project-specific noise levels, further analysis isadvisable to determine whether the actual occurrenceof temperature inversions in the area issignificant enough to warrant inclusion in theassessment.Determining the extent of impactW here further analysis is required, meteorologicaldata need to be analysed to determine whethertemperature inversions occur often enough to causea significant noise impact. Existing weather datamay be used, provided the subject area is within aradius of 30 km of the collection point and in thesame topographical basin. Table C4 in Appendix Coutlines the methods that may be used to analysethe data, as well as the type of data and length ofmonitoring time (3 months in winter) required. Byapplying any one of the methods described in detailin Appendix E, one can predict both the type ofinversion (in terms of a stability class) and thepercentage occurrence of inversions. Where F or G-class inversions, or a combination of both, arepredicted to occur for at least 30 per cent of the totalnight-time in winter, this is considered to be significantand warrants assessment. But if these inversionsare predicted to occur for less than 30% of thetotal night-time, temperature inversions do not needto be included in the noise assessment. For thepurposes of this case study, it is assumed that the F-class inversion occurs for more than 30 per cent ofthe night-time in winter and so requires assessment.Using default or alternative parameters to predictnoiseNow that it is certain that temperature inversionswill be occurring for a significant amount of time,they should be included in the noise assessment.The noise assessment may use the default parametersspecified (that is, 3°C/100 m and 2 m/s),avoiding the need for on-site measurements. Otherwiseon-site measurements may be conducted todetermine the exact temperature inversion parametersto use in the prediction model. (See Appendix Cfor guidance on how to conduct on-site measurements.)Assessing the impactThe predicted increase in noise level is then comparedwith the project-specific noise levels derivedin the previous section for case (a) to determinewhat the noise impact will be.(b)Area affected by gradient windsThe effects of gradient winds need to be consideredin all assessments. Generally, wind-speed data needto be analysed to determine whether wind enhancementof noise warrants inclusion in the analysis.The assessment of wind effects needs to considereach season and, within those seasons, each assessmentperiod (that is, day, evening and night).Significant wind enhancement is deemed likelywhere a source-to-receiver component wind of up to3 m/s occurs for 30 per cent of the time in anyassessment period. However, this step may bebypassed by first assuming that wind is a feature ofthe area and then applying a default wind-speed of3 m/s (in the direction from source to receiver) to thenoise prediction to estimate the upper level ofimpact.If no significant increase in noise is shown, thenthis should be reported, and no further analysis ofthe effects of gradient winds will be needed. Whereimpacts are significantly increased with this windspeed it is advisable to conduct a more detailedanalysis of wind-speed data to confirm whetherwind is a feature of the area.Existing or measured data may be used as fortemperature inversions. Wind roses offer a goodvisual interpretation of the occurrence of wind at alocality. These should be inspected to determinewhat, if any, wind speeds travel in the directionfrom source to receiver (or have components thattravel from source to receiver); and of these, whatwind speeds occur for more than 30 per cent of thetotal time in each season for each assessment period(day, evening, night). Where wind does not travelfrom source to receiver, it is not considered likely tohave a significant impact and does not need to beincluded in the assessment. Similarly, if windspeeds up to 3 m/s do not occur for 30 per cent ormore of the time in any season for any assessmentperiod, gradient winds do not need to be includedin the assessment. If a wind speed of up to 3 m/soccurs for 30 per cent of the time or more, then adefault wind speed of 3 m/s can be used in theassessment. The following examples show whatwind speed should be used in the assessment,when using data displayed by the wind rose.• If the wind rose shows that a source-toreceivercomponent of wind up to 3 m/soccurs for at least 30 per cent of the time inNSW industrial noise policy64
any assessment period in any season, thenthe default wind speed of 3 m/s should beused in the assessment for the particularassessment period/s and season/s where itoccurred.• If the wind rose shows that a source-toreceivercomponent of wind less than3 m/s occurs for at least 30 per cent of thetime in any assessment period in any season,then the highest wind speed (below 3 m/s)may be used instead of the default.• If the wind rose shows that there is less than30 per cent occurrence of source-to-receivercomponent wind speeds up to 3 m/s, thenwind is not included in the noise predictioncalculation.Assessing the impactThe parameters for wind direction and speed arethen used to predict increased noise levels, whichare then compared with the project-specific noiselevels derived in the previous section for case (a) todetermine what the noise impact will be.A4Existing premises case studyIn dealing with noise impacts from existingpremises it needs to be recognised that the meansavailable to mitigate noise may be more limited thanfor new premises. For example, planning approachesthat could avoid impact are generally notavailable at this stage (for example, spatial separationbetween source and receiver and attention tonoise reduction in designs for residential andindustrial buildings). Measures such as consideringnoise impacts when laying-out buildings on theindustrial premises and when selecting equipmentwill be more limited, and may be a feasible considerationfor existing premises only in the longer term.In treating existing cases it is often necessary to takea structured approach to mitigation that starts withdefining the problem, then identifies the range offeasible and reasonable controls that can be applied,then implements noise controls over time.The following example provides a summary of howan existing premises with a noise problem can beaddressed.Example:An existing scheduled premises is receiving complaintsabout noise from neighbouring residences.Currently the premises has no noise limits on itslicence.The EPA and the company have contacted complainantsand it appears the complainants mayhave a genuine grievance.The next step is to ascertain whether in fact there isa noise impact (that is, noise levels exceedingapplicable project-specific noise levels for the site).After discussions with the residences the companyagreed to the EPA’s request to conduct a noisesurvey. The survey, conducted by an accreditedacoustics practitioner, covered the following items:• identified the most affected residences• identified the activities on the site that causedthe reported annoyance• conducted a noise assessment without theexisting premises operating, includingbackground (L A90, 15 minute) and ambient(L Aeq, 15 minute) noise levels. (The short-term orlong-term background and ambient noisemethods may be used, depending on theextent of noise impact.)• determined project-specific noise levels basedon EPA noise criteria and applying theexisting background and ambient noiselevels• measured noise levels from the site at theidentified affected residences. (Whereapplicable, include adjustments to takeaccount of any annoying characteristics ofthe noise.)• compared measured noise levels (adjusted,where necessary) against the project-specificnoise levels to determine the extent of impact• discussed whether noise impacts may beincreased by weather patterns (for example,winds and temperature inversions) typical tothe area and estimated the extent of theincrease• discussed mitigation measures and theamount of noise reduction expected fromeach measureNSW industrial noise policy65