Any news on 9Cr failures in CC plants ??

[MJCronin]

To all,

Any news on recent failures of 9Cr materials in Combined Cycle power plants intalled over the past 7 -10 years ?

It was my expectation that we were due for at least a few failures from the plants that were slapped together by Calpine and others since 1995.

Type IV cracking failures from improper post-weld heat treatment take a few years to get going, but there were plenty of HRSGs that were welded together by a large group of inexperienced US and Mexican welding shops.

I was just hoping that perhaps a couple of US Senators or Congressmen could be taken out by the unintended launch of a HRSG steam drum.....

Oh, well....

-MJC

 

[davefitz]

a) plenty of reports can be found by searching eng-tips archives

b) not too many steam drums made of P91. maybe a total of zero.

c)the best shops as well as the worst shops have been fabricating this alloy incorrectly. The techniques needed to correctly fabricate this alloy are "foreign" to the prevailing culture of most fab shops.

d)politics don't mix well in this forum.

 

[Welder4956]

All of our CC steam drums are carbon steel - HP, IP and LP. God help whoever rolls the dice and makes one out of Grade 91.

We have been fortunate with P91 so far (knock on wood). One failure in middle of pipe due to manufacturing flaw, a few socket welds that were "stress relieved" by hand torch or not at all, some IP Bypass welds that cracked due to hammer events. We're having more problem with attemperators causing thermal fatigue in P22 and header nozzle welds in P22 than any problems with P91.

 

[MJCronin]

Could you tell us more detail about the attemporator failures ?

Where are the failures occuring and what knid of a fix are you using ?

EPRI has identified attemporator nozzle welds and downstream piping as locations that must be evaluated in any "boiler life extension" studies


-MJC

 

[davefitz]

We have also had fatigue failures on nozzles installed immediately downstream of attemporator sprays . IN our case, the nozzle was a 1.5" thick f22 weldolet saddle welded to a thinner 0.75" thick walled P91 pipe ( HRSG reheater interstage spray).

Good design practice ( per Fisher and also Deutche Babcock) is to allow at least enough straight run downstream of the spray nozzle for a steam /water mixing transit time of at least 0.15 sec to the next bend, and at least 0.3 sec to the downstream distribution header and monitoring thermocouple.

IN our case , the HRSG mfr had close coupled the spray nozzle immediately upstream of the distributing header with less than 10% of the recommended transit time for mixing/evaporation of spray water. We had severe fluctuating metal temps of the thin walled P91 pipe ( 1000F to 480 F to 1000F ) due to impacting water droplets, while the thicker walled F22 weldolet saddle would remain at an intermediate temp of about 750 F ( twice the thickness implies the mean metal temp will fluctuate at a rrate that is 4 times slower). Failure after 1.5 yrs .

Imrpovements to reduce thjese fatigue issues inlcude (a) adding a 40 ft long straight run of pipe between spray nozzle and downstream header and (b) using hotter water ( from economizer outlet ) as source of spray water instead of cold water from the BFP discharge.

 

[stanweld]

Add failures in weld fabricated laterals. Too little reinforcement in the crotch at the highest stress levels. Also type IV cracking in the elipical weld closest to resembling a longitidinal seam. Laterals only post weld heated at 1400 F; normalizing and tempering afer welding would have delayed cracking in this area but not in the crotch. Root cause of problem is design. FEN, after the fact, predicted actual time to failure.

 

[Welder4956]

Our biggest problem is thermal fatigue cracking in the first elbow downstream of the reheat attemperator. I suspect the operators are spraying too much for temperature control or there is a problem with the spray water valves. We have had through-wall circ cracks in the elbow weld on the downstream side of the elbow at the intrados and multitudes of small cracks on the I.D. of the intrados. We have also found a couple of SH and RH attemperator liners broken loose and had to drill and pin them in place.

 

[davefitz]

we had a failed P91 fabricated wye as well.

Bad news : From the 2 Oct EPRI conferneece on advanced alloys, paper titled "advances in welded creep resistnat 9-12% Cr steels" by DJ Abson et al:

"It appears from the information presented above that premature wledment failure is likely wherever weldments in fabrications from 91 grade and similar steels operating at temperatures around 600 C are sbjected to stresses that are of simialr magnitude to the parent steel design stresses. It is likely that utilities that have used these steels in high pressure steam service at such temperatures with inadequate allowance for thepresence of weldmnets are " sitting on a time-bomb", and that monitoriing and perhaps in-service stress measurements are rquired to urgently avoid catastrophic premature failures"

Good news: The sumitomo paper by Abe shows that seam welded P91 pipe that is N+T'd has weld properties that match the parent properties in all respects.