Why alarm philosophy matters
Poor alarms create noise, bad decisions, and compliance risk. Good alarms guide the technician to the right action quickly.
Three alarm categories (practical)
Process / Measurement Alarms
High/low readings, instability, drift trend warnings.
- Action: Verify signal + compare with calibration
- Likely causes: process upset, leaks, restriction
System / Utility Alarms
Low sample flow, heater fault, chiller high temp, pump fault.
- Action: Fix utilities and sample system first
- Likely causes: plugging, heater failure, drains
Data / Communication Alarms
DAHS offline, comm error, invalid flags, time sync issues.
- Action: Verify network + mapping + timestamps
- Likely causes: resets, IP changes, scaling
Good alarm rules (field-friendly)
- Alarms must be actionable — each alarm should suggest first checks
- Avoid duplicate alarms — one root alarm should drive others
- Use delays — ignore short spikes during switching or warm-up
- Separate “warning” and “shutdown” — not everything is critical
- Link alarms to decision trees — reduce guesswork
Example: “Sample Flow Low” should come before “Unstable Reading”.
Fix flow first, then evaluate reading stability.
Alarm response method
- Confirm alarm is real (not maintenance / calibration state)
- Check utilities (power, air, temps, flow)
- Inspect sample system (probe, heaters, filters, condensation)
- Run a quick zero/span verification
- Verify DAHS flags and reporting status
Alarm floods usually mean missing delays, missing state logic (calibration/warm-up),
or poor prioritization — fix philosophy, not only symptoms.