Removing the residual heat from a nuclear reactor is an important safety aspect of thermal hydraulic analysis. In this study, a typical VVER-1000 reactor residual heat removal system has been evaluated using RELAP5 thermal hydraulic loop code during cool-down. Reactor cooling down starts from hot state temperature and then continues to the cool-down stages with 130°C and 70°C, respectively. The second stage of cooling down is the boundary of the reactor repair condition. Main cooling pump head, steam generator (SG) water level, system pressure, the level of coolant in the pressurizer (PRZ), and the temperature of fuel element are examined in a steady state condition. PRZ level, primary circuit and secondary circuit pressure/temperature, and SG water level are evaluated during 32,000 s after cool-down scenario. By comparison, it concluded that the results of RELAP5 code are in agreement with plant experimental data and final safety analysis report (FSAR). Thus, it is proved that the studied reactor is capable to remove the residual heat generated during shutdown. Moreover, RELAP5 is properly recommended for analysis of the VVER-1000 pressurized water reactor during cool-down. Ó 2017 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

No abstract

This paper presents a neural network based fault diagnosing approach which allows dynamic crack and leaks fault identification. The method utilizes the Principal Component Analysis (PCA) technique to reduce the problem dimension. Such a dimension reduction approach leads to faster diagnosing and allows a better graphic presentation of the results. To show the effectiveness of the proposed approach, two methodologies are used to train the neural network (NN). At first, a training matrix composed of 14 variables is used to train a Multilayer Perceptron neural network (MLP) with Resilient Backpropagation (RBP) algorithm. Employing the proposed method, a more accurate and simpler network is designed where the input size is reduced from 14 to 6 variables for training the NN. In short, the application of PCA highly reduces the network topology and allows employing more efficient training algorithms. The accuracy, generalization ability, and reliability of the designed networks are verified using 10 simulated events data from a VVER-1000 simulation using DINAMIKA-97 code. Noise is added to the data to evaluate the robustness of the method and the method again shows to be effective and powerful.

No abstract

No abstract