The Reliability Numbers Behind Your Backup Plan
SAIDI, SAIFI, CAIDI, and MAIFI describe your outage risk. The numbers your utility leads with are filtered, averaged, and stripped of their worst days. Here is how to read them and what to demand.
Who this is for
- ■Plant managers at manufacturers whose production stops when the lights go out
- ■Facility and operations leaders at hospitals, cold storage, and critical infrastructure
- ■Finance leaders evaluating backup power, BESS, or microgrid capital requests
- ■Any C&I operator whose electric bill runs five figures or more a month
Does the reliability number your utility is showing you actually reflect your facility's exposure, or a sanitized version of it?
SAIDI measures how many total minutes per year the average customer loses power. SAIFI measures how many times that happens. Your utility reports both numbers. Those numbers are filtered, averaged across a customer population dominated by residential meters, and stripped of their worst days. The question for your operation is whether the number your utility is showing you reflects your actual exposure, or a sanitized version of it.
The core four
| SAIDI | Total outage duration per average customer per year | Minutes per year | Lower |
| SAIFI | How often outages happen per average customer | Interruptions per year | Lower |
| CAIDI | Average duration of an outage once it starts (SAIDI divided by SAIFI) | Minutes | Lower |
| MAIFI | Frequency of momentary interruptions under the MAIFI threshold | Interruptions per year | Lower |
32%of this guide, read. The rest of it is below.
- 02 The mechanism How the numbers get built and where they leak
The math is straightforward. SAIFI is total customer interruptions divided by total customers served. SAIDI is total interruption minutes divided by total customers served. CAIDI is SAIDI divided by SAIFI. If SAIDI is 120 minutes and SAIFI is 1.5 outages/year, CAIDI works out to 80 minutes. That is the average time to restore power once an outage starts.
How CAIDI is derived120 minutes/1.5 outages/year=80 minutes = 80 minutesCAIDI is not measured independently. It is a ratio of the two headline indices your utility reports.Two blind spotsWhere the reporting breaks down
Blind spot oneMajor Event Days
A Major Event Day is a day when severe storms or natural disasters cause widespread outages outside normal operating conditions. IEEE 1366 sets the criteria. Utilities report two sets of numbers, one with MEDs included and one without. The MED-excluded number is what usually leads.Blind spot twoMAIFI
Momentary interruptions under the MAIFI threshold do not count toward SAIDI or SAIFI. A three-second interruption can stop a CNC machine, trip a variable frequency drive, or corrupt a PLC process the same as a three-hour outage in terms of production loss.The EIA reports national averages excluding MEDs at SAIDI of 123.6 minutes, SAIFI of 1.5 outages/customer, and CAIDI of 131.6 minutes. Those are the numbers that show up in utility performance decks. To see what MED inclusion actually does, consider one major regional utility that carries a five-year average SAIDI of 5.1 hours/customer/year when MEDs are included. The gap between MED-included and MED-excluded is the storm exposure that standard reporting obscures. For many utilities that gap is a multiple of the number they lead with.
Reported reliability versus storm-inclusive reliabilityThe number your utility leads with sits below the national average. The number a continuity plan must address sits well above it. MAIFI is the other half of the story. Momentary interruptions never appear on a standard SAIDI or SAIFI report, but they land the same hit on a sensitive load. The consequence: you can be told your utility improved and still see your production line trip more often. The industry itself is trending the wrong way, with a 71% increase in momentary outages during the period documented, and 20% of C&I customers experiencing outages on a monthly basis.
- 03 What it does to you The gap between reported and experienced
The EIA reports national averages excluding MEDs at SAIDI of 123.6 minutes, SAIFI of 1.5 outages/customer, and CAIDI of 131.6 minutes. Those are the numbers that show up in utility performance decks. To see what MED inclusion actually does, consider one major regional utility that carries a five-year average SAIDI of 5.1 hours/customer/year when MEDs are included. The gap between MED-included and MED-excluded is the storm exposure that standard reporting obscures. For many utilities that gap is a multiple of the number they lead with.
Reported reliability versus storm-inclusive reliabilityThe number your utility leads with sits below the national average. The number a continuity plan must address sits well above it. MAIFI is the other half of the story. Momentary interruptions never appear on a standard SAIDI or SAIFI report, but they land the same hit on a sensitive load. The consequence: you can be told your utility improved and still see your production line trip more often. The industry itself is trending the wrong way, with a 71% increase in momentary outages during the period documented, and 20% of C&I customers experiencing outages on a monthly basis.
System-wide averages are population-weighted and dominated by residential meter counts. Your service point may sit on a circuit that performs 2 x or more worse than the reported system average. That is the number that belongs in your capital request, not the utility-wide headline.
What the utility number implies What your facility actually faces The system meets a SAIDI target of about two hours per year Your specific feeder may run several multiples of that, and no one has told you Reliability improved because SAIDI and SAIFI went down MAIFI may have gone up as reclosers converted sustained outages into momentary trips Storm risk is an act of God, outside normal reporting MED-inclusive figures are the design basis for your continuity plan The national average is a floor for utility performance Best-in-class utilities operate well below it, so laggard status is measurable SAIFI benchmarksStrong performance sits below one outage per customer per year. Best-in-class sits well under that. A laggard utility is measurable against both. - 04 The trap Averages hide your feeder
System-wide averages are population-weighted and dominated by residential meter counts. Your service point may sit on a circuit that performs 2 x or more worse than the reported system average. That is the number that belongs in your capital request, not the utility-wide headline.
What the utility number implies What your facility actually faces The system meets a SAIDI target of about two hours per year Your specific feeder may run several multiples of that, and no one has told you Reliability improved because SAIDI and SAIFI went down MAIFI may have gone up as reclosers converted sustained outages into momentary trips Storm risk is an act of God, outside normal reporting MED-inclusive figures are the design basis for your continuity plan The national average is a floor for utility performance Best-in-class utilities operate well below it, so laggard status is measurable SAIFI benchmarksStrong performance sits below one outage per customer per year. Best-in-class sits well under that. A laggard utility is measurable against both. Each index maps to a specific design input. SAIFI drives how many times your emergency systems need to activate per year. SAIDI drives your annualized production loss calculation. CAIDI gives you the average restoration time, a starting point, not a design target. Roughly half of sustained outages will exceed CAIDI, so backup autonomy should be sized to a higher percentile of the outage duration distribution, or to a worst-credible-event assumption tied to your MED-inclusive data. MAIFI drives your UPS sizing and ride-through requirements for sensitive equipment.
SAIFIActivation frequency
How many times per year the transfer switch has to work, the generator has to start, the batteries have to discharge.SAIDIAnnualized loss
Total downtime hours multiplied by hourly cost of outage. This is the top line of your avoided-cost math.CAIDIAutonomy starting point
Average restoration time. Size backup to a higher percentile than the average, or to the MED-inclusive worst case.MAIFIRide-through requirement
Drives UPS sizing and sensitive equipment protection for VFDs, PLCs, CNC controllers, and process instruments.What to demandThe three data requests
- 1 MED-inclusive and MED-excluded SAIDI and SAIFI for your utility, side by side.
- 2 Circuit-level SAIDI, SAIFI, and MAIFI for the feeder serving your facility.
- 3 Ranking of your feeder among the utility's worst-performing circuits, and the specific investment plan for it.
The trend backs the case. 72% of C&I customers who purchased alternative energy solutions cited improved reliability as the primary reason, and 25% of companies doubt their utility can support their long-term growth. If you have already been thinking about backup power, a BESS, or a microgrid, these indices are what turn the intuition into a defensible number.
- 05 Your leverage From index to capital decision
Each index maps to a specific design input. SAIFI drives how many times your emergency systems need to activate per year. SAIDI drives your annualized production loss calculation. CAIDI gives you the average restoration time, a starting point, not a design target. Roughly half of sustained outages will exceed CAIDI, so backup autonomy should be sized to a higher percentile of the outage duration distribution, or to a worst-credible-event assumption tied to your MED-inclusive data. MAIFI drives your UPS sizing and ride-through requirements for sensitive equipment.
SAIFIActivation frequency
How many times per year the transfer switch has to work, the generator has to start, the batteries have to discharge.SAIDIAnnualized loss
Total downtime hours multiplied by hourly cost of outage. This is the top line of your avoided-cost math.CAIDIAutonomy starting point
Average restoration time. Size backup to a higher percentile than the average, or to the MED-inclusive worst case.MAIFIRide-through requirement
Drives UPS sizing and sensitive equipment protection for VFDs, PLCs, CNC controllers, and process instruments.What to demandThe three data requests
- 1 MED-inclusive and MED-excluded SAIDI and SAIFI for your utility, side by side.
- 2 Circuit-level SAIDI, SAIFI, and MAIFI for the feeder serving your facility.
- 3 Ranking of your feeder among the utility's worst-performing circuits, and the specific investment plan for it.
The trend backs the case. 72% of C&I customers who purchased alternative energy solutions cited improved reliability as the primary reason, and 25% of companies doubt their utility can support their long-term growth. If you have already been thinking about backup power, a BESS, or a microgrid, these indices are what turn the intuition into a defensible number.
When these indices should drive a capital decision
✓ Act on the numbers- Your hourly cost of an outage is high enough that a handful of events per year moves the P&L
- Your utility reports SAIDI well above the national average, or refuses to provide MED-inclusive figures
- Your feeder ranks near the bottom of the utility's worst-performing circuits
- You run sensitive equipment (VFDs, PLCs, CNC, process instruments) exposed to momentary interruptions
- You are already evaluating backup power, BESS, or a microgrid and need a defensible avoided-cost number
✗ Not the right lever yet- Your loads tolerate short interruptions and you have no production or safety consequence from an outage
- You have not yet quantified your hourly cost of an outage, so any capital math is speculative
- Your utility publishes strong MED-inclusive numbers and your feeder ranks in the top tier
- You have no path to request circuit-level data and no willingness to engage the regulator
- Decision matrix
When these indices should drive a capital decision
✓ Act on the numbers- Your hourly cost of an outage is high enough that a handful of events per year moves the P&L
- Your utility reports SAIDI well above the national average, or refuses to provide MED-inclusive figures
- Your feeder ranks near the bottom of the utility's worst-performing circuits
- You run sensitive equipment (VFDs, PLCs, CNC, process instruments) exposed to momentary interruptions
- You are already evaluating backup power, BESS, or a microgrid and need a defensible avoided-cost number
✗ Not the right lever yet- Your loads tolerate short interruptions and you have no production or safety consequence from an outage
- You have not yet quantified your hourly cost of an outage, so any capital math is speculative
- Your utility publishes strong MED-inclusive numbers and your feeder ranks in the top tier
- You have no path to request circuit-level data and no willingness to engage the regulator
Questions for your morning huddle- What is our utility's SAIDI with Major Event Days included, and how does it compare to the MED-excluded number they usually report?
- Where does our specific feeder rank among the utility's worst-performing circuits?
- If reclosers have been installed on our feeder, do we have MAIFI data, and have we tested our sensitive equipment's ride-through capability against it?
- Using MED-inclusive SAIDI and our facility's hourly cost of an outage, what is the maximum justifiable capital investment in backup or on-site generation?
The one thing to rememberThe number your utility leads with is a filtered, residential-weighted, storm-excluded average. Your capital decision needs the MED-inclusive figure, the circuit-level figure, and MAIFI, and you are entitled to request all three.
This week, send your utility account manager a written request for MED-inclusive and MED-excluded SAIDI and SAIFI, circuit-level performance for your feeder, and MAIFI data. Bring the response to the next capital planning meeting.
Know if the numbers actually pencil out before you sign anything.
A written second opinion on the project in front of you, whether that is a rate change, new equipment, or a renewable installation.
- 01A short call, to figure out quickly whether we can actually be helpful. If we can't, we'll say so on the spot.
- 02We pull the data, your bills, your rate structure, vendor proposals, project specs.
- 03You get the verdict in writing: whether the payback will materialize, and the opportunities or risks nobody has raised.
Get a Blueprint at blueprint.tac-nrg.com Free for Indiana-based operations spending five figures or more a month on electricity. No obligation. You keep the write-up either way. The number your utility leads with is a filtered, residential-weighted, storm-excluded average. Your capital decision needs the MED-inclusive figure, the circuit-level figure, and MAIFI, and you are entitled to request all three.
This week, send your utility account manager a written request for MED-inclusive and MED-excluded SAIDI and SAIFI, circuit-level performance for your feeder, and MAIFI data. Bring the response to the next capital planning meeting.
The Energy Decision BlueprintKnow if the numbers actually pencil out before you sign anything.
A written second opinion on the project in front of you, whether that is a rate change, new equipment, or a renewable installation.
- 01A short call, to figure out quickly whether we can actually be helpful. If we can't, we'll say so on the spot.
- 02We pull the data, your bills, your rate structure, vendor proposals, project specs.
- 03You get the verdict in writing: whether the payback will materialize, and the opportunities or risks nobody has raised.
Get a Blueprint at blueprint.tac-nrg.com Free for Indiana-based operations spending five figures or more a month on electricity. No obligation. You keep the write-up either way.- SAIDI
- System Average Interruption Duration Index. Total minutes of outage experienced by the average customer over a year.
- SAIFI
- System Average Interruption Frequency Index. Number of sustained interruptions experienced by the average customer over a year.
- CAIDI
- Customer Average Interruption Duration Index. Average minutes to restore power once an outage starts. Calculated as SAIDI divided by SAIFI.
- MAIFI
- Momentary Average Interruption Frequency Index. Frequency of short-duration interruptions, typically under five minutes, that do not count toward SAIDI or SAIFI.
- Major Event Day (MED)
- A day of extreme conditions, such as severe storms or natural disasters, that produces outages outside normal operating conditions. Identified using the IEEE 1366 standard.
- IEEE 1366
- The industry standard that sets the criteria utilities use to classify a day as a Major Event Day for reliability reporting.
- Recloser
- An automated device on a distribution line that detects a fault, opens briefly, and attempts to restore power. Improves SAIDI and SAIFI on transient faults but adds momentary interruptions that count in MAIFI.
- Ride-through
- The ability of equipment such as VFDs, PLCs, and UPS systems to tolerate a brief voltage sag or interruption without tripping or resetting.
- Circuit-level data
- Reliability performance reported for the specific distribution feeder serving your facility, rather than a system-wide average. The number that belongs in a capital request.