Introduction

It has always been recognised that environmental noise pollution, and the industrial noise level aspects of occupational health and safety, have had the potential to place a heavy burden upon any company. One reason for this burden has been in the technical approach to dealing with these issues. Noise problems have been attached after they occur rather than before they arise. The cost difference between solving existing noise problems, and the cost of avoiding those problem by informed design, has been estimated as being an order of magnitude or more. This approach of finding noise solutions after installation has, until recently, been a "fact of life" imposed by the technology of the time.

The obvious, but until recently elusive, solution to this burden is the ability to predict, and to predict accurately, the acoustic impact on the entire industrial plant of any design changes before the actual installation of a new item of equipment or process. For larger industrial complexes, this implied the need for a prediction model accurate over an area of several kilometres. The limits to the early introduction of prediction models were : a need for large computational power; limitations on the ability to obtain accurate noise source information in industrial environments; and a lack of reliability of then existing acoustic propagation algorithms.

Prediction models have been around for a number of years. The accuracy of the earlier models was limited and these earlier models were based on empirically derived data from limited site specific field studies. It was not until the 1970s that a clear understanding of the physical mechanisms of sound propagation emerged. Unfortunately, many of these international noise prediction models currently in use are based on research that pre-date this period.

In Australia, acoustic engineering underwent a dramatic evolution. The State Pollution Control Commission (now the Environment Protection Authority) instigated the development of an Australian noise prediction model based on the many of the breakthroughs in acoustic research of the 1970s. The result of this initiative was the PC based Environmental Noise Model.

Camets Acoustics personnel have had a long involvement with the Environmental Noise Model reaching back to the very start of its development. Since the commercial release of the model in 1987, Camets Acoustics personnel have been intensive users of the model on a number of very large industrial projects.

Accuracy in Modelling

As part of our policy to ensuring a consistently high level of quality in acoustic engineering, Camets has undertaken extensive cross checks of the calculated and field measured noise emissions from industrial projects over extended time periods, and on large sites.

While the Environmental Noise Model has proven to be an effective tool in acoustic modelling the key areas, improved accuracy has centred in the sourcing and preparation of input data, and in the detailed interpretation of the model output.

To enhance acoustic modelling accuracy, Camets has needed to :

• Develop programs to filter and collate field monitoring data into a reliable format suited to model cross checks, and fine tuning of the acoustic model.
• Develop programs to accelerate the processing of the necessary large acoustic sets of on-site measurements essential for thorough coverage to minimisation sources of error.
• Develop the field measurement procedures to ensure there is acquisition of sufficiently acoustic data to fully characterise the noise emission source regions. This must includes all potential noise source emissions, even those that appear not to be obvious contributor.
• Obtain the latest acoustic measurement equipment and become fully conversant with range of capabilities limit to use. Based on the capabilities of the instrumentation, Camets has developed new and improved modelling techniques to obtain appropriate acoustic data in the difficult acoustic environments encountered in industry

Camets Acoustics has been able to provide a proven prediction accuracy of between plus or minus 1dB(A), and plus or minus 3dB(A).

On a frequency spectrum basis, Camets have been able to obtain excellent correlation between the computer model generated one third octave frequency band sound pressure spectrums, and the actual one third octave frequency band sound pressure spectrums measured at company designated monitoring locations for a number of large industrial complexes, in both the petrochemical and metal smelting industries at widely diverse locations.

Processes of Modelling

To consistently achieve this accuracy requires an integrated and disciplined approach to quality control over every step of the modelling process, from the data acquisition, through model operation, to final data interpretation. The Environmental Noise Model forms only one component in the chain of acoustic modelling prediction chain.

The full chain requires :

• Selection of the overall measurement strategy appropriate for each unique noise source and situation based upon the statutory requirements and the complexity of the acoustic environment.
• Use of appropriate measurement instrumentation , and methodologies to obtain discrete, item specific, noise spectrums within the complex industrial noise environment. Different techniques and/or instrumentation are required depending on the characteristics of the noise emission sources (e.g. continuous, intermittent, repetitive, impulsive, transient); on the type of noise source and end use of the noise emission measurement (e.g. railway noise, road noise, industrial noise intrusion, hearing conservation, construction noise, communication systems); and on the acoustic environment surrounding the emission sources to be measured (e.g. highly reverberant or crowded with sound sources: or acoustically isolated and quiet). In the area of industrial noise, there are usually many competing production-related noise sources, which must be kept operating and it is often necessary to cleanly extract the acoustic emission of a single process, or machine, without interrupting the overall process. At other times it is necessary to keep the process operating to obtain a specific noise measurement.
• A methodology which incorporates acoustic cross checks to validate the measurement methodology in use.
• Wide ranging experience to correctly interpret the acoustic surroundings to ensure both signal discrimination and the collection of all relevant noise emission data.
• Appropriate selection of the necessary noise descriptor, instrument response, and the level of acoustic detail needed, taking into account the process type, the process characteristics, and the end use of the model output data. Depending on the noise descriptor settings, it is not always possible to interchange measurement types. The indiscriminate use of acoustic data, such as the use of hearing conservation readings inserted into environmental studies, lead to major errors which can be very misleading.
• Appropriate interpretation of the raw field data to ensure that the locations, numbers, and orientations of equivalent source points within the model provide a true representation of the actual noise emissions within the plant.
• Choice of the most appropriate model settings and ranges, based on previous acoustic engineering experience in atmospheric and environmental acoustics. This is to ensure the model accurately reproduces, in an acoustic sense, the appropriate micro-climates for the area under investigation.
• Correct choice of internal reverberation calculation theory for the accurate calculation of acoustic fields within buildings. This is for machinery and processes which are fully, or partially, enclosed within buildings. The Environmental Noise Model can only deal accurately with noise emissions from either unenclosed noise sources, or the external surfaces of buildings. Camets Acoustics has developed a full three dimensional approach to the calculation of reverberant fields in buildings, and has achieved accuracies of plus or minus 1dB(A) in this area.
• Engineering experience in the interpretation of the computer generated outputs. This requires a balancing of the individual attenuated sound sources based upon an appreciation of the cost/benefit and practicality of the various individual noise reduction strategies available, and their realistic integrated into the overall plant wide noise emission patterns

The Environmental Noise Model is only one part of the overall array of programs and expertise to ensure prediction accuracies better than plus or minus 3dB(A).

Conclusion

Without a high standard of data acquisition and data manipulation, both before and after the computing components of the Environmental Noise Mode, the model's accuracy and its usefulness for planning and engineering design testing is greatly compromised.

If the raw model is operated, using industry standard acoustic data acquisition methods, and the data is prepared without due regard to the complexity of the sound fields that can, and do, occur throughout industrial plants, the level of model accuracy that can be expected fall to plus or minus 7 dB(A), or worse.

The consequences of lowering the quality of acoustic the acoustic prediction modelling for the environmental and internal noise predictions, to greater than plus or minus 7dB(A), becomes obvious in terms vastly increased cost through over design the noise control measures. With the lowering of the level of quality there is an increase in the uncertainty of the model. This uncertainty must then be incorporated into the pricing of environmental controls in the assessment of project feasibility.

To more than halve this error margin, Camets strives to obtain and retain a consistent high level of quality field measurements, data processing, analysis, design, and interpretation. When the accuracy is consistently better than plus or minus 3dB(A), the noise control measures can be designed with right to the limit of the noise criteria imposed by the Environment Protection Authority. This minimises cost per noise problem and permits either a lower expenditure on noise related issues or a greater number of noise issues to be dealt with for the same expenditure.

Put us to the test and ask for some free advice. For further information about Camets Acoustics please contact us.