If Reliability is the Absence of an Outage, Then What’s an Outage?

Back to Top

Power-interruption problems on utility systems are often called “faults.” How often faults occur and how long they last can vary widely based on the age of the grid’s infrastructure, local geography and climate, storm frequency, and whether distribution lines are overhead or underground.  

The wind in Arizona’s deserts, for example, often slaps conductors together. And on the tree-lined streets of New England, foliage frequently brushes against power lines. When these scenarios occur, faults are gone quickly. These are considered “temporary faults.”

Sadly, in most areas of the country, auto accidents can knock a pole to the ground in an instant. This type of fault would be “permanent” because it would take a longer time to repair the damage and return power to the affected lines.

Power outages caused by faults are what “reliability” boils down to: the fewer and shorter the outages, the more reliable the grid. Utilities are measured on their reliability, and there are ways to track outages. Most deal with the duration, frequency, and impact on the customer. The most common are the System Average Interruption Duration Index (SAIDI), which is an average measure of outage duration across a system, and the System Average Interruption Frequency Index (SAIFI), which measures how often outages occur.

However, most utilities don’t use these metrics to track outages that last less than 5 minutes. They’re not required to. Today, Energy Information Administration data suggest that only 20% of the major utilities recognize this gap and report SAIFI and SAIDI outages that are less than 5 minutes to the Department of Energy.  

Metrics drive change to utility system design, so are we settling for the grid’s status quo by only requiring utilities to report outages longer than 5 minutes? Think about what it’s like to rely on power. At locations such as manufacturing plants or military bases, a one-second power outage or a sag in voltage can be a serious issue. And, as our economy becomes more digitized, the impacts are equally serious and costly.

Those of us who have had the joy of living in an older house or staying in a summer cabin know how turning the washing machine or a hair dryer on can affect power quality: the lights dim or go out for a moment, clocks blink, and maybe the computer or cable box requires resetting. Was that an outage?  

This is a fitting analogy of what utility engineers face today. Conventional causes of temporary outages, such as strong gusts of wind or overgrown trees, aren’t being accounted for in SAIDI and SAIFI. And the grid has newer challenges, too, such as solar or wind, that can drop off or come back on with a passing cloud or change in wind speed. This means load can appear or disappear in an instant, drawing down on the power grid.

This is a gross oversimplification of the real impact done for brevity’s sake. But the result is that momentary outages, even voltage dips, are a norm we’re not measuring.

There is an Institute of Electrical and Electronics Engineers-defined metric for momentaries called MAIFI, or the Momentary Average Interruption Frequency Index, which counts all losses of voltage within a service territory. Some utilities are adding MAIFI or are changing their outage definition to account for momentary losses of voltage.

So what really is an outage? Our grid, our customers, and the dispersion and speed of our economy are all changing rapidly. If we’re truly measuring reliability, we all need to start defining an outage in timeframes that are less than 5 minutes and track ANY momentary absence of power.

I’d be interested in learning your thoughts on whether the power industry is properly measuring utility performance.