How to use alarms and trends from BAS or SCADA for predictive work

Alarms and trends from BAS or SCADA systems can support predictive work when teams separate noise from meaningful signals, compare patterns over time, and turn verified issues into work orders. The key is disciplined alarm management and trend review, not simply collecting more points.

Controls Data Snapshot

TL;DR: Start with critical systems, clean up nuisance alarms, trend the points that reveal performance drift, and assign responsibility for review. Predictive value appears when data leads to timely maintenance action.

BAS and SCADA data are not automatically predictive

Building automation systems and SCADA platforms already collect operational information: temperatures, pressures, valve positions, pump status, runtimes, faults, energy use, tank levels, and more. That data can reveal deterioration before failure, but only if it is accurate, trended, reviewed, and connected to maintenance response.

ASHRAE Guideline 13 covers building automation system specification topics such as architecture, installation, training, testing, documentation, performance monitoring, and fault detection. That scope is a reminder that useful BAS data begins with how the system is specified and commissioned, not just how dashboards look after occupancy.

Facility teams learning a myth-busting guide to predictive maintenance hype should apply the same discipline to BAS and SCADA: do not confuse a platform feature with a working maintenance process.

Start by reducing alarm noise

Too many alarms can make teams less responsive. If every minor deviation creates an alert, operators learn to silence or ignore the system. Predictive work requires a smaller number of meaningful alarms with clear response rules.

Alarm cleanup should be documented. If a threshold is changed, the reason should be recorded. If an alarm is disabled, the risk should be reviewed. This prevents alarm tuning from becoming quiet risk acceptance.

Trends that can reveal early problems

Useful trends often compare related points rather than one point alone. For HVAC systems, examples include supply-air temperature against valve command, fan speed against static pressure, chilled water valve position against zone temperature, or runtime against outdoor temperature. For utility or process systems, examples may include pump starts, tank level changes, motor current, pressure drift, or repeated trips.

The U.S. Department of Energy describes monitoring-based commissioning as the continuous application of the commissioning process to building or energy systems. That idea fits predictive work well. Trends are not only for troubleshooting after complaints. They can be reviewed regularly to find performance drift.

BACnet documentation from ASHRAE describes features such as event logs and trend logs, which are part of why many modern building systems can support operational analysis when configured properly.

Turn signals into work orders

A predictive signal should have a path to action. That path might be manual at first: a building engineer reviews weekly exceptions, validates the issue, and creates a work order. More mature teams may integrate fault detection with CMMS workflows.

A simple workflow is:

• System generates an alarm or trend exception.
• Operator verifies whether the signal is valid.
• Issue is assigned a priority and asset.
• Work order is created with trend screenshot or point history.
• Technician investigates and records findings.
• Controls team confirms whether the trend returned to normal.

Without the last step, teams may fix symptoms without confirming the outcome. Access and cybersecurity also belong in the workflow. The people who can change setpoints, disable alarms, export trend data, or connect integrations should be clearly authorized. Predictive maintenance should not create a new path for accidental control changes or unauthorized system access.

This process connects with construction takeoffs: common mistakes that distort estimates when repairs become capital scope. Accurate asset history and trend evidence can help estimators define quantities and avoid rough allowances.

Common mistakes with BAS and SCADA predictive efforts

One mistake is treating every point as equally valuable. Trend storage, review time, and analytics effort should focus on systems where failure or drift matters.

Another mistake is failing to maintain sensors. A predictive program built on inaccurate temperature, pressure, flow, or current readings can send teams in the wrong direction.

A third mistake is ignoring controls sequences. A trend may look abnormal because the sequence is wrong, the schedule is outdated, or a setpoint was overridden months ago.

A fourth mistake is not involving technicians. Operators may see the alarm, but technicians see the equipment. Predictive work improves when controls data and field observations meet.

A practical starter list for facilities

Start with a small set of high-value conditions:

• Simultaneous heating and cooling.
• Pumps or fans running outside schedule.
• Valves commanded open with no temperature response.
• Rapid cycling or excessive starts.
• Repeated manual overrides.
• Abnormal energy use after occupancy changes.
• Recurring fault alarms on critical assets.

Review these conditions weekly at first. If the same issue repeats, address root cause rather than simply resetting alarms. A short monthly controls review can also help. Invite operations, maintenance, and the controls vendor when needed. Look at the top repeating alarms, unresolved overrides, comfort complaints, and energy anomalies. The meeting should produce decisions, not just screenshots. Assign one owner, one due date, and one confirmation method for every selected issue so the trend review does not become another passive report or dashboard ritual.

Turning Trends Into Work Without Creating Noise

BAS and SCADA systems can support predictive maintenance, but only when teams clean data, tune alarms, choose meaningful trends, and assign action. The best early wins often come from fixing obvious drift and recurring faults, not from complex analytics.

For assets that require field verification, pair controls data with what causes concrete cracking and how to minimize it style documentation discipline: location, pattern, timing, evidence, and qualified interpretation.

Informational note: This article is educational only and does not replace controls engineering, cybersecurity, electrical, mechanical, legal, compliance, or project management advice. BAS and SCADA changes should be reviewed by qualified professionals and authorized system owners.