Reliability Prediction Models

Note: This is the third part of a three-part article. It provides guidelines for making a sound judgment when deciding which reliability prediction standard to apply to a particular analysis. Part I identified the major factors in reliability prediction models that contribute to predicting component failure. Part II, described the most widely accepted reliability prediction standards applied in current projects.

Part 3 of 3


MTBF vs Environment The selection of a prediction standard to use for analysis is a critical one. Except in cases where the standard has been specified by your customer, there is no single correct answer for every situation. Many factors must be taken into account when determining which standard is best for your specific needs.

One effective way to decide which prediction standard to employ is to understand the underlying similarities and differences between the standards. Armed with this information, you can select the standard that best fits a particular situation.

Understanding Similarities

Analysis Methodology

No matter which standard is chosen, the MTBF prediction assumes that all components in the system are in series, and this a failure of any component will result in a system failure. The authors of each standard have used either test data or field data from manufacturers to determine empirical equations for predicting the failure rate of a particular part type.

Contributing Factors

Pi Factors The standards use the same factors to calculate the failure rate of a part: part type, environment, temperature, quality, and electrical stress. These factors have different weights in different standards, but the failure rate is based on these factors. To improve the reliability of a system, these five factors could be improved. For example, the failure rate will be lower for a system operating at 20 degrees Celsius than it will be for a system operating at 40 degrees Celsius.

"What-If?" Studies

One of the most common and advantageous uses of predictions is to compare two different designs of a product. It is beneficial to know the impact of different designs on the reliability of a product before finalizing the design. If a warranty will be offered, it could save money to determine which of the designs is predicted to be more reliable. MTBF predictions can be used for a trade-off study, regardless of which standard is used.

Global Use

Each standard was written to be used in a variety of different applications and industries. Each standard supplies an assortment of environments and allows for a large range of temperatures so that it can be utilized for applications worldwide. Translations give the standards a global presence and expand their use to many countries.

Understanding Differences

Failure Rate Adjustment Techniques

Process Grade Questions Some standards offer additional techniques to adjust the predicted MTBF values based on available data. Descriptions of some of these techniques follow.

  • Laboratory and field data can be factored into the analyses in Telcordia, PRISM, and 217Plus.
  • Burn-in time can be used in Telcordia.
  • Process Grades factors are utilized within PRISM and 217 Plus to adjust the predicted MTBF based on design, manufacturing, part quality, system management, cannot duplicate (i.e., maintenance and tracking of errors), induced (i.e., outside failures such as static discharge), wearout, growth, infant mortality, and environment.
  • Predecessor data can be used in PRISM and 217Plus if there is data available from a very similar product such as a previous generation.
  • Cycling profiles can be set up in RDF2000 and IEC TR 62380 when conditions are not constant for the life of a product.

Components Modeled

While a large number of components are modeled across all prediction standards, not all part types are modeled in each standard. Some examples follow:

  • PRISM does not have models for switching devices, relays, or connectors.
  • Optical modules are not covered in MIL-HDBK-217 or GJB/z 299B.
  • Many mechanical components, such as seals and springs, are covered only in NSWC-98/LE1.

Environments Supported

In every standard, the MTBF is directly impacted by the surroundings in which the device will be used. For example, a product will have a lower MTBF if it is used in an airborne environment than if it were used in a fixed ground environment because of the extra stress and vibration. Each standard has its own list of supported environments. Telcordia has only five supported environments, but they are the five most used in the telecom industry: Ground Fixed Controlled, Ground Fixed Uncontrolled, Ground Mobile, Airborne Commercial, and Space. MIL-HDBK-217 offers all of these environments plus more choices useful in military environments, such as Naval Sheltered for equipment inside submarines. PRISM and 217Plus provide the most choices, with 37 environments that are a mix of commercial and military locations. It is important to check that a standard has the environment for your product before choosing it for your MTBF prediction analysis.

Target Industry

Many standards were developed for a specific industry, and because of that, may be more aptly suited to model the environments and/or part types found in that industry. However, it should be noted that all standards are in use in a variety of industries. While MIL-HDBK-217 and 299B were developed for military applications, they are often used in commercial applications. Telcordia, HRD5, RDF 2000, and IEC TR 62380 were developed specifically for the telecom industry but are also used to model products in many other industries. PRISM and 217Plus have been designed from their inception to support both military and commercial products.

Failure Rate Units

The units for failure rate differ between standards. For example, MIL-HDBK-217 uses failures per million hours (FPMH), whereas Telcordia uses failures in time (FITs), i.e., failures per billion hours. The conversion factor between FPMH and FITs is 1,000. Another difference would be PRISM, which is expressed in failures per million calendar hours rather than failures per million operating hours. Although the units for failure rate differ between standards, MTBF values are most commonly expressed in hours.


Windchill Prediction Parts Library Windchill Prediction can use any of the standards described in this article to assess the reliability of your system design. Its superior integration enables you to mix calculation models within a single project, allowing you to select the model most appropriate to each part or component in your design. With Windchill Prediction, the additional analysis capabilities available in one model can typically be applied to any model you use. For example, Telcordia calculation methods for taking field and test data into account in the prediction can also be used with MIL-HDBK-217. Part 3 of this three-part article provides guidelines for making a sound judgment decidingwhich reliability prediction standard to apply to a particular analysis. Additional information about Windchill Prediction, supported models, and predictive modeling in general can be found by visiting our Windchill Prediction product page.