|
|
|
Failure Analysis Is A Continuous Process: Tools That Can Be Utilized
By Rick Fioravanti
|
|
|
An essential part of any AMI business case analysis is the failure rates and replacement factors assumed for the AMI meters and system components. Over the 20-year lifetime typically used for AMI systems, swings in estimated percentage rates for component replacements and failures can have dramatic effects on the estimated paybacks and net present values (NPV) that have been assumed in the analysis and accepted by program stakeholders.
Though manufacturers employ rigorous techniques in order to ascertain these estimates, the uncertainty in some of the numbers is highlighted by the fact that, though there is a great deal of historical data on electro-mechanical meters and communications systems, there is very little operational data on the new AMI meters that are being proposed in today’s systems. The combination of communications systems being integrated into meters for “underglass” products has the potential to reduce overall meter costs; however, such combinations also introduce new components into the meter that do not have a field history on which to base performance and reliability metrics of the final AMI meter.
This article will focus on two methodologies that can be utilized to assist utilities in removing some of these uncertainties and, consequently, help ensure a successful AMI implementation. These methodologies are (1) accelerated life testing – used prior to implementation to test whether components will be able to meet their advertised lifetime thresholds; and (2) Weibull Analysis – used in reliability testing once the components begin to be implemented into the field for actual operation.
|
|
|
Accelerated Life-Cycle Testing
There are a number of variables that can impact the expected lifetime of a unit. Though manufacturers are required to operate over a number of ranges, accelerated life-cycle testing is designed to simulate the possible failure modes that a product would experience in its normal operating life. Accelerated factors commonly used in simulating wear-out failures for solid-state meters are the following:
- Thermal
- Electrical
- Humidity
The approach that is typically used is to put the meter through extremes of one of the factors listed above to allow days to become equivalent to years. The knowledge gained by potential failures modes discovered in the process becomes useful to determine initial corrective actions that may need to be incorporated before large-scale installations occur. This can help in the formulation of maintenance schedules, field training, and spare parts requirements.
As important as the testing process is, choosing the parameters and test plans are equally important in focusing the confidence and accuracy of the test. Soon, IEC 62059-31, the new standard for accelerated reliability testing of energy meters will be published and will be able to provide a standard tool for manufacturers to substantiate claims of expected lifetimes.
Testing facilities can work with utilities to determine the proper testing plans, factors, and samples size in order to provide the best estimates of lifetimes before components reach large-scale installations.
|
|
|
Weibull Analysis:
As field data and operating hours begin to be collected on early installations, continued efforts should be made to assess the lifetimes and failures of the AMI meters. One of the best means to conduct this type of analysis is with Weibull Analysis.
Weibull Analysis is not a new distribution function. The distribution is named after Waloddi Weibull, a Swedish engineer who first popularized its used in the early 1920s for reliability analysis. Weibull Analysis is a form of mean time between failure (MTBF) analysis. However, as meters are first deployed and failures first witnessed, data can be utilized to further refine the failure rates and lifetime expectancies that were estimated prior to installations.
Weibull is a distribution that takes into account information from units that are still operating as well as units that have failed. For example, if three units out of 10 fail in the field, standard MTBF analysis would only utilize information from the failed units. Weibull utilizes the information from the operating units as well to bias the distribution curves to form more accurate assessments of lifetimes.
As each installed component continues to accumulate operating hours, Weibull Analysis provides the opportunity for utilities to continually refine and adjust their failure analysis and replacement estimates. As with accelerated life cycle testing, the data is essential for maintenance schedules and estimating spare part expenditures. However, Weibull also allows for a greater focus on components that make up the AMI meter as well. The analysis not only can be used as a predictive tool to estimate the expected lifetimes of the units operating in the field, but also can be extended to the component level to determine whether any “weak” points in the meter exist.
Extending Weibull Analysis to the component level has the advantage of allowing utilities to forecast areas of improvements for the meter. If unexpected failures are occurring in the field, the analysis will allow utilities to identify the areas that are causing the problems, accurately estimate when issues will affect the greater installed population, and enable manufactures to focus their analysis on single point areas that can improve the overall product lifetime.
Quality and testing groups can help utilities to monitor and record the on-going performance of their meters and then run the distributions to make performance estimates.
|
|
|
Conclusion
These recommended tools allow a utility to make failure and performance analysis a continuous process. Swings in estimated failure and replacement rates can have a dramatic impact on the overall paybacks on AMI systems. However, with very little operating history, utilities can accurately estimate and react to potential issues before the problems impact the overall implementation process.
Accelerated life cycle testing allows utilities to assess whether factors unique to their service territory can have an adverse impact on meter life as well as to confirm the expected lifetimes promised by the manufactures.
Weibull Analysis is a tool that allows utilities to take limited operational data and extrapolate the information into accurate predictions of future performance of the meter. This information can be used to examine the potential replacement cost of an installed AMI base. In addition, it allows utilities to monitor the individual components of the AMI system and employ continuous improvement techniques to circumvent larger issues that may be encountered at later dates in the program.
These preventative maintenance techniques will ensure that utilities are able to focus on the benefits being provided by the AMI systems and not expending time, energy, and money keeping the system operating at its expected levels.
|
|
|
Download the May 2008 Issue
Use the link below to download the PDF of the full issue of the May 2008 Automation Insight for the complete print versions of the articles.
|
|
|
|
[download] Automation Insight - May 2008 (.pdf 130 kb)
|
|
|
|
|
|
|
|
|
