BDI: 1,842 ▼ 1.2%
COTTON NO.2: 84.12 ▲ 0.4%
LME COPPER: 8,432.50 ▲ 2.1%
FOOD SAFETY INDEX: 94.2 ARCHIVE_SECURED
OPTICAL INDEX: 11,204.09 STABLE
BDI: 1,842 ▼ 1.2%
SECTOR INDEX
V.24.08 ARCHIVE
Metro station noise criteria rarely reduce to one permissible decibel value. An acceptable result depends on where the measurement is taken, which operating condition it represents, the acoustic metric specified and the authority or contract that sets the limit. A platform measured during train arrival cannot be assessed on the same basis as a quiet concourse measured between services unless the applicable criterion explicitly allows that comparison.
For technical assessors, the first task is to establish a complete test condition: a numerical limit, a descriptor such as LAeq or LAFmax, a location, a measurement period and the operating scenario. A reading can be accurate and still be unsuitable for a compliance decision if any of those elements differs from the requirement.
There is no single international noise limit that automatically applies to every metro platform and concourse. Requirements may come from national or local law, a transit authority’s design standard, an environmental approval, an occupational health regulation or a project specification. More than one may apply, and they may govern different receptors or conditions.
A station assessment should distinguish at least three questions. First, does the passenger environment meet the acoustic requirement for each public area? Second, is staff exposure controlled over the relevant working period? Third, does noise transmitted outside the station meet the limits applicable at nearby receptors? A platform reading alone does not answer all three.
Passenger-area criteria may also differ between normal operation and periods when trains are absent. A specification might control the continuous background created by ventilation while separately limiting train events. In other projects, it may prescribe an average that includes a representative mix of train arrivals, departures and announcements. The correct interpretation rests on the wording of the governing document.
Do not compare a measured value with a limit until their conditions match. A stated limit in dB(A), by itself, leaves open the averaging time, whether train movements are included, and whether the value applies during peak service or another defined period. Those details can change the outcome of an assessment.
Acoustic standards often describe measurement methods or instrument performance; they do not necessarily establish a pass/fail limit for a metro station. For example, the IEC 61672 series addresses sound level meter performance. ISO 1996 addresses the description, measurement and assessment of environmental noise, while ISO 9612 provides a method for determining occupational noise exposure. Their relevance depends on the question being assessed and the documents incorporated into the project or jurisdiction.
An assessor should identify the applicable edition and any amendments, together with the clauses adopted by the authority or contract. A project may specify its own positions, operating conditions or acceptance procedure. Where documents conflict or leave a material gap, obtain an interpretation from the responsible authority or contracting parties before testing; choosing a convenient method after collecting data weakens the result.
Metro noise changes quickly. Trains, ventilation, crowd activity and public-address announcements contribute in different ways, so a single metric seldom describes every concern. The following measures answer different questions:
The duration attached to LAeq matters. A measurement that includes several train movements may give a different result from one taken during a single approach or an empty interval. Neither is inherently better; each is useful only for the question and criterion it represents. Similarly, an event maximum should be reported with enough context to identify whether it came from wheel–rail interaction, an announcement or another source.
A numerical example shows why the descriptor matters. An LAeq over a complete observation period and an LAFmax from the loudest train can have very different values at the same microphone position. Comparing both with a limit written only for LAeq would produce an invalid conclusion for the maximum. The measurement report should reproduce the criterion’s descriptor and time basis alongside the result.
Start with a plan of the spaces, the relevant criteria and the operating timetable. Select positions that represent occupied areas and locations where high levels or important variations are plausible. The plan should explain why each point was chosen and how it relates to a particular requirement.
On a platform, sound can vary along its length and across its width. Train speed, braking, track geometry, the position of equipment and the proximity of reflecting surfaces can affect a reading. A single point may be insufficient to represent the whole platform, especially where different train stopping positions or entrances create distinct conditions. Placement must also respect safe separation from the platform edge and any operational restrictions.
In concourses and circulation routes, consider ventilation outlets, ticket barriers, escalators, retail activity and the paths used by passengers. A measurement immediately beside a loudspeaker or grille may be useful for diagnosing a source, but it should not automatically be presented as representative of the wider space. Conversely, avoiding a persistently occupied area because it is loud can understate exposure.

At each location, document the microphone position and height, nearby surfaces and obstructions, and whether the location represents a passenger position, a staff workstation or a source investigation. Follow the applicable method for placement. If site constraints require a departure from it, record the departure and assess its effect on comparability.
The measurement schedule should match the acceptance scenario. Record the number and direction of trains, whether they stop or pass, service frequency, relevant ventilation settings and the presence of announcements. Passenger activity can also influence the sound field, so note whether the station is lightly occupied, busy or closed to the public.
Where a specification requires several conditions, measure them separately. A ventilation-only interval can help establish the contribution of building services, while train-event measurements can reveal short-duration noise. An operational LAeq that includes both may be appropriate for a separate requirement. Combining the data without preserving those distinctions can make the results hard to interpret or reproduce.
Repeat observations where train events or operating conditions vary. The number of samples and the way they are combined should follow the governing procedure; there is no universal sample count that makes every station survey representative. If the observed service pattern differs materially from the specified one, report the limitation and arrange additional measurements or an agreed assessment method.
Use instrumentation that meets the performance class required by the specification. Record the sound level meter, microphone and acoustic calibrator identifiers, relevant calibration records, frequency weighting, time weighting, logging interval and measurement range. Check the measurement chain with an acoustic calibrator before and after the survey, following the applicable procedure and manufacturer’s instructions. An unacceptable change between checks calls for investigation under that procedure; it should not be silently corrected.
Keep a synchronized field log. A time-stamped record of each train movement, announcement, unusual disturbance and equipment change makes it possible to associate peaks with events and to decide whether an interval belongs in the assessment. Preserve original logged data as well as processed results. If a reading is excluded, explain the reason and identify the rule that permits exclusion.
Background sound deserves particular care. When assessing an individual source, other sounds may affect the measured level. Any background correction must follow the applicable method and depends on the relationship between the source-on and background measurements. When levels are close, the source contribution can become highly uncertain; simple subtraction of decibel values is not valid because decibels are logarithmic.
For measurements near outdoor entrances or at external receptors, document relevant weather and other site conditions required by the method. For underground locations, focus instead on the conditions that can alter the station sound field, including operating modes and temporary works. In either case, the report should make clear whether the recorded conditions represent the condition being judged.
Measurement uncertainty can arise from instrumentation, microphone placement, changing train operations, background noise and limited sampling. Its treatment matters most when a result is close to a limit. Some requirements specify how uncertainty is to be handled in a compliance decision; others leave the decision rule to an agreement or governing framework. Do not assume that uncertainty is automatically added to or subtracted from the measured value.
A defensible report presents the measured result, the applicable limit, the decision rule and the uncertainty information required by the assessment method. It also identifies any condition that prevents a firm decision. For example, an LAeq measured during a reduced service cannot establish compliance with a peak-service criterion merely because its numerical value is below the limit.
If a location exceeds a requirement, separate the compliance finding from the diagnosis. Time histories and event logs can show whether the dominant contribution coincides with trains, ventilation or announcements; frequency data may help narrow the source further. Those observations guide the next investigation, but they do not by themselves prove the effectiveness of a proposed acoustic treatment.
A useful final assessment allows another reviewer to trace each conclusion back to a requirement and a defined observation: where the microphone stood, what the station was doing, which metric was calculated and how the result was compared with the limit. Where that chain is incomplete, further targeted measurement is more valuable than a broader collection of decibel readings.
Recommended for You