Skip to main content.

4 Odour Assessment

4.1 Why assess odour discharges?

The effects of odour emissions are assessed for a wide range of reasons, including complaint investigation and resource consent applications. The amount and type of information required for an assessment depends on the circumstances of the odour discharge and the reason for undertaking the assessment. Usually the aim of the assessment is to determine whether the odour is offensive and objectionable and therefore causing adverse effects on the local community.

Odour assessments can generally be categorised as being needed for one of five reasons:

  1. investigating odour complaints
  2. resource consent applications or consent review to renew an existing activity
  3. resource consent to modify an existing activity
  4. resource consent for a proposed activity
  5. monitoring compliance with resource consent conditions.

4.2 What information is required?

Odour assessments can be conducted based on a combination of approaches and information from a range of sources, including:

  • the odour complaint history
  • experience with the discharge and past compliance
  • community consultation
  • odour diaries, community surveys, and other surveying tools such as field investigations
  • dynamic dilution olfactometry (DDO) measurements
  • odour dispersion modelling
  • site management and contingency plans, and whether the best practicable option (BPO) is being applied
  • process controls and design, including records of emission control improvements undertaken or proposed and engineering risk assessment for system failures
  • analysis of site-specific wind and topographical features
  • experience and knowledge from other sites of a similar nature, scale and location.

Advice on each of these methods is contained in the following sections. Guidance on choosing which ones are most useful for a particular assessment is provided in the assessment decision trees in Appendix 1 and below. The types of assessments are divided into three categories:

  • category 1: existing activity - a resource consent renewal, compliance monitoring or complaint response
  • category 2: modifications to an existing activity
  • category 3: a proposed activity.

Selecting the appropriate assessment method should include consideration of whether the effects are likely to be acute or chronic. The techniques and information sources recommended are generally applicable to either chronic or acute effects. However, the use of dynamic dilution olfactometry (DDO) measurement and dispersion modelling to investigate acute effects is not recommended in most circumstances due to its limitations (see sections 4.10 and 4.11).

Please note that the decision trees in Appendix 1 should only be considered as a guide. Assessment methods may deviate from those outlined on a case-by-case basis depending on the circumstances.

4.2.1 Category 1: existing activity

Category 1 assessments are for a consent renewal for an existing activity with no process changes, or for compliance monitoring. The primary information source for assessing category 1 odours are council experience with the site (compliance monitoring reports, etc) and community feedback, such as from complaints, community groups or consultation, and/or surveys. If significant adverse odour effects are found to be occurring, then changes will be required and the activity will fall into category 2.

Process controls and operations, including emission control improvements and documented management systems, can also be important information for consent renewals. This kind of information may also be required for monitoring and complaint investigation depending on the requirements of any consent held.

Dispersion modelling is not generally recommended as a tool for assessing whether significant adverse effects are occurring from category 1 activities.

4.2.2 Category 2: modifications to an existing activity

Where modifications are proposed to an existing activity, the existing level of odour effect should usually be assessed to provide baseline information. The baseline should be established using the methods for category 1 activities. Odour dispersion modelling should not be used to try to prove the absence of an adverse effect when community data can be collected or is already available to demonstrate the current level of effect. Where modifications such as a plant expansion or installation of odour controls are proposed, the effect of the changes on the potential for odours must be assessed. The assessment of whether the proposal is likely to create odour effects and/or whether mitigation is sufficient should be based on information from other sites, known performance and proven control technology, and/or dispersion modelling. Modelling can be useful as a tool for undertaking a comparative assessment of the significance of various odour sources, or for identifying the degree and range of community impacts so that community surveys and diaries can be designed.

4.2.3 Category 3: a proposed activity

For a new activity in a location where there is presently no activity of the type proposed, the assessment must be based on dispersion modelling results or past experience with the same activity and the proposed controls in other locations. Modelling can be difficult where odour emission-rate data are not readily available, or the use of data from another site cannot be readily transferred or easy to justify. For new activities conservative assumptions are often required in the modelling, and in some circumstances it may not be appropriate to model at all because of the lack of justifiable data for the model. In these cases the assessment should rely on experience from other sites in conjunction with site-specific considerations.

Recommendation 7

Before commencing an odour assessment:

  • determine which assessment tools are the most effective for the specific discharge and situation using the categories above and the decision trees provided in Appendix 1
  • determine whether the potential effects are likely to be caused by continuous/semi-continuous process emissions (chronic effects), or by infrequent emission events (acute effects).

Use the decision trees in Appendix 1 to determine appropriate assessment methods depending on the category into which the activity falls.

Only use odour dispersion modelling for new activities where the predominant odour effect is due to normal process discharges that are continuous or semi-continuous and reliable odour emissions data are available.

Do not use odour dispersion modelling to try to prove the absence of an adverse effect when community data can be collected, or are available to demonstrate the current level of effect.

Do not use dynamic dilution olfactometry (DDO) measurement and dispersion modelling to investigate potential acute effects of odour discharges.

4.3 Complaint investigation and analysis

4.3.1 General principles

Odour complaint data can be a good indicator of the effect of an odour discharge, particularly where there is a relatively dense population. But complaint records do not necessarily indicate the full degree of adverse effects because many people will not complain even if they are very annoyed, and others will give up making complaints if they feel they are not making any difference.

Sometimes complaints are made for vexatious reasons, or by people who are sensitised or have vested interests. These factors can reduce the overall usefulness of the complaint records because they may skew the complaint frequency data compared to other evidence of adverse effects.

Sometimes there is a lower complaint rate than would otherwise be expected because the population exposed to the odour is reduced when people are away from their homes while the odour effects are occurring. They may, for example, be at work.

Odour complaint data may be of less value:

  • in areas with low population density
  • when other sources of odour are present
  • when complaint records cannot be validated with wind data.

Complaints that have been validated during an inspection by a council officer and/or cross-checked against wind direction are extremely useful, regardless of population density or other odour sources.

Councils have a duty under section 35(5)(i) of the RMA to record a summary of all written complaints received concerning alleged breaches of the RMA and the details of how the complaint was dealt with. When a complaint is received, the details should be recorded in a complaint database or log, as discussed in section 4.3.2. If a site inspection was not possible, date, time and location information can be used with the operating status of the alleged source and data on wind conditions at the time (from monitoring records) to help determine whether the complaint was valid.

Complaints should always be recorded even where the complaint cannot be investigated by a site inspection, such as when staff are unavailable outside normal working hours. Complaint incidents can be used to build up a long-term picture of odour effects and provide a measure of the cumulative effects of repeated incidents.

Recommendation 8

Odour complaint data should be given more weight in odour assessments when validated by a council officer or validated against wind direction data at the time of the complaint.

Complaint data alone should not be relied upon to assess the significance of adverse effects, particularly where:

  • there are low population densities
  • there are other similar sources of odour
  • the complaint records cannot be validated against wind conditions and site operations at the time.

4.3.2 Investigations by council officers

Council officers will usually carry out a site investigation in response to an odour complaint. The officer must form an opinion as to whether the odour is having an objectionable or offensive effect on that specific occasion, and determine the cumulative effect of the odour. Officers should consider the legal interpretation of 'objectionable or offensive effect' to determine whether an adverse effect is occurring (section 3.4).

Consistent procedures for odour complaint investigation and reporting are needed to ensure that the data captured are as useful as possible for assessing the effect of the odour. The procedure followed is particularly important if the complaint records are likely to be used as evidence for enforcement action under the RMA.

It is difficult to validate complaints in every circumstance because odour emissions are often highly variable with time. For example, an odour's intensity may lessen or disappear by the time an officer arrives to investigate a complaint. This may be due to varying wind speed or atmospheric stability between the time the odour complaint was received and when the officer makes it to the site.

Measurements of plume width can help identify sections of the community that are likely to be affected and whether complaints could be expected from elsewhere. Plume width assessment can help determine if odour is fluctuating due to plume movement or emission variation. This helps to develop a general understanding of the nature of dispersion from the odour source, and may be used to identify whether there is odour from other sources. Plume width should be assessed by moving at right angles to the wind direction through the anticipated plume of odour.

When investigating a complaint it is important to complete all off-site investigations before going onto the site of the alleged odour source. This prevents an officer from becoming desensitised from exposure to strong odours at the source before investigating the nature of the effects in the receiving environment. The recommended complaint investigation and recording procedure is provided in Table 4.1. There will be circumstances where following each step in the procedure is unnecessary. Officers should use their judgement to decide what is appropriate to the circumstances; for example, when an odour is extremely intense, 30-minute observations may not be required to determine that an adverse effect is occurring. In such a case it is more important to go on site to determine the source of the odour, perhaps with an expectation that it may be able to be stopped. Another reason not to follow the procedure could be where it is clear that multiple complaints relate to one incident and a detailed investigation has already been undertaken.

Note that where an assessment is being carried out as part of routine monitoring and there is no objectionable or offensive odour, it is sufficient to do a traverse across the wind direction and note the direction, time and location of the inspection.

Table 4.1: Complaint investigation and recording procedure

Step Action

Step 1: Receive the complaint

  1. Record the date, time and location of complaint and the complainant's description of the alleged odour event, including strength, duration and character of the odour.

Step 2: Visit the location of the complaint for at least 30 minutes

  1. Record the time of arrival.
  2. Assess and record the strength/intensity, character and duration of the odour using the intensity descriptors in Table 4.2 and character descriptors in Table 4.3 (section 4.3.6).
  3. Record the wind direction and strength, and weather conditions throughout the investigation and how these were determined.
  4. Record the type of impact that the odour has from Table 4.4, considering the location and observations recorded from action 3 above.
  5. Assess the width of the odour plume by moving at right angles to the wind direction, where possible.
  6. Record the time of departure from the complainant's location.

Step 3: If there is an effect from odour and the source is identified

  1. Assess the odour upwind of the suspected source. Where practicable, conduct a 360o sweep around the source to eliminate other possible sources of odour.
  2. Record any observations of recognisable odour at other locations surrounding the alleged source, including times of observations at each location.
  3. Visit the suspected site causing the odour and explain the findings of the investigation to site staff.
  4. Confirm the site operations taking place at the time of the complaint.
  5. Request an explanation for the odour discharge (if appropriate, warn that their statement may be used in evidence).
  6. Record the name(s) of persons spoken to at the site and their comments.
  7. Investigate whether odours are from abnormal or normal operations and record evidence to support the conclusions made.

Step 4: Make overall assessment

  1. Make an overall assessment of adverse effects beyond the boundary, as illustrated in Figure 4.2.

When making any contact with site operators/owners, if a council has decided on prosecution, the officer must inform the operator that their statement may be used in evidence.

Where there is an obvious shift in wind direction it may be more appropriate to go to the current position of the plume. If the wind is fluctuating, remain at the complainant's location and carry out the 30-minute intensity assessment to obtain a picture of frequency and intensity as the complainant experiences it.

Council officers should always provide a copy of their report to the site management of the alleged odour source. This allows site management to check the details of the report, note the problem and make any response necessary. Complainants often want to remain confidential, so this needs to be considered when passing information to site management. Complainants should be encouraged to be identified and reassured that complaints are a means of gathering information that can help to diagnose a problem on site.

Some councils have adopted an approach of carrying out proactive investigations. This involves visiting a site at times when odours are likely to occur, and is based on previous complaint records, weather conditions and/or time of day when odour effects are more likely to occur. This may be particularly useful in situations where the officer has been having difficulty validating complaints due to response time after a complaint is logged. Usually validation problems are due to changing weather conditions or short-duration odour events. The approach is also useful for determining whether complainants may be being vexatious. Some examples of proactive monitoring are given in Case Study 1.

Case Study 1: Proactive monitoring

In one case where the council was having difficulty validating complaints, proactive monitoring gave the council confidence to go to the parties with a case that they did not consider the odour was causing an objectionable effect. A review of this case by the Parliamentary Commissioner, the local MP and the ombudsman resulted in the council's position being accepted and no further action was needed, saving resources in the long run.

In another case proactive monitoring allowed the council to quickly confirm that there was a legitimate problem and the council was able to convince the discharger to take action to resolve the problem, again saving time and resources.

Recommendation 9

Council officers should follow the complaint investigation and reporting procedure specified in Table 4.1 to investigate reported odour complaints.

Officers should complete all beyond-the-boundary observations before going on to the site of the odour source.

4.3.3 Independent assessors

In some circumstances independent assessors may be used instead of council officers to investigate complaints. An independent assessor may be used, for example, when the travel distance to the odour complaint is too far for the officer, or where council staff are not trusted, as can happen if a council officer does not agree with the position of one of the parties. In certain situations it may be appropriate for a Justice of the Peace (JP), who has the trust of both the community and company, to be used. They can also be viewed as being entirely independent of the situation, and this often brings about a resolution that is accepted by all parties involved.

4.3.4 Investigation by site workers

People who work at the odour source will often undertake investigations themselves, particularly if the council officers are unable to respond quickly or complaints come directly to site staff.

Where the site has a resource consent to discharge contaminants to air, a condition of the consent will often be to keep records, and to investigate and report any odour complaints received. In this case, the information recorded by the site workers should generally follow the same approach as outlined in Table 4.1.

Records should also be kept of complaints investigated by council officers. Site staff should consider whether an adverse effect is occurring and if so undertake a course of action to remedy the problem.

Recommendation 10

Consent holders should keep records of odour complaints and should follow the procedures outlined in Table 4.1.

4.3.5 Sensitivity of investigators

The variation in people's sensitivity to odour should be considered when investigating odour complaints. Dynamic dilution olfactometry (described in section 4.10) is sometimes used to assess whether a council officer undertaking an investigation is within the normal range of the population and can therefore be considered to be representative. The limitations of this method are that people vary in their sensitivity from day-to-day, the test involves only one chemical (n-butanol), and it is not necessarily representative of sensitivity to all odours in environmental situations. DDO is, however, generally accepted as being useful for screening to assess whether an officer has a highly sensitive or particularly insensitive sense of smell. But the test should not be seen as a 'calibration' of the officer according to a specific standard.

An officer who has a particularly sensitive or insensitive sense of smell can still carry out odour investigations provided they are aware of any 'bias' that may affect their conclusions regarding adverse effects. In cases of ongoing complaint or where prosecution is being considered, a number of council officers should assess the odour to account for varying sensitivity.

When assessing the significance of any off-site odour effects, site workers need to be aware that they may be desensitised to an odour and should be careful to consider the odour from the point of view of someone who does not work at the site.

Recommendation 11

DDO should be used to indicate to staff investigating odours whether they are normal, highly sensitive or insensitive to odour.

Site staff involved in assessing odour incidents should consider whether they might be desensitised to the odour when making their assessment.

4.3.6 Field observations of intensity, character and impact

Details about the intensity, character and adverse effect rating of the odour as recommended in actions 3 and 5 of Table 4.1 should be recorded using the methods described below.

A scale for describing odour intensity during field observations is detailed in the German Standard VDI 3882 (I) (1992): Olfactometry Determination of Odour Intensity [VDI 3882 - Part (I), Pub. Verein Deutscher Ingenieure, Dusseldorf. Available from Beuth Verlag GmbH, Berlin.] (Table 4.2). This scale is used in Europe and Australia for grading intensity. Experience using this scale has shown that observations have a good degree of consistency between observers.

Table 4.2: German VDI 3882 odour intensity scale

Odour intensity
Intensity level

Extremely strong

A

Very strong

B

Strong

C

Distinct

D

Weak

E

Very weak

F

Not perceptive

G

The German Standard VDI 3940 Determination of Odorants in Ambient Air by Field Inspection procedure for logging observations in the field involves recording the odour intensity every 10 seconds over a 30-minute period in one location. The VDI scale should be used for recording odour intensity versus time information spanning a desirable period of around 30 minutes. This provides short-term information on frequency, intensity and duration factors.

The observations should be carried out for at least 30 minutes. Shorter time periods may result in the observer missing the extent of the effects.

An exception to the 'every 10 seconds for 30 minutes' rule is needed when the odour plume is strong and constant, such as in stable, drainage flow conditions. Staying permanently in the plume will result in the observer becoming desensitised to the odour, so it is appropriate in this case to drive or walk through the plume once every 5-10 minutes, then repeat over a period of at least 30 minutes.

Objective recording of FIDOL factors needs to include the intrinsic nature of the odour, which is often referred to as the odour character (such as fishy, sewage, bakery, etc). A suggested table of general odour character descriptions is given in Table 4.3. Councils may wish to add other descriptions to this table for field use.

Table 4.3: Odour character descriptors

001

Fragrant

021

Like blood, raw meat

002

Perfumy

022

Rubbish

003

Sweet

023

Compost

004

Fruity

024

Silage

005

Bakery (fresh bread)

025

Sickening

006

Coffee-like

026

Musty, earthy, mouldy

007

Spicy

027

Sharp, pungent, acid

008

Meaty (cooked, good)

028

Metallic

009

Sea/marine

029

Tar-like

010

Herbal, green, cut grass

030

Oily, fatty

011

Bark-like, birch bark

031

Like gasoline, solvent

012

Woody, resinous

032

Fishy

013

Medicinal

033

Putrid, foul, decayed

014

Burnt, smoky

034

Paint-like

015

Soapy

035

Rancid

016

Garlic, onion

036

Sulphidic

017

Cooked vegetables

037

Dead animal

018

Chemical

038

Faecal (like manure)

019

Etherish, anaesthetic

039

Sewer odour

020

Sour, acrid, vinegar

   

The investigator should summarise the overall impact of the odour on the complainant. An example impact scale is shown in Table 4.4. This covers a range of impacts that refer to chronic through to acute effects and should not be used at the end of the assessment.

Table 4.4: An example of a scale for rating odour impact

a) The odour can be detected but is not annoying under normal conditions.

b) The odour can be detected but is not annoying, unless it is continuous.

c) The odour is moderately strong and is annoying if it is continuous or if its occurrence is very frequent.

d) The odour is moderately strong and is annoying if it occurs for periods of more than 5 to 10 minutes. Shorter, infrequent occurrences are not annoying.

e) The odour is strong and is annoying even in periods of short duration.

Recommendation 12

The FIDOL factors and impact rating of an odour should be recorded during investigation of an odour incident/complaint based on Tables 4.2, 4.3 and 4.4.

4.3.7 Summarising complaint data

A chronological summary of odour complaints can be used to indicate changes in long-term odour exposure. Trends can illustrate seasonal changes in complaint frequency, which may be due to changes in plant production or in the prevailing meteorology. An example summary of complaints received following the commencement of a new odour-producing process is shown in Figure 4.1.

Figure 4.1: Example of a complaint frequency graph

See figure at its full size (including text description).

It can also be useful to plot wind direction and the frequency of complaints registered from particular locations. The analysis of weather conditions and other parameters, such as time of day and character of the odour, can be very important for identifying the odour source and validating the complaint.

4.3.8 Determining an adverse effect based on complaint data and analysis

Following a complaint investigation or series of investigations, an officer must decide whether the odour is objectionable or offensive to the extent that it was having an adverse effect. The effect could be either chronic or acute, or possibly both. For acute odour events the opinion will be formed during observation in the field. For long-term chronic effects, an overall assessment of repeated observations is required.

Objectionable or offensive effects should be determined on a case-by-case basis by considering the FIDOL factors and other issues discussed in section 2.5. Figure 4.2 illustrates the factors that need to be taken into account when deciding whether there is an adverse effect or not. The methods described in this guide will assist investigating officers to determine the degree of effect, but it will be up to the officer and council (or ultimately the Environment Court) to evaluate each factor and draw conclusions on whether an 'ordinary reasonable person' would consider that the odour is having a significant adverse effect or not.

If decisions cannot be made about the significance of effects based on complaint response and investigations, other methods such as odour surveys and odour diaries may be useful. Some further guidance on using assessments to instigate action is contained in section 5.

Figure 4.2: Factors in considering if there is an adverse effect from odour

See figure at its full size (including text description).

4.4 Odour surveys

Odour surveys typically measure 'population annoyance' due to all sources of odour. The results can be used to rank odour sources according to their contribution to the cumulative stress within a community. Odour surveys are limited in that they are only useful in areas where there is sufficient population density to achieve statistically significant results. In low population areas, odour diaries, odour-modelling assessments and complaint records are the preferred methods of assessment.

Odour surveys directly measure the extent of adverse effects resulting from repeated odour impacts in a community. A standard questionnaire for surveys in New Zealand is recommended to ensure that the results between surveys are comparable (see Appendix 2). The survey has questions about the state of the environment in general - not just odour - in order to minimise bias. The questionnaire may be modified to suit particular circumstances or objectives but the Technical Report (Ministry for the Environment, 2002c) should be referred to for further information.

Survey respondents are asked about the level of annoyance they experience from odour. Responses are classified according to the scale in Table 4.5. The key statistic used from odour surveys is the 'percent at-least annoyed' category, which is made up of the responses ranging from annoyed to extremely annoyed.

Table 4.5: Annoyance levels used in odour surveys

Definitely not annoying

Very little annoyance

Little annoyance

Some annoyance

 

Annoying

Quite annoying

Very annoying

Extremely annoying

Percent at least annoyed

The steps when undertaking an odour survey are as follows.

  1. Identify sub-areas within the community to be surveyed by reviewing historical complaint records and wind data. Within each sub-area the population should be exposed to a similar amount of odour, considering distance from the source and/or prevailing winds. If the sub-area is too large, exposure will vary too much among the group and there will be a wide range of responses that may dilute the results from the affected area with those unaffected. The survey areas should be agreed in consultation with council air-quality staff.
  2. Select the population size for the survey. A minimum target of 50-70 respondents from any one sub-group of the community provides a margin of error for the survey in the order of 5-10%.
  3. Conduct the survey by a telephone interview. Randomly select phone numbers from a number listing for the area being surveyed. It is usually necessary to obtain three times the number of phone numbers as the sample size. Phone numbers can be purchased from Telecom to order, or for a small township can be obtained by scanning the white pages. If possible, conduct the survey during a two-hour period in the evening. This prevents people in the neighbourhood talking about the survey, which may skew the results.
  4. Calculate the percentage of people who were 'at-least annoyed' by odour for each sub-area. Where there are multiple sources, break the survey results down according to the main source(s) identified by respondents.
  5. Compare the survey results to those for a control population.
  6. Calculate survey margins of error using statistical methods as described by McCullagh and Nelder (1983) or by Perry and Green (1984).

Control data should be gathered by surveying a population that is not affected by any significant odour, in parallel with the affected population. The control population should have similar demographics and culture to the affected population being surveyed and similar exposure to any background odours such as mudflats or geothermal odours.

Typical results from control surveys in New Zealand show that 5-15% of the community report being 'at least annoyed' by industrial odours, even when there are no significant odours (Aurora Environmental, 2000). Fifteen percent at least annoyed is therefore taken to be representative of a community unaffected by odour.

The guideline criterion recommended for assessing survey results is '20% at least annoyed' based on New Zealand experience from control populations.

When interpreting the results of the survey, if the control population has a level of less than 15% at least annoyed, this indicates that the survey methods were correct. If the control population returns a level of more than 15%, the control population may be subject to some odours and is not a true control group.

The 20% at least annoyed criterion allows for a small level of effect that would not be considered objectionable by reasonable people. Twenty percent is the cumulative effect from all sources. The 20% criterion should not be compared to the upper percentile value but to the mid-point of the confidence range, provided that the survey has been carried out according to the appropriate and statistically significant sampling methods.

Recommendation 13

Odour annoyance surveys should be conducted using the questionnaire from Appendix 2.

Target and control populations should be identified in consultation with council air-quality staff.

The acceptable level of cumulative odour impacts due to all sources is = 20 % at-least annoyed.

4.5 Odour diaries

Odour diaries are used by people in affected communities to record their daily exposure to odour. Diaries can be useful for determining particular conditions under which people are affected by odour from a particular source or sources. An example odour diary record sheet is provided as Appendix 4.

A diary programme can be useful for collecting data on the frequency and strength of odour impacts at various locations over a given period. The resulting data can be used to calculate the percentage of time (hours/year) that people are exposed to odours from a specific source, as well as the typical strength and character of the impacts. The information recorded in a comprehensive diary programme includes:

  • date and time of day
  • duration of the event
  • continuity of the odour during the event
  • character and strength of odour
  • likely source of odour
  • wind direction and strength.

Diarists should be given instructions on how to record information so that it is as consistent as possible. They should also be given feedback on the programme to help maintain their enthusiasm to continue keeping the records.

A less comprehensive diary programme may be sufficient. For example, where the aim is to investigate whether an odour source is still creating some impacts in a community following some improvement in odour control, such as in Case Study 2.

Case Study 2: AFFCO NZ Ltd odour diary programme

A diary programme was used by AFFCO NZ Ltd as a monitoring tool for odour at its Imlay plant in Wanganui. Imlay is a meat-processing plant with a large rendering facility that had a history of odour problems. AFFCO undertook substantial upgrades to its extraction and odour treatment system in 1998 and wanted to establish whether the new system was effective in eliminating rendering plant odours.

AFFCO had undertaken an odour annoyance survey during 1997 to establish the extent of adverse effects on the neighbouring community. An odour diary programme was considered an appropriate tool for the second study in 1998 because a further survey could be affected by a lag between the reduced odour levels and the level of annoyance likely to be measured in the community. A diary programme was used to establish if rendering odours were still occurring as a result of AFFCO's activity, bearing in mind there were other sources of industrial and commercial odour present in the community.

Five residential properties downwind of the rendering plant were used for the programme. Diarists were selected in consultation with the community liaison group that had existed for a number of years.

Six months of diary records were used, along with plant operating status and wind records, which confirmed that the predominant rendering odours had been eliminated. The results also helped to identify other sources of odour within the community and their relative significance.

4.6 Repeat questioning

The German VDI standard 3883 (Part 2) for assessing annoyance by repeat questioning involves using selected individuals within a community to routinely record their annoyance rating from odour at a pre-specified time of the day at a given location. An annoyance index is calculated from the observations and is used to monitor the relative change in the extent of odour impacts over a period of time. There is, however, little experience of this method in New Zealand, so evaluation of the results would be difficult. This method is therefore not recommended for use without further research.

Industries or councils that wish to pursue this method are encouraged to do so in parallel with other survey methods described in this guide. Findings should be made available to other New Zealand practitioners so that the future use of this method and its ability to measure adverse effects can be evaluated.

4.7 Community meetings

Community meetings are often used to gauge the extent of any dissatisfaction due to exposure to odours. Holding an open public meeting is generally a first step and a community liaison sub-group is often established from this meeting. This group can be used as a forum to negotiate solutions and to provide direct and ongoing community input on odour issues. Membership of the liaison sub-group should be decided in a democratic and transparent manner. It must be noted that the views of the group are only indicative of those in the wider community. Other tools such as newsletters may be useful to ensure the wider community is kept informed on an ongoing basis.

In situations where there are only one or two complainants, open public meetings can be used to see whether there is a more widespread problem. This can indicate whether complainants may be vexatious or are particularly sensitive; i.e. not representative of the 'ordinary reasonable person'.

Community consultation is useful to investigate whether people consider that any odours are of an acceptable level. Sometimes concerns are raised during consultation such as at the consent renewal time, even though there have been no formal complaints made. Ongoing dialogue between odour producers and potentially affected communities is recommended to allow dischargers to deal with issues as they arise. This can prevent ill feelings building up in the community.

Community liaison groups normally include management and engineering staff from the site producing the odour, members of the local community, and council officers. Group meetings may be chaired by an independent mediator/chairperson. Normal meeting rules and standard procedures should be followed to ensure order is maintained. Minutes and matters arising from the minutes should be recorded and discussed.

Further guidance on running community consultation can be found in Striking a Balance: A Practice Guide for Consultation and Communication for Project Advocates (Ministry for the Environment, 1999).

Recommendation 14

Community meetings and liaison groups are strongly recommended to provide community input into odour issues.

Community liaison groups should include management and engineering staff from the site, members of the local community, and council officers.

4.8 Experience from other sites and best practice

The potential for objectionable or offensive odours can be assessed using knowledge and experience from similar sites. Assessing technology and management systems from other similar sites can indicate the effects expected from a new or upgraded activity. Assessment of this type generally requires input from an air pollution control expert, and/or a person with experience of a particular odour source.

Best practice guides or industry codes of practice recommend technology and practices for specific industry types. But be aware of the objective of the code when applying the recommendations. Codes may be designed to protect from odour, noise, health effects or safety hazards, but the recommendations will depend on the type of land use at the boundary of the site and the processes in practice when the code of practice was developed.

Site-specific considerations must be taken into account in any assessment based on experience from another site, including:

  • meteorology, such as the occurrence of calm conditions and down-valley air flows (katabatic winds)
  • surrounding land use and population density
  • terrain and the effect on dispersion
  • the degree of similarity of the processes (e.g. methods and raw materials).

If any of these factors vary significantly between the sites being compared, then the experience of impacts from a particular operation my not be readily transferable.

4.9 Management and contingency plans

Many odour emissions are strongly related to site procedures and process management. If good management is not in place then adverse effects are more likely to occur. Therefore information on management and contingency is needed to assess the potential for adverse odour effects. Management procedures should be documented in an odour management plan, which is usually submitted at the time of lodging an application for a resource consent.

A management plan should clearly describe the systems required to ensure the reliable operation of odour controls, performance monitoring, and the reporting of these.

Matters that should be included in an odour management plan are outlined in Appendix 3.

4.10 Measuring odour by olfactometry

Odour emissions can be measured in odour units (OU) using dynamic dilution olfactometry (DDO). DDO is a laboratory measurement of the concentration of an odour. The method uses a panel of observers to identify whether an odour is present through sniffing ports. The concentration of the odour is determined by using odour-free air to dilute the sample to a level where 50% of a panel of people smelling the odour can just detect it. This point is given the concentration of 1 OU and the number of dilutions required to reach 1 OU determines the original concentration of the sample. The concentration of odour in air, as measured by DDO, is expressed as the number of odour units per cubic metre (OU/m3). These data are then used as an input into atmospheric dispersion modelling to predict downwind odour effects (see section 4.11).

The standardisation of DDO in Australasia and Europe has only occurred in the past five years. The variability in the measurement method before standardisation means that earlier data are not necessarily comparable to the current measurements. The recommended method for DDO in New Zealand is AS/NZS 4323.3:2001 Stationary Source Emissions - Determination of Odour Concentration by Dynamic Olfactometry, which is based on the European draft standard.

There are two methods for conducting DDO measurements: the 'yes/no' and the 'forced choice' methods. The AS/NZ standard applies to both methods, but the two current laboratories in New Zealand use the forced choice method. There are also two ways of reporting measurements: detection, and certainty or recognition odour thresholds. The detection threshold is the lowest concentration of a compound that can just be detected by a certain percentage of the population, while the certainty or recognition threshold is the lowest concentration of a compound that can be recognised with certainty as having a characteristic odour quality. In general, recognition thresholds are higher than the detection threshold.

DDO and other techniques for odour measurement are described in detail in the Technical Report (Ministry for the Environment, 2002c), along with other less commonly used techniques, such as electronic instruments and chemical measurement of odorous compounds. The use of odour thresholds in odour assessment is discussed in Appendix 6.

4.11 Odour modelling

Odour dispersion modelling predicts the concentration of an odour downwind of the source using a computer programme. Modelling inputs include the characteristics of the discharge, local terrain heights, meteorological conditions, the location of downwind receptors, and odour emission rates. Odour dispersion modelling is one of the only tools that can predict the potential effects of a new odour-emitting activity.

This section briefly covers the key issues and limitations of odour modelling methods and uses, and highlights some things to watch out for when preparing or auditing an assessment of odour effects based on dispersion modelling. Specific advice on odour modelling is contained in the Technical Report, and detailed guidance on general dispersion modelling is provided in the Good Practice Guide for Atmospheric Dispersion Modelling (Ministry for the Environment, 2002b,) currently under development. It is important to recognise that odour modelling is a complex technique and those intending to carry it out should be appropriately trained.

Dispersion modelling of odour emissions should only be used where the emission sources can be quantified, and where the discharge is continuous or semi-continuous. In other words, modelling should only be applied to discharges with potential chronic odour effects rather than acute odour effects, such as may occur from abnormal operations.

4.11.1 Odour emission rate measurement

Odour emission rate estimates are needed for dispersion modelling. Both the concentration and the volumetric flow of the emission must be measured to estimate the odour emission rate. For a point source, the odour emission rate is expressed as odour units per second (OU/s) and for area sources the rate is expressed per unit area per second (OU/m2/s).

It is often difficult to determine odour emission rates for diffuse or fugitive sources of odour and/or where the flow rate is low, because it is hard to estimate the discharge flow rate and the limit of detection for DDO measurement is relatively high. Fugitive sources may have a very high concentration that is noticeable close to the source, but a very low flow rate so that the odour dissipates rapidly with increasing distance from the source. Subjective assessments of the significance of sources may be the only option where the emission rate cannot readily be measured.

Where only one chemical is causing the odour and the expense of DDO is not justified, it may be appropriate to measure the concentration of the particular chemical alone and compare the result to its odour threshold to estimate its concentration in odour units. The odour threshold approach is, however, limited because most odours are caused by a complex interaction between different chemicals, and the chemicals responsible are present in very low concentrations, often making interpretation difficult. It is therefore generally more appropriate to use DDO odour measurement data in dispersion modelling. Experience indicates that for an odour that contains several compounds the combined odour measurement cannot be readily related to the concentrations of specific chemical compounds.

Standards Australia has published an Australia/New Zealand standard: Stationary Source Emissions - Determination of Odour Concentration by Dynamic Olfactometry, code AS/NZS 4323.3:2001. The standard was prepared by the Joint Standards Australia/Standards New Zealand Committee EV/7, Methods for Examination of Air. The objective of AS/NZS 4323.3:2001 is to provide a method for determining the odour concentration of a gaseous sample using DDO with a panel of human assessors as the sensor. The standard can be downloaded from the Standards Australia website (http://www.standards.com.au). AS/NZS 4323.3:2001 applies to both yes/no and forced-choice response methods.

Odour thresholds for individual chemicals and a list of recommended odour threshold references are given in Appendix 6. More detailed information and specific recommendations about odour emission measurement are available in Chapter 7 of the Technical Report (Ministry for the Environment, 2002c).

Recommendation 15

Odour emission rates should be measured using dynamic dilution olfactometry carried out in accordance with the joint Australian/New Zealand standard AS/NZS 4323.3:2001 Stationary Source Emissions - Determination of Odour Concentration by Dynamic Olfactometry. [The standard can be downloaded (for a fee) from the Standards Australia website,] [http://www.standards.com.au.]

4.11.2 Model applicability versus community feedback

For existing operations, atmospheric dispersion modelling predictions should generally be given less weight than community feedback on odour effects. In particular, if sufficient community data are available to demonstrate that there is an odour problem, odour modelling should not be used to try to argue the contrary. Dispersion models can be used diagnostically to identify sources of odour from existing activities that are contributing to off-site effects. Models allow individual sources of odour to be 'switched off' to investigate the contribution of the remaining sources to the overall odour impact, and help to identify which sources should be controlled and to what level.

4.11.3 Multiple sources and background odour

For dispersion model scenarios with one or two sources, the maximum measured emission rate from each source is typically used for dispersion calculations. For multiple sources, however, this may result in overly conservative and unrealistic results. To prevent this, assessors should consider whether or not the different sources are likely to be additive, or if one is likely to mask others when more than one is present. Average emission rates are sometimes preferred for multiple sources when not all the sources discharge at the peak rate at the same time. The assessors must understand the emission characteristics of the processes they are modelling, and if peak emissions are likely to coincide, these scenarios should be accounted for in the model set-up.

The effect of background odours and multiple sources should be considered on a case-by-case basis. If the odour being modelled is quite different to, or much stronger than, any background odour (e.g. a strong odour from a chemical manufacturing plant in a rural area), then background odour should probably not be included in the model but considered subjectively in terms of its potential influence.

4.11.4 Odour-modelling guideline values

Dispersion model outputs in odour units per cubic metre (OU/m3) can be compared to odour-modelling guideline values to determine whether objectionable or offensive effects are likely to occur. The comparison should take into account the sensitivity of the proposed receiving environment, as described in Table 2.2.

The recommended odour-modelling guideline values are summarised in Table 4.6. Other values can be used on a case-by-case basis where they are justified for specific odour sources and the work has been adequately peer reviewed.

Table 4.6: Recommended odour-modelling guideline values

Sensitivity of the receiving environment (refer to Table 2.2)
Concentration
Percentile

High (worst-case impacts during unstable to semi-unstable conditions)

1 OU/m3

0.1% and 0.5%

High (worst-case impacts during neutral to stable conditions)

2 OU/m3

0.1% and 0.5%

Moderate (all conditions)

5 OU/m3

0.1% and 0.5%

Low (all conditions)

5-10 OU/m3

0.5%

Note that:

  • atmospheric stability has been accounted for in high-sensitivity receiving environments (stability refers to the degree of mixing that occurs)
  • the percentile allows for a small level of exceedance of the predictions, to account for worst-case meteorological conditions, at which objectionable odours are unlikely because the conditions occur infrequently
  • the 'baseline' percentile is 0.5%, although 0.1% will also be used to assist in the evaluation of model results depending on the type of source and consistency of emission data; further discussion of percentile selection is given in the Technical Report (Ministry for the Environment, 2002c)
  • the concentration components in the table already include the peak-to-mean ratio adjustment for all source types, and should be used as design ground-level concentrations for one-hour modelling averages.

Two approaches were used to develop the modelling guideline values for New Zealand: the annoyance threshold method and the dose-response method. The former is more theoretically based whereas the latter is empirically based using odour surveys. More detailed information on how surveys can be used to define guideline values and the rationale behind the recommended guideline values is provided in Chapter 9 of the Technical Report (Ministry for the Environment, 2002c).

The Ministry for the Environment will update the modelling guideline values as necessary when more empirical research of the effects of odours on communities emerges from odour dose-response studies (discussed in section 4.4).

Recommendation 16

The guideline values in Table 4.6 should be used along with Table 2.2 to assess modelling results and to determine whether the odour is likely to cause an adverse effect.

The guidelines already include the peak-to-mean ratio adjustment for all source types and should be treated as design ground-level concentrations for one-hour modelling averages.

4.11.5 Model interpretation and limitations

Odour-modelling guideline values should not be interpreted as a 'pass or fail' test. The evaluation of the potential for objectionable or offensive effects must be on the basis of probability. The conservatism in the model predictions should be considered. Factors influencing the level of conservatism include the odour emission rate data, land use and activities where guideline exceedances are predicted to occur, the model assumptions, and the meteorological data file used.

There are a number of limitations inherent in the modelling approach. Depending on the level of conservatism, predicted guideline value breaches do not necessarily mean that adverse odour effects will occur. Likewise, being within the guideline value does not mean there will be no adverse effects. Any calculated breach occurs for a whole hour according to the model, which predicts hourly averages, but in practice peaks will only occur for short periods. The model assumes that the wind direction remains constant throughout the hour, but wind directions can fluctuate within an hour. The model also assumes that the rate of odour emission from each source is constant from hour-to-hour, but the emission rate will vary over time. Factors called peak-to-mean ratios are applied to models to help account for the short-term peaks versus the hourly average model outputs, but the science is uncertain. Using annoyance surveys (odour dose-response studies) to calibrate the model accounts for many of the limitations inherent in the theoretical approach to developing guidelines.

Odour modelling may be limited in its application due to:

  • the variability in odour emission rates, which may not be adequately characterised by 'one-off' odour measurements
  • lack of a meteorological data set representing local conditions
  • the fact that odours are not simply additive in their effect - there are complex masking and synergistic effects that vary for each mixture of odorants
  • the fact that intensity of odour does not vary linearly with concentration.

Model results should therefore be just one of the indicators of the potential for adverse effects, and other tools should be used in conjunction with modelling when assessing potential effects.

Recommendation 17

Modelling should be given less weighting in situations where reliable community assessment data are available.

Modelling is best applied to situations where the odour emission rate can be measured and where the odour emissions are reasonably constant, causing potential chronic effects.