DESIGN FACTORS FOR AVOIDING FAC EROSION IN HRSG LOW PRESSURE EVAPORATORS

摘要

Two types of flow-driven wear phenomena are a frequent source of failures in Heat Recovery Steam Generators (HRSGs): Flow-Accelerated Corrosion (FAC) and Liquid-Droplet Impingement (LDI). The two mechanisms combined are also known by an older term, erosion-corrosion. This better describes what can be viewed as a continuum of mechanisms, driven by chemistry and/or fluid velocity, which lead to wear and consequent thinning of pipe walls. One common failure location in the HRSG is in Low Pressure (LP) Evaporator circuits, both in regions with two-phase and with single-phase flow conditions. Replacement and/or redesign of the LP Evaporator system is often required. Water chemistry is known to be a key factor in determining the risk of FAC; it can be modified if required during operation to potentially affect wear rates. Another key factor, determined at the design stage, is the chrome content of the steel. The emphasis here is on evaluating other design factors, particularly local process conditions throughout the LP Evaporator circuit, that affect susceptibility to FAC and LDI. LP systems today often are designed to operate at very low pressures (＜ 3 barg, 45 psi) with the goal of extracting the last practical amount of thermal energy from gas turbine exhaust gas. Furthermore, many combined cycle power plants (such as those in combined power and desalinization service) have variable loads that lead to wide variations in LP operating pressures. The overall bulk fluid circulation in natural circulation LP Evaporators of several different HRSG designs is evaluated over a range of operating conditions. These circulation calculations are sensitive to small changes in heat transfer and load conditions at low pressures, due to the high void fraction in the two-phase region. An accurate estimation of void fraction is therefore required to determine local process conditions (flow velocities and flow regimes) in this region. The limitations of some of the common methods for estimating void fractions, such as asymptotic values for high void fractions at low pressures, are considered. The evaporator process conditions determined by modeling are used as inputs to assess the potential for wear by FAC and/or LDI using several established methods (Kastner, Sanchez-Caldera). The results are compared with case studies from field investigations to assess if wall thinning encountered in actual HRSG service can be correlated with certain design factors.