Coincidentally, a superb article has been published in Chemical Engineering Magazine on this specific topic.
The article was written by V. Ganapathy, who arguably is the best boiler/heat transfer authority in the world today.
If you are a newbie engineer and your MBA boss has assigned you to develop superheater repair/replacement methods, you can do no better than to purchase Ganapathy books. Ask the MBA for a budget to purchase books ...
Stainless steel liners generally fail due to an initiating crack caused thermal stress, followed by complete filure due to flow induced vibration. The crack growth can be accelerated if there are traces of chlorides in the spray water. The solution to that particular issue is to use ferritic material ( P22/P91 or inconel) for the liner material , and to avoid the flow induced vibration by kinetically forming ( detna forming) the liner to the header ID.
Nozzle failure has a similar damage mechanism, where the initiating crack forms by thermal fatigue , accelerated by trace chlorides, and ultimate failure by flow induced vibrations . De-tuning the PID temperature control loop to reduce the number of cycles by a factor of 5 would extend its fatigue life by a similar amount. A different style nozzle can be used, where the multiple nozzles are arranged tangentially around the ID of the header have a longer fatigue life, and not using stainless as the nozzle material will also improve its fatigue life.
Failure of downstream elbows and tube stubs are indicative of unacceptable spray water atomization and/or insufficient straight pipe downstream of the spray nozzles. Newer style nozzles that can atomized with very little pressure drop ( eg tangential water spray patern nozzles) are available. Latest research into spray atomization finds that secondary atomization is a function of the Weber number ( We > 14) , see papers by Pentair /Tyco
