All calculators

LoadCairn / Reliability planning

Availability and downtime calculator.

See what 'four nines' really means in minutes, and how adding a redundant unit changes the numbers.

Section 1

Availability to downtime

Convert a service availability percentage into annual, monthly, and weekly downtime.

Quick availability choices

Planning estimate

Downtime by period

The estimate uses a 365-day year, 8,760 hours, 525,600 minutes, and 10,080 minutes per week.

Year
52.56 min
Month
4.38 min
Week
1.01 min
Calculation details

Formula

  1. Unavailability = 1 − availability ÷ 100.
  2. Annual downtime = unavailability × 8,760 hours = unavailability × 525,600 minutes.
  3. Monthly downtime = annual minutes ÷ 12.
  4. Weekly downtime = unavailability × 10,080 minutes.

Current calculation

  1. 1 − 99.990 ÷ 100 = 0.000100 unavailability
  2. 0.000100 × 525,600 = 52.56 minutes/year (52.56 min)
  3. 52.56 ÷ 12 = 4.38 minutes/month
  4. 0.000100 × 10,080 = 1.01 minutes/week

Section 2

Redundancy comparison

Compare all-needed N, one-spare N+1, and two-system 2N availability.

For example a UPS module, CRAH or generator.

Enter a whole number from 1 through 20.

Planning estimate

N, N+1, and 2N

Each result shows system availability and annual downtime.

N
99.8001%
17 h 31 min/year
N+1
99.9997%
1.58 min/year
2N
99.9996%
2.10 min/year

In this simple model, N+1 can score as well as or better than 2N, because it counts units only. Real 2N designs earn their keep through separate distribution paths, fault isolation, and maintaining one side while the other carries the load, which this model doesn't include.

Calculation details

Formula

  1. N = A^k, where A is one-unit availability and k = 2.
  2. N+1 = Σ from i = 2 to 3 of C(3, i) × A^i × (1 − A)^(3−i).
  3. 2N = 1 − (1 − A^k)^2 for two independent systems.

Current calculation

  1. A = 0.999000; k = 2
  2. N = 0.999000^2 = 99.8001%; 17 h 31 min/year
  3. N+1 = binomial sum = 99.9997%; 1.58 min/year
  4. 2N = 1 − (1 − 0.998001)² = 99.9996%; 2.10 min/year

Assumptions

  • Unit failures are independent; identical units; no common-mode failures.
  • Switching, transfer and distribution are treated as perfect.
  • Availability is a long-run average and doesn't predict when an outage happens.

Sources

  • IEEE 3006.7, Recommended Practice for Determining the Reliability of 7x24 Continuous Power Systems in Industrial and Commercial Facilities
  • Standard k-out-of-n reliability formula (binomial)

Planning estimate only. Real availability depends on design, maintenance, human error and common-cause failures that this model doesn't capture.

Calculated in your browser · No values leave this page