Abstract
Walsh et al. (1994) reported measurements of erosion of carbon steel by fly ash and unburned char particles in the convective heat transfer section of an industrial boiler cofiring coal–water fuel and natural gas. Erosion was enhanced by directing a small jet of nitrogen, air, or oxygen toward the surface of a test coupon mounted on an air-cooled tube. Ash and char particles that entered the jet from the surrounding flue gas were accelerated toward the surface of the specimen. Samples were exposed for 2 hours with metal temperature at 450, 550, and 650 K (350, 530, and 710°F). Changes in shape of the surface were measured using a surface profiler. Time-averaged maximum erosion rates were obtained from the differences between the original surface height and the lowest points in the profiles. Erosion was slowest at the lowest metal temperature, regardless of the jet gas composition. When the oxygen partial pressure at the sample surface was very small, under the nitrogen jet, erosion increased with increasing temperature over the range of temperatures investigated. At the intermediate oxygen level, in the air jet, erosion was most rapid at the intermediate temperature. In the presence of the pure oxygen jet erosion was slow at all three temperatures. A model was developed by Xie (1995) to describe wastage of tube material in the presence of the erosion by particle impacts and oxidation of the metal. The observed changes in erosion rate with temperature and oxygen concentration were consistent with a mechanism based upon the following assumptions: (1) Metal was eroded as a ductile material, at a rate that increased with increasing temperature. (2) Oxide was eroded as a brittle material, at a rate independent of temperature. (3) The oxide scale was strongly attached to the metal. (4) The erosion resistance of metal and scale was a linear combination of the resistances of the individual components. (5) Oxide formed according to the parabolic rate law, with a rate coefficient proportional to the square root of the oxygen partial pressure. (6) Erosion resistance from particles sticking to, or embedded in, the surface was negligible. Using the model and rate coefficients for metal and oxide erosion derived from the measurements, estimates were made of the erosion rate of a boiler tube as functions of impaction angle and gas velocity. Under the conditions of metal temperature, gas composition, particle size, particle concentration, and particle composition investigated, erosion of carbon steel is expected to be slower than 0.05 μm/h when the gas velocity in the convection section is less than approximately 8 m/s.