Simply click on the title to reveal the abstract for the publication. If you are interested in reading the full report, please contact us.
Air Quality and Odour Impact Assessment of an Estate Containing Noxious and Offensive Industry (2011)
The Narangba Industrial Estate, located approximately 35 km north of Brisbane, is a designated hazardous and noxious industry site with greater than 70 individual enterprises ranging from leather tanning, hot dip galvanising, liquid and solid waste treatment, swimming pool, inorganic chemical, organic/inorganic fertiliser and aquaculture feed production. In 2005, a fire at a chemical manufacturing facility in the estate prompted the Queensland Government to commission an investigation into the risk to public health associated with air contaminants emitted by industrial activities operating in the NIE. People in the surrounding communities have complained to regulatory authorities about odour and health concerns that they associated with activities within the estate. This paper discusses the objectives, methods and findings of the air quality and odour impact assessments of 11 businesses in the NIE, including the assessment of 153 individual hazardous air pollutants and odours from 25 individual sources. Key air pollutants were identified using a hazard index approach and assessed for the cumulative impact of all sources monitored in the estate. The assessment found that the air quality criteria for all air pollutants were met at all sensitive places beyond the NIE boundary with the exception of three substances emitted by three individual businesses. The predicted ground-level concentrations of these three substances were highly sensitive to assumptions made in the calculation of the emission rates, and the impact was likely to be over-estimated. While standard methods were employed for the study, several issues were identified including the assessment of pollutants with an emission rate based on a concentration at the test method limit of detection, identification and monitoring of sources during routine operations, emission variability during routine operations and emission variability due to non-continuous, batch or cyclical production processes.
Prepared by: Andrew Balch, Andrew Wiebe, Alex Schloss, Andrew Vernon, Christine Killip and Simon Welchman
Investigation of Regionally Specific PM10 and PM2.5 Signatures for Developing a Technique for use in Cumulative Impact Assessment (2011)
An analysis of ambient PM10 and PM2.5 concentrations in multiple urban, industrial and rural regions of Queensland is presented based on DERM monitoring information. Monitoring data has been handled in accordance with the NEPM procedure for data analysis of TEOM measurements. The analysis characterises the statistical distributions of ambient particulate concentrations in various airsheds including Brisbane/South East Queensland; the industrial town of Gladstone, the rural city of Toowoomba; and an undeveloped region of the Bowen Basin representative of a rural, green-field location. Several techniques are discussed for the derivation of an ambient background concentration for use in cumulative impact assessment. A statistical technique is suggested for the assessment of cumulative impacts from dust generating sources with the inclusion of a representative ambient background. The technique incorporates a risk assessment approach to estimate the probability of an exceedence of the PM10 and PM2.5 air quality objectives at a sensitive receiver through the convolution of distributions of model predictions and observed ambient concentrations that are based on a distribution representative of the airshed.
Prepared by: Andrew Wiebe, Andrew Balch, Frank Quintarelli, Michael Burchill, Christine Killip and Simon Welchman
An investigation into the potential implications for regional photochemical ozone due to the expected significant future growth of population in Southeast Queensland (SEQ) has been conducted using CSIRO’s advanced numerical model TAPM-CTM. Results suggest that existing rural and urban fringe regions of Ipswich and Toowoomba to the west and the current urban areas of Greater Brisbane and the Gold Coast have a high risk of exposure to ozone based on projected population growth. Recommendations for air quality monitoring to assist in management of airsheds are given.
Prepared by: Christine Killip, Natalie Leishman, Darlene Heuff, Alex Schloss and Simon Welchman
Presented at the 14th IUAPPA World Congress in Brisbane, Queensland, Australia, 2007.
Modelling of Sulphur Dioxide Levels from Coal-burning Industrial Sources Using a Stochastic Estimation Technique (2007)
The use of a stochastic (Monte Carlo) technique when modelling the potential impact of sulphur dioxide emissions from major industrial coal-burning sources has been investigated. The likelihood that the 1-hour average ground-level concentration of sulphur dioxide will exceed the NEPM(Air) standard of 570µg/m 3 as a result of emissions from a generic coal-fired power station, was predicted. The flexibility to use a different sulphur content distribution for each source allows for a more realistic estimation of the potential impacts of multiple coal-burning sources on the airshed which is important if local and regional airsheds are to be managed effectively.
Prepared by: Darlene Heuff, Lena Jackson, Christine Killip, Frank Quintarelli.
Odour intensity has recently found favour in some Australian state odour policies. We consider the implications of adopting odour intensity rather than odour concentration criteria for estimating the necessary separation distances to avoid odour nuisance, using source measurements at a modern integrated meat processing plant. The sources have quite different concentration-intensity relationships. Concentration criteria from various states yield a four-fold difference in the recommended buffer distances if linear addition of odour components is assumed. Odour intensity can be used as a design goal for new industries to limit the occurrence of distinct or weak odours in sensitive areas. We have found that using this distinct criterion yields generally smaller separation distances than conventional concentration criteria. Choosing weak rather than distinct odour almost doubles the separation distance. The results for the case study are relatively insensitive to the assumptions of source additivity or the use of individual or facility-average intensity relationships, as results are dominated by one source.
Whilst there are some advantages to odour intensity, the coping ability of the community and hedonic tone are also likely to be critical (otherwise laboratory intensity measurements may be misleading). An alternative approach based on community-dependent response functions and hedonic tone is outlined.
Prepared by: Simon Welchman, Abbie S. Brooke and Peter R. Best
Fragrance, Fluctuations, Facts or Fantasy – The Use, Abuse and Disuse of Extreme Statistics in Odour Evaluations (2004)
Odour science is exasperatingly incomplete but rapidly evolving, yielding problems for integration with environmental management techniques. Odour risk management involves both deterministic and statistical features evaluated over a wide range of space and time scales. Detection and recognition occur within seconds; transparent and transferable models must therefore include emission and dispersion variability. Discomfort measures (e.g. annoyance, health impacts) involve the repeated interaction of physical measures with a social network and are influenced by both past history and social dependencies. Incorporation of these features and their quantification should emphasise stable and meaningful measures for odour response within a given type of community.
Prepared by: Peter Best, Christine Killip and Abbie Brooke
Application of Various Dispersion Models to Management of Local and Regional Air Quality in Australia (2003)
This paper explores the use of air dispersion models for planning industrial developments within an existing multi-source complex or as part of real-time forecasting tools for intermittent control of emissions. A brief review is given of our recent experience in using TAPM, Calpuff and AERMOD in sub-regional airsheds in coastal locations in Australia, New Caledonia and Vietnam. These studies cover various scenarios of medium-tall stacks with a range of buoyancies in near-coastal terrain ranging from relatively simple plains to moderately complex topography. The focus is on producing a decision-support system for each application that utilises the more sophisticated approaches to the generation of four-dimensional meteorological fields and pollutant dispersion in a quality-controlled manner, yet allow sensitivity studies for conservative pollutants to be undertaken by non-specialist modellers or planning professionals. The meteorological fields have been generated from the CSIRO TAPM model or the Mesolaps high-resolution forecasting scheme run by the Australian Bureau of Meteorology. Dispersion simulations can utilise one of several schemes with emphasis on Calpuff, TAPM and recently AERMOD, depending on the user’s requirements for spatial applicability and computational efficiency. The multi-model GUI yields the relevant statistics, time series and contour plots for a wide range of averaging times and is readily generalised to treat stochastic emissions or to undertake a health risk evaluation. In general, there is no preferred modelling scheme for many applications because of the wide range of spatial scales required for air quality management. Some model performance evaluations for the multi-industry sub-region of Gladstone show reasonable accuracy is obtained for the robust extreme statistics.
Prepared by: Christine A. Killip, Abbie S. Brooke, Peter R. Best
Previous studies have demonstrated that TEOM instruments using heated inlets may underestimate PM10 levels in urban areas due to the loss of semi-volatile secondary aerosols and organic compounds. Whilst temperature-dependent or seasonal adjustment factors have been suggested as suitable procedures to bring closer agreement between traditional gravimetric methods such as high volume air samplers and other measurement techniques, considerable variability may still occur. Detailed compositional measurements are rarely available in Australasian situations that can provide a better approach for standardisation of TEOM or beta-attenuation techniques. Recent measurements at 2 sites in the M5 East tunnel air quality monitoring network (Sydney) allow reasonable seasonal factors to be defined but have much smaller adjustment factors than at a nearby EPA site. The implications for routine measurement campaigns and setting of project performance measures are discussed with reference to recent recommendations from Australian and European agencies.
Prepared by: Peter R Best1, Lena Jackson1, Jane Barnett2, Eloise Duguid2, Melissa Hart2, Kerry Holmes2
2 Holmes Air Sciences
Accurate prediction of horizontal and vertical windfield structure is required for a variety of air quality and energy development assessments. Spatial and temporal interpolation of available measurements and/or numerical prognostic modelling offer practical approaches but may differ considerably when predicting near-surface conditions. This has implications for the validation of prognostic models and the dispersion treatment of normal anthropogenic pollution sources, especially in urban areas where microscale turbulence due to sub-grid scale roughness in homogeneities may be considerable. A review of performance measures for windfield prediction suggests additional considerations beyond the normal indices, including the comprehensive treatment of synoptic weather types for a region and a weighting for information content. Extensive use of TAPM Versions 1 and 2 at a variety of inland and coastal sites suggests some initial rules to aid the non-specialist modeller.
Prepared by: Lena Jackson, Natalie Leishman, Christine Killip and Peter Best
Major air quality field and modelling studies were carried out for the Gladstone region in the early 1990’s to assist in land use planning for prospective major industrial developments. Recent air quality monitoring and modelling studies have focussed on ratifying a suitable scheme for estimating short-term concentrations of conservative pollutants in the region. A review of the air quality measurements in the light of historical monitoring since 1980 confirms the importance of seabreeze and convective conditions for significant pollution events from major emission sources and allows some definition of complex terrain influences. A coupled TAPM – CALPUFF scheme has been produced for event evaluation and production of air quality statistics. These advances have been encapsulated in a graphical user interface for regional air quality management, suitable for non-specialist use. This tool allows rapid evaluation of cumulative air quality impacts and industry planning, and is open to a wide range of future applications such as source-specific air quality models, complaint evaluation and resource scheduling.
Prepared by: Christine A. Killip1, Abbie S. Brooke1, Lena D. Jackson1, Peter R. Best1, Ken Verrall2, David Wainwright2
2 Queensland Environmental Protection Agency
Downscaling of Weather Prediction Models for Short-Term Alert Systems for Air Quality and Other Decision-Support Systems (2002)
Current advanced numerical modelling schemes cannot yet routinely produce reliable forecasts of basic weather variables out to a horizon of 6-10 days at a half hourly time resolution and at spatial scales of 1 km. Yet many applications involving weather-related impacts (e.g. smog forecasting, agricultural management, industrial scheduling) require this facility at a reasonable cost. Should a real-time local weather station be available, the coupling of monitoring information and even coarse-scale (75 km, 6 hourly) numerical forecasts can be “downscaled” to give robust and relatively inexpensive point forecasts of winds, temperatures and other derived variables. A variety of tools now available to perform such point forecasts in Australasia and elsewhere includes statistical, time series, neural network and other artificial intelligence methods. Experience for many sites in eastern Australia using relatively simple techniques has been very encouraging (and useful for energy trading). Extreme events may require a more holistic approach. The implications for urban and industrial air quality management are discussed.
Prepared by: Peter R. Best, Karen E. Lunney, Lena D. Jackson, Damien J. Cresta, Christine A. Killip and Nigel D. Ellis
Development & Optimisation of Predictive Relationships between Corrosivity and Environmental Parameters for Power Station Microclimates (2001)
Corrosivity measurements at eight sites over a one year period near a coal-fired power station in Central Queensland have been interpreted in light of detailed, concurrent air quality and meteorological measurements, and with the aid of a comprehensive dispersion model for sub-tropical regions. The (limited) measurements suggest that there may be a doubling of short-term corrosivity on mild steel plates in areas close to the power station that experience enhanced ground-level concentrations of combustion gases. The mild steel corrosivity at the control site, measured over three months, was 23 µm/year.
As corrosion in a sub-tropical environment 50-80 km from the coast can have several causes, evaluation of exposures to microclimate and pollutants utilised detailed numerical modelling schemes in conjunction with emission information and site monitoring. Meteorological variability between control and test sites was predicted to be small. Chloride contributions were expected to be uniform and relatively small, based on evidence from similar sites in Australia. Treatment of repeat, three-month corrosivity field data allowed first-order regression equations to be developed between measured corrosivity and atmospheric pollutant exposure.
Careful experimental measurement, in conjunction with a reliable dispersion model, can lead to the development of first-order corrosivity maps for a microclimate. The maps can be useful for the assessment of future industrial scenarios in a region.
Prepared by: David Druskovich, PhD, School of Chemical and Biomedical Sciences, Central Queensland University
Peter Best, PhD, Katestone
There has been considerable recent pressure in most Australian State for major point sources to minimise their emission levels and emission rates of nitrogen oxides NOx. This arises from a desire to minimise local air quality (from a health impact and odour perspective) and avoid increasing smog levels on those few days per year when the photochemical nature of the Australian urban air-shed is NOx-limited. Strict emission controls or load-based licensing are currently in place or being advocated for new point sources, often with little demonstration that significant benefits will thus accrue.
This paper provides an concise summary of the key parameters and characteristics for the evaluation of NOx impacts, the current regulatory requirement, the nature of smog generation within industrial plumes and the benefits of NOx-control. Some discussion is given of future trends and drivers for improved NOx-control, either via source emission technology or via temporary NOx reductions. Recent work in Australia, United States and Europe suggest that local air quality may be as strong a driver as regional considerations. Recent Queensland studies also suggest that near-sources odour impacts may be quite important for coal-fired power stations in various environmental and gas-fired power stations in densely populated urban areas.
Prepared by: Peter Best and Christine Killip
Two key elements in predicting odour response are the estimation of peak (few seconds) ground-level concentrations and the evaluation of the likelihood of a consequent adverse response. Peak rather than ensemble-average concentrations are not easily predicted by current dispersion models. The required peak-to-mean ratios depend on source characteristics, downwind distance and atmospheric stability. Using recent wind tunnel simulations for four types of sources and two atmospheric stabilities and various measures of intermittency and non-stationarity, different regimes of behaviour are resolvable. A set of peak-to-mean ratios for a specified probability of exceedance is recommended and their practicality discussed.
Prepared by: Peter R. Best, Karen E. Lunney and Christine A. Killip
The past forty years have produced a multitude of deterministic and statistical modelling tools for predicting the temporal and spatial distribution of air pollutants around a variety of source structures. Long-memory time series analysis and pattern recognition techniques are now an attractive alternative or supplement to numerical modelling for identifying significant trends in long-term air quality and providing short-term air quality forecasts.
Recent statistical analysis of Australian capital city air quality information has demonstrated the relative importance of different source types, the role of bushfires in ozone production and the changing environment of monitoring locations in a rapidly expanding metropolis. The same techniques form the basis of user-friendly decision-support systems for local/regional air quality. With realistic forecasts of regional meteorological states now available up to 7 days in advance, predictive and reactive control philosophies can be integrated. Particular applications are discussed for providing a 4-48 hour forecast of regional photo-oxidant and PM10 distributions in the Brisbane urban area and the predictive/reactive control strategies for a major inland industry.
Prepared by: C. Killip, K.E. Lunney, P.R. Best, M. Kanowski
Available literature yields a variety of estimation techniques and parameterisations for estimating short-term or long-term exposure from, say, extreme value statistics for measured or predicted hourly ground-level concentrations of key air quality indicators close to a given set of emission sources. The simplest case with concentration ratios depending on a power law of the ratio of time periods is an expected form for a wide range of distributions for low probability events. Power-law dependencies are very useful as long as proper consideration is given to source structure, terrain characteristics between source and receptor, pollutant nature, timescales and overlap between adjacent sources. Experimental information for a given source and downwind distance suggests 2-3 regimes of behaviour with differing values of power law exponent p. The influence of source structure is now well documented. Tall stack sources are relatively well parameterised with p H 0.4 holding for time periods from 1-2 minutes to many hours. For surface non-point sources, values of p H 0.1—0.2 are supported by field, laboratory and theoretical considerations.
Recent field results for an elevated buoyant source show a dependence of peak-to-mean ratios (and hence p) on stability, exceedance probability and averaging time. We suggest practical approximations for determining peak-to-mean ratios for averaging times from 1 second to several hours that reflect source structure adequately.
An approach based on universal multifractals suggests a powerful relationship between p and the tail index of an extreme-value probability distribution. The power law exponent is naturally linked to the degree of intermittency, threshold crossing, resolution time and long memory parameter of a given time series. This framework produces consistent results for available wind tunnel information and an indicator of the type of probability distribution for a given source structure and downwind distance.
Prepared by: Peter R. Best, Karen E Lunney and Christine Killip
A broad range of tools is needed to assess and manage urban air. We have shown that detailed statistical analysis of the relatively substantial database of various Australian cities facilitates considerable insights into the spatial and temporal variability of urban air quality, beyond that determined by the usual numerical modelling. It was found that a focus on pollution-conducive synoptic types allows a much easier determination of air quality trends. Ozone (as well as fine particulate) levels are significantly increased if bushfire activity coincides with adverse meteorological conditions.
Prepared by: Karen E. Lunney, Peter R. Best, Chris B. Watson and Damien J. Cresta.