A hundred alarms on screen: turning BMS notifications into an action plan

Consequences determine priority, not colour

A faulty room sensor, a dirty filter and loss of server room cooling should not have the same response priority. For each alarm, the consequences, permissible response time and available backup capacity need to be assessed.

A BMS can sort alarms and route notifications by priority. Metasys, for example, supports separate configuration of importance and routing; the manufacturer recommends a consistent priority scheme. ¹

We would first agree this scheme with the facilities team. The numbers in the settings should reflect actual responsibilities, rather than being chosen at random during programming.

The notification should explain what to do

‘AI-17: Alarm’ does not give the operator enough context. It is more useful to specify the equipment, parameter, measured value, agreed limit and active operating mode.

For a hypothetical pump, the notification could explain: a start command is present, no run confirmation has been received, and the standby pump has been enabled to start. This is an illustrative example of a notification structure for a design, not a description of a completed project.

There is no need to fit a complete procedure into an SMS. But the operator should be able to navigate from the notification to the relevant system component, trend and agreed response procedure, rather than searching through every screen.

A delay must not conceal a hazard

After start-up, parameters may take some time to reach normal operating conditions. If every expected deviation immediately triggers an alarm, unnecessary noise results.

Informational and operational signals are therefore evaluated with equipment status, hysteresis and justified delays taken into account. However, delays must not be applied arbitrarily to frost protection, overheating protection or other critical functions. These must remain consistent with the design and equipment requirements.

It is important to distinguish between local protective action and notification to the operator. A problem delivering an email must not disable the protection the unit needs.

Loss of communication does not mean twenty zero temperatures

If a gateway stops transmitting twenty parameters, substituting zeros for them makes analysis unreliable. It is better to generate a source-unavailable event, mark dependent data as invalid and avoid generating numerous false temperature alarms from those data.

This is proposed logic for a project. It must not conceal a separate, independent critical alarm that continues to arrive from another source.

Acknowledgement is not a repair

Acknowledgement means the operator has noted the notification. It does not, in itself, confirm that the physical parameter has returned to normal. In a BMS, acknowledgement and event display management are defined separately. ²

It is therefore useful to see, at a minimum, when the event occurred, whether it was acknowledged and when normal conditions were restored. A comment on the action taken should also be added if required for the service history.

Who receives notifications outside working hours?

Before handover, we check not only that an alarm appears on screen, but that it reaches the correct recipient. Filters by equipment, priority and time are supported in Metasys Remote Notifications, among other systems. ³

We separately agree a backup channel and escalation procedure if the responsible person does not respond. This is an organisational task that even the most expensive platform cannot replace.

A good alarm system does not force people to watch the BMS constantly. It makes deviations visible, responsibilities clear and the outcome of the response verifiable.