Given the high cost of replacing the steam drum, it is very important to understand its structural integrity. The main concern with regard to drum integrity is widespread fatigue cracking of large nozzle welds or seam welds. Fatigue in the steam drum is primarily caused by the combined action of pressure and temperature cycling. The effects of thermal cycling are always a concern, but also difficult to quantify by means of hand calculations. This contributes to uncertainty regarding the drum’s integrity. Small cracks can usually be ground out and is consequently not a major risk. However, in the event of a large crack in the drum, the client will be forced to do a high risk weld repair. Therefore, fatigue assessment was focused on features of which failure would lead to high-risk repairs, e.g. downcomer nozzle welds.
The Plant
The fatigue assessment of a 200MW boiler steam drum in the context of a life extension study to determine if the boiler could be used until 2035. The boiler under consideration is a Babcock & Wilcox El-Paso unit that was commissioned around 1970. It burns pulverized coal and it is rated at 200MW. Since commissioning, the unit has done approximately 300 000 turbine run hours.
The Fatigue Assessment Method
Throughout this work, the method of PD5500a was used to calculate and evaluate fatigue damage. The assessment was not aimed at calculating the remaining life of the drum. Rather, the assessment was conducted in a conservative manner and consequently demonstrated that estimated fatigue damages were insignificant.
Given the complexity of geometry and loads, stresses were calculated using the finite element method. Processing of the stresses to derive stress ranges was done according to the prescriptions of PD5500. For unwelded material, the principal stress with the largest range was cycle counted; for weld toes, stress linearization was done according to the method of EN13445b in order to arrive at structural hot spot stresses which could be used for cycle counting. Since stress cycles turned out to be relatively simple, cycle counting was done by inspection.
The Conclusion
The following conclusions were drawn from the work conducted:
- Current ramp rates not problematic in view of life extension until 2035.
- The main shell with its seam welds and internal fillet welds at very little risk of cracking.
- The most fatigue damage will accumulate at the saturated steam nozzle openings, with estimated damage of 0.32 at 2035.
- Drumhead at very little risk of cracking, with total damage of 0.05 estimated for the safety valve nozzle bore at 2035.
Download the detailed white paper