Numerical Simulation of Multiphase-Flow-Enhanced Erosion-Corrosion Problems, 20-R8088
Inclusive Dates: 10/01/10 – 09/01/11
Background — Fluid flow enhances corrosion by accelerated transfer of corrosion products from the reaction site. Any dissolved chemical species in the fluid also affects the water chemistry and corrosion environment by controlling the pH, electrochemical corrosion potential and reaction rate. On the other hand, erosion is a mechanical process that may take place either due to presence of a secondary solid phase such as sand in the flow or due to fluid shear that breaks away the layer of corrosion products. The detached corrosion product behaves like a solid particle and may cause further erosion upon wall impact. The erosion process can be influenced by fluid flow parameters such as velocity or turbulence. It also may depend on the physical characteristics of the solid particles, such as their diameter and shape, and their dynamic characteristic such as impingement velocity and angle. Erosion is known to enhance wall corrosion by peeling the wall protective layers and exposing the base metal to further corrosion. The combined effect of the erosion-corrosion process is greater compared to each individual process acting independently. Flow-enhanced erosion-corrosion problems are encountered in almost every engineered system including power plants, process industry and pipeline networks. The present study focuses on the erosion-corrosion encountered in the secondary cooling circuit of nuclear power plants (NPP), where a number of fatal accidents have occurred because of erosion-corrosion damage of pipelines. Modeling the erosion-corrosion process is highly challenging, as it could be affected by a large number of flow parameters that are difficult to control and quantify simultaneously. Existing system level tools are currently used to predict erosion-corrosion processes based on empirical correlations that yield reasonable results for simplified flow conditions. The objective of the present study is to develop a methodology to calculate combined erosion-corrosion damage in NPP secondary coolant circuits for relatively complex flow conditions such as occur at T-joints, expansions and bends.
Approach — In the current study, the flow field and particle motions will be solved using a commercial computational fluid dynamics package, ANSYS-FLUENT. Initially, the flow modeling will simulate a multicomponent continuous phase that includes water and the dissolved corrosive gas. Detailed simulation of water chemistry will be done to understand the effect of oxygen scavengers such as hydrazine in the wall mass transfer process. The discrete phase simulation will model the behavior of solid particles such as sand or debris. This will be useful in location-critical surface areas where the protective corrosion product layer has been removed because of erosion. The chemical corrosion process will be modeled as a mass transfer process that depends on diffusion of corrosive agents through a boundary layer and diffusion of metal through an oxide layer. Specialized water chemistry, corrosion kinetics, and erosion models are being implemented as customized field functions.
Accomplishments — A bend pipe section has been identified as the workbench for the current research, because this configuration is regularly encountered in secondary coolant circuits of nuclear power plants. Detailed hydrodynamic simulations revealed the complex secondary flow patterns that occur in the bend section and how the flow pattern affects the wall mass transfer. The study also showed the Reynolds Stress Model to be most effective in predicting secondary flows and surface mass transfer rates. In addition, a detailed water chemistry study examined the effect of oxygen scavengers in the wall mass transfer process. Simulation showed the effect of surface reaction on the overall corrosive species distribution. Discrete solid phase sand particle trajectory calculations elucidated the solid particle impingement locations. Three different erosion models have been implemented and tested in the solver. Currently, project staff are developing the detailed corrosion model and are coupling the corrosion and erosion models.