Liquid droplet impingement erosion mechanism of low-alloy steels in the secondary side of pressurized water reactors

다국어 초록 (Multilingual Abstract)

Reliable operation of nuclear power plants (NPPs) is necessary for producing stable electricity and reducing the carbon emission. In this post-Fukushima era where the safety enhancement of NPPs has emerged as the most mandates, aging management of operating NPPs emerges as an imminent challenge. There are several material degradation issues relating to pressurized water reactors (PWRs) that constitute most of operated fleets in Korea. In order to ensure the safety of NPPs throughout a long-term operating period, the integrity of the structural materials within the NPPs must be secured.
Past studies on the material degradation issues have been focusing on the safety-critical problems which may arise with the structural materials constituting the primary pressure boundary (i.e. primary circuit) of PWRs. In contrast, issues with the secondary side of PWRs have received relatively insufficient attention. However, ruptures of piping systems in the secondary side of PWRs caused by the pipe wall-thinning phenomenon have persisted safety impact both within and outside Korea including losses of life and costly damages. Flow-accelerated corrosion (FAC) and liquid droplet impingement erosion (LDIE) have been identified as key causes of pipe wall-thinning phenomenon. While significant understanding has been made for the former, little research has been conducted on the latter.
Interestingly, ruptures caused by LDIE have only recently been reported and no systematic research has been conducted on mechanisms or remedies. For long-term operation, it is thus essential to understand on the damage mechanism of LDIE. Difficulties in carrying out investigations arise from the fact that it is difficult to (i) simulate at a lab scale a phenomenon slowly develops in a time scale of over several decades; and (ii) identify the detailed micro-processes for such phenomenon due to the complex interactions among several key variables.
In this thesis, an accelerated test method has been developed and fundamental mechanisms as well as damage rates are examined for LDIE in simulated physico-chemical environments that are relevant to the secondary side of PWRs. In order to accelerate the LDIE phenomenon, experiments have been conducted under reducing conditions with; (i) the pH level is adjusted to a level lower than those of the secondary side of PWRs; and (ii) the velocity and size of water droplets are adjusted to levels higher than those of the secondary side of PWRs. LDIE experiments by controlling the pH level and velocity and size of water droplets under water chemistry conditions typical of actual plant have been made as a unique attempt to understand mechanisms.
A unique high temperature test apparatus has been developed in this thesis starting from the ASTM G-73-10 standard testing method which requires the control the velocity and size of water droplets. Water droplets having uniform density and size are sprayed through a nozzle and impinged upon specimens that are attached to the rim of rotating disc. Water droplets are produced by using chemically controlled water as a function of DO/DH concentrations, pH level and time.
Two types of typical low-alloy steels for PWRs have been studied, including A106 Gr.B (UNS K03006) and A335 P22 (UNS K21590). The latter with higher Cr contents has lower general corrosion rate than the former. High sulfur grade low-alloy steel, SUM24L (UNS G12144), is also tested to examine the effect of MnS inclusions. All tested materials have microstructure consisted of relatively equiaxed ferrite subgrains and lamellar pearlite regions.
Then, LDIE tests were conducted on test coupons by using the specially developed apparatus. After measuring the accumulated mass-loss over testing time, it has been observed for all materials that: (i) there exists a transition point at which the damage rate thereafter increase significantly; (ii) prior to the transition point, damages tend to be initiated from a particular region of the surface of the specimen; and (iii) after the transition period, the test specimen, while under the same experimental conditions, displayed a higher damage rate than the rate displayed prior to the transition point.
By examining microstructure and nanostructure of damaged coupons, it has been observed that for all tested materials early damages tend to be mainly inflicted within the pearlite structure whereas the ferrite structure remained relatively intact. Microscopic observations of damaged area revealed that the LDIE phenomenon occurred in the following sequence: the selective dissolution of ferrite layers within the pearlite structure leaving cementite plate behind. The selective dissolution of ferrite has been previously reported to have been caused by a galvanic effect arising from the electrochemical disparity between alternating layers of cementite and ferrite, acting as the cathode and anode, respectively. Subsequent erosion by impinged water droplets of the cementite layers within such pearlite structure leads to the detachment of the top surface layer and reveal underlying layer with high roughness. Water droplet impingement on the roughened surface thereafter significantly increase momentum transfer and the erosion rate.
Similar behavior was observed with A335 P22. Microscopic observations of A335 P22 displayed the same phenomenon where damages were mainly inflicted on the pearlite structure and the ferrite structure remained relatively stable until a certain point in time, as well as the same sequence of events observed on A106 Gr.B.
Therefore, this research proposes a damage mechanism for liquid droplet impingement erosion (LDIE), an identified cause of the pipe wall-thinning phenomenon in the secondary side of PWRs, which is explained by the first step of corrosion-induced roughing of surface layer and the second step of erosion-induced material removal. The transition from the first to the second step occurred when the surface grains are removed.
Based on the results of this research, a bi-linear prediction model has been proposed, as functions of time, pH, and water droplet momentum to delineate the corrosion-induced period and the erosion-dominating period. The model was applied to predict one known case of a field failure to fail that the developed model over-predict the damage rate and under-predict the failure time by about 30%, indicating a fair agreement.

목차 (Table of Contents)

• Chapter 1. Introduction 1
• 1.1 Background 1
• 1.2 Issues of material aging in the secondary side of PWRs 3
• 1.2.1 Overview of aging issues in PWRs 3
• 1.2.2 Pipe wall thinning of the secondary side of PWRs 6
• 1.2.3 Pipe wall thinning due to LDIE 10
• 1.3 Objectives 17
• Chapter 2. Literature Review 18
• 2.1 Erosion of low-alloy steel (LAS) 18
• 2.1.1 Theories of LDIE 18
• 2.1.2 Low-alloy steel (LAS) 20
• 2.1.3 Key variables 22
• 2.2 Water chemistry in PWR secondary side 23
• 2.2.1 pH 23
• 2.2.2 Dissolved oxygen (DO) 24
• 2.2.3 Temperature 24
• 2.2.4 Velocity 26
• 2.3 Previous models for liquid droplet impingement erosion (LDIE) 33
• 2.3.1 Experimental models 33
• 2.3.2 Corrosion-erosion combined models 35
• Chapter 3. Rationale and Approach 40
• 3.1 Problem statement 40
• 3.2 Goals 41
• 3.3 Approach 42
• Chapter 4. Experimental Design and Procedure 44
• 4.1 Microstructure and chemistry of materials 44
• 4.2 Development of the LDIE accelerated test apparatus 50
• 4.3 Test matrix and test procedures 59
• Chapter 5. Experimental Results 65
• 5.1 Analysis of experimental results 65
• 5.2 Microstructural evolution of damages 67
• Chapter 6. Mechanism of LDIE Damage 77
• 6.1 Corrosion mechanism 77
• 6.2 Erosion mechanism 84
• 6.3 Microstructural damage propagation process 88
• Chapter 7. LDIE Prediction Model 92
• 7.1 Semi-empirical model for pipe wall thinning 92
• 7.2 Model benchmark with field data 109
• Chapter 8. Conclusions 114
• 8.1 Conclusions 114
• 8.2 Future work 118
• References 119
• Appendix 126
• A1. Mass loss database of LDIE acceleration test 126
• A2. Microstructure images for LDIE damaged specimens 128
• 초 록 131