Reevaluating CAIDI’s Relevance: Does it Measure Up to Modern Reliability Expectations?

Some utilities consider achieving a low Customer Average Interruption Duration Index (CAIDI) value an appropriate indicator of success in delivering reliable power. However, the index is not always properly understood, and its use may be overplayed.

CAIDI represents the ratio between the System Average Interruption Duration Index (SAIDI) and the System Average Interruption Frequency Index (SAIFI). As such, CAIDI can be misleading because it can stay the same even when the SAIDI and SAIFI values decrease and there is an improved customer experience.

For example, if a utility improves its grid and decreases its SAIDI to three hours and its SAIFI to one outage, the utility’s CAIDI value would be 3 (three hours per outage). However, the CAIDI value would still be 3 for a utility with a SAIDI value of 12 and a SAIFI value of 4. It can be challenging to see the value in grid modernization when the CAIDI stays flat.

In fact, grid improvements can even cause the CAIDI value to increase. Improving overcurrent coordination between devices, adding more sectionalizing devices, or deploying automatic restoration systems all improve SAIFI and SAIDI ratings. But doing so also tends to produce a higher CAIDI metric because the simpler outages are avoided while the difficult ones that require repairs remain. As a result, a CAIDI value may not accurately represent the reliability state of a utility’s entire system.

Moreover, CAIDI may not offer a meaningful measure of overall service performance for individual customers, regardless of whether they’re experiencing an interruption. As customer satisfaction becomes increasingly driven by the expectation of continuous uninterrupted power, CAIDI’s value in benchmarking overall service performance loses relevance compared to other metrics.

Indeed, the Momentary Average Interruption Frequency Index (MAIFI), which measures occurrences of annoying temporary outages lasting 5 minutes or less for a given period, is all the more important to consider because any variance of power reliability has become so intolerable to customers that they take to social media to express their frustration when power is lost.

Utilities can best evaluate whether their service performance measures up to modern power-reliability expectations by improving their worst performing feeders. These efforts will reduce the number of customers affected by all interruptions, regardless of duration, and improve all relevant measurement values as a result.

I’d be interested in hearing about steps you’re taking to meet customer reliability expectations and how they are has reflected positively on your measurement values. Please feel free to leave a comment below.