[en] This paper provides an exact quantification algorithm dealing with shared signals in the GO methodology. The exact calculation formulas of the output signal state probability of a logic operator, which includes the input signals with the shared signals, have been derived. All the shared signals should be transferred to the output signal of this logic operator and in the following calculation the formulas are still exact. This exact algorithm permits the direct calculation of all signal state probabilities of the systems. The result of an example shows that the exact algorithm is available and meaningful for more development of the GO methodology, it will be useful for more practical applications

[en] The GO methodology is an effective method of system reliability analysis. It has been applied to non-repairable systems. This paper discusses the application of the GO method to a repairable system which is described by a Markov model and presents the quantification algorithm of the steady characteristics of the repairable system. The calculation formulas of the ordinary operators and the logical gates are derived and the steady reliability parameters of the system such as average operation probability and average failure frequency can be directly computed by the GO method. The result of an example shows that the algorithm is correct. The algorithm will be useful for the safety analysis of most engineering repairable systems

[en] This paper provides a new quantification algorithm in the GO methodology, which includes three parts: (1) A signal state cumulative probability is introduced and defined based on the definitions of the signal states in the GO method. (2) The quantitative formulas of signal state cumulative probabilities have been derived, so it is unnecessary to calculate the joint probability of the complex combination of states. (3) The calculation process and the transfer rules about the shared signal have been presented, therefore, the new algorithm permits the direct calculation of all signal state cumulative probability. The new algorithm is very easy for programming. It is an improved algorithm in the GO methodology instead of the joint probability algorithm. The result of an example shows that the algorithm is practical and meaningful for development and application of the GO methodology and useful in practice

[en] The general principle of ergonomic in design of NPPs is given and some causes of TMI and Chernobyl accidents from the view point of human factor engineering are reviewed. The paper also introduces some Ergonomic problems in design, operation and management of earlier NPPs. Some ergonomic principles of man-machine systems design have been described. Some proposals have been suggested for improving human reliability in NPPs

[en] Human factor event analysis (HFEA) is important for system reliability assessment, especially for probabilistic risk assessment (PRA) in nuclear power stations. There are two analytic methods, THERP (Technique for Human Error Rate Prediction) and HCR (Human Cognitive Reliability), which are generally used separately in HEFA. These methods have their own features. HCR method is better for determining human errors during diagnosis stage of an accident, whereas THERP provides human event tree model and huge amount of data for determining personnel operational errors. Because there are two phases, i.e. diagnoses and manipulations in a human factor event, it is necessary that the THERP with HCR is integrated into a new HFEA model. This paper introduces the analysis procedure and the report format on this new model, and gives an example in one nuclear power station. (authors)

[en] With the development of protein and polysaccharides pharmaceuticals, there has been a renewed interest in analytical ultracentrifugation. Many protein pharmaceuticals exist in the form of glycosylated proteins. The standards used to calibrate chromatography and SDS-PAGE and not accurate enough to determine the molecular weight of the glycosylated protein in solution. The apparent molecular weight determined may lead to the erroneous conclusion that the protein exists in an associated state in solution. Our experiments show three glycoproteins (envelope glycoprotein of human immunodeficiency virus type 1 rgp 120IIIB, envelope glycoproteins of Sindbis virus E1 and E2) and two polysaccharides (from Poria cocos and Papaya). Gel permeation chromatography suggests that these molecules exist as dimers in solution. However, molecular weights, determined by analytical ultracentrifugation, showed these conclusions to be incorrect

[en] As the increasing role of the probabilistic safety assessment (PSA) in the plant regulation, the uncertainties of PSA need to be narrowed as possible. Human reliability analysis (HRA), as an essential part of PSA, is an important source of the uncertainties of the PSA results. The paper discusses the influence of the uncertainties in HRA upon the PSA results. Then through the discussion of the HRA models, data, and process, the possible causes of the HRA uncertainties are identified, and some measures are forwarded to reduce the uncertainties. In view of the impossibility to fully avoid the uncertainties, the treatment method could be adopted in the PSA is discussed. (authors)

[en] GO methodology is a method of system reliability with success-oriented, it has been applied in non-repairable system. The author studies application of GO methodology in repairable system. Considering dependence of components, the quantitative formulas of some GO operations for repairable system have been derived. According to this study, the GO program has been developed. Reliability parameters such as steady availability and failure number can be calculated directly from GO figure and a High Pressure Injection System of Nuclear Power Plant example is given. The study is available for developing and application of GO methodology in repairable system such as nuclear engineering pipeline systems or chemical process systems

[en] The authors provides a number of important principles relevant to human reliability modeling that have emerged from cognitive models, and presents generic error modeling system (GEMS) and its application in nuclear power plants incidents, and addresses the effort of GEMS on human error predication. Last section presents the principle of the method of computer simulation of GEMS

[en] The author presents a methodology of coded fault tree to analyze the various disasters, which links all failures, as far as possible back to some human errors, such as design error, maintenance error, operator error, etc. It also shows that human reliability analysis and the factors of organizational and management have played very important role in the safety of complex system