Overview: Noise monitoring
Managing noise in construction primarily involves monitoring noise levels in line with regulatory and/or contractual requirements, reporting on those measurements, and working with stakeholders who may be impacted by the noise. Where stakeholders are likely to be impacted, proactive noise management is much more important.
Elements of noise management often include:
- Identifying baseline noise levels
- Establishing thresholds in line with the project requirements
- Monitoring and managing to those thresholds
- Identifying and explaining noise events or exceedance
- Keeping stakeholders informed
Understanding pre-construction noise
Most projects require baseline noise measurements to be taken and submitted for approval before construction works can go ahead. Baseline measurements create an evidence base with which to determine the scale of any noise being caused by works.
Noise is regulated by local councils, who each have their own policies. However, when the capital value of the project reaches a certain size, the State regulations become active. In NSW, for example, any projects of $250m or over become a State Significant Development and must be conducted in line with State regulations.
The State baseline measurement requirements also differ from State to State. In NSW, for example, even small Development Applications (DAs) require multiple days of baseline data to be submitted with the application.
The noise background level is often referred to as a Rating Background Level (RBL), and has a specific methodology for calculation. This methodology has been automated in SiteHive, making calculating RBLs a quick and easy process when the appropriate data has been captured.
Noise monitoring requirements
Most Construction Environmental Management Plans (CEMPs), which are approved during the planning phase of the project, specify the noise monitoring that is required during the course of the project. Although noise measurements can be collected using either attended or unattended monitoring methods, an increasing number of projects now conduct unattended monitoring (even when it is not specifically required) as it mitigates the risk of not having data to refer to if/when required.
Unattended monitoring can be continuous and provides a complete record of noise levels at multiple locations. It also means that attended monitoring, which provides a snapshot measurement with subjective expert assessment, only needs to be done when required for specific issues or situations (such as complaints).
In NSW for example, the noise criteria for projects are typically based on the NSW Noise Policy for Industry: Noise Policy for Industry (2017), where Noise Management Levels (NMLs) are set for various Noise Catchment Areas (NCAs), over day/evening/night periods. These are based on the Rating Background Levels (see above). Alternatively, the EPA Construction Noise Guideline: Construction noise is used, which sets a guideline for ‘Highly Affected Noise Level’ of 75dB (15 min LAeq) during ‘Site Hours’. (This latter document has been in draft form only for several years.)
Noise monitoring should not be conducted (or the data should be excluded) when average wind speeds (for 15-minute periods or shorter) at microphone height are greater than 5 metres per second, or when rainfall occurs. Source: NSW Noise Policy for Industry: Noise Policy for Industry (2017).
SiteHive automatically integrates weather data from the nearest Bureau of Meteorology weather station for your site, enabling the identification of weather-affected noise measurements in background level calculations and events investigations.
Sound levels, A/C/Z weighting, fast/slow filtering
The sound level, expressed in decibels, or dB, is the basic measurement of any sound level meter. A Frequency Weighting (A/C/Z) and Time Weighting will be applied to the sound signal that is picked up by the microphone. Parameters such as Max (Lmax), Min (Lmin) and Ln (percentiles) are all based on the Sound Level.
The most commonly used measurement is LAeq15, which means an A weighted equivalent level, over 15 minutes.
The most common weighting that is used in noise measurement is A-Weighting. Like the human ear, this effectively cuts off the lower and higher frequencies that the average person cannot hear. The frequency weightings used in sound level meters are often related to the response of the human ear, to ensure that the meter is measuring pretty much what you actually hear.
Sound level measurements using any grade of sound level meter can be Fast, Slow, or Impulse time weighted. These weightings date back to the time when sound level meters had analogue meters and defined the speed at which the meter moved. Under Fast the needle would move fast to show quickly varying noise and under Slow the needle would be damped to smooth the noise out to be easier to read.
Fast corresponds to a 125 ms time constant. Slow corresponds to a 1 second time constant. Impulse has a time constant of 35 ms.
Types of noise monitoring
Unattended monitoring is the practice of deploying automated noise monitoring devices that continuously monitor noise. The benefit of unattended monitoring is that it provides a far richer and more representative dataset of environmental conditions than attended measurements alone. Unattended monitoring can be continuous, over a much longer time period, and multiple locations on the same site can be measured simultaneously for comparisons. This allows a true noise baseline to be identified, and means the impact of works can be quantified.
SiteHive Hexanode providing unattended continuous noise monitoring at the Sydney Fish Market redevelopment:
Unattended monitoring also provides a complete and continuous record of noise levels that is useful for reporting and stakeholder management. It also means monitoring can be undertaken at inconvenient times, such as during night works.
Unattended monitoring does not provide the same subjective assessment of noise contributions that can be gained from attended monitoring. However, it can be used effectively to determine when/where to focus more human-intensive measurements such as attended monitoring.
Attended monitoring is the practice of a qualified practitioner undertaking noise measurement in person, using a hand-held sound level meter (usually IEC 61672 Class 1 - see below on standards). The measurements are accompanied by a subjective assessment of contributions.
Attended monitoring is generally used for specific situations, such as in response to complaints; at the start of works likely to generate high noise levels; or periodically to ensure unattended monitoring is accurate.
Attended monitoring results can be captured in SiteHive, whilst in the field, to provide a complete system of record for all project noise data.
Acoustic consultant undertaken attended monitoring in the field, next to a SiteHive Hexanode and dust deposition gauge:
Noise modelling and prediction
A large number of planning applications require that noise modelling is undertaken before any works commence. The purpose of the modelling is to predict, plan and mitigate the risk of high noise impacts.
Noise modelling involves measuring background noise levels as a model input; developing a 3D model of the site; and then testing a range of different activity scenarios (e.g., five excavators working at once in a particular location vs three during the day and two at night), to understand the predicted noise impacts of the works.
There are a number of industry standard modelling packages used for this such as:
Example noise model (CADNA):
Noise modelling is almost exclusively carried out by consultants, as it involves providing advice on the best approach. Most consultants also have their own software, which uses the outputs from standard modelling software to develop scenarios specific to the project plan. Using the consultants’ software, users are able to play with the proposed plan and adjust the metrics until it conforms to planning guidelines. For example, if the model shows work will be over the prescribed Noise Management Levels, they can adjust the scenarios until the plan conforms and can be submitted for planning approval. Modelling is required at the start of major projects, and also then for every new out-of-hours work permit.
Examples of the software used by consultants include:
The accuracy of sound level meters, and the microphones within them, are governed by international standards maintained by the IEC (International Electrotechnical Commission). The main relevant standard is IEC 61672, which has 3 parts:
- AS IEC 61672.1 - Specifications
- AS IEC 61672.2 - Pattern evaluation tests
- AS IEC 61672.3 - Periodic tests
Part 1 gives the precision requirements, part 2 specifies how these should be tested, and then part 3 specifies how to check that the sound level meter is still performing. Pattern-approved sound level meters are designated as Class 1 or 2, depending on the precision, and have passed all of the tests stipulated in the standards. Class 1 sound level meters need to measure sound over a wider frequency than Class 2 meters (10KHz vs 20KHz) and meet narrower tolerances for all performance criteria.
The IEC 61672 standards are becoming outdated as they only refer to condenser microphones, which produce an analogue signal. Modern MEMS (micro-electromechanical system) microphones provide digital signals and are rapidly becoming an accepted alternative for environmental monitoring. Driven by innovations in consumer electronics, the high performance, reduced cost and increased stability of MEMS microphones allows more of them to be deployed for an overall lower cost than a single, traditional sound-level meter.
The SiteHive Hexanode, for instance, has 11 different MEMS microphones measuring both overall sound level and the direction of arrival of sound:
Although they provide more data, MEMS devices can’t achieve full pattern approval as the standard requires the injection of a signal into an analogue signal chain (MEMS are digital). MEMS devices can still be calibrated though, and certified in line with the precision requirements of the standard. This provides confidence in the accuracy of the recorded measurements.
SiteHive works closely with the Federal Government’s National Measurement Institute’s (NMI) acoustic, ultrasound and vibration measurement team. NMI’s services are the most accurate in Australia, and include both testing against the precision requirements of the standard (IEC 61672) and the NATA (National Association of Testing Authorities) calibration and certification of sound level meters.
The performance of the SiteHive Hexanode sound level meter is shown below, as measured by NMI, compared with the tolerances of a Class 2 sound level meter (note Class 2 devices are only required to perform up to 10KHz frequency range, the Hexanode actually measures beyond this, to 12.5KHz, Class 1 devices go to 20KHz):
Calibration of sound level meters is recommended at set intervals, and for assurance can be undertaken at NATA-accredited facilities. NATA facilities undertake a set of tests as outlined in IEC 61672.3 - Periodic tests, and provide traceable and certified calibrations for sound level meters.
A NATA calibration certificate is sometimes a stated requirement for all monitoring devices within a Construction Noise and Vibration Management Plan. The SiteHive Hexanode can be provided with a NATA calibration certificate upon request.
Managing noise with SiteHive
SiteHive provides noise monitoring devices and software that help project teams to manage noise and meet compliance requirements across the full project lifecycle. It automates processes and rules, making it easier to manage and report on sound level activity. SiteHive includes:
- A complete system of record of noise captured for the whole project lifecycle
- Unattended noise monitoring data from all types of devices managed within SiteHive software
- Attended noise measurements captured live on site via the SiteHive app
- Noise analysis tools, including RBL calculation
- SiteHive Hexanodes provide real-time, continuous unattended noise monitoring; directional monitoring; directional images and audio of noise events