Answered

Alarm priorities: how do you decide what counts as High?

LADX · 2d ago · 1 reply

Accepted answer, by LADX

The question that kills most proposed High alarms is the second one. "What happens if they do nothing for ten minutes" is very often "nothing", and that is a Low.

LADX asked · 24 Aug 2026

By consequence and by the time available to act, not by how important the equipment feels. ISA-18.2 asks three questions of every alarm, and they settle nearly every argument: 1. What happens if the operator does nothing? 2. How long do they have before that happens? 3. What action do they take? If there is no answer to question 3, it is not an alarm. It is a status, and putting it on the alarm list is how alarm floods start. A workable split: - High: injury, environmental release, or major loss. Action needed in under about three minutes. - Medium: product loss or equipment damage. Action needed this shift. - Low: efficiency or minor quality. Action needed, not urgent. The distribution matters as much as the individual decisions. Aim for roughly 5 percent high, 15 percent medium, 80 percent low. If a third of your alarms are High, none of them are, because the operator cannot tell which to look at first. Then measure it in service: average alarms per hour per operator should be around six. If it is sixty, the priorities are not the problem, the alarm list is.

1 reply

  • LADX · 2d ago

    Answer
    The question that kills most proposed High alarms is the second one. "What happens if they do nothing for ten minutes" is very often "nothing", and that is a Low.

Reply

Be specific. Platform and version help.