Preventive vs predictive vs corrective maintenance
Corrective maintenance repairs after failure, preventive maintenance follows a fixed schedule, and predictive maintenance acts on measured condition. Preventive work is done whatever the equipment's state; predictive work waits for a warning sign such as rising vibration. Reliability-centered maintenance chooses among the three for each failure mode, including deliberate run-to-failure.
By Jun Zhong, Editor·Researched and drafted by Beacon's AI newsroom·Checked by Jun Zhong, Editor·Last checked ·Next review
Why does it matter?
A fixed schedule can hide how fast equipment is wearing. At a refinery in Superior, Wisconsin, a catalyst slide valve was replaced at each turnaround, and turnarounds came every five years from 2003 instead of every two to three years as before 1. The valve removed in 2013 had lost about 85% of the metal and hard facing in part of its seat, and on 26 April 2018 the eroded valve in service let air pass while closed during a shutdown, contributing to an explosion and fire 1. Employees had not seen the erosion affect normal operation, and the US Chemical Safety and Hazard Investigation Board (CSB) recommended monitoring the valves between turnarounds, with quarterly reviews that could lead to a shorter turnaround cycle 1.
Preventive vs predictive vs corrective at a glance
| Corrective | Preventive | Predictive | |
|---|---|---|---|
| What triggers the work | A failure, or equipment drifting out of its required performance 4 | Calendar time, operating hours or cycles, whatever the condition 2,4 | A measured sign of degradation crossing an alert level 2,3 |
| Other names | Reactive, breakdown, run-to-failure, fix-when-fail 2 | Time-based or interval-based maintenance 2,3 | Condition-based maintenance; predictive testing and inspection 2,3 |
| What it assumes | Failure consequences are acceptable, where chosen deliberately 2 | Failure becomes more likely with age, and a useful life can be set 5 | The failure gives a detectable warning early enough to act 5 |
| Typical fit | Small, non-critical or redundant items; equipment unlikely to fail 3 | Items that wear, consumables, known failure patterns, manufacturers' recommendations 3 | Critical equipment, random failure patterns, equipment not subject to wear 3 |
| Main drawbacks | Unplanned downtime, overtime, damage to other equipment, large spare parts stock 3 | Unneeded work, damage caused during it, and failures still possible 3 | Cost of instruments and training; not every failure mode gives warning 2,3 |
| Savings quoted by the DOE guide (2010) | Baseline | 12% to 18% over reactive 3 | 8% to 12% over preventive 3 |
| Average share of maintenance, US discrete manufacturing, 2016 | 45.7% 4 | 31.8% 4 | 17.3% 4 |
Sources: NASA RCM guide, 2008 2; US Department of Energy (DOE) O&M Best Practices Guide, Release 3.0, 2010 3; NIST AMS 100-34, 2020, survey excluding petroleum and chemical plants 4; DoD Manual 4151.22-M 5. The DOE savings figures are estimates from earlier studies cited in the guide.
When is corrective maintenance used?
- As a deliberate choice. NASA's reliability-centered maintenance guide accepts run-to-failure once an analysis has weighed what a failure would risk and cost against the price of preventing it; its examples include non-critical electric motors under 7.5 hp and restroom exhaust fans 2.
- When failures happen anyway. No program prevents every failure. NASA notes that a purely reactive program brings a high share of unplanned work, large stocks of spare parts and inefficient use of staff 2.
- On regulated process equipment, within limits. US process safety rules require deficiencies outside acceptable limits to be corrected before further use, or in a safe and timely way with measures to keep operation safe 6.
When is preventive maintenance used?
- For wear and known patterns. The DOE guide lists equipment subject to wear, consumables, known failure patterns and manufacturers' recommendations 3.
- Where rules set frequencies. US process safety rules require inspection and test frequencies consistent with manufacturers' recommendations and good engineering practice, and more often if operating experience shows the need 6.
- For degradation that builds with time. NASA keeps interval-based tasks for abrasive, erosive or corrosive wear and for fatigue or embrittlement 2.
Its weakness is the interval itself. NASA notes that a mean time between failures gives only an average, so a fixed time is hard to choose for an item that fails at random, and an overhaul can add early-life failures 2.
When is predictive maintenance used?
- When the failure gives warning. The US Department of Defense (DoD) manual requires that the warning sign be detectable, that the time from warning to failure be reasonably consistent, that checks fit inside that time, and that it leave room to act 5.
- For critical equipment and random failures. The DOE guide places critical equipment and equipment with random failure patterns here 3.
- With the right instruments. Typical techniques are vibration analysis on rotating machinery, oil analysis, thermography and thickness measurement for erosion and corrosion 2. DOE notes that much of the equipment costs more than $50,000 3.
A common mix-up: is a scheduled inspection preventive or predictive?
Condition checks are often done on a calendar, which makes them look preventive. The DoD manual's on-condition tasks, such as visual or nondestructive inspection for cracks and corrosion, are carried out at set intervals, but repair follows only if a potential failure is found 5. What separates the strategies is what triggers the repair, not whether a calendar is involved. NASA also names a fourth element, proactive maintenance, which uses root cause analysis and feedback to designers to remove the causes of repeat failures 2.
Where integrity problems show up
| Element or step | Integrity link | In the Library |
|---|---|---|
| Setting inspection and test frequencies for covered process equipment | US rules tie them to manufacturers' recommendations and good engineering practice, and to operating experience 6 | How are inspection intervals set for vessels, piping and tanks? |
| Lengthening a fixed overhaul interval | At Superior, turnarounds moved from every two to three years to every five, while slide valve erosion exceeded the 1/8 inch allowance in the refinery's own standard 1 | Corrosion vs erosion vs erosion-corrosion: what is the difference? |
| Choosing run-to-failure | Defensible only after weighing failure risk and cost 2; protective devices that act on demand need failure-finding tests instead 5 | What is reliability-centered maintenance (RCM)? |
| Letting corrective work on safety-critical equipment build up | UK HSE found offshore installations that began a shutdown with several hundred hours of overdue safety-critical maintenance and finished it with several thousand hours outstanding; HSE's target for such backlog is zero 7 | What are safety and environmental critical elements (SECEs)? |
| Treating thickness surveys as condition checks | Nondestructive inspection for corrosion is an on-condition task, and NASA lists ultrasonic thickness readings for erosion and corrosion 2,5 | What are condition monitoring locations (CMLs)? |
Sources: CSB Superior report, 2022 1; NASA, 2008 2; DoD Manual 4151.22-M 5; 29 CFR 1910.119(j) 6; HSE Key Programme 3, 2007 7.
Common questions
- What is run-to-failure maintenance?
- It is corrective maintenance chosen in advance. NASA treats it as a conscious decision suited to items whose failure has negligible consequences, made once all functions and consequences are understood.
- Which strategy costs least?
- It depends on the equipment. DOE quotes study estimates of 12% to 18% savings for preventive over reactive work and 8% to 12% for predictive over preventive. In NIST's 2016 survey, among plants with less than half reactive work, the half using more predictive maintenance had 15% less downtime and an 87% lower defect rate, an association NIST notes may not be wholly caused by the strategy.
- Where does reliability-centered maintenance fit?
- It is the method that chooses among the three. NASA describes reliability-centered maintenance as combining preventive, predictive, reactive and proactive maintenance to reach the required reliability at the lowest life-cycle cost.
Beacon explains these concepts in its own words for education. It does not reproduce or replace the standards cited, and Winda is not affiliated with or endorsed by the organizations that publish them. Engineering decisions need the current edition of the standard and a qualified engineer.
Sources
- 1.Investigation Report: FCC Unit Explosion and Asphalt Fire at Husky Superior Refinery (No. 2018-02-I-WI) · US Chemical Safety and Hazard Investigation Board (CSB)
- 2.Reliability-Centered Maintenance Guide for Facilities and Collateral Equipment · US National Aeronautics and Space Administration (NASA)
- 3.Operations & Maintenance Best Practices: A Guide to Achieving Operational Efficiency, Release 3.0 (PNNL-19634) · Pacific Northwest National Laboratory for the US Department of Energy, Federal Energy Management Program
- 4.Economics of Manufacturing Machinery Maintenance: A Survey and Analysis of U.S. Costs and Benefits (NIST AMS 100-34, D. S. Thomas and B. A. Weiss) · US National Institute of Standards and Technology (NIST)
- 5.DoD Manual 4151.22-M, Reliability Centered Maintenance (RCM), incorporating Change 1 · US Department of Defense (via Whole Building Design Guide)
- 6.29 CFR 1910.119, Process safety management of highly hazardous chemicals, paragraph (j) Mechanical integrity · eCFR, US Office of the Federal Register
- 7.Key Programme 3: Asset Integrity Programme · UK Health and Safety Executive (HSE), Offshore Division