Alarm values are not answers set by the manufacturer

When a new gas detector is switched on, low and high alarms are often already entered. But the manufacturer’s values are only initial values for product testing and general use; they are not the result of assessing a particular workplace’s materials, process, ventilation, working hours, governing laws, and emergency-response capability.

The Ministry of Employment and Labor’s Technical Guidelines on Installation of Gas-Leak Detectors provide direction. Flammable-gas alarms must sound at no more than 25% of the lower explosive limit of the gas detected, and toxic-gas alarms at no more than the permissible concentration for that gas. The guidelines also require suitable equipment selected with full regard to gas characteristics and give priority to toxicity for substances that are both flammable and toxic. They provide important upper limits and selection principles, but do not fix every site’s low alarm at one number. Each workplace must decide whether earlier action is required, whether the device can detect that value reliably, and whether real action can follow an alarm.

Thus, the short answer to “who decides?” is the employer or organization that operates the workplace. It should not be entered arbitrarily by one safety officer. The employer bears final responsibility for risk assessment and legal duties, and must approve the rationale after gathering input from industrial-hygiene specialists; process, production, facilities, and instrumentation staff; emergency responders; field workers; and the detector manufacturer or a qualified service provider.

Write the purpose of the alarm in one sentence first

Setting work begins not with “how many ppm for CO?” but with “what does this alarm protect, and what action does it start?” If its purpose is unclear, exposure limits, early leak warnings, evacuation criteria, and process-shutdown criteria become mixed together. Common purposes can be separated as follows.

  • An alarm telling workers to evacuate, use respiratory protection, or stop work to reduce personal exposure

  • An alarm that finds a small equipment leak early and starts site checks and stronger ventilation

  • An alarm that starts ignition-source isolation, valve closure, or process shutdown before fire or explosion risk grows

  • An alarm that checks whether entry to a confined space is possible and warns of deteriorating atmosphere during work

  • An alarm enabling a control room or disaster-prevention center to view trends from several detectors and convene the emergency organization

If one detector has low and high alarms, the action at each stage must differ. For example, if low alarm means “site check and ventilation” and high alarm means “immediate evacuation and process isolation,” decide who does what within how many seconds or minutes. If both alarms only operate the same beacon and change no action, reconsider the operating purpose of both values. IEC 62682 describes the core function of process-industry alarms as informing operators of abnormal process conditions or equipment faults and supporting response. An alarm is a management function connected to human action, not a numeric display.

TWA, STEL, Ceiling, and %LEL are not interchangeable

The most dangerous error in alarm setting is treating different kinds of limits as the same. TWA, STEL, and Ceiling are primarily time concepts for health hazards, whereas %LEL is a concentration ratio related to whether a flammable mixture can ignite. Their meanings must be kept separate.

The time-weighted average exposure limit (TWA) averages concentration and exposure time over a specified working period. The Korean Ministry of Employment and Labor notice defines it on an 8-hour day. Unless otherwise stated, NIOSH REL assumes up to a 10-hour workday and 40-hour week, so TWAs from different sources must not be combined by looking only at the number. A TWA alarm calculates cumulative exposure; it does not guarantee that the present air is immediately safe. It must not be used to continue work after high exposure while waiting for the average to fall in clean air.

The short-term exposure limit (STEL) normally manages acute effects and brief high concentrations as a 15-minute time-weighted average. The Korean notice specifies 15 minutes per exposure, intervals of at least 1 hour, and at most 4 exposures per day. Unless otherwise stated, the NIOSH Pocket Guide also describes STEL as a 15-minute average. Check the manufacturer’s manual: whether the displayed STEL is a rolling 15-minute average or calculated from session start, and whether power cycling resets it. A 15-minute average can look low despite a very high instantaneous peak, so STEL alone cannot monitor a sharp rise.

The Ceiling (C) is an upper-limit concept that must never be exceeded at any time during work. On a continuously measuring device it serves a purpose closer to an instantaneous or short-response high alarm, but sensor response delay and data averaging can make the displayed value lower or later than the actual peak. Confirm whether the limit document requires an instantaneous value and whether it allows an assessment period when measurement is impossible.

The lower explosive limit (LEL or LFL) is the lowest concentration at which gas or vapor can ignite in air. A detector’s %LEL indication is a proportion of that limit. For example, 10% LEL does not mean that the substance concentration is 10% vol. LEL differs by substance; if an instrument is calibrated with methane while the actual gas is propane, hydrogen, or solvent vapor, response differences and correction factors arise. A substance that is both toxic and flammable can reach a health-hazard limit at a concentration far below its explosion alarm. That is why %LEL alarms alone cannot manage toxic exposure.

How to read legal limits and industrial-hygiene recommendations together

First confirm applicable laws, notices, permit conditions, and client standards. Korean workplaces should first review the current Exposure Limits for Chemical Substances and Physical Agents, the Rules on Occupational Safety and Health Standards, and separate laws applying to the process and facility. The Ministry notice expresses exposure limits as TWA, STEL, or Ceiling and directs users to verify each substance’s form of limit in its annex. Confirm the equipment and scope and apply the gas-leak detector guideline’s requirements of no more than 25% LEL and no more than the permissible concentration.

Second, review health hazards that legal limits alone may miss from an industrial-hygiene perspective. Compare NIOSH REL, toxicity data for the substance, SDSs, and reliable epidemiological and toxicological material. ACGIH states that TLVs are health-based guidance to assist industrial-hygiene judgment, not legal limits or absolute lines between safety and hazard. Therefore, do not apply a TLV automatically only because it is lower than a legal limit, nor ignore it because it has no legal force. An industrial-hygiene specialist should assess exposure routes, work intensity, shift length, heat, skin absorption, mixed exposure, and potentially sensitive workers, then recommend whether the organization’s internal control limit should be stricter.

Risk assessment adapts numbers to the site

Legal limits are a starting point, not a complete design. OSHA’s confined-space standard requires employers to identify and evaluate space hazards and establish acceptable entry conditions; Appendix B recommends that technically qualified persons perform or review data interpretation and entry procedures based on all significant hazards. Legal systems differ in wording, but share the principle that the employer operating the site assesses hazards and documents control conditions.

The assessment team must look beyond normal operation. Include raw-material changes, cleaning, drain opening, commissioning, power loss, ventilation failure, purging, agitation, tank opening, welding, and emergency-leak scenarios. Review leak rate and space volume, ventilation rate, worker location, gas buoyancy and temperature, and travel time to the detector. If a fixed detector is far from the leak source, the sensor may receive the gas late even after concentration has become dangerous. Lowering a setpoint alone cannot remedy a wrong installation location or too few detectors.

Can the device actually meet that setpoint?

If the desired value lies outside sensor capability, the alarm exists only on paper. NIOSH guidance for evaluating direct-reading gas and vapor monitors identifies response time, calibration, stability, range and detection limit, environmental effects, interfering substances, and reliability as evaluation factors. Before setting, confirm the following through manufacturer performance data and field tests.

  • Do range and resolution distinguish adequately around the setpoint?

  • Are accuracy and uncertainty not too large relative to the control limit?

  • Even adding response time such as T90 and sample-hose transit time, is evacuation or isolation time secured?

  • Do temperature, humidity, pressure, oxygen concentration, and condensate distort the indicated value?

  • Are cross-sensitivities or catalyst poisoning from expected coexisting gases present?

  • Can response differences between the actual target gas and calibration gas be corrected by manufacturer-approved factors?

  • Do data averaging, latching, delay, and hysteresis suit the alarm purpose?

  • Are sensor failure, loss of power, and communication loss delivered as separate fault alarms rather than appearing normal?

IEC 60079-29-2 treats selection, installation, safe use, and maintenance of flammable-gas and oxygen detectors as one lifecycle. IEC 62990-2 likewise links selection, installation, use, and maintenance for toxic gas and vapor detectors. A setpoint is therefore not something entered once at purchase and forgotten; it is part of a detection system that includes device performance and maintenance.

Who proposes and who approves?

In practice, an approval table separating responsibility and review roles is useful. One person may hold several roles at a small workplace, but recording the person who proposed the basis and the final approver reduces arbitrary changes.

  1. Safety and health or industrial-hygiene personnel review applicable law and exposure limits, toxicity data, working hours, and personal-exposure scenarios.

  2. Process and production personnel present normal and abnormal operating conditions, leak scenarios, safe operating limits, and available actions after an alarm.

  3. Instrumentation and facilities personnel verify sensor range, accuracy, response time, cross-sensitivity, voting logic, communications, and maintainability.

  4. Emergency-response personnel confirm evacuation, isolation, rescue, notification, and return conditions for each alarm stage.

  5. Field workers and worker representatives provide actual work locations, whether alarms can be perceived, PPE use and evacuation time, and recurring false-alarm experience.

  6. An authorized manager reviews residual risk and approves the setpoints, response procedures, training, test cycle, and effective date.

Manufacturers and vendors can explain sensor technology and recommended ranges, but cannot replace the employer’s risk assessment. Even when an external industrial-hygiene specialist performs calculations and validity review, the organization retains responsibility for the scope of application and approval. ISO 45001’s emphasis on leadership, worker participation, hazard identification, legal requirements, and performance evaluation aligns with this joint-review structure.

A setting change is management of change, not a small instrumentation job

Raising a threshold at the site by entering a password because “there are too many false alarms” is dangerous. OSHA’s process-safety-management standard requires written procedures to manage changes to process chemicals, technology, equipment, procedures, and facilities in covered processes. Nonmandatory Appendix C explicitly lists changes to alarms and interlocks as examples of equipment and instrumentation changes. Even where a Korean workplace is not directly subject to this U.S. rule, managing alarm changes with technical basis and approval is practical.

A change request records the current and proposed values, reason, technical basis, affected detectors and work, risk-assessment results, sensor-performance review, post-alarm actions, approver, trainees, test method, and effective date. Temporary changes need an end date and a person responsible for restoration. After a change, verify actual alarm, beacon, control-room indication, data record, relay, and shutdown logic using calibration gas or a suitable test method. It is insufficient merely to confirm that the screen’s setting menu changed.

Workers also need training before and after a change. Explain the meaning of low and high alarms, prohibited actions during an alarm, evacuation direction, supervisor call, and conditions for re-entry approval. Update connected work permits, emergency plans, checklists, drawings, setpoint lists, and maintenance systems as well. Even an equipment replacement that appears “like for like” may in fact be a change if sensor principle, range, firmware averaging, or default alarm values differ.

Periodic review begins with alarm history

Approved values are not permanent. Review them when laws or exposure limits change; materials, process, or ventilation change; a sensor is discontinued; working hours extend; or an accident or near miss occurs. Regularly analyze alarm frequency, duration, maximum value, worker response time, false alarms and missed alarms, sensor faults, and calibration drift. IEC 62682 treats alarm history and performance indicators as part of the system because they show whether alarms are actually managed.

Repeated alarms may be data indicating a problem with process, ventilation, work method, or sensor, rather than a signal that the setpoint is inconvenient. Investigate the cause and improve controls before judging the setpoint’s suitability. The absence of alarms does not prove safety: detector placement may be wrong, the sensor may have become insensitive, or test gas may not have reached it. A bump test checks sensor and alarm function; calibration adjusts indicated accuracy against a known concentration. Neither replaces the validity of the setting rationale, but both are necessary for the setpoint to operate as real protection.

One setpoint-rationale sheet to retain at the site

Good alarm management can start with a table that traces the decision at each detection point rather than a complex report. Record detector-point number; location and protected target; target gas; sensor principle and range; calibration gas; low/high alarm and TWA/STEL settings; source for each value; measurement uncertainty and response time; action by alarm; interlock logic; reviewer and approver; last review date; and next review conditions. Show legal limits and internal company control limits separately if they differ.

In conclusion, an alarm setpoint is not a number taken from an online “recommended ppm” list or a factory default. It is an operating criterion approved after the employer meets applicable law, industrial-hygiene specialists interpret health limits, process and instrumentation personnel verify leak scenarios and device limits, and field workers and emergency responders confirm that action is feasible. Only when TWA, STEL, Ceiling, and %LEL are distinguished and changes are treated through formal management of change does an alarm become a safety device that prompts timely action rather than noise.

Sources

Cover photo

Original Pexels photo