Analytics and Diagnostics – Proactive vs. Reactive Troubleshooting

Problem

You only find out about problems when someone complains ("Why is the office cold?"). By then, the system has been malfunctioning for hours or days, and the issue is hard to trace.

Solution

1. Implement Basic Fault Rules

Don't rely on manual oversight—automate alerts:

• High outdoor temp + supply fan off = alert
• Zone temp >78°F for >30 minutes = alert
• Communication loss to device = alert
• These run continuously in the background

2. Use Historical Trending

• Log key points (supply air temp, return air temp, damper position, fan speed) to database
• Query trends to find patterns:
  • "Why does AHU 1 always overshoot temp by 2°F?"
  • "When did this valve start leaking (gradual temp rise)?"
  • "Is fan ramp-up time unusually slow?"

3. Set Performance Baselines

• After commissioning, record "normal" operation (baseline)
• Alert if performance deviates >10% from baseline
• Example: Normal supply air temp: 52°F ± 1°F. If it's now 54°F, alert technician.

4. Proactive Maintenance Rules

• Temperature sensor drift detection: If a sensor reads 5°F higher than adjacent sensor for >1 hour, likely failing
• Valve hysteresis: If valve moves too slowly or erratically, actuator may be dying
• Compressor cycling: If cycling too frequently (>5 times/hour), maintenance needed soon

5. Build Fault Hierarchy

Not all faults are equally urgent:

• Critical: HVAC offline, security breach → immediate notification
• High: Zone temp >80°F for >1 hour → notify within 15 minutes
• Medium: Sensor drift detected → log for next service visit
• Low: Trending data gap for 1 hour → log but don't alert

6. Implement Scheduled Self-Tests

• Weekly: Verify all devices still respond to pings
• Monthly: Spot-check trending data for anomalies
• Quarterly: Full system commissioning verification