No description available.

Yes / Fault Tree Analysis (FTA) is a well-established and well-understood technique, widely used for dependability evaluation of a wide range of systems. Although many extensions of fault trees have been proposed, they suffer from a variety of shortcomings. In particular, even where software tool support exists, these analyses require a lot of manual effort. Over the past two decades, research has focused on simplifying dependability analysis by looking at how we can synthesise dependability information from system models automatically. This has led to the field of model-based dependability analysis (MBDA). Different tools and techniques have been developed as part of MBDA to automate the generation of dependability analysis artefacts such as fault trees. Firstly, this paper reviews the standard fault tree with its limitations. Secondly, different extensions of standard fault trees are reviewed. Thirdly, this paper reviews a number of prominent MBDA techniques where fault trees are used as a means for system dependability analysis and provides an insight into their working mechanism, applicability, strengths and challenges. Finally, the future outlook for MBDA is outlined, which includes the prospect of developing expert and intelligent systems for dependability analysis of complex open systems under the conditions of uncertainty.

Current regulations specify the minimum number of operators required per nuclear power plant. However, these requirements are based on the operation of large nuclear power plants, which are not inherent safe and can result in a meltdown. For newly developed small nuclear reactors, the current number of operators seems to be excessive causing the technology to be less competitive. Before the number of required operators can be optimised, it should be demonstrated that human errors will not endanger or cause risk to the plant or public. For this study, a small pebble bed High Temperature Reactor (HTR) Nuclear Power Plant (NPP), the Th-100, was evaluated. The inherent safety features of this type of nuclear reactor include independent barriers for fission product capture and passive heat dissipation during a loss of coolant. The control and instrumentation architecture include two independent protection systems. The Control and Limitation System is the first protection system to react if the reactor parameters exceed those of the normal operational safe zone. If the Control and Limitation System fail to maintain the reactor within the safe zone, the Reactor Protection System would at that time operate and force the reactor to a safe state. Both these automated protection systems are installed in a control room local to the reactor building, protected from adverse conditions. In addition, it is connected to a semi-remote control room, anticipated as a multi-unit control room to include the monitoring and control of the auxiliary systems. Probable case studies of human error associated with multi-unit control rooms were evaluated against the logic of the Control and Limitation System. Fault Tree Analysis was used to investigate all possible failures. The evaluation determined the reliability of the Control and Limitation System and highlighted areas which design engineers should take into account if a higher reliability is required. The scenario was expanded, applying the same methods, to include the large release of fission products in order to verify the reliability calculations. The probability of a large release of fission products compared with studies done on other nuclear installations revealed to be much less for the evaluated HTR as was expected. As the study has proved that human error cannot have a negative influence on the safety of the reactor, it can be concluded that the first step has been met which is required, when applying for a waiver to utilise a multi-unit control room for the small pebble bed HTR NPP. Also, from the study, it is recommended that a practical approach be applied for the evaluation of operator duties on a live plant, to optimise the number of operators required. This in turn will position the inherent safe HTR competitively over other power stations. / MIng (Nuclear Engineering), North-West University, Potchefstroom Campus, 2015

Current regulations specify the minimum number of operators required per nuclear power plant. However, these requirements are based on the operation of large nuclear power plants, which are not inherent safe and can result in a meltdown. For newly developed small nuclear reactors, the current number of operators seems to be excessive causing the technology to be less competitive. Before the number of required operators can be optimised, it should be demonstrated that human errors will not endanger or cause risk to the plant or public. For this study, a small pebble bed High Temperature Reactor (HTR) Nuclear Power Plant (NPP), the Th-100, was evaluated. The inherent safety features of this type of nuclear reactor include independent barriers for fission product capture and passive heat dissipation during a loss of coolant. The control and instrumentation architecture include two independent protection systems. The Control and Limitation System is the first protection system to react if the reactor parameters exceed those of the normal operational safe zone. If the Control and Limitation System fail to maintain the reactor within the safe zone, the Reactor Protection System would at that time operate and force the reactor to a safe state. Both these automated protection systems are installed in a control room local to the reactor building, protected from adverse conditions. In addition, it is connected to a semi-remote control room, anticipated as a multi-unit control room to include the monitoring and control of the auxiliary systems. Probable case studies of human error associated with multi-unit control rooms were evaluated against the logic of the Control and Limitation System. Fault Tree Analysis was used to investigate all possible failures. The evaluation determined the reliability of the Control and Limitation System and highlighted areas which design engineers should take into account if a higher reliability is required. The scenario was expanded, applying the same methods, to include the large release of fission products in order to verify the reliability calculations. The probability of a large release of fission products compared with studies done on other nuclear installations revealed to be much less for the evaluated HTR as was expected. As the study has proved that human error cannot have a negative influence on the safety of the reactor, it can be concluded that the first step has been met which is required, when applying for a waiver to utilise a multi-unit control room for the small pebble bed HTR NPP. Also, from the study, it is recommended that a practical approach be applied for the evaluation of operator duties on a live plant, to optimise the number of operators required. This in turn will position the inherent safe HTR competitively over other power stations. / MIng (Nuclear Engineering), North-West University, Potchefstroom Campus, 2015

Les véhicules aériens sans pilote (UAVs), utilisés et développés pour la première fois dans le domaine militaire, ont connu de profonds changements ces dernières années et sont de plus en plus utilisés dans le domaine civil. Etant plus connus sous le nom des drones, ils sont le plus souvent utilisés dans les domaines civiles et militaires. Ils sont employés pour : la lutte contre les incendies, le sauvetage ainsi que dans des applications spécifiques comme la surveillance et l’attaque. Le vol en formation est de loin le plus utilisé car il permet une répartition judicieuse des tâches et améliore grandement l’efficacité des drones (principe de l’attaque en meute, des animaux carnassiers). Cela pose alors la problématique de la coordination et de la stratégie, ainsi que du type de fonctionnement (maitre/esclave,…).Le type et la qualité d’informations optimums restent aussi à définir.L'utilisation accrue de ces systèmes coopératifs dans des environnements dangereux rend leur fiabilité essentielle pour prévenir tout événement catastrophique. Une performance globale de la flotte des drones doit être garantie, malgré une possible dégradation des composants ou de toute modification du réseau et de l'environnement. Il est nécessaire de détecter les comportements anormaux pouvant contribuer aux collisions et ainsi affecter la mission. Compte tenu des performances et du coût, les systèmes à tolérance de pannes et à redondance ne représentent pas toujours la solution la plus efficace pour ce type de vol de flotte en formation. Différentes méthodes telles que l'analyse par arbre de défaillance (ADD), l'analyse des modes de défaillance, de leurs effets et de leurs criticités (AMDEC) ont été utilisées dans le monde des hélicoptères.Dans une première partie, une méthode statique basée sur l’ADD est proposée, pour assurer la fiabilité de la communication entre les drones d’un côté et entre les drones et la station de base de l’autre côté en accentuant l’échange de flux d’informations. Nous utilisons des arbres de défaillance pour représenter les différentes conditions d’erreur de ce système complexe.Dans une deuxième partie, nous analysons les différents états de défaillance des communications et leurs probabilités. Ce processus étant stochastique, une approche par chaîne de Markov absorbante est développée. L’approche proposée peut être utilisée pour trouver les scenarios les plus risqués et les éléments à prendre en compte pour améliorer la fiabilité.Enfin, dans une troisième partie, nous étudions le problème de réception des messages d’un drone en proposant un protocole basé sur le nombre de retransmissions. La réception est assurée avec une certaine probabilité de fiabilité, en fonction de plusieurs attributs tels que la modulation, le taux d’erreur des bits (BER) caractérisant les drones. / Unmanned aerial vehicles, used and developed initially in the military field, have experienced profound changes in recent years and are increasingly used in the civilian field. Recognized as drones, they are most often used in the civil and military domains. They are used for firefighting, rescue as well as in specific applications such as surveillance and attack. The formation flight is the most used because it allows a judicious distribution of the tasks and greatly improves the efficiency of the drones (principle of the attack in pack, carnivorous animals). This will raise the issue of coordination and strategy, as well as the type of operation (master /slave, ...). The type and quality of optimal information also remain to be defined.The increased use of these cooperative systems in hazardous environments makes their reliability essential to prevent any catastrophic event. Overall performance of the drone fleet should be ensured, despite possible degradation of components or any changes that occur to the network and the environment. It is necessary to detect the anomalous behaviors that might contribute to collisions and thus affect the mission. Taking into consideration performance and cost, the fault-tolerant system and redundant systems are not always the most efficient solution for the formation fleet flight. Different methods like the fault tree analysis (FTA), Failure Modes and Effects Analysis (FMEA) have been used in the helicopter field.In the first part, we propose a static method based on FTA, to ensure a successful communication between the drones from one side, and between the drones and the ground station from the other side by emphasizing on the exchange of information flows. It uses various fault trees to represent the different error conditions of this complex system.In the second part, we analyze the different fault states and their probabilities. As this process is stochastic, an absorbing Markov chain approach is developed. The proposed approach can be used to find the most risky scenarios and considerations for improving reliability.Finally, in the third part, we put the emphasis on the message receipt problem in a drone’s communication network by proposing a protocol based on number of retransmissions. The reception of a message is provided with a certain probability of reliability depending on several attributes such as modulation and bit error rate (BER) characterizing the UAVs.

Prescriptively, the requirement of fire safety protection systems for distribution substations is not provided in the compliance document for fire safety to the New Zealand Building Code. Therefore, the New Zealand Fire Service (NZFS) has proposed a list of fire safety protection requirements for distribution substations in a letter, dated 10th July 2002. A review by Nyman [1], has considered the fire safety requirements proposed by the NZFS and discussed the issues with a number of fire engineers over the last three years. Nyman concerned that one of the requirements regarding the four hour fire separation between the distribution substation and the interior spaces of the building may not be necessary when considering the risk exposure to the building occupants in different situations, such as the involvement of the sprinkler systems and the use of transformers with a lower fire hazard. Fire resistance rating (FRR) typically means the time duration for which passive fire protection system, such as fire barriers, fire walls and other fire rated building elements, can maintain its integrity, insulation and stability in a standard fire endurance test. Based on the literature review and discussions with industry experts, it is found that failure of the passive fire protection system in a real fire exposure could potentially occur earlier than the time indicated by the fire resistance rating derived from the standard test depending on the characteristics of the actual fire (heat release rate, fire load density and fire location) and the characteristics of the fire compartment (its geometric, ventilation conditions, opening definition, building services and equipment). Hence, it is known that a higher level of fire safety, such as 4 hour fire rated construction and use of sprinkler system, may significantly improve the fire risk to health of safety of occupants in the building; however, they could never eliminate the risk. This report presents a fire engineering Quantitative Risk Assessment (QRA) on a transformer fire initiating in a distribution substation inside a high-rise residential and commercial mixeduse building. It compares the fire safety protection requirements for distribution substations from the NZFS to other relevant documents worldwide: the regulatory standards in New Zealand, Australia and United States of America, as well as the non-regulatory guidelines from other stakeholders, such as electrical engineering organisation, insurance companies and electricity providers. This report also examines the characteristics of historical data for transformer fires in distribution substations both in New Zealand and United States of America buildings. Reliability of active fire safety protection systems, such as smoke detection systems and sprinkler systems is reviewed in this research. Based on the data analysis results, a fire risk estimate is determined using an Event Tree Analysis (ETA) for a total of 14 scenarios with different fire safety designs and transformer types for a distribution substation in a high-rise residential and commercial mixed-use building. In Scenario 1 to 10 scenarios, different combinations of fire safety systems are evaluated with the same type of transformer, Flammable liquid (mineral oil) insulated transformer. In Scenario 11 to Scenario 14, two particular fire safety designs are selected as a baseline for the analysis of transformer types. Two types of transformer with a low fire hazard are used to replace the flammable liquid (mineral oil) insulated transformer in a distribution substation. These are less flammable liquid (silicone oil) insulated transformers and dry type (dry air) transformers. The entire fire risk estimate is determined using the software package @Risk4.5. The results from the event tree analysis are used in the cost-benefit analysis. The cost-benefit ratios are measured based on the reduced fire risk exposures to the building occupants, with respect to the investment costs of the alternative cases, from its respective base case. The outcomes of the assessment show that the proposed four hour fire separation between the distribution substations and the interior spaces of the building, when no sprinkler systems are provided, is not considered to be the most cost-effective alternative to the life safety of occupants, where the cost-benefit ratio of this scenario is ranked fifth. The most cost-effective alternative is found to be the scenario with 30 minute fire separation and sprinkler system installed. In addition to the findings, replacing a flammable liquid insulated transformer with a less flammable liquid insulated transformer or a dry type transformer is generally considered to be economical alternatives. From the QRA analysis, it is concluded that 3 hour fire separation is considered to be appropriate for distribution substations, containing a flammable liquid insulated transformer and associated equipment, in non-sprinklered buildings. The fire ratings of the separation construction can be reduced to 30 minute FRR if sprinkler system is installed. This conclusion is also in agreement with the requirements of the National Fire Protection Association (NFPA).

Made available in DSpace on 2014-10-09T12:27:12Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:56:51Z (GMT). No. of bitstreams: 0 / Dissertacao (Mestrado) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN-CNEN/SP

The main goal of this diploma thesis "IT systems reliability analysis" is to analyse and suggest the best PC component seller according to criterion of minimum costs for guarantee services. Theoretical part deals with general importance of technical systems reliability and brings methodological basis for reliability analysis and describes methods and ways how to increase reliability in practise. Following part is focused on identification computer most critical component through Failure Tree Analysis method and modelling its failure-free time by using Weibull analysis. On the base of maintenance costs analysis was chosen the most appropriate distributor.

Made available in DSpace on 2014-10-09T12:27:12Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:56:51Z (GMT). No. of bitstreams: 0 / Neste trabalho é aplicada a metodologia da Análise Probabilística de Segurança nível 1 ao reator IEA-R1. Inicialmente são descritos os eventos iniciadores de acidentes identificados no reator para duas categorias: perda de vazão e perda de refrigerante primário. Dentre eles foram escolhidos dois eventos iniciadores para análise mais detalhada do acidente e obtenção da estimativa da freqüência de danos ao núcleo devido a sua ocorrência. Foram selecionados os seguintes eventos iniciadores: bloqueio de canal de refrigeração (maior probabilidade) e perda de refrigerante por grande ruptura da tubulação do circuito primário (maiores consequências). Para modelar a evolução do acidente a partir da ocorrência do evento iniciador e da atuação ou não dos sistemas de segurança utilizou-se Árvore de Eventos. Através de Árvore de Falhas, também foi avaliada a confiabilidade dos seguintes sistemas: sistema de desligamento do reator, isolamento da piscina, sistema de resfriamento de emergência (SRE) e sistema elétrico. Como resultados foram obtidas as estimativas das frequências de danos ao núcleo do reator e as probabilidades de falha dos sistemas analisados. As freqüências de danos ao núcleo mostraram-se dentro das margens esperadas, sendo da mesma ordem de grandeza que os encontrados para reatores similares. As confiabilidades dos sistemas de desligamento do reator, de isolamento da piscina e do SRE foram satisfatórias para as condições em que estes sistemas foram exigidos. Todavia, para o sistema elétrico seria recomendável uma análise para verificar a possibilidade de modernização a fim de aumentar a sua confiabilidade. / Dissertacao (Mestrado) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN-CNEN/SP

Submitted by Claudinei Pracidelli (cpracide@ipen.br) on 2016-11-11T09:45:23Z No. of bitstreams: 0 / Made available in DSpace on 2016-11-11T09:45:23Z (GMT). No. of bitstreams: 0 / Um desafio recorrente em processos produtivos é o desenvolvimento de sistemas de monitoração e diagnóstico. Esses sistemas ajudam na detecção de mudanças inesperadas e interrupções, prevenindo perdas e mitigando riscos. Redes Neurais Artificiais (RNA) têm sido largamente utilizadas na criação de sistemas de monitoração. Normalmente as RNA utilizadas para resolver este tipo de problema são criadas levando-se em conta apenas parâmetros como o número de entradas, saídas e quantidade de neurônios nas camadas escondidas. Assim, as redes resultantes geralmente possuem uma configuração onde há uma total conexão entre os neurônios de uma camada e os da camada seguinte, sem que haja melhorias em sua topologia. Este trabalho utiliza o algoritmo de Otimização por Colônia de Formigas (OCF) para criar redes neurais otimizadas. O algoritmo de busca OCF utiliza a técnica de retropropagação de erros para otimizar a topologia da rede neural sugerindo as melhores conexões entre os neurônios. A RNA resultante foi aplicada para monitorar variáveis do reator de pesquisas IEA-R1 do IPEN. Os resultados obtidos mostram que o algoritmo desenvolvido é capaz de melhorar o desempenho do modelo que estima o valor de variáveis do reator. Em testes com diferentes números de neurônios na camada escondida, utilizando como comparativos o erro quadrático médio, o erro absoluto médio e o coeficiente de correlação, o desempenho da RNA otimizada foi igual ou superior ao da tradicional. / Dissertação (Mestrado em Tecnologia Nuclear) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN-CNEN/SP