Reliability & Risk Models
Reliability Analysis Models
30 Days FREE TRIAL on all the Models 
Field Trial Model (FTM) - This model calculates the expected number of field failures against a set of Confidence Levels. For a Time/Failure Truncated Test Plan testing is continued until a predetermined time is exceeded (accept) or a predetermined number of failures is exceeded (reject). This plan is based on Chi-Squared distribution model. This model can be used to determine the confidence level with which the item under test would be above a certain MTBF. A typical Test Plan based on 95% confidence level, which can be applied to demonstrate the reliability of products with small MTBFs, is outlined below:
3 x MTBF = N x T
N = 3 x MTBF / T ...........(If Test Duration ‘T’ is fixed)
T = 3 x MTBF /N ...........(If Sample Size ‘N’ is fixed);
If no failures are observed in time ‘T’ then there is 95% confidence that the product has met its specified MTBF. This type of Test Plan can be used to demonstrate the MTBF of LRUs provided suitable sample size and sufficient time for testing can be made available. [Ref: Reliability Guide - 7 (Issue 2) Reliability Requirements and R&M Plans Edited by D S Ubhi]
View Demo Reliability_Models.xls
Chi-Square Model
Sequential Test Plan (STP) - Sequential Test Plans are used to decide whether to accept, reject or continue testing an item. A sequential test plan is an alternative to the Failure Truncated or Time Truncated test plan. Sequential Test Plans are produced based on agreed risk for both the Consumer and the Supplier with a Discrimination Ratio between the contractual and offered MTBFs. The test may be run with or without replacement. In majority of cases the producer's risk and the consumer's risk are assumed to be equal. In practice, the items are tested and the item test-hours are accumulated. The observed failures are recorded as they occur.
As the failures occur with time, the staircase plot of failures may either cross the Accept Line or the Reject Line. If the staircase plot crosses the accept line indicating that the item has proved its reliability and it would be reasonable to stop the test. (Refer to Reliability Guide - 7 (Issue 2) Reliability Requirements and R&M Plans Edited by D S Ubhi).
Spares Allocation Model (SAM) - What’s happened to the way we manage spares? We never have the ones we need but there are lots lying at the wrong sites. The answer to this lies in the introduction of Spares Optimisation Algorithm (SOA) with a view to resolving some of the issues being faced through uncoordinated approach.
The SOA provides an easy tool for calculating optimum number of spares required to meet specified requirements for Availability. A simplified model based on Poisson distribution with Lookup Tables has been produced for efficient calculations. Consider all the relevant factors that influence the optimum number of spares
- Working population of FRUs (Field Replaceable Unit)
- FRU MTBF, Turnaround Time dependent on the distance and travel, Stock Out Risk, etc.
- Criticality - What would be the impact (focus on frequency & duration of outages, number of end-users affected etc.?)
- Classification of FRUs (Critical, Essential & Low Priority)
- Cost – CBA (Cost Savings Analysis)
- Geography - Location of Main Stores & Service Points
- Resilience verses Spares holdings
- On-site or off-site spares holding. Other factors that influence the inventory are also considered - Infant Mortality Failures,
- Environmental impact,
- Dead on Arrivals (DOA),
- NFF (No Fault Found) & RWT (Right When Tested),
- Wear-out failures,
- Repair Policy; Swap/Repair, Non-Repairable items
- Delivery of spares,
- Collection of faulty items.
Total Cost of Ownership (TCO) – (TCO) project is very wide ranging and ambitious in its scope, with the goal of producing:
• An integrated over-arching model which will incorporate financial cost modelling, business scenario modelling, and business process modelling in addition to the more common elements of network design modelling.
• It is intended to investigate a wide range of scenarios, and identify radical but cost effective design options for a future multi service combined telecoms and data access network.
Laser Reliability Model (LRM)- Predicting Device Reliability & Lifetime based on Accelerated Life Test data. This model determines Device Acceleration Factor, MTBF & Failure Rate based on empirically calculated Ea (Activation Energy).
Reliability & Survivability Model (RSM) - EXPECTED MTBF FOR A SPECIFIED RATE OF SURVIVAL FOR A GIVEN PERIOD IS SHOWN AS A LOOKUP TABLE FOR EASY NEGOTIATIONS OF CONTRACTUAL REQUIREMENTS AND MANAGEMENT DECISION MAKING. the MODEL ALSO HELPS FOR DETERMINING FAILURE FREE PERIOD FOR THE WARRANTY PERIOD.
Equipment Replacement Model (ERM) - Dynamic Model Based on Cost Minimising Criteria. The model takes age of equipment into account as calendar time in years. The aim is to Minimise total future cost, disregarding discounting.
Risk Models
Refined Risk Analysis Model (RRAM) - This model helps with the Analysis and Quantification of Risk for decision making. It shows a Risk Register and typical Risk categories. The following diagram depicts different techniques for analysis.
The following Diagram depicts how to assess various risks and some case studies.
