For US Utilities serving 3,000-100,000 meters and for operations team, billing team and utility managers. For Heads of Billing who own collections accuracy and revenue leakage.
A SAIDI score (System Average Interruption Duration Index) is the total number of minutes of sustained power interruption the average customer of an electric utility experiences in a year. It is calculated by adding up customer minutes of interruption and dividing by the total number of customers served. SAIDI is usually reported alongside SAIFI, which counts how many interruptions the average customer experiences, and together they are the two most common measures of distribution reliability.
SAIDI is one of the distribution reliability indices defined in IEEE 1366, the IEEE guide that sets out how electric utilities should calculate and report reliability for systems, substations, circuits, and regions. The index answers one question: over a year, how long was the average customer without power?
The score is expressed in minutes per customer per year, although some reports convert it to hours. A SAIDI of 120 means the average customer was without service for two hours in total across the year. That total might come from one long outage or from several short ones; SAIDI does not distinguish between the two, which is why it is read alongside SAIFI.
SAIDI counts only sustained interruptions. Under the IEEE definition, a sustained interruption lasts longer than five minutes; shorter ones are momentary and tracked separately (T&D World). A breaker that trips and recloses in a few seconds does not add to SAIDI, even though the customer sees the lights blink.
For a distribution utility, SAIDI is both a performance measure and a record-keeping test. The number is only as accurate as the utility's records of which customers were affected and when service was restored. Those records sit across the outage, field, and customer systems in an electric utility management platform, which is why reliability reporting often exposes data problems that have nothing to do with the grid itself.
IEEE 1366 defines several indices. Four appear in most utility and regulatory reports, and each answers a different question about the same outage record.
The relationship between the first three matters most in practice. SAIDI equals SAIFI multiplied by CAIDI, so a utility can see whether a poor SAIDI comes from too many outages, slow restoration, or both. A falling SAIFI with a rising CAIDI, for example, usually means fewer outages are occurring but crews are taking longer to restore the ones that do.
The SAIFI score works the same way as SAIDI but counts events instead of minutes. A SAIFI of 1.2 means the average customer experienced just over one sustained interruption in the year.
The arithmetic is simple. The work is in collecting accurate inputs for every interruption in the reporting period.
The worked example below uses an illustrative utility serving 10,000 customers with four sustained interruptions in a year.
For this utility, SAIDI is 343,500 divided by 10,000, or 34.4 minutes. SAIFI is 4,750 divided by 10,000, or 0.48 interruptions. CAIDI is 343,500 divided by 4,750, or about 72 minutes per interruption.
There is no single number that counts as a good SAIDI score, because the result depends heavily on geography, weather, and how the utility treats major events. The most useful references are national averages and peer comparisons.
The U.S. Energy Information Administration reported that customers averaged about five and a half hours of interruptions in 2022, with about 1.4 interruptions per customer, and that the average excluding major events has stayed around two hours a year since 2013 (EIA). In 2024, hurricane damage pushed the national average to 11 hours, of which major events accounted for about 9 hours, and the average customer experienced 1.5 interruptions (EIA).
A utility whose SAIDI excluding major events sits well under 120 minutes is performing better than the national norm. The American Public Power Association awards reliability certificates to member utilities whose SAIDI falls in the top quartile nationally, and it reported that more than 90% of utilities that verified their 2024 data had a SAIDI below the 2019 to 2023 national average (APPA).
Comparisons between utilities need care. Several factors move SAIDI regardless of how well a utility operates:
For these reasons, the most reliable test of a SAIDI score is the utility's own trend over several years, calculated the same way each year.
Most utilities report two SAIDI figures: one including every interruption, and one excluding major event days. Without the split, one hurricane can hide years of steady improvement, or a mild year can hide worsening equipment.
IEEE 1366 separates the two with a statistical method known as the 2.5 beta method. In summary, it:
Because the threshold is calculated from each utility's own history, a larger or more weather-exposed utility gets a threshold that reflects its normal conditions. IEEE working group guidance presented to state regulators recommends setting performance targets on day-to-day operations and reporting both data sets separately (NARUC).
The practical point for a utility is consistency. A SAIDI trend is only meaningful if major events are identified the same way every year.
Reliability indices serve several audiences at once.
EIA publishes utility-level reliability data in the Reliability data file of Form EIA-861 (EIA). Requirements from state regulators vary; investor-owned utilities typically report to their commission, while many municipal and cooperative utilities report to their own boards. Our overview of electric utility compliance software covers how reliability reporting fits with other regulatory obligations.
Internally, SAIDI and SAIFI are most useful when broken down by circuit, cause, and season. A utility-wide SAIDI shows whether reliability is improving; circuit-level results show where to spend.
SAIDI depends on three inputs: when each interruption started and ended, which customers it affected, and how many customers the utility serves. Each comes from a different system, and the handoffs between them are where most reporting errors begin.
An outage management system usually produces the event record; our guide to what an outage management system is explains how it models the network and tracks restoration. The customer and field records it relies on are a separate responsibility.
Can you list every account connected to a given transformer, and is that list current?
Does your customer count for reliability reporting match your active billing accounts?
When a crew restores service, is the restoration time recorded separately from the time the work order is closed?
If the answer to any of these is no, the SAIDI score reflects record quality as much as grid performance.
SMART360 is not an outage management system, and it does not calculate reliability indices or model the distribution network. It is the platform where the customer and field records behind those indices are kept, on one database:
Improving the indices comes down to two levers: fewer interruptions, which lowers SAIFI, and faster restoration, which lowers CAIDI. SAIDI falls when either improves.
Much of this work starts with knowing what equipment is in the field, how old it is, and what has failed before. Our guide to electric utility asset management software covers keeping those records current across the distribution system.
Small utilities often see the fastest gains from data rather than equipment. Correcting customer-to-transformer mapping and recording restoration times accurately can change reported SAIDI before any capital project begins, because the utility is finally measuring what actually happened.
There is no universal threshold, but the U.S. average excluding major events has stayed around two hours (about 120 minutes) a year since 2013, according to EIA. A utility consistently below that level is performing better than the national norm. The fairest comparison is against the utility's own history and peers of similar size, region, and weather exposure.
SAIFI, the System Average Interruption Frequency Index, is the number of sustained interruptions the average customer experiences in a year. It is calculated by dividing the total number of customers interrupted by the total number of customers served. A SAIFI of 1.0 means the average customer lost power once.
SAIDI spreads interruption minutes across all customers served, so it measures the average customer's total outage time. CAIDI spreads the same minutes across only the customers who were interrupted, so it measures average restoration time. CAIDI equals SAIDI divided by SAIFI.
SAIDI includes only sustained interruptions, those longer than five minutes. Momentary interruptions are tracked separately with MAIFI. Major storms are included in the all-events SAIDI, but most utilities also report a SAIDI that excludes major event days identified with the IEEE 2.5 beta method.
No. SMART360 is not an outage management system and does not calculate reliability indices. It keeps the customer accounts, service points, asset records, and work orders that reliability calculations depend on, and exports that data to the systems that produce SAIDI and SAIFI.
A SAIDI score is only as accurate as the records behind it. SMART360 keeps customer accounts, service points, asset records, work orders, and customer communication on one database, so the customer count, customer-to-equipment mapping, and field records used in reliability reporting stay consistent. A South Asian electric distribution utility has run SMART360 across 133,000 accounts for seven years with zero platform failures.
Bynry Inc.
8 The Green STE.R,
Dover, Kent,
Delaware 19901
+1 732-630-0285
Bynry Inc.
Serene Tower, Pakharbaug
Pune, Maharashtra 411021
+91 93252-51217
