Pervasive tube failures in heat-recovery steam generators (HRSG) have severely impacted performance and efficiency. Several fundamental failure mechanisms have been identified and their metallurgical causes are well understood. However, current practice to increase the thermal efficiency and flexibility of HRSG operations has contributed to these tube failures. Substantial rises in flue gas temperature, units designed for base load operation switched to cyclic load units, and more aggressive start- up loading rates are just some of the constant demands placed on HRSG' s. In this report, case histories are presented which illustrate the common HRSG failure modes that can occur as a consequence of the increasingly demanding service.
A tube failure event is a principle hazard to reliable and durable HRSG operations. Corrosion related problems including flow-accelerated corrosion (FAC) and steam blanketing, or predominately mechanical problems such as fatigue and overheating, are common challenges to the designers and operators of HRSG's. A recent survey (Figure 1) of failures analyzed in the GE Betz laboratory indicates that while these failure mechanisms are well understood and their root causes can be determined in most cases, the same problems are repeating. Figure 1 shows the relative number of failures by type occurring in economizer, evaporator and superheater sections. From this survey, it is clear that flow-accelerated corrosion, fatigue and steam blanketing account for over half the analyzed failure types. A review of the literature concerning these failure mechanisms shows that flow-accelerated corrosion and fatigue damage have been studied in a number of HRSG types and designs.
FLOW-ACCELERATED CORROSION (FAC)
This mechanism is often referred to as erosion corrosion, flow-assisted corrosion, or even velocity-induced corrosion. This mechanism involves the destruction or dissolution of a protective metal oxide on the metal surface producing general and widespread wall thinning. (1) This mechanism has been identified in single-phase aqueous environments and in two-phase steam and water flow situations. In single-phase flow, FAC is differentiated from erosion corrosion by the impact of flow velocity. For erosion corrosion, the surface shear stress must increase to a level that causes the oxide film to break. Also, variations in fluid velocity must exist. It is believed that when there is no evidence that the oxide film has broken by shear force, the damage is essentially caused by chemical dissolution. (2)
The features observed in the attacked areas may not readily determine the exact threshold that separates FAC from erosion corrosion. However, it is possible with some knowledge of the operating parameters to determine if the dominant mechanism of dissolution is chemical or mechanical. Figures 2 a, b, and c are photographs of low pressure (LP) riser tube sections near the upper header. In each of these cases the tubes are oriented vertically and only very thin layers of oxide were observed in the attacked areas. From general observation and comparison it is possible to discern an increasing degree of chemical dissolution dominance from damage caused by FAC. Figure 2a exhibits a scalloped directional attack suggesting it was strongly influenced by velocity (velocity 2 dominated). Figure 2b shows significantly less directionality to the attack. Figure 2c, which exhibits a smooth transition and initiation from non-affected areas, and then mostly featureless uniform metal loss, suggests the FAC process was chemical dissolution dominant.
FAC produces damage over a fairly wide area of water-wetted surfaces. It has often been observed that scal
