TIME DEPENDENT RELIABILITY OF EMRGENCY DIESEL GENERATOR STATION

Transcription

1 Indian J.Sci.Res.1(2) : , 2014 ISSN: (Online) ISSN : (Print) TIME DEPENDENT RELIABILITY OF EMRGENCY DIESEL GENERATOR STATION SIMA RASTAYESH 1a, SAJJAD BAHREBAR b, AND KAMRAN SEPANLOO c ab Science and Research Branch Islamic Azad University of Tehran c Daneshestan Institute of Higher Education ABSTRACT As a consequence of Emergency diesel generators (EDGs) are in stand-by status through normal operation of nuclear power plants (NPPs), it is required to make certain about their start-up in demand. Due to the fact that they are in standby mode for a long period of time, it is essential to consider time effect in system reliability during reactor age. This (Ageing) would be influence on component s reliability for plants, which are designed for more than 25 years operation. In this paper, dynamic reliability block diagram (DRBD) is used as one of time dependent methods in reliability assessment of emergency diesel generator station (EDGS) which that its unavailability would lead to a severe accident such as Fukushima. In order to consider that an appropriate time dependent reliability distribution has been selected. Timedependent reliability models have been utilized to study EDGS s reliability. All mechanical subsystems properties including: fuel storage & pumping, lubrication, cooling, starting air and air cooling system compared with each other by taking into account scheduled, corrective and preventive tasks and dependency of all sub sub systems in point of the view of considering time dependent behaviour of all components of EDGS. At the end, by applying discrete event simulation and considering maintenance policy, availability of the system is calculated for 30 years operation of a typical NPP EDGS. KEYWORDS: Ageing, Dynamic Reliability Block Diagram, Discrete Event Simulation, Maintenance, Time Dependent Reliability. Reliability assessment is a vital step, in designing and analyzing (critical) systems. In the last years, reliability assessments have gained widespread importance in many technological areas such as, the nuclear, aerospace and chemical industries (Marseguerraa, 1998). In such cases, accurate models are required to adequately take into account the impact of external environment interference or dependent behaviors affecting the system, especially if the system experiences high complexity (Distefano, 2012). An important characteristic of many engineering systems is that they behave dynamically, i.e. their response to an initial perturbation evolves over time as system components interact with each other and with the environment (Siu, 1994). Conventional reliability assessment methodologies are static in nature and have potential weaknesses when applied to dynamic process systems (Siu, 1992). Conventional methods for reliability assessment such as fault tree are designed to illustrate static relations between logical variables, and do not explicitly include time, process variables, or human behavior (which affect the system dynamic response). effects in the calculations by choosing an appropriate dynamic reliability distribution. Dynamic reliability models have been used to investigate EDGS s reliability. All mechanical subsystems properties including: fuel storage & pumping, lubrication, cooling, starting air and air cooling system compared with each other by taking into account scheduled, corrective and preventive tasks and dependency of all sub sub systems in point of the view of considering time dependent behaviour of all components of EDGS. Finally, by applying discrete event simulation and considering maintenance policy, availability is calculated for 30 years operation of Boushehr NPP EDGS. In recent years, a new generation of dynamic modeling tools for reliability evaluation of complex systems has been developed. Initially, state-space methodologies were successfully used to model systems including interdependencies that could not previously be handled by traditional combinatorial techniques (Manno, 2012). One of these methods is DRBD which can model: load-sharing, standby redundancy, interferences, dependencies, common mode failures. Furthermore, the configuration of a system considering various reliability aspects: failed component /repaired subsystem, the changed systems phases could vary (Distefano, 2007). In this paper, DRBD is used as one of dynamic approaches for reliability assessment of a critical component EDGS, which its unavailability was highlighted in Fukushima accident. One of principal aspects that could affect system s reliability is ageing. Ageing becomes an important issue as average age of operating nuclear power plants is passing over 25 years (Poghosyan). It is necessary to consider ageing INTRODUCTION OF EDGS The EDGS is intended for independent power supply to essential station services on the case of occurrence of loss of off-site power (LOOP). The EDGS incorporates four physically separated trains independent from each other. Each one provides electric power to one safety train. Each train has two diesel generators (DGs). DGs are connected in parallel. In case of power loss, two DGs of each EDGS train are startedup simultaneously and are connected to an emergency bus bar when achieving nominal voltage. It is assumed that the emergency trains won t fail in case of occurrence of initiating events requiring EDGS operation. In 1 Corresponding author

2 NPP normal operation condition, DGs are in stand-by status in a continuous readiness for start-up. By initiating event occurrence, requiring operation of DGs, the start-up of DG is performed automatically. EDGS HAVE FOLLOWING SUPPORT SYSTEMS: i. EDGs Fuel Storage & Pumping System is intended for fuel storage, preparation (sediment, filtration) and supply of fuel to the diesels. ii. EDGs Lubrication System is intended for oil storage, preparation (sediment, filtration) and supply of lubrication oil to the diesel as well as maintaining the diesel in a state of readiness for startup within 15 s. iii. EDGs Cooling Water System (cooling system) is intended for maintaining the predetermined oil and water temperature in diesel standby mode (in a state of readiness for start within 15 s) and cooling the DG and compressor during DG operation on LOOP at the NPP or in its testing mode. iv. EDGs Air Intake & Exhaust System (suction exhaust system) provides intake, cleaning and supply to the diesel of air necessary for combustion process and removal of exhaust gases during DG operation. v. EDGs starting air system is intended for compressed air generation, storage, preparation and supply of air to the diesel. vi. EDGs Air Cooling System is intended for removal of excessive heat releases and maintaining air temperature within the specified limits in DG machine rooms (IAEA-TECHDOC- 643,1992). DYNAMIC MODELING PROCESS Following introducing EDGS system, the relationship among their components is delineated by means of DRBD. Investigation was done through the whole reliability conditions and traced the results for each component, applying for probabilistic risk assessment of Boushehr nuclear power plant (Atom energoproekt, 2003) using generic data (Eide). Distribution selection Reliability model of each component is modeled through Weibull distribution taking into account shape factor (beta) equal to 2. This shape factor is used to consider ageing effect on components by increasing failure rates of them. β in Weibull distribution: β=1 Exponential distribution β 1 Decreasing Failure Rate (DFR) β 1 Increasing Failure Rate (IFR) Scale factor (alpha) was calculated utilizing goodness of fit testing method formulated by using constant failure rate (λ ) in the exponential distribution. In this way: 1 β Mean Time to Failure = = αγ( ) = αγ( ) λ β 2 (1) RESTORATION FACTOR The effect of repair and inspection on restoration of components is calculated by using a restoration factor (RF). In classic assessment, it was assumed that a repaired component is as good as new. This is usually the case when the component is replaced with a new one. The concept of a RF may be used in cases in which one wants to model imperfect repair, or a repair with a used component. The best way to indicate that a component is not as good as new is to give the component some age. A RF concept is used to better describe the existing age of a component. The RF is used to determine the age of the component after a repair or any other maintenance activity. The RF in this paper is defined as a number between 0 and 1 and has the following effect: RF value of 1 (100%) implies that the component is as good as new after repair, which in effect implies that the starting age of the component is 0. RF value of 0 implies that the component is the same as it was prior to repair, which in effect implies that the starting age of the component is the same as the age of the component at failure. RF value of 0.25 (25%) implies that the starting age of the component is equal to 75% of the age of the component at failure. Experience has showed that on the average the RF for EDGS is about 85%; this is an important part for modeling dynamic behavior of a component. These considerations are applied for both corrective tasks and scheduled task properties of each block in DRBD. Standby modeling During this process, for achieving better result for dynamic modeling through DRBD, standby containers is used since the system is considered in standby phase. In standby redundancy, the redundant components are set to be under a lighter load (or no load) condition while not used and under the operating load when activated. In standby redundancy, the components are set to have two states: active state and standby state. Components in standby redundancy have two failure distributions, one for each state. When in the standby state, they have a quiescent (or dormant) failure distribution and when operating, they have an active failure distribution. Indian J.Sci.Res.1(2) : ,

3 As an example, DRBD of Emergency DGs Fuel Storage & Pumping System is shown in (Fig. 1). Figure 1. DRBD of EDGs fuel storage & pumping system including standby containers Sub diagram modeling Furthermore, in the process of modeling sub diagram blocks are employed. Sub diagram blocks represent other diagrams within the project. Utilization of sub diagram blocks allows us to maintain separate diagrams for portions of a system and to incorporate those diagrams as components of another diagram. This also allows generating and analyzing extremely complex diagrams representing many subsystems, sub of subsystems, etc. DRBD of one EDGS train is illustrated in (Fig. 2). Figure 2. DRBD of one EDGS train Including Sub diagrams Mirror group modeling Another methodology for modelling a better dynamic behavior of system is using mirror group. Mirrored blocks allows to place the same block in more than one location. In other words, mirrored blocks allows to represent a single component more than once in a diagram or in multiple diagrams within a project. This is useful, for example, when simulating bi-directional paths within the diagram or if the same item performs more than one function and each function has a different reliability-wise impact on the system. The ability to mirror across diagrams allows to represent situations where a single component s failure affects multiple subsystems that are represented in separate sub diagrams. Any changes made to the properties of a block in a mirror group will apply to all other blocks. Switchgear in (Fig. 3) is an example of mirror group. It can also be seen in (Fig. 2). Indian J.Sci.Res.1(2) : ,

4 Figure 3. DRBD of EDGs lubrication system including mirrored blocks DISCRETE EVENT SIMULATION This is the average time the system was up and operating. This is obtained by taking the sum of the uptimes for each simulation and dividing it by the number of simulations. Obtaining from (Fig. 5) fuel storage and pumping system has the most uptime compare with other sub systems. At the end, discrete event simulation is performed as significant part of dynamic reliability analysis. In contrast to classic FT, DRBD can take into account repair and restoration actions. These actions mean that the age of system components is no longer uniform and the operation of the system is not continuous. If the information on the repair and maintenance characteristics of the components is available in the system, other information can also be obtained, such as system availability. This can be accomplished through discrete event simulation. In this simulation, random failure times are generated from each component's failure distribution. These failure times are then considered in accordance with the way the components are reliability-wise arranged within the system. The overall results are analyzed in order to determine the behavior of the entire system. (ReliaSoft, Weibull) RESULTS AND DISCUSSION The DRBD of EDGS is simulated for 30 years. Simulation results for each subsystem is shown in following figures, number of simulation is assumed Number of failures This is the average number of system downing failures, this number does not include failures with zero duration. It can be undrestood from (Fig. 4) that high percentage share of failures occur in air cooling system. So that focous on improve designing of it should be considered to eliminate numbers of failures of air cooling system. CM Downtime, T CM Down Figure 5. Uptime This is the average time the system was down for corrective maintenance actions (CM) only. This is obtained by taking the sum of the CM downtimes for each simulation and dividing it by the number of simulations Uptime, Figure 4. Number of failures Inspection Downtime This is the average time the system was down due to inspections. This is obtained by taking the sum of the inspection downtimes for each simulation and dividing it by the number of simulations. PM Downtime, This is the average time the system was down due to preventive maintenance (PM) actions. This is obtained by taking the sum of the PM downtimes for each simulation and dividing it by the number of simulations. Indian J.Sci.Res.1(2) : ,

5 Mean Availability (w/o PM, OC & Inspection), This is the mean availability due to failure events only. Downtimes caused by PM and inspections are not included. However, if the PM or inspection action results in the discovery of a failure, then these times are included. As an example, consider a component that has failed but its failure is not discovered until the component is inspected. Then the downtime from the time failed to the time restored after the inspection is counted as failure downtime, since the original event that caused this was the component's failure. Point Availability (All Events), Figure 6. Comparison of corrective maintenance,inspection and preventive downtime Mean Availability (All Events), This is the probability that the system is up at time. As an example, to obtain this value at = 300, a special counter would need to be used during the simulation. This counter is increased by one every time the system is up at 300 hours. Thus, the point availability at 300 would be the times the system was up at 300 divided by the number of simulations. For example, if this is 0.930, or 930 times out of the 1000 simulations the system was up at 300 hours. (Fig. 7) indicates the comparison among all subsystems. This is the mean availability due to all downing events, which can be thought of as the operational availability. It is the ratio of the system uptime divided by the total simulation time (total time). Indian J.Sci.Res.1(2) : ,

6 Figure 7. Comparison of availability of each subsystem MTTFF MTTFF is the mean time to first failure for the system. This is computed by keeping track of the time at which the first system failure occurred for each simulation. MTTFF is then the average of these times. Figure 8. Mean time to first failure Reliability of each sub-diagram decreases over time (30 years). This trend for air cooling system has a rapid rate which indicate that this subdiagram possess more failures than the others. As a result, design of this subdiagram should be done more carefully. For instance using redundant systems could be a solution. It can be seen in (Fig. 9): It is important to note that for each simulation run, if a first failure time is observed, then this is recorded as the system time to first failure. If no failure is observed in the system, then the simulation end time is used as a right censored (suspended) data point. MTTFF is then computed using the total operating time until the first failure divided by the number of observed failures (constant failure rate assumption). Furthermore, and if the simulation end time is much less than the time to first failure for the system, it is also possible that all data points are right censored (i.e., no system failures were observed). In this case, the MTTFF is again computed using a constant failure rate assumption, or: where is the simulation end time and is the number of simulations. One should be aware that this formulation may yield unrealistic (or erroneous) results if the system does not have a constant failure rate. If you are trying to obtain an accurate (realistic) estimate of this value, then your simulation end time should be set to a value that is well beyond the MTTF of the system (as computed analytically). As a general rule, the simulation end time should be at least three times larger than the MTTF of the system. Mean time to first failure according to (Fig. 8) for startup air system is the highest. Figure 9. Reliability vs. time for each subsystem Point reliability for each sub diagram is illustrated in Fig. 9, which shows its independency to corrective and scheduled tasks. It is shown that reliability of a system the main contribution is made by only failure model and the resulting RBD. The availability of system over 30 years decreases. As we can see from the ( Fig. 10), the maintenance policy, shows its effect here. That mean, ageing has effect on availability because it decreases by decreasing reliability and constant maintenance. Indian J.Sci.Res.1(2) : ,

7 Indian J.Sci.Res.1(2) : , 2014 ISSN: (Online) ISSN : (Print) Figure 10. Availability vs. Time for one train of EDGS over 30 years In our assumptions, upper band is 24 hours and its availability calculated , by passing time about 2 years availability become 0.8 which is reasonable because in the availability, corrective tasks and preventive maintenance (inspections) are taken into account and it is assumed that EDGS operates continuous for more than 30 years. IMPORTANCE AND EVALUATION OF RESEARCH Calculations and results of this research are prerequisite for PRA (Probabilistic Risk Assessment) which unavailability of EDGS can lead to core melt down as an example Fukushima accident. Finally, unavailability of Boushehr EDGS NPP including two parallel diesel generators calculated from the mentioned REFERENCES Marseguerraa M., Zioa E., Devooght J., Labeaub P.E.; A concept paper on dynamic reliability via Monte Carlo simulation, Mathematics and Computers in Simulation, 47: Distefano S., Puliafito A., Trivedi K.S.; Dynamic aspects and behaviors of complex systems in performance and reliability assessment, PSAM11-ESREL2012 Proceeding. Siu, N.; Risk assessment for dynamic systems: An overview, Reliability Engineering and System Safety, 43: methods discussed previously is compared with a similar EDGS from generic data (NUREG/CR 6928). It is found that the order of magnitude are at the same range. All preceding calculation are followed by classic methods, which could not calculate a long period of time such as 30 years, which is calculate in this paper. In order to compare unavailability, 24 hour of EDGS operation assumed to validate our results. Table 1. Comparison of unavailability Our calculation 1.1 According NUREG/CR In 24 hour unavailability [Siu N.0.; Dynamic Approaches - Issues and Methods: An Overview. Proceedings of the NATO Advanced Research Workshop on Reliability and Safety Assessment of Dynamic Process Systems, August: 3-7. Manno G., Chiacchio F., Compagno L., D Urso, D., Trapani N.; RAATSS, an extensible Matlab toolbox for the evaluation of repairable dynamic fault trees, PSAM11- ESREL2012 Proceeding. Distefano S., Puliafito A.; Dynamic reliability block diagrams: Overview of a methodology. Risk, Reliability and Societal Safety, ANS PSA 2011 International Topical Meeting on Probabilistic Safety Assessment and Analysis, Wilmington, NC, March 13-17, 2011, on CD-ROM, American Nuclear Society, LaGrange Park, IL (2008). 1 Corresponding author

8 Poghosyan Sh., Amirjanyan A.; Investigation of Ageing Impact on Safety Systems Reliability. Nuclear and Radiation Safety Center. Atom energoproekt.; BVPP Probabi Listic safety Assessment, level 1, revision 3, state resarch, Design and Engineering survey Institute, Atomener goproekt. Eide S.A et al, Industry Average Performance for components and Initiating Events at U.S. Nuclear Regulatory Commission, NUREG/CR BlockSim user's guide; ReliaSoft Corporation; Indian J.Sci.Res.1(2) : ,