments.

Athomas236
I recommend that the low end of the PWHT temperature range be increased to 740 C to mitigate the high hardnesses experienced. I would increase the hold time to 2 hrs minimum for 16mm < T </= 50mm. For T > 50 mm, 1 hr/25 mm, 2 hrs min.

careful now. The spec called for hi allowable levels of Ni+Mg. If nickel exceeds 0.4% , PWHT temp and pre-cooling temp may need to be lowered.

Athomas236
Nothing against your weld procedure, I would suggest you follow Section I, PW-39 Note (3) for P-No 5B, Group 2 base material PWHT requirments. This is the best advice for welding consumables for Grade 91 material. Plus, it is based on past expierence.

Gentlemen,

Thanks again.

I have looked at ASME 1 2007, PW-39 note 3, with thicknesses greater than 13mm and less than 125mm, it looks like the PWHT complies with ASME 1, bearing in mind max Ni content in A335 P91 is 0.4%

I understand that max PWHT temperatures based on Ni content were proposed in a draft paper by Henry and Tanosh in a draft paper for the 2004 Piping and Pressure Vessel conference but these do not seem to have been adopted by ASME.

Best regards,

athomas236

I agree with alot of the comments made above. We have performed alot of testing on P91 for HRSG applications and I have had a few sleepless nights over it. These are my snippets of advice:

*P91 is the Devils material, is difficult material to heat treat and the heat treatment range between good and bad is narrow. The material has no margin for error and take it above the AC3 temperature and it loses its tensile and creep properties. The AC3 is affected by composition, so it is a good idea to know what the AC3 temperature is by testing before heat treatment is commenced.

*Make sure there are an adequate number of thermocouples on the pipes. The bigger the pipe section the more thermocouples you need to control/ monitor.

*Make sure that the company doing the hardness tests has some proper hardness blocks and the equipment is checked daily (minimum) against the standard. We could not get any suitable P91 standards and we had to manufactured are own specials. The hardness block standards also needs to be appropriate for the equipment you are using either bouncing ball or Equotip types.

* I prefer Equotip for general hardness readings survey and Microdur for any areas which need special attention.

*A great deal of care needs to be taken to ensure adequate surface preparation/ finish for hardness testing otherwise you will get erroneous readings.

*Material is very prone to decarb so follow Metengr advise and in cases of low hardness carefully remove surface layers until hardness is stable. Make sure you don't grind below minimum thickness!!

*Anything below 190HB should be assessed by replication for appropriate microstructure.

Metaljon,

Thanks for your advice.

Best Regards,

athomas236