Detective maintenance is the periodic testing of devices whose failure does not show in normal operation: protection systems, alarms, interlocks and standby equipment. It hunts the hidden failure, the one that already happened and nobody noticed, because the device is only called upon when something goes wrong. It is the fourth type in the classic classification, next to corrective, preventive and predictive, and the least known of the four.
The problem it solves: the hidden failure
A stuck safety valve does not stop production, makes no noise, generates no heat: the plant runs normally for months with its protection dead. The risk only materializes on the day overpressure arrives and the valve does not open, turning a controllable event into an accident. That is a hidden failure: a protective function lost without symptoms. The concept came from aviation reliability analysis and is central to RCM (reliability centered maintenance): for hidden failure modes, the only strategy that works is actively looking for the failure.
Where detective maintenance applies
The candidates are the devices that spend their lives waiting for their moment: safety and relief valves, protective pressure and level switches, safety interlocks, emergency stop systems, alarms, emergency lighting, standby pumps and generators. The typical test is functional: provoke the condition (or safely simulate it) and check whether the protection responds. Starting the standby generator on load every month is detective maintenance; checking its oil level is preventive.
Detective is not predictive
The confusion is common and the difference is clean: predictive tracks the degradation of an active function (vibration rising on a running bearing); detective tests a hidden function that is not being called upon (the valve that must open if pressure ever rises). In predictive, the equipment gives signals; in detective, maintenance asks the question. The four types and their triggers are in the types of maintenance guide.
How to plan the tests
Three definitions per device: the interval (the longer the gap between tests, the longer a hidden failure can exist unseen; higher-consequence devices test more often), the procedure (how to provoke or simulate the condition safely, who authorizes, what to record) and the pass criterion (did it open at the right pressure? start within how many seconds? pick up the load?). Each test becomes an order in the plan, with the result recorded: the history of failed tests is the data that calibrates the interval. A device that never fails its test can space out; recurring failures demand shorter intervals or a review of the device itself.
The record that protects twice
Testing protection without an auditable record is a liability in safety audits. An order with the procedure checklist, result, executor and date plays two roles: it feeds the protection's reliability statistics and documents the plant's diligence. In the digital flow, the checklist runs on the phone during the test and the result enters the system on the spot, without depending on a clipboard surviving the shift.
Frequently Asked Questions
What is detective maintenance?
It is the periodic testing of protective and standby devices to find hidden failures: failures that have already happened but do not show, because the device is only called upon in emergencies.
What is the difference between detective and predictive maintenance?
Predictive tracks the degradation of an active function in operation; detective tests a hidden function that is not being exercised. In predictive the equipment gives signals; in detective, maintenance provokes the answer.
Which equipment needs detective maintenance?
Safety and relief valves, protective switches, interlocks, emergency stop systems, alarms and standby equipment (pumps, generators). Everything that spends its life waiting for its moment to act.
How often should a protective device be tested?
It depends on the consequence of failure and the test history: the higher the consequence, the shorter the interval. Recurring test failures demand more frequent testing or a device review; a clean history may justify spacing out.
