Failure rate estimating

The Poisson distribution for observing M events in time r is given by equation 2.5-1, where / is the failure rate estimated as M/i. This model may be used if the failure rate is time dependent rather than demand... 

Over 100 data sheets, conta/ning failure rate estimates for various failure modes and some repair times. 

For catastrophic demand-related pump failures, the variability is explained by the following factors listed in their order of importance system application, pump driver, operating mode, reactor type, pump type, and unidentified plant-specific influences. Quantitative failure rate adjustments are provided for the effects of these factors. In the case of catastrophic time-dependent pump failures, the failure rate variability is explained by three factors reactor type, pump driver, and unidentified plant-specific Influences. Point and confidence interval failure rate estimates are provided for each selected pump by considering the influential factors. Both types of estimates represent an improvement over the estimates computed exclusively from the data on each pump. The coded IPRDS data used in the analysis is provided in an appendix. A similar treatment applies to the valve data. 

Appendix III contains failure rate estimates for various genetic types of mechanical and electrical equipment. Included ate listings of failure rates with range estimates for specified component failure modes, demand probabilities, and times to maintain repair. It also contains some discussion on such special topics as human errors, aircraft crash probabilities, loss of electric power, and pipe breaks. Appendix III contains a great deal of general information of use to analysts on the methodology of data assessment for PRA. 

WASH-1400 is a fundamental document for PRA methodology. The data appendixes contain a great deal of useful information on methods of data assessment. A large number of sources for data are considered, and very general failure rate estimates will produce only gross approximations. Since the advent of data collection schemes across and within plants, the WASH-1400 data are solely useful as a constituent to a data aggregation process or as widely bounded figures that provide a basis for comparison. 

Consider the data from Table 3-1. If it is assumed that the first four failures would be detected by the manufacturer or would occur during the first two weeks of installation and commissioning, they can be excluded from a "useful life" failure rate calculation. If only the data from the first year of operation (8760 hours) is available for failure rate estimation, the failure data is presented in Table 3-2. 

Kuball, S., J. May, and G. Hughes (1999). Building a system failure rate estimator by identifying component failure rates. Proceedings of the 10th International Symposium on Software Reliability Engineering (ISSRE 99), 32-41. 

The structure of the paper is as follows In Section 2, some of the factors that may influence the predicted and the actual safety integrity are discussed. Section 3 gives a brief overview of some of the challenges related to failure rate estimation, including an overview of relevant literature. The new approach, comprising four steps, is presented in Section 4. Some case examples are also included here. In Section 5, the authors share some experience from some industry projects involving analysis of SIS related failures. Section 6 gives some conclusions and final remarks. 

Step 3 Updating failure rate estimates The initial rate of DU-failures can be updated based on operational experience in two different ways (i) by estimating 3.du,i solely based on observed failures, or (ii) by combining operational experience with the initial estimate by a Bayesian updating procedure. 

A very high initial DU-failure rate The fact that no failure has occurred during the observation period indicates that the initial failure rate estimate may be too high. 

When the criteria for sufficient operational experience are not met, we suggest a Bayesian update of the failure rate estimate. 

The recorded experience data, i.e., the number, x of DU-failures during the aggregated time in service ti, can now be used to calculate the new (posterior) failure rate estimate ... 

We will next discuss how the updated failure rate estimate is used as basis for evaluating the functional test interval. 

The primary benefit to be derived from reliability and safety engineering is the reliability and integrity growth which arises from ongoing analysis and follow-up as well as from the corrective actions brought about by failure analysis. Reliability prediction, based on the manipulation of failure rate data, involves so many potential parameters that a valid repeatable model for failure rate estimation is not possible. Thus, failure rate is the least accurate of engineering parameters and prediction from past data should be carried out either ... 

This independence between the upper bound on failure rate and the number of accidents is particularly useful in cases where the failure rate bound has not been estimated correctly, e.g. due to a flaw in the specification. Such a flaw would invalidate any failure rate estimate based on testing, but the accident bound derived from equation (6) would still be valid provided N included an estimate for dangerous specification flaws. This differs from hardware where the instantaneous failure rate is often assumed to be constant, so expected accidents always increase with fleet usage. 

Example 4. A particular microprocessor (MPU) is assigned for a fuel-injection system. The failure rate must be estimated, and 100 MPUs are tested. The test is terrninated when the fifth failure occurs. Failed items are not replaced. This type of testing, where n is the number placed on test and ris the number of failures specified, is termed a Type II censored life test. 

Human error probabilities can also be estimated using methodologies and techniques originally developed in the nuclear industry. A number of different models are available (Swain, Comparative Evaluation of Methods for Human Reliability Analysis, GRS Project RS 688, 1988). This estimation process should be done with great care, as many factors can affect the reliability of the estimates. Methodologies using expert opinion to obtain failure rate and probability estimates have also been used where there is sparse or inappropriate data. 

Confidence Estimation for the Constant Failure Rate Model... 

A pressure vessel designer, on the basis of test, inspection, and experience, says that a vessel has a failure rate less than 1E-6A Y, Experience has already demonstrated 9.3E5VY without failure. What is the estimated failure rate for 95% confidence ... 

The objective is to estimate, numerically, the probability that a system composed of many components will fail. The obvious question is, "Why don t you just estimate the failure rate of the system from operating experience " There are three reasons IJ the system may not exist, so new data are not available, 2) the injuries and fatalities from the developmental learning experience are unacceptable - the risk must be known ahead of time, and 3) by designing redundancy, the probability of the system failing can be made acceptably remote in which case system failure data caimot be collected directly. The only practical way uses part failure statistics in a system model to estimate the system s reliability. 

Perhaps the simplest way to assess the reliability of a system is to count the active parts, flic 1C liability estimate is the product of the number of parts and some nominal failure rate for the parts. Ill the design phase, two competing designs may be compared on the basis of the numbei of parts but several cautions are in order. 

We previously encountered failure modes and effects (FMEA) and failure modes effects and criticality analysis (FMECA) as qualitative methods for accident analysis. These tabular methods for reliability analysis may be made quantitative by associating failure rates with the parts in a systems model to estimate the system reliability. FMEA/FMECA may be applied in design or operational phases (ANSI/IEEE Std 352-1975, MIL-STD-1543 and MIL-STD-1629A). Typical headings in the F.Mld. A identify the system and component under analysis, failure modes, the ef fect i>f failure, an estimale of how critical apart is, the estimated probability of the failure, mitigaturs and IHissihiy die support systems. The style and contents of a FMEA are flexible and depend upon the. ilitcLiives of the analyst. 

The overall system failure rate including common modes was estimated to be the geometric mean of these extremes (equation... 

The distribution disk with this book includes a folder entitled BNLDATA which contains the file bnlgener.xls, a spreadsheet in EXCEL format. It is a collection of failure rate data drawn from many sources. The file refem.txt contains the references to the table s data. There are 1,311 data entries many cite different estimates by different organizations for the failure rate of the same... 

Appendix HI, of WASH-1400 presents a database from 52 references that were used in the study. It includes raw data, notes on test and maintenance time and frequency, human-reliability estimates, aircraft-crash probabilities, frequency of initiating events, and information on common-cause failures. Using this information, it assesses the range for each failure rate. 

The numerator is a random normally distributed variable whose precision may be estimated as V(N) the percent of its error is f (N)/N = f (N). For example, if a certain type of component has had 100 failures, there is a 10% error in the estimated failure rate if there is no uncertainty in the denominator. Estimating the error bounds by this method has two weaknesses 1) the approximate mathematics, and the case of no failures, for which the estimated probability is zero which is absurd. A better way is to use the chi-squared estimator (equation 2,5.3.1) for failure per time or the F-number estimator (equation 2.5.3.2) for failure per demand. (See Lambda Chapter 12 ),... 

Table 4.1-5 shows in these records, there has been 3 control rod drive failures. Assume 100 plants in the U.S. with an average of 30 control rods/plant and 10.7 years of experience in this database. Estimate, the mode, 90% and 10% confidence limits for the failure rate. 

The accident sequence frequencies are quantified by linking the system fault tree models together as indicated by the event trees for the accident sequence and quantified with plant-specific data to estimate initiator frequencies and component/human failure rates. The SETS code solves the fault trees for their minimal cutsets the TEMAC code quantitatively evaluates ihe cm sols and provides best estimates of component/event probabilities and frequencies. 

Answer You can treat them as one component in the component column and prepare the FMECA for ways they can fail together. For random failure of both valves in a mission time, estimate the failure probability as one-half the failure rate of one times the probability of ailing in the mission time of the other. 

Confidence that the calculated failure rate is a good estimate of the true rate can be increased by lengthening the study or sample time. Adding another population of the same equipment under the identical circumstances to the original population will reduce uncertainties and increase confidence in the calculated failure rates. 

Event frequencies estimated from historical data failure rates from System Reliability Service... 

In April 1982, a data workshop was held to evaluate, discuss, and critique data in order to establish a consensus generic data set for the USNRC-RES National Reliability Evaluation Program (NREP). The data set contains component failure rates and probability estimates for loss of coolant accidents, transients, loss of offsite power events, and human errors that could be applied consistently across the nuclear power industry as screening values for initial identification of dominant accident sequences in PRAs. This data set was used in the development of guidance documents for the performance of PRAs. 

---