Survey findings link higher redundancy to lower outage risk (Owen Rogers, Douglas Donnellan)

Overview

This article, written by Owen Rogers and Douglas Donnellan in the Uptime Institute Journal, is an analysis that combines more than 1,200 responses from the 2025 and 2026 annual Uptime surveys. Its aim is to correlate power and cooling redundancy configurations (N, N+1, N+2, 2N) with outages reported over three years.

The core finding is that outage incidence and serious (category 3-5) outages fall consistently as redundancy increases. The share reporting at least one impactful outage in three years is 70% with N and 39% with 2N for cooling, and 68% with N and 42% with 2N for power. Category 3-5 outages, regarded as financially meaningful, occur at about 8-11% of 2N respondents against about 19-22% of N respondents.

In quantifying the gains, versus N the likelihood of a serious outage was about 36% lower for N+1 cooling, 53% for N+2 and 63% for 2N; for power the figures are 23%, 37% and 44%. Using N+1 as the baseline, 2N cooling is 42% and N+2 cooling 26% less likely to have an outage. In addition, power and cooling together cause more than two-thirds of impactful outages in the 2026 survey.

The authors stress that this is correlation and should be used as an indicative benchmark and not a prediction for an individual site. Survey data rests on reporting; besides redundancy, factors such as facility age, operating maturity and maintenance quality may also influence outage frequency. The ratios should therefore be read as relative risk multipliers when comparing design options and not as a substitute for a reliability calculation.

Key points

1. Share reporting at least one impactful outage in 3 years: cooling 70% with N vs 39% with 2N; power 68% with N vs 42% with 2N.
2. Category 3-5 outages: about 8-11% of 2N respondents vs about 19-22% of N respondents.
3. Versus N, serious outage likelihood was about 36% lower for N+1 cooling, 53% for N+2 and 63% for 2N.
4. Power, same comparison: 23%, 37% and 44%.
5. With N+1 as baseline, 2N cooling is 42% and N+2 cooling 26% less likely to have an outage.
6. Power and cooling together cause more than two-thirds of impactful outages in the 2026 survey.
7. Based on over 1,200 responses from the 2025 and 2026 surveys; findings are correlation, not prediction.

Using it in design

• Use these ratios as indicative risk multipliers in an option-comparison view, for example showing relative outage likelihood when switching cooling from N+1 to 2N.
• Label them clearly as survey correlation, not a reliability calculation.
• Assess power and cooling redundancy together, noting in the rationale that most outages originate in these two systems.

Who should read it

Designers, operators and investment decision makers deciding on redundancy level.

Publication details

Authors

Owen Rogers; Douglas Donnellan

Publisher

Uptime Institute Journal

Year

2026

Document language

English

Source type

Industry body

Access

Free to read

Licence

Copyright retained; citation only — the publisher’s terms apply to the original document

Citation