
When Should BESS Alarms Activate in Australia?
A BESS incident rarely begins with visible smoke. Long before flame, a failing lithium-ion cell can release hydrogen and electrolyte vapours, generate volatile organic compounds, and create abnormal heat within a rack, module or enclosure. For operators asking, “when should BESS alarms activate?”, the answer is: well before conditions approach smoke, fire or a flammable atmosphere.
The most effective alarm strategy is not a single threshold connected to a single action. It is a staged response that identifies early battery degradation, confirms escalation, directs personnel action and, where required, initiates protective controls. The aim is to detect danger before disaster while avoiding nuisance alarms that erode confidence in the safety system.
When should BESS alarms activate?
BESS alarms should activate at the earliest credible indication that battery conditions are departing from normal operation. In lithium-ion systems, this may be an off-gas signature, a sustained rise in hydrogen or VOC concentration, an unexpected humidity change, or a rate-of-rise in temperature that does not match the operating profile.
This point is deliberately earlier than a conventional fire alarm event. Smoke detection and heat detection remain valuable parts of a fire protection strategy, but they may only respond after a failure has advanced significantly. By then, operators may have little time to investigate, isolate the affected system or protect adjacent assets.
Early warning sensors should therefore be treated as a condition-monitoring and risk-mitigation layer. Their first alarm does not necessarily mean thermal runaway is underway. It means the system has evidence of abnormal battery behaviour that requires a defined response.
The exact activation point depends on battery chemistry, enclosure volume, ventilation rate, rack configuration, sensor location, normal background conditions and the consequence of failure. A utility-scale container, a commercial indoor BESS and a UPS battery room should not share identical alarm thresholds simply because they use lithium-ion batteries.
Use staged alarms, not one critical setpoint
A single alarm at a high concentration can be simple to configure, but it wastes the most valuable part of the incident timeline. A better design uses several alarm levels, with each level tied to an action that is practical, proportionate and documented.
Level 1: Early warning and investigation
The first level should activate when gas readings, temperature trends or humidity data indicate a meaningful departure from the established baseline. This alarm can be sent to the site BMS, SCADA platform or remote monitoring team through relay outputs or Modbus RTU.
At this stage, the purpose is to prompt verification. Operators may review battery management system data, inverter status, charging and discharging activity, HVAC operation, recent maintenance and sensor readings from nearby zones. If the system is attended, trained personnel can inspect from a safe position under the site’s emergency procedures.
An early warning must be noticeable and actionable, but it should not automatically create a full emergency response every time. That balance is what makes baseline commissioning and trend analysis essential.
Level 2: Confirmed abnormal condition
The second level should activate when the abnormal condition persists, increases, appears across more than one sensing parameter, or is corroborated by BMS fault data. For example, a continuing rise in hydrogen and electrolyte vapours combined with local temperature elevation is a more serious signal than a short, isolated sensor fluctuation.
This level commonly triggers escalation to responsible personnel, enhanced logging, an operational review and a controlled reduction in battery activity. Depending on the approved cause-and-effect matrix, it may command the BESS to stop charging, stop discharging, or place the system in a safe state.
The correct response is site-specific. In some installations, stopping energy transfer is appropriate. In others, the battery manufacturer’s safety logic, fire engineering design and grid obligations must determine the sequence. Alarm logic should never override OEM requirements without engineering review.
Level 3: Critical escalation and emergency action
A high-high alarm should activate when readings indicate rapidly escalating cell failure risk, such as strong and rising off-gas concentrations, concurrent heat increase, multiple affected zones, or a confirmed battery fault that is consistent with internal cell damage.
This is the point for emergency actions defined by the facility’s fire safety and incident response plan. These may include battery isolation, emergency notification, control room escalation, evacuation of affected areas and engagement with emergency services where required. The goal is to protect people first, then prevent propagation and reduce damage to surrounding infrastructure.
A critical alarm should not wait for a sensor to approach hydrogen’s lower explosive limit. Waiting for flammability conditions is not early warning. The detection system should provide time for informed action while concentrations remain well below that point.
Set thresholds around risk, baseline and rate of change
There is no universal parts-per-million setting that is suitable for every BESS. Fixed thresholds without context can either generate nuisance alarms or respond too late to be useful.
A sound commissioning process starts by establishing normal conditions. Record sensor readings during typical charging, discharging, idle periods, HVAC cycles and expected ambient temperature changes. This provides a baseline against which abnormal gas concentration, humidity and temperature behaviour can be assessed.
Thresholds should then consider both absolute concentration and rate of rise. A modest reading that is stable may reflect background conditions or a known site influence. The same reading climbing quickly can indicate an active fault and justify immediate escalation. Trend-based logic is particularly valuable in ventilated containers, where dilution can reduce absolute gas concentration while a cell failure continues to develop.
Sensor placement is equally important. Gas detectors need to be located where off-gas is likely to accumulate or travel, considering enclosure geometry, natural buoyancy, mechanical ventilation paths, rack layout and obstructions. A detector installed only near an extraction point may miss a developing localised event until gases have mixed throughout the enclosure.
For critical infrastructure, off-gas detection should complement rather than replace the BMS. A BMS can identify electrical and thermal anomalies at cell, module or string level, while gas detection can identify physical battery degradation and venting that may not be fully represented in electrical data. Together, these systems produce a stronger decision basis than either layer alone.
Connect every alarm to a clear operational decision
An alarm with no owner, no escalation path and no defined action is only a notification. Before commissioning, asset owners should establish who receives each alarm, how quickly it must be acknowledged, what information is reviewed and who has authority to change BESS operating state.
The cause-and-effect matrix should account for local versus remote operation, after-hours coverage, communications loss and false alarm handling. It should also define what happens if gas sensing, SCADA communications or BMS data becomes unavailable. A detector fault deserves attention because it removes an early-warning layer when the site may need it most.
For Australian projects, this process should sit alongside the relevant electrical, fire safety, planning, insurer and emergency-service requirements. Site-specific fire engineering, battery OEM instructions and applicable standards remain the governing references. Detection equipment should support those controls, not be treated as a substitute for them.
Systems such as the Evikon E2673 are designed for this operational role, detecting hydrogen, VOCs, electrolyte vapours, humidity and temperature changes while providing relay outputs and Modbus RTU compatibility for integration with BMS and SCADA architectures. For constrained BESS rooms and containers, that integration can turn early-stage sensor data into a timely, controlled response.
Test the alarm sequence before an incident tests it
Alarm thresholds are not set once and forgotten. Changes to battery type, rack density, ventilation settings, enclosure modifications, control software or duty cycle can alter how an event develops and how sensors respond.
Commissioning should include functional testing of each alarm level, relay output, SCADA point, notification route and shutdown or safe-state command. It should also verify that operators understand the difference between an early warning, a confirmed abnormal condition and a critical incident. Periodic testing and documented review keep the alarm philosophy aligned with the asset as it operates.
The best time for a BESS alarm to activate is when there is still time to make a safe decision. Design for the first signs of battery failure, not the moment smoke makes the risk impossible to ignore.




Comments