Barac v Strathbogie SC [2011] VCAT 1146

Barac v Strathbogie SC [2011] VCAT 1146 | Case summary | Zoe

[7]

In April 2009, the applicants applied for two 320,000-bird broiler farms on two lots abutting an existing cluster of four broiler farms each growing 320,000 chickens or 1.28 million birds in total. A 320,000-bird broiler farm required no Odour Environmental Risk Assessment ('Odour ERA') under the 2001 Broiler Code which applied at the time the application was lodged.
On 21 September 2009, the Broiler Code 2009 ('Code') was gazetted as an incorporated document into all Planning Schemes. The application has since been amended to two broiler farms of 400,000 birds each and sheds relocated to comply with new 2009 Code setback measures[3]. In April 2010, the applicants lodged this review to the Tribunal against the failure of the Responsible Authority to make a decision on the application within statutory time limits.
Two broiler farms in the existing four-farm cluster had also sought to increase their bird numbers and sheds and these other two applications were listed for hearing together with this review. The two other broiler farms have since withdrawn their applications for expansion.
Broiler farms are known to emit odours that can be offensive to people. Aside from odours emitted during the ventilation of the broiler sheds, other normal operations on a broiler farm can include the composting of dead birds and spent litter on site and the opening of shed doors during litter removal at the end of each growing cycle. Such operations can result in odour emissions, depending upon wind dispersion conditions.
Two neighbours oppose the proposed broiler farms: Mr Perry who owns land abutting lot 55A and Mr Robinson whose dwelling is 3.7 km to the south, but who owns and farms a number of properties in the area.
Apart from the impact of odours, Mr Perry fears that chemical contamination of a watercourse that passes through his and the subject property will contaminate his grazing cattle and adversely affect his overseas cattle market. His cattle need to be chemical free[4].
To prevent heavy transport frightening Mr Perry's cattle, which was also of concern, the Baracs agreed to re-route the access driveway to the southern most broiler so it would be 50 metres away from Mr Perry's cattle yards.
At over 100 hectares, Mr Perry's combined abutting lots can contain a future dwelling. In distinguishing a single Class B farm[5] from a farm cluster, Measure E1 M5.1 of the Code aims for a single Class B farm to not utilise more than 50% of the neighbour's farm provided that 50% is suitable for a dwelling. Compliance with a Measure is deemed to be compliance with the Objective and Standard[6]. No approved setback or buffer area Measures exist for broiler farm clusters which must be assessed under an Odour ERA.
Mr Robinson, who grows hay milled into chaff for the horse feed industry, already experiences odours from existing industries, particularly the mushroom composting facility. Mr Robinson seeks the incorporation of expensive, cutting edge, state of the art odour reducing technologies into the broiler sheds to reduce the risks of odour emission.
The EPA was notified of the application and objects to the grant of a permit. The EPA has commissioned its own Odour ERA which it uses to support its grounds of objection, which essentially distil to the issue that the new farms would add to existing adverse odour conditions.
The Responsible Authority advised the Tribunal that had it been able to make a decision within time limits, it would have refused a permit for similar reasons to the EPA.
The key issue for determination is the acceptability of the risk of cumulative odour emanating from the addition of the two 400,000-bird broiler farms to an existing cluster of four broiler farms. The risk is measured to sensitive receptors who claim they already experience adverse odour impacts from the existing cluster of four broiler farms as well as from other odour sources, such as the mushroom compost farm, waste compost operations and wastewater treatment plant. Expressed differently, the key issue is whether it is reasonable to increase odour emission in this already impacted neighbourhood.
Subsidiary issues are:
1. the different modelling inputs and risk matrices used by the experts in arriving at their odour risk assessment;
2. whether both experts' sensitive receptor no. 1 at 464 Ballantynes Road Nagambie is a dwelling rather than a 'vacant dwelling';
3. whether Dr Cowan's sensitive receptor no. 2 on Ballantynes Road near the corner of McDonalds Road Nagambie is legitimate because it is used in part as a residence despite the fact that it is a shearing shed;
4. encroachment of odour separation distances onto neighbouring properties limiting the potential future use and development of those properties.
The Tribunal must decide whether a permit should be granted in the context of these issues and, if so, what conditions should be applied. Having considered all submissions and evidence presented with regard to the applicable policies and provisions of the Strathbogie Planning Scheme, we have decided not to support the grant of a permit. Our reasons for this decision follow.

[14]

We are told that each grower[11] rears an average of approximately five and a half batches a year. The chickens are grown to an age of five to eight weeks. They are collected for processing at intervals within this period, depending on their weight and the requirements of the chicken meat processor.
The shed floor is covered with litter, a bedding material, usually rice hulls and/or wood shavings. Feed is provided automatically via an auger system in regularly spaced pans, and water is most commonly supplied through a nipple drinking system. Following the removal of a complete batch of chickens, the litter bedding is removed, sheds are cleaned and disinfected, and fresh litter is applied to decrease the risk of disease, and to prepare for the following batch of chickens.
Odour is produced by the aerobic and anaerobic decomposition of manure, spilt feed and other organic matter, and also from the birds and bird respiration. High moisture content in the litter furthers anaerobic digestion, which results in stronger, more offensive, odours. Generally, the greater the frequency, intensity, duration and offensiveness of an odour, the more likely it is to cause annoyance and lead to complaints[12].
Spent litter disposal and dead bird disposal are, inter alia, two matters influencing odour emission. Spent litter and dead birds can both be composted on site and used as fertiliser on site or on nearby farms[13]. We understand that the existing sheds are cleaned under contract with the contractor undertaking some composting of the litter on nearby land. The composted material is sold as a fertiliser. One farm currently supplies its litter and dead birds for use as fertiliser on a farm in another district.
A number of inter-related factors can influence odour emissions from broiler farms including the number of birds, broiler farm management and operation, disease and digestive upsets in the birds, feed formulation, the amount of faecal material in the litter, the moisture content of the litter, the shedding, equipment and other technologies (for example, odour reduction technologies) employed and waste management practices.
In addition, the likelihood of odour emissions impacting nearby sensitive uses is influenced by:
1. local meteorological conditions (wind, temperature);
2. topographical features (terrain, vegetation) that govern the transport and dispersion of odorous emissions;
3. the distance of the receptor (sensitive use) from the odour source(s); and
4. the nature and sensitivity of the receptor.
Odour emissions primarily originate from the broiler sheds, although some generate from inappropriately sited or managed temporary litter stockpiles, compost piles, or from the spreading of litter.

[16]

There is very strong policy support for broiler farms in this Farming Zone subject to their compliance with the Code and consistent with orderly and proper planning.
At the level of state policy, clause 13.04-2 seeks suitable separation between land uses that reduce the amenity of sensitive land uses. Clause 14.01-1 seeks to protect productive farmland whereas clause 14.01-2 promotes the establishment and expansion of poultry farms consistent with orderly and proper planning and protection of the environment.
At the level of local policy, clauses 21.01 and 21.03-9 strongly support agribusinesses and intensive farming.
A large area of the northern sector of the Shire particularly to the east and north east of Nagambie is targeted as particularly suited to enterprises requiring buffer distances, such as broiler farms. This is because the area is accessible to employment sources, is well served by access to arterial roads, and contains large property holdings and a limited number of residences. The area is able to support uses requiring buffer distances because it can meet the requirements for extensive open areas with no dwellings[16].
A planned precinct, currently known as the Strathbogie Special Use Precinct, is nominated under local policy for investigation by the Shire to determine its suitability for intensive forms of agriculture and horticulture. The proposed area already contains agribusinesses and uses which require buffers that need to be protected from the encroachment of dwellings. Under the proposed Strathbogie Special Use Precinct, the further development of this area for rural dwellings (except Caretakers houses) is to be discouraged while encouraging intensive agricultural and rural industry uses that require substantial buffers.
Council was going to investigate this Special Use Precinct[17] over a 437km2 triangular area in the northern sector of the Shire for uses requiring buffer distances, but has now resolved to abandon this investigation. The subject land is within this now abandoned investigation area. It remains to be seen whether a review of the Municipal Strategic Statement will continue to include an investigation area.
Under the heading "Economic Development" at Clause 21.01 council notes:

[18]

The 2009 Code sets out that separation distances are designed to provide sufficient space to minimise the risk of offensive odour and dust emissions under both routine and upset conditions adversely affecting the amenity of existing sensitive uses[18]. Separation distance requirements assume that the design and ongoing management of broiler farms employ best practice[19].
After the application was lodged in 2009 the Code[20] was amended[21] to move the measurement of separation distances from the centroid of sheds to the outer edge of sheds and to deal with 'farm clusters'. Consequently, substituted plans relocated Unit 2's broiler farm to the east as well as relocating stormwater retention basins and both dwellings.
All broiler farms (be it single farms or farm clusters) must meet the objectives and standards of six best practice Farm Design and Operation Elements set out in the Code[22]. The Code sets objectives, standards and approved measures for each element.
Under the Code this application is characterised as a Farm Cluster because the minimum separation distance requirement overlaps with the minimum separation distance requirement of existing broiler farms and the combined farm capacity of the broiler farms with overlapping minimum separation distances (as defined by Formula 1 of the Code) is greater than 400,000 birds. Consequently, an Odour Environmental Risk Assessment ('Odour ERA') must be undertaken[23].
An Odour ERA is a staged process of modelling and analysing odour emissions from broiler farms that enables a responsible authority to assess the acceptability of the risk of offensive odour adversely impacting beyond the broiler farm boundary. An Odour ERA will not determine an acceptable level of risk. The responsible authority and the Tribunal on review needs to decide on the acceptability of the risk.
An Odour ERA for Farm Clusters must assess cumulative odour emissions from all the broiler farms within the cluster. We have more to say about the Odour ERA later in these reasons.

[20]

Understanding what a sensitive receptor is, is essential in assessing risk. There is disagreement between the parties about Receptor 1 (R1), which is a farmhouse (dwelling) on the property of Mrs Tranter. Mr Peake submitted that Mrs Tranter's farmhouse has been vacant for more than two years after her tenant, Mr Ray Boyer, shifted out in 2009. Because of this vacancy period and because it is located on a lot of less than 100 hectares, Mr Peake submitted that her dwelling had lost its existing use rights and is not a sensitive receptor.
We note that a dwelling that cannot be built as-of-right under section 1 of the Farming Zone because it occupies a lot of less than 100 hectares becomes a section 2 use subject to the grant of a permit. It does not become a prohibited use.
Nevertheless, when it became apparent during the hearing that the status of the Tranter farmhouse as a sensitive receptor was in issue, Mrs Tranter came to Melbourne to give evidence and be cross-examined.
Mrs Tranter of 1200 Kilmore Road Riddell's Creek grazes cattle, sheep and grows crops there and at the Ballantynes Road property. She explained she had not objected to this planning application because she had not received notice of it. But Mrs Tranter could not emphatically say notice was not served on her at the address registered with Strathbogie council for service of notices, because she did not empty the Post Office Box to which all mail was sent. It was her daughter's job to empty the Post Office Box as part of her administrative office duties.
As for the issue of vacancy or abandonment of her farmhouse at 464 Ballantynes Road, Ms Tranter gave the following evidence. She owns Talana where she grazes cattle, sheep and grows crops. She regularly occupies the dwelling on a periodic basis. She commutes back and forth from Riddell's Creek and stays overnight. She works a day, stays overnight, works the next day and then goes home. She estimates that she occupies the property on each alternate week.
She stays according to need, at alternating times during the week or weekend. No one knows when she is there. Her sojourns are intermittent but necessary in order to run the farm.
Because Talana is not a fully furnished home, she brings her own bedding and a portable wardrobe and she uses the farmhouse rooms and facilities. The internal utilities are still connected and used for shower, toilet, water, cooking, tea and coffee making facilities.
No evidence to the contrary was presented to the Tribunal. Therefore, we are satisfied that the Tranter farmhouse is a dwelling that is neither vacant nor abandoned, following Nunawading CC v Harrington[25] because Mrs Tranter still uses it to live in when attending to her Ballantynes Road farm. The fact that it may not be her principal place of residence does not prevent its use as a dwelling.

[23]

We now address an issue raised by Mr Peake on behalf of the Broiler Industry. Mr Peake submitted that there exists an anomaly between the odour impact outcomes deemed acceptable for a Class B farm and for clusters where an Odour ERA is required in accordance with the SEPP (AQM). The anomaly is that a single farm separation distance (in this case 686 metres for a 400,000 bird farm) is not expected to meet the SEPP (AQM) design criterion of 5OU[26] (99.9 percentile, 3 minute averaging time) at the farm boundary. So, for example, two new 400,000-bird broiler farms could establish on land without an Odour ERA, albeit subject to planning permission, provided they are not less than 1,372 metres apart from each other and 686 metres from the nearest resident.
Mr Peake submits however that under the Code, a cluster farm ERA requires the farms to meet the SEPP (AQM) 5 OU criterion at the farm boundary.
Our reading of the Broiler Code 2009 together with the SEPP (AQM) is that neither single farms nor clusters are required to meet the SEPP (AQM) design criteria of 5OU (99.9 percentile, 3 minute averaging time) at the farm boundary. Although odour contours for cluster farms must be modelled at a design criterion of 5OU (99.9 percentile, 3 minute averaging time) at the farm boundary in accord with modelling practices under the SEPP (AQM), if that design criterion is not met, a risk assessment must be undertaken. The Code, however, does not require a risk assessment of cluster farms to meet that design criterion.
The Code requires a risk assessment to take into account the existing and likely future use and development of the land including any approved sensitive use, the purpose and decision guidelines of the zone applying to the land and any other relevant decision guideline in the planning scheme and any strategies and policies in the planning scheme which may affect the use and development of land. Thus the Code goes beyond mere odour criteria.
That single farms which meet their separation distances do not require an Odour ERA is clearly a policy decision taken at State Government level balancing the needs of the community for meat chickens and the operational requirements of the broiler industry against the legitimate expectations of residents surrounding broiler farms[27]. Lemic v Colac-Otway SC[28] recognised that if the 5OU limit were required at the farm boundary, it would require vast tracts of land to comply together with the concomitant expense of infrastructure and road transport costs.
However, farm clusters are a different matter. Their clustering as well as their cumulative impact must be modelled and their impacts analysed by way of a risk assessment, although no guidelines are provided as to what measure of risk is acceptable.
In support of farm clustering, Mr Peake submits that as a matter of orderly and proper planning it is preferable to co-locate like uses. Mr Peake submits a separation-distance-compliant farm would still cause odour to dwellings because it does not meet the design criterion of 5OU (3 minute averaging time, 99.9 percentile) at the farm boundary. From this, Mr Peake submits the intention of an Odour ERA for cluster farms is not that there will be no impact - but not to create a worse case than a single Class B farm. That is to say, the risk assessment should only assess the incremental impact that these two farms make over and above that made by the four existing farms. If that increment is not worse than that emitted by a single 400,000-bird Class B farm located 686 metres from a sensitive receptor, then both farms should be allowed to establish. The question we have to ask ourselves is whether that is the correct test or assessment to apply when the test is that set out in Section 6 of the Code.

[28]

Mr Peake throws down the challenge of determining whether it is better to establish a series of farms that each meet separation distances but disperse odour more widely in the community compared with co-locating broiler farms where impact will occur in a more confined manner.
It is clear enough from a reading of section 6 of the Code that the purpose of the Odour ERA is to allow a responsible authority to assess and understand the level of environmental risks on a case-by-case basis. The Odour ERA is not intended to determine separation distances nor describe an acceptable level or risk. Thus, whether one is better than the other, as suggested by Mr Peake's challenge, is not for us to say. It is clearly a case of each application being assessed on its merits, whether by the Tribunal, a responsible authority or by a potential applicant in formulating his or her proposal.
The Baracs submit the Odour ERA by their air quality scientist Mr Ormerod demonstrates the two farms will not have a significant and adverse odour impact on nearby sensitive receptors. Based on EPA's air quality scientist Dr Cowan's odour ERA, the EPA submit the contrary.
Dr Cowan and Mr Ormerod have produced tables of risks associated with the levels of odour emissions projected to occur at specified receptors, which they have derived from their respective models.
We have considered both the manner in which these risk levels have been derived and the projected emissions by each expert. For the reasons that we explain, we cannot adopt one approach over the other. Instead, we find it appropriate to draw upon each expert's evidence to assess the level of odour impact risk. In doing so we have formulated what we consider to be the level of acceptable or unacceptable odour impact.
The requirements of the Code are for a staged process of odour risk assessment. Stage 1 is for air dispersion modelling to be undertaken in accordance with the SEPP (AQM). Stage 2 is for an assessment of the risk of odour impacts predicted to occur by the air dispersion modelling. Stage 2 need only be undertaken if the modelling indicates that the SEPP (AQM) design criterion is exceeded at the broiler farm boundary.
Stage 3 of the Odour ERA is required to address any unacceptable levels of risk identified in Stage 2. Stage 3 requires further information from the applicant that would satisfy a responsible authority that:
1. the sources resulting in the identified unacceptable levels of risk can be managed; or
2. by amending the application, the risk can be acceptably reduced.

[30]

The Code directs that the air dispersion modelling be completed in accordance with the SEPP (AQM). This should be read as being a reference to Schedule C of the SEPP (AQM) (modelling emissions to air) and more specifically Parts A and B of that Schedule. Part A applies when a proponent is uncertain of the modelling requirements and a modelling proposal is to be put forward to the EPA to ascertain the suitability of the proposed approach.
Part B of Schedule C sets out the modelling procedures. These procedures are most relevant in this application and we will consider them later in these reasons.
However, we do not consider that the direction to undertake the modelling in accordance with the SEPP (AQM) includes undertaking the risk assessment under Part C of schedule C. The assessment of the modelled emissions and their impact is instead to be taken from the Code itself and the matters set out under Section 6 as we have set out earlier.
Clause 52.29 reminds us that it is the Code that must be met and not the SEPP (AQM).
Having completed the air emission modelling, Stage 1 of the Code directs that the predicted impact of the odour emissions be compared against the SEPP (AQM) design criteria at and beyond the boundary of the broiler farm. If this design criterion is met, the Code says the responsible authority should accept that the risk of odour amenity impacts is low and no further assessment is required.
The design criteria contained in Schedule A of the SEPP (AQM) is in fact an integrated set of criteria that can be applied to intensive animal husbandry sites[30]. The integrated set of criteria includes:
1. The site's location in a low density of sensitive land uses (i.e. rural zone);
2. The site's location in an area where the long term planning strategy is to maintain appropriate separation from sensitive uses;
3. The farm is designed in accord with an approved set of 'industry standards', which in this case we take to mean design accords with best practice for broiler farms and hence the Broiler Code;
4. Operation of the site is in accord with an approved EMP; and
5. A risk assessment that includes a model that shows the predicted maximum odour level does not exceed five times the odour detection threshold for the 99.9 percentile 3-minute averaging time at and beyond the boundary.
We accept that the reference to the design criteria under Stage 1 of the Code is more specifically a reference to 5 OU (as measured for the 99.9 percentile over the 3 minute averaging time) as predicted by the emissions model. This is self evident from the manner by which the reference is made within the Code to the 'design criteria at and beyond the boundary'. It is also self evident that a location in a rural zone and a farm design that is in compliance with the Broiler Code is, save for the issue of 5 OU concentration at the boundary, in general compliance with the other integrated criteria under the SEPP Schedule.
While the reference to the design criteria in the Code is somewhat obtuse and requires interpretation, we also acknowledge that to be specific about the 5 OU locks in a criterion which may be subject to revision under the SEPP. The manner is awkward but builds in the adaptability of any revision to the SEPP (AQM) being incorporated automatically into the Code.

[34]

We are persuaded however that the level of impact attributed to odour concentrations put forward by Mr Demetriou is too low. Mr Demetriou considers that an odour concentration of around 5 OU would be perceptible but not recognisable, while an odour between 7 OU - 10 OU would be recognisable and therefore of potential impact to amenity.
Mr Ormerod agrees that a concentration of 5-10 OU would be detectable above most background odours but this would not lead to annoyance or adverse impact on amenity. He suggests that an odour level of 10-20 OU would be recognisable above background, but would need to be greater than 40 OU to be annoying.
Dr Cowan has approached this task based on his own judgement with guidance provided from the Department of Environmental Protection in Western Australia though its odour assessment guidelines. He suggest the following ranges:
- 0-5 OU may be detectable but not recognisable;
- 5-10 OU is detectable and recognisable;
- 10-20 OU is distinct;
- 20-30 OU is a strong odour; and
- 40 OU or more would be a very strong odour.
Much of the verbal evidence of the experts focussed on what level of odour would generate complaints; thereby indicating a level of unacceptable impact on amenity.
Given the evidence about the complaints that we set out later in these reasons, we do not consider that a level that generates complaints is the proper test of an adverse impact on amenity. We are of the view that it is likely that a level of annoyance that adversely impacts on amenity will be reached at a lower threshold of odour concentration.
We are persuaded that an odour level of 5 OU to 10 OU is one that would be perceptible in the outdoor environment but not necessarily so distinguishable or recognisable as to be annoying. Similarly, we do correlate recognition of an odour with annoyance. However, at a concentration when an odour is distinct we do not consider it to be unacceptable. We are persuaded that an odour needs to be at a strong level in order to generate annoyance and hence an adverse impact on amenity. The evidence suggests that this may occur at odour levels of more than 20OU and will definitely occur at concentrations of more than 40OU. In this respect we have regard to the relationship between odour concentrations and intensity that is set out in Odour Methodology Guideline[38] which uses the Weber-Fechner Law for a logarithmic relationship between odour concentration and intensity.
Thus the Tribunal concludes that in the assessment of impacts the following bands should apply:
- 0-5 OU will be so unlikely to cause annoyance as to be discounted, as this would represent an odour of imperceptible to very weak odour;
- 5-10 OU may be detectable and recognisable but very unlikely to lead to annoyance and adverse impact on amenity, representing a weak to distinct odour;
- 10-20 OU is likely to be perceived as a distinct to strong odour and may be possible but not very likely to have an impact on amenity;
- 20-40 OU is a strong to very strong odour that is likely to have an adverse impact on amenity if it occurs often enough; and
- 40 OU or more would be a very strong odour that if it were to occur would be an adverse impact on amenity.

[40]

Dr Cowan and Mr Ormerod have each produced tables of the frequency of certain risked based odour events based on their respective odour emission models. Following revelation of errors in the initial input of broiler farm shed coordinates in Dr Cowan's model, his assessment was corrected, albeit in a more limited fashion due to time constraints. Nevertheless, we have the benefit of each expert's assessment of risk under present and proposed broiler farming levels.
We confine our discussions to the impact on a selected number of receptors, noting however that we have considered the range of risks to all receptors modelled by each of the experts. Those selected receptors are those that lie closest to the broiler farm cluster and hence are most likely to be impacted. We reference these receptors on the basis of the identifiers adopted in Dr Cowan's evidence. Our selected receptors are:
- Receptor 1 (R1) - the Tranter farmhouse to the south, which we consider to be a sensitive receptor for reasons we have set out. Mr Ormerod also designates this dwelling as R1;
- Receptor 2 (R2) lying to the south-west of the cluster just west of the corner of Ballantynes and McDonalds Roads. Mr Ormerod has not identified this receptor as he was not advised of its location. While we concur with Mr Peake that its legitimacy is questionable, we have nevertheless used this location as a useful indicator of the extent of impact that may extend west of the cluster towards Nagambie.
- Receptor 6 (R6) which lies to the north/north-east of the cluster on Wormangal-Wahring Road, which Mr Ormerod identifies as his R5; and
- Receptor 10 (R10) which lies to the west/north-west of the cluster on Racecourse Road. Mr Ormerod identifies this dwelling as R2.
From Dr Cowan's corrected odour modelling[42] we note that whereas the 99.9 percentile 3 minute averaging odour concentrations at R1 increase by 33% from 33 OU to 44 OU due to the two new farms, odour concentrations at the remaining receptors increase between 1 OU to 5 OU. The latter increases represent only a slight increase in sensory terms, i.e. a noticeable but unidentifiable odour increase in intensity. Analysis of modelling in this way directs attention toward the impacts to the Tranter property.
By applying our risk assessment matrix to the outputs of each of the experts' emissions modelling we have compared the level of risk for each model to each receptor. The outcomes are as follows.

[47]

As has been set out, the Broiler Code directs that the air dispersion modelling is to be completed in accordance with the SEPP (AQM). We have considered the modelling of both experts against the requirements of Part B of Schedule C. We accept that both models have been generally completed in accord with these requirements. However, it is apparent from the above risk analysis of each model's outputs that there are significant differences in the projected odour emissions.
In order to resolve this significant difference we have considered the approach adopted by each expert and other outputs from their modelling.
In doing so we have compared the odour contours produced for the 99.9 percentile 3 minute average, i.e. the 9th highest predicted odour concentrations. The plotting of values at the 99.9 percentile, 3 minute average is a required output under the SEPP (AQM) Modelling procedures. These values represent the ninth highest odour levels for the years(s) modelled. In other words, 8,751 times per year, the odour concentration will be lower.
Dr Cowan's corrected modelling indicates that under 2008 meteorological conditions, the 10 OU's contour would extend through the town of Nagambie, while the 20 OU's contour would lie approximately halfway between the western edge of the cluster and the town. The 20 OU contour would also extend beyond the dwelling of Receptor 8, which lies to the north of the cluster.
Dr Cowan's corrected modelling shows that whereas the Tranter farm R1 currently lies on his 40 OU contour, when both proposed farms are added, it will lie on his 50 OU contour.
In comparison, Mr Ormerod's 99.9 percentile, 3 minute average 5 OU contour passes through the town of Nagambie, his 10 OU contour lies approximately halfway between the cluster and the town and extends just to the north of where receptor 8 would be located. Whereas the Tranter farm currently lies between his 10 OU and 20 OU odour contours, when both proposed farms are added it will lie on the 20 OU contour.
In order of magnitude terms, Dr Cowan's modelling shows twice the level of impact compared to Mr Ormerod's.
Dr Cowan's base case modelling (i.e. modelling of existing conditions) displays a similar trend of predicting almost twice the level of impact over Mr Ormerod's.
We therefore question why such a difference should occur between the model outputs and which is to be adopted in assisting the consideration of the risk assessment outcomes we have set out earlier.
In considering this question, we have identified that there is a significant difference in a key model input, as is submitted by Mr Ormerod in his Response to Requested Information ('Response') dated 14 June 2011. That input is the time varying odour emission rate from each shed.
Dr Cowan has prepared an odour emission rate algorithm based on data taken from a RIRDC[43] study of odour emission control technologies.[44] We do not repeat all that Dr Cowan has completed in this analysis and the formulation of this algorithm. He has set out in his evidence that to derive a more realistic approach to odour emissions from broiler sheds, he has reviewed data from this RIRDC study, taking account of such factors as broiler age, shed air flow requirements, ambient air temperature and shed temperature. The end result has been to derive a relationship between ventilation rates and odour emission rates that provide hourly inputs to his emission model.
His figure C-5 shows a plot of odour emission rate for a 35,000 bird shed over an eight week growth cycle. We have compared this plot of odour emission rates to that of Mr Ormerod's plot of the same odour emissions rates over a shed growing cycle (figure 3-3), although we note the latter is for a 50,000 bird shed. Notwithstanding the difference in bird numbers, we find Dr Cowan's calculated odour emission rates are comparable to Mr Ormerod's for only the first three weeks of the growth cycle. Thereafter, while peak daily emission rates are approximately comparable, the baseline or lower range of daily emissions estimated by Dr Cowan are 50% to 100% higher. This notwithstanding Dr Cowan standardising his calculations to a lesser number of birds, which should result in slightly lower baseline odour emission rates compared to Mr Ormerod's 50,000-bird shed.
Mr Ormerod's Response of 14 June 2011 points out that Dr Cowan's analysis of the RIRDC data assumed it was representative of 30,000 birds per shed - whereas three of the four farms assessed in the RIRDC Report had between 38,000 - 46,000 birds per shed. This would lead to an overestimation of odour emission rates if it were not adjusted to represent 30,000 birds.
The difference in odour emission rates is likely to be a major factor in the discrepancy between the two modelled projections of odour units beyond the farm boundary.
Another factor in the discrepancy will be the fact that Dr Cowan used Calpuff while Mr Ormerod used the Ausplume model. Both experts however agree that in broiler farm cases such as this, the Calpuff model is more appropriate. Both experts agree that Ausplume is more conservative as it does not account for buoyancy effects, thus over-predicting the impacts closer to the source.
Dr Cowan was also critical of the mixing heights and concurrent cycling of all shed batches adopted in Mr Ormerod's Ausplume model. He contended that there were significant components of the modelling.
It was Mr Ormerod's evidence that sensitivity testing of some of these components shows that they would not significantly influence the model outputs.
The evidence of Dr Cowan also demonstrates the influence that the odour emission rates have on the model outputs as demonstrated by comparisons of the EPA dataset model results with the RIRDC dataset results.
Both experts agree that the EPA dataset is too conservative and that the use of datasets that account for the variations in odour emissions over the growth cycle is more appropriate. In this respect, Dr Cowan and Mr Ormerod use different datasets, which has likely led to a difference in odour emission rates as discussed above.
It follows from the above that this doubling in the OU contour lines beyond the farm boundary carries over into the risk assessment, with the consequential differences in the level or risks we have observed.

[49]

Dr Cowan's evidence is that any model of emissions should be tested against ground truthing, where available. Mr Ormerod more or less agrees although he cautions about the use of complaint records alone, as there may be various community, perception, political and psychological factors involved in making complaints. Indeed both experts agree that community members are generally hesitant to complain. They both caution that any complaint record should be taken to indicate the worst cases of impact, unless dealing with a sensitised community. Mr Ormerod relies on overseas studies and his own work to identify that complaint making will be influenced by the community's attitude to the source, which in turn can be linked to a community's economic dependence on the source, and perception of benefits or adversity deriving from the source of the odour.[45]
Two sets of complaints records have been tabled in this proceeding, a set of records from the Council and another from the EPA. Odour logs recorded by community members who were contacted by the Council and EPA after making complaints have also been included in the Council records. These logs required the community members to record any instance of odour they detected. We are cautious to assign all of the logs as complaints in the normal sense, as odour events recorded in the logs may not have given rise to actual complaints in the normal circumstance.
Dr Cowan has compared the number of complaints recorded that are or have been attributed to the broiler farms, to his projection of likely unacceptable level of risks projected (albeit for 2008 conditions) under his model for the existing conditions. His evidence is that the projected number of 110 unacceptable events is comparable in magnitude to the 68 complaints in 2010 (this being the only full year of broiler odour complaints under existing operations). He acknowledges that the 2009 figures are due to the mushroom composting operations and illegal composting at the Nagambie Goldmine and are thus not a reliable comparison to modelled broiler farm impacts.
We have considered such material along with our own assessment of the records. We draw the following information from the two sets of complaint records:
1. In 2009, odour complaints received by the Council and EPA are attributable to the Chiquita mushroom compost production site and chicken litter stockpiling / composting on the Nagambie Goldmine site.
2. In or around 2010, complaints were being received by the Council about odour thought by the complainants to be from the broiler farms.
3. A peak for complaints occurred over 2009, largely attributable to mushroom composting operations and chicken litter stockpiling / composting on the Nagambie Goldmine site together with an increase in complaints about broiler farms odours.
4. The volume of complaints has decreased since 2009. Complaint records continue to show that the reporters identify the broiler farms as the source of odour but limited EPA field surveys have not confirmed this to be the case.
5. Inspection of the broiler farms has identified some unacceptable composting practices on at least two of the farms, with one the subject of a clean-up notice from the EPA.
6. Unauthorised composting and storage of chicken litter and manure has and continues to occur in the general area of the broiler farms. These activities continue to be a source of possible odour impact to the community, especially those reporting odour impacts that are located well away from and sometimes upwind of the broiler farms.
7. A reasonable proportion of the complaints recorded by the Council identify complainant locations to be upwind rather than downwind of the broiler farms yet the complainants identify the broiler farms as the source of poultry odours.
8. A small number of complaints to the south of the broiler farm cluster in fact seem to correlate with poultry odours being wind driven from the breeder farms.
It is apparent from our review of the complaints records and other evidence produced at the hearing that other activities are occurring in the area, some of which are unauthorised, that have or could continue to result in odour emissions characteristic of broiler farms. These include the use of chicken manure and chicken litter in the mushroom farm-composting complex, stockpiling and composting of chicken litter and manure well to the south as well as to the north-east of the broiler farms and composting of dead chickens in similar locations. One Council complainant located well to the south, nearer to Mangalore, identified a possible source as being another poultry farm further to the south. While not expected to be a source of odours, we do not discount that from time to time the breeder farms may also be a source of 'poultry' odour.
Thus while we do not discount reports that poultry and broiler farm odours are being experienced, we are mindful of the fact that very few events have been specifically identified as emanating from the existing farm cluster except under certain specific conditions (e.g. shed clean out) or because of specific events such as the poorly managed composting of dead chickens.
Further, we are conscious of the fact that some reports describe odour character, specifically an ammonia smell, which the experts agree would be very unlikely to carry a long distance from any broiler shed.
We are also puzzled about some of the reports of poultry odours from locations to the north and northeast of the present farms at times when reported wind directions are from the northerly sectors.[46] We do not fully discount that there may be local wind direction variations from that being recorded at the Mangalore meteorological station (to the south). However, we accept the evidence of the experts that there is little by way of topography or other factors to generate major differences in wind direction from that recorded at this meteorological station and the subject area. It is our view that perhaps not all odour sources have been identified in this area.
We note that one complaint record followed up by EPA identifies the storage of chicken litter to the northeast of the cluster on farm land following the cessation of such activity on the Nagambie Goldmine site.
That said we find that there are a number of broiler farm odour events that have led to complaints that correlate well with the wind directions for the odour source to be the cluster of farms. These have occurred at periods that both experts agree represent the higher risk periods of late in the day or early morning. In many such instances we have identified that more than one community member has been affected at the time of such events giving further strength to the conclusion that the cluster of farms was the source of these events.[47]
We consider that the records indicate that from time to time there are odour emissions from the broiler farms that are noticeable and sufficiently strong enough to generate complaints. We consider such reports to be an indicator of adverse amenity impact but that not all adverse impacts necessarily mean a complaint will be made. Some community members may well be just bearing the brunt of the occasional odour impact.
That said we do not think it is reasonable to consider that all of the poultry odour complaints recorded by the Council or EPA should be assigned to the broiler farm complex. There remain a number of questions about the possible sources of these odours in light of our findings about poor correlation with wind direction and the location of the existing broiler farms.
In light of this, we do not consider that Dr Cowan's predicted base line impacts of 110 odour complaint events for 2008 produces an acceptable match to the existing on-ground conditions.
In reaching our conclusions on the reliability of the modelling, we have not relied on the records of complaints alone. It is also important to consider what complaints have not been made. By this, we mean that Dr Cowan's predicted levels of impact on the town of Nagambie, as evidenced by the 99.9 percentile odour contours, would suggest some at least occasional report of complaint from town residents. If his model projections represent the actual current situation, the town, or at least a reasonable proportion of it would be subject to noticeable odours at reasonably frequent intervals. There are, however no complaint records to substantiate such a frequency of impact.
We find that the complaint records confirm that Dr Cowan's assessment has overestimated the degree of odour emissions. This adds weight to preferring Mr Ormerod's modelled emissions outputs over Dr Cowan's.
However we remain concerned that Mr Ormerod's modelling may under-predict the level of impacts. This is perhaps best evidenced by a comparison of his model projections to the north-east of the site, the area most likely to be impacted by prevailing winds, and the record of complaints from sensitive receptors in this sector. Here we have focussed on the records for 2010, as issues at the mushroom compost facility and unlawful composting at the Nagambie Goldmine affect 2009.
Mr Ormerod's modelling predicts that at his receptor 5[48] there would be one odour event above 20 OU and perhaps six odour events between 10 and 20 OU's. A comparison with the 2010 EPA records for complaints in this area indicate well over this amount, with some 19 separate reports being made. Some of the reports coincide with a time when poor composting of dead birds was found to be occurring, around July 2010. Others are not tied to any particular event other than the complainants identify the broiler farms as the likely downwind source.
Some complainants report to the EPA that noticeable but tolerable levels of odour are detected when the sheds are being cleaned out. However, they also report offensive odours at other times that prevent them from being outdoors. Weighing up the evidence about odour levels, we would expect such odour levels to be in excess of 10 OU and likely be in excess of 20 OU.
The poor correlation of Mr Ormerod's model projections with such reports gives us reason to be cautious in adopting his model outright.
Another aspect that gives us some reason to be cautious is the record of some complainants about the impact from clean-out phases of each shed. Mr Ormerod's model does not include this phase as a source of odour.
We conclude therefore that the comparison of complaints data to modelling projections indicates that the likely impact lies somewhere between that of Dr Cowan's model and that of Mr Ormerod's.

[55]

The Broiler Code 2009 requires us to assess the acceptability of the risk of offensive odour adversely impacting the amenity of existing sensitive uses and the orderly and sustainable use and development of land beyond the broiler farm property boundary.
In the broader planning context, our decision must take into account:
1. The extra broiler odour impact on the Tranter farmhouse at Receptor 1;
2. The extra broiler odour impact on neighbours who can erect a dwelling as-of-right on farmland in excess of 100ha to the east (Mr Perry), north and north-east;
3. Additional or cumulative odour impacts to be experienced by all receptors in a context where there are other odour-emitters such as Chiquita mushroom composting, the waste water treatment plant and the Nagambie Goldmine;
4. That the Broiler Code 2009 odour risk assessment assumes best practice farm management although the evidence we have heard indicates this is not always the case in this district;
5. That it is normal and acceptable broiler farming practice under the Code to compost spent litter and dead birds on site and use it a fertiliser on site or on nearby farms[50].
From Dr Cowan's corrected odour modelling[51] we note that whereas the 99.9th percentile 3 minute averaging odour concentrations at the Tranter farm R1 increase by 33% from 33 OU to 44 OU due to the new farms, odour concentrations at the remaining receptors increase between 1 OU to 5 OU.
From Dr Cowan's corrected odour modelling, all our representative receptors R1 to the south, R2 to the west, R6 to the north and R10 to the north-west are at a high risk of experiencing concentrations and frequency of odour that amount to annoying levels, while in addition they are at a high risk of experiencing identifiable odours from the broiler farm cluster.
Dr Cowan's corrected modelling shows that whereas the Tranter farm R1 currently lies on his 40 OU contour, when both proposed farms are added, it will lie on his 50 OU contour. Mr Ormerod's modelling shows that whereas the Tranter farm currently lies between his 10 OU and 20 OU odour contours, when both proposed farms are added, it will lie on the 20 OU contour. We have determined that such predictions are an over-estimation and an under-estimation of what is to be the likely impacts. However, in any event, both models predict what we find to be an unacceptable risk of annoying or otherwise unacceptable level of odour impacting events.
As for the abutting vacant land, even if Dr Cowan's modelling is an over-prediction and Mr Ormerod's is an under-prediction, both models confirm that there is potential to seriously constrain the future use and development of the Perry property and of the two properties to the north and northeast for a dwelling.
When we look at Mr Ormerod's comparison of events that exceed 5 OU at all hours, including sleep hours, there is almost a doubling of such events between existing conditions and future conditions for Receptor 1[52]. Apart from his Receptor R4 (Dr Cowan's R9) that is the most significant increase in odour events of all receptors modelled by Mr Ormerod.
Taking into consideration that Mr Ormerod's model underestimates the impacts, we expect that this doubling of odour events would be the minimum, and with actual conditions on the ground the number of events would be more than double.
In the light of our analysis of Mr Ormerod's risk matrix and odour events in the non-sleep hour and sleep-hour periods, we consider that some of these additional odour events, which contribute to a doubling in numbers, would fall within the strong to annoying odour categories.
Again we note that Mr Ormerod's modelling is likely to be an underestimation, so the 99.9 percentile 3 minute averaging contour is likely to be higher than the 20 OU indicated in his analysis, which pushes the impact on R1 into the strong/annoying odour risk category.
These impacts, together with those identified to Mr Perry and the two properties to the north and north-east, lead us to conclude that the addition of two more farms to the existing cluster will adversely impact on nearby receptors and neighbours.
Considering the risks of these odour impacts in the broader planning context of the area, which includes the presence of other odour emitting rural industries, we conclude that the cumulative impacts are unacceptable.
The Code assumes best practice farm management, which includes composting dead birds and litter on the site. Such activities in this area have led to a series of complaints from neighbouring receptors. Had we approved these two farms, we would have imposed conditions prohibiting these activities.
In determining the acceptability of the risk of cluster broiler farm odour adversely impacting on the amenity of existing sensitive uses and the orderly and sustainable use and development of land beyond the broiler farm property boundary we have come to the conclusion that adding two 400,000-bird farms to the existing four-farm cluster is an unacceptable risk. Accordingly, the decision of the responsible authority will be affirmed.

At a glance

Court

Decision date

Source

Judgment (88 paragraphs)

Parties

Legislation Cited (1)

Cases Cited (1)

Barac v Strathbogie SC [2011] VCAT 1146 (20 June 2011)