Latest / NOXTalks / A 90-Day BAS Alarm Cleanup Plan for Commercial Facilities
Transcript
- 0:07SPEAKER_00: Welcome to Knox Talks, the AI-generated podcast exploring the groundbreaking ways technology shapes our built environment.
- 0:14SPEAKER_00: Today we're diving into a problem almost every facility manager or building engineer has felt at some point.
- 0:21SPEAKER_00: You open your BAS, and instead of clarity, it feels more like a slot machine.
- 0:26SPEAKER_00: Alarms scrolling, pop-ups firing, notifications everywhere.
- 0:30SPEAKER_00: And somehow the one alarm that actually matters is buried under a pile of noise.
- 0:36SPEAKER_00: This episode is based directly on a practical article that lays out a clear 90-day plan to clean this up.
- 0:43SPEAKER_00: Without buying new software or ripping out your system, it's about discipline, not dollars.
- 0:48SPEAKER_00: Here's a fun but slightly scary fact to start.
- 0:52SPEAKER_00: Industry studies referenced by groups like Ash Ray consistently show that operators start ignoring alarms when more than about 30 to 40 alarms a day hit the console.
- 1:02SPEAKER_00: Many commercial buildings generate hundreds.
- 1:05SPEAKER_00: At that point, alarms stop being decision support and become background noise.
- 1:10SPEAKER_00: That's when real failures slip through.
- 1:12SPEAKER_00: First, an important line in the sand.
- 1:15SPEAKER_00: BAS alarms are not life safety alarms.
- 1:18SPEAKER_00: Fire and life safety systems live in a governed, code-driven world with authorities having jurisdiction, testing regimes, and strict boundaries.
- 1:26SPEAKER_00: BAS alarms are operational tools.
- 1:29SPEAKER_00: They exist to help people make faster, smarter maintenance, and operations decisions.
- 1:34SPEAKER_00: Mixing those responsibilities is how confusion and risk creep in.
- 1:37SPEAKER_00: Now here's the core issue.
- 1:39SPEAKER_00: Most things we call alarms in a BAS aren't really alarms at all.
- 1:43SPEAKER_00: They're status updates, notifications, or logs that ended up in the wrong place.
- 1:48SPEAKER_00: A real alarm is something abnormal that needs timely action.
- 1:52SPEAKER_00: A pump failure, a main AHU supply fan down, a sustained freeze risk.
- 1:59SPEAKER_00: If it fires over and over and nobody ever acts, that's your signal that it doesn't belong in the alarm console or nobody knows what to do with it.
- 2:08SPEAKER_00: So why does alarm chaos happen so consistently across buildings?
- 2:13SPEAKER_00: One big reason is alarm flooding.
- 2:15SPEAKER_00: One root cause, like a failed AHU, can trigger dozens of downstream zone alarms.
- 2:21SPEAKER_00: By the time someone finds the real issue, they've wasted half an hour chasing symptoms.
- 2:26SPEAKER_00: Another is nuisance thresholds, no deadbands, no delays, set points that don't reflect reality.
- 2:33SPEAKER_00: A sensor bouncing around a limit can generate an alarm every 30 seconds and slowly train operators to ignore everything.
- 2:40SPEAKER_00: Then there's flat priority.
- 2:42SPEAKER_00: When everything looks critical, nothing is.
- 2:45SPEAKER_00: Add inconsistent naming across systems and buildings, and operators lose pattern recognition completely, and finally, no playbooks.
- 2:54SPEAKER_00: An alarm fires, someone acknowledges it, and that's the end of the story.
- 2:58SPEAKER_00: The fix isn't complicated, but it does require structure.
- 3:02SPEAKER_00: The first two weeks are about baseline and triage.
- 3:06SPEAKER_00: Export 30 to 90 days of alarm history.
- 3:09SPEAKER_00: Look at your top 20 alarms by frequency.
- 3:12SPEAKER_00: Find alarms that never clear.
- 3:14SPEAKER_00: Look hard at after hours alarms because those are the ones burning trust and sleep.
- 3:20SPEAKER_00: Most teams get their first big aha moment right here.
- 3:23SPEAKER_00: The next step is deceptively powerful.
- 3:26SPEAKER_00: Write a one-page alarm philosophy, just one page.
- 3:30SPEAKER_00: What counts as an alarm?
- 3:32SPEAKER_00: How many priority levels you have, how alarms are named, who gets notified, and when.
- 3:37SPEAKER_00: What happens if no one responds?
- 3:39SPEAKER_00: How planned work is handled.
- 3:41SPEAKER_00: This single document turns alarms from anxiety into a shared language.
- 3:45SPEAKER_00: Then comes the payoff phase.
- 3:47SPEAKER_00: Days 31 to 60 are about fixing the top 20 offenders, add deadbands and delays, fix bad sensors and flaky comms, correct schedules, and critically implement root cause grouping.
- 4:00SPEAKER_00: If an AHU is down, surface that as the primary alarm and suppress the downstream noise, while still logging it for diagnostics.
- 4:09SPEAKER_00: This alone can cut alarm volume dramatically during failures.
- 4:12SPEAKER_00: The final month is about locking it in.
- 4:15SPEAKER_00: Only critical alarms leave the BAS after hours.
- 4:18SPEAKER_00: Escalation rules are clear.
- 4:20SPEAKER_00: Maintenance modes suppress expected noise.
- 4:23SPEAKER_00: Repeat alarms become work orders, not endless notifications.
- 4:27SPEAKER_00: And once a week, for just 30 minutes, the FM, chief engineer, and SI review what's bubbling to the top.
- 4:34SPEAKER_00: Here's what not to do.
- 4:36SPEAKER_00: Don't just disable alarms to make the screen look clean.
- 4:40SPEAKER_00: Don't use alarms as maintenance reminders.
- 4:42SPEAKER_00: Don't notify everyone.
- 4:44SPEAKER_00: And don't try to alarm your way out of missing trends.
- 4:48SPEAKER_00: Prevention lives in trending and analysis, not paging.
- 4:52SPEAKER_00: Alarm fatigue is fixable.
- 4:54SPEAKER_00: It doesn't require a capital project.
- 4:56SPEAKER_00: It requires a mindset shift.
- 4:59SPEAKER_00: Treat alarms as decision support, not background noise.
- 5:02SPEAKER_00: Start small, build momentum, and within 90 days your operators will trust the system again.
- 5:20SPEAKER_01: Innovations are shadows of night.