EPA (2001) Odour Impacts and Odour Emission Control Measures For Intensive Agriculture

E N V I R O N M E N TA L R E S E A R C H
R&D REPORT SERIES No. 14
Odour Impacts and Odour Emission Control

Measures for Intensive Agriculture
F I N A L R E P O RT
EUROPEAN COMMUNITY
European Regional
Development Fund
© Environmental Protection Agency 2001
All or part of this publication may be reproduced without further permission, provided the
source is acknowledged.
Although every effort has been made to ensure the accuracy of the material contained in this
publication, complete accuracy cannot be guaranteed. Neither the Environmental Protection
Agency nor the author(s) accept any responsibility whatsoever for loss or damage occasioned
or claimed to have been occasioned, in part or in full, as a consequence of any person acting,
or refraining from acting, as a result of a matter contained in this publication.
Project information
This report was commissioned by:
Environmental Protection Agency
Johnstown Castle Estate, Wexford, Ireland
Contact: Dr. Vera Power / Dr. Tom Stafford
and was issued by:

OdourNet UK Ltd
5 St Margaret’s Street, Bradford-on-Avon, Bath BA15 1DA
Contacts: Mr. Ton van Harreveld, Mr. Nick Jones
Funding and Copyright

The Environmental Protection Agency took the initiative to carry out an R&D project
entitled ‘Odour Impacts and Odour Emission Control Measures for Intensive Agriculture’.
The project was co–funded by the Environmental Monitoring, R&D Sub–programme of the
Environmental Services Operational Programme 1994 – 1999 (50% grant–aid).
The partner consultant in the project, OdourNet UK Ltd, a member of the OdourNet group,
prepared the report.
Management and organisation and consultation

The project was managed by the Environmental Protection Agency. A steering group was
formed, with the following members:
• Dr. Vera Power, Mr. David Maguire, Ms. Becci Cantrell,
Dr. Tom Stafford, Ms. B Rooney, Mr. Dara Lynott
Technical comments were contributed by individuals involved in the
Irish pig production sector.
Acknowledgements
The authors wish to thank all those who contributed their information, opinion, advice,
criticism, time and wisdom to this report.
ISBN 1-84095-075-7
Price Ir£15 / €19.05 11/01/500
Contents
1. EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. SCOPE OF THE STUDY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1 SCOPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1.1 Study objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1.2 Structure of Part A of the report. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.3 Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3. INTRODUCTION TO THE CHARACTERISTICS OF PIG PRODUCTION IN IRELAND . . . . . . . . . . . . 9
3.1 OVERVIEW OF THE PIG PRODUCTION SECTOR AND ITS ECONOMICS . . . . . . . . . . . . . . . . 9
3.1.1 Distribution, size and type of pig production units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.1.2 Strengths of the Irish pig industry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.1.3 Operational characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4. INTRODUCTION TO ODOURS AS AN ENVIRONMENTAL STRESSOR . . . . . . . . . . . . . . . . . . . . . . . 13
4.1 ODOUR PERCEPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2 CHARACTERISING ODOURS: PSYCHOPHYSICAL DIMENSIONS OF
ODOUR PERCEPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.2.1 The common traits of units used for odour and noise assessment . . . . . . . . . . . . . . . . . . . . . 14
4.3 THE MECHANISM LEADING FROM PIG SMELL TO ODOUR NUISANCE . . . . . . . . . . . . . . 15
5. IMPACT ASSESSMENT METHODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.1 ASSESSMENT OF EFFECTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.1.1 Direct measurement of percentage of people annoyed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.1.2 Complaints analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.2 DIRECT ASSESSMENT OF ODOUR EXPOSURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2.1 Direct field methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.3 ASSESSMENT OF ODOUR IMPACT BY MEASURING EMISSIONS AT SOURCE,
FOLLOWED BY DISPERSION MODELLING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.1 Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3.2 Odour concentration analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.3.3 Modelling of atmospheric dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.3.4 On the choice of meteorological data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.3.5 On the choice of percentile values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
6. DOSE–EFFECT RELATIONSHIP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.1 THE RELATIVE PROPERTIES OF PIG ODOURS COMPARED TO OTHER
ENVIRONMENTAL ODOURS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.2 RELATION BETWEEN ODOUR EXPOSURE AND PERCENTAGE OF
POPULATION ANNOYED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7. COMPARISON OF REGULATORY APPROACHES IN DIFFERENT COUNTRIES . . . . . . . . . . . . . . . . 35
7.1 GENERIC APPROACHES TO LIVESTOCK ODOUR MANAGEMENT AND REGULATION . . 35
7.2 BACKGROUND TO THE USE OF QUANTITATIVE ODOUR EXPOSURE CRITERIA . . . . . . . 36
7.2.1 The reasoning underpinning the choice of odour exposure limit values varies. . . . . . . . . . . . 36
7.3 OVERVIEW OF LIVESTOCK ODOUR GUIDELINES IN DIFFERENT COUNTRIES . . . . . . . . 37
7.3.1 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
7.3.2 The Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
7.3.3 United Kingdom. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
7.3.4 United States of America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
7.3.5 New Zealand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
7.3.6 Comparison of setback distances in different countries for fictitious pig units. . . . . . . . . . . . 40
8. PROPOSED FRAMEWORK FOR IMPACT ASSESSMENT WITH A VIEW TO LICENSING . . . . . . . . 43
8.1.1 Scope of impact assessment framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
8.1.2 Legal framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
8.1.3 The proposed assessment framework: general principles and odour exposure criteria . . . . . . 45
8.1.4 Definition of sensitive receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
8.1.5 Emission factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
8.1.6 Cumulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
8.1.7 Practical application of the assessment framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
8.1.8 The general, simple case: assessment by screening using standard contour overlays . . . . . . . 53
8.1.9 Full impact assessment using atmospheric modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
8.1.10 Specific terms and definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
9. METHODS FOR REDUCTION OF ODOUR IMPACT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
9.1 WHAT CAUSES ODOURANTS TO BE PRODUCED? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
9.2 THEORETICAL OPTIONS FOR REDUCING ODOUR EMISSIONS FROM
PIG PRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
9.3 GOOD OPERATIONAL PRACTICE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
9.3.1 Slurry removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
9.3.2 Cleanliness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
9.4 HOUSING DESIGN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
9.4.1 Standard housing systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
9.4.2 Low-emission housing systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
9.5 OPTIMISATION OF VENTILATION AND ATMOSPHERIC DILUTION . . . . . . . . . . . . . . . . . . . 63
9.5.1 Ventilation in livestock housing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
9.5.2 Optimisation of atmospheric dilution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
9.6 FEED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
9.7 ADDITIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
9.7.1 Feed additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
9.7.2 Slurry additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
9.7.3 Odour counteractants and masking agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
9.8 EXTRACTION AND TREATMENT OF VENTILATION AIR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
9.8.1 Chemical scrubbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
9.8.2 Bioscrubbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
9.8.3 Biofilters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
9.8.4 Ozone treatment of ventilation air. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
9.9 SLURRY STORAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
9.9.1 Odour emission reduction in open slurry storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
9.9.2 Enclosed slurry storage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
9.9.3 Operational aspects of slurry storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
10. TWO CASE STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
10.1 EXECUTIVE SUMMARY PART B: CASE STUDIES ASSESSING
THE ODOUR EMISSIONS AND IMPACT OF TWO PIG
PRODUCTION UNITS IN THE IRISH SITUATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
11. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
ANNEX A. ODOUR REGULATIONS FOR INTENSIVE LIVESTOCK IN OTHER COUNTRIES . . . . . . . . 77

A.1 GERMANY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
A.1.1 General regulatory framework and relevant guideline documents . . . . . . . . . . . . . . . . . . . . . 77
A.1.2 The distance graph in VDI3471 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
A.2 NETHERLANDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
A.2.1 Guideline documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
A.3 UNITED KINGDOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
A.4 UNITED STATES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
A.4.1 Setback distances: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
A.5 NEW ZEALAND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
ANNEX B. METHODS FOR ODOUR ASSESSMENT AND UNITS OF MEASUREMENT. . . . . . . . . . . . . 87

B.1 DETECTABILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
B.1.1 Measurement of odour concentration using olfactometry . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
B.1.2 The unit of measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
B.1.3 Odour concentration measurement using quantitative olfactometry . . . . . . . . . . . . . . . . . . . . 88
B.2 INTENSITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
B.3 ODOUR QUALITY (DESCRIPTIVE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
B.4 HEDONIC TONE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
B.5 ANNOYANCE POTENTIAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
B.6 CHARACTERISATION OF ODOURS USING CHEMICAL ANALYSIS . . . . . . . . . . . . . . . . . . . 91
ANNEX C. GLOSSARY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
ANNEX D. REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
ANNEX E. CONTOUR PLOTS FOR SCREENING ASSESSMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

E.1 STANDARD CONTOURS OVERLAY, SCALE 1:50,000, FOR THE LIMIT
VALUE FOR NEW PIG PRODUCTION UNITS, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
E.2 STANDARD CONTOURS OVERLAY, SCALE 1:50,000, FOR THE TARGET
VALUE FOR ALL PIG PRODUCTION UNITS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
E.3 STANDARD CONTOURS OVERLAY, SCALE 1:10,560, FOR THE LIMIT
VALUE FOR EXISTING PIG PRODUCTION UNITS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Man is the measure of all things
Protagoras, sophist, (c. 485-410 BC)
‘The question whether something is true or false, good or bad, should always be
considered in relation to the needs of that person’
Odour Impacts and Odour Emission Control Measures for Intensive Agriculture
Executive Summary
Close to 200 pig units will be applying for IPC licences €3,200 and profitability is variable. After a period of
in Ireland in the next few years, as a result of the EPA generally low prices in 1998 and 1999, leading to losses
Act of 1992 and European IPPC licensing requirements. per pig produced, profitability improved in 2000 due to
The assessment of the odour impact of these pig units relatively strong demand and strong sterling. Sixty five
will be an important element of the licensing process, percent of production is exported to the UK.
which will be carried out by the Environmental
Protection Agency. Suitable criteria for ‘acceptable Chapter 4 sets out an introduction to odours as an
exposure’ to odours are required, in order to set, limit environmental stressor and provides operational
and target values for odour impact. Appropriate odour knowledge for professionals involved in odour
assessment methods are required to quantify emissions annoyance licensing and management. The evolutionary
and, if required, to control and enforce licence origins of our sense of smell, its function and relevance
conditions. To allow pig producers to manage the impact to our social behaviour are explained. The dimensions
of odours, information is required on the relationship for characterising odours are identified: detectability,
between production practice, housing types and odour intensity, hedonic tone, odour quality and the recently
emissions, as well as on suitable approaches to the proposed dimension of odour annoyance potential. In
abatement of odour emissions. assessing environmental odours, detectability is
generally the only dimension used. A European draft
This report provides an overview of these issues, based standard is in preparation by CEN/TC264/WG2 Odours
on literature review, a limited programme of (EN13725) for measuring odour concentration in
measurements and experience gained from the European odour units per cubic metre (ouE/m3). An
regulatory practices in other countries. This information odour that is just detectable by 50% of selected panel
will assist the EPA in formulating its approach to members is described as having an odour concentration
processing licence applications and in achieving of 1 ouE/m3. It must be noted that the relation between
transparent and uniform decision-making on odour perceived intensity and odour concentration is not linear
issues for that purpose. but logarithmic. A useful similarity is that of noise
where the linear measure of energy of the noise in
This report was prepared by OdourNet UK Ltd. as part Watt/m2 is translated to intensity using the logarithmic
of the Environmental Monitoring R&D sub programme, unit dB. Like the odour unit, the dB is based on a
supported by an ERDF grant. sensory detection threshold: at 0 dB 50% of young
people can detect a sound at a frequency of 1000 Hz.
Chapter 3 gives an overview of the characteristics of pig However, in contrast to common practice for noise,
production in Ireland. While pig production capacity in odour concentration is typically expressed in the linear
Ireland is limited relative to other EU countries, there is unit (ouE/m3) rather than a logarithmic one such as the
a high proportion of large pig units. Approximately 190 dB.
units exceed current licensing thresholds and will
require an IPC licence. The majority of these (143) are The mechanism that leads from the production of pig
integrated breeding and finishing units. The average size odours via release and dispersion in the atmosphere to
of a pig unit in Ireland is 316 sows/unit, which again is causing odour nuisance in a specific population is
relatively large compared to other EU countries. The complex and is discussed in some detail. Odour
density of pigs is relatively low, ranging from 4 to 69 nuisance is a result of long-term, intermittent exposure
hectares per sow, with an average of 26 hectares per sow. to an environmental stressor, in a complex context of
The Irish pig production sector employs approximately physical, physiological, social and psychological factors
6000 people in pig production, slaughtering, processing, that determine the behavioural response of the
feed production and facilities supply. The investment per individual. Odour nuisance is not a linear push-button
sow in an integrated unit ranges between €2,000 and response to a particular intensity of exposure at any
1
moment by a particular smell. The exposure history is a out for the Ministry of Public Planning and the
major factor in the appraisal of the impact of Environment and used approximately 2,300
environmental odours. An epidemiological approach is, standardised telephone questionnaires collected from
therefore, the most appropriate tool with which to study householders living in the vicinity of pig units.
the relationship between a source, the dispersion
characteristics of a site and the long-term effects on the Chapter 8 sets out a proposed framework of target and
population in terms of annoyance. limit values based on the results of the Dutch study. This
framework provides a starting point for a licensing
Once that relationship is known, odour impact can be procedure to be used in Ireland. The proposed structure
assessed in a more straightforward manner, using source of target and limit values is:
emission measurements combined with dispersion
modelling. The results can be assessed using the • Target value: C98, 1-hour ≤ 1.5 ouE/m3
epidemiological dose-effect relationship, or exposure The target value provides a general level of protection
criteria derived from such a relationship. Assessment of against odour annoyance for the general public,
odours is typically undertaken by measurement of aiming to limit the percentage of people experiencing
emission rates at source, followed by dispersion some form of odour-induced annoyance to 10% or
modelling. Assessment in the field is more difficult, less. The target value is to be used as an
because of the large variations in momentary environmental quality target for all situations.
concentration caused by atmospheric dilution; other
background odours (e.g. soil, vegetation) and the The target value is achieved when the calculated
practical problems associated with measuring very low odour exposure for all locations of odour sensitive
odour concentrations (≤ 20 ouE/m3). receptors is less than an hourly average odour
concentration of 1.5 ouE/m3 in 98% of all hours in an
In Chapter 6 the dose-effect relationship for odour average meteorological year.
annoyance as a result of long-term intermittent exposure
to odours is discussed in detail. For a number of • Limit value for new pig production units:
industries in the Netherlands, specific targets for air C98, 1-hour ≤ 3.0 ouE/m3
quality have been defined as a certain 1-hour average The limit value for new pig production units provides
odour concentration that should not be surpassed in a minimum level of protection against odour
more than 2% of all hours in an average meteorological annoyance, aiming to limit the percentage of those
year. This criterion, commonly expressed as C98, 1-hour = experiencing some form of odour-induced annoyance
x ouE/m3, is assessed using a measured source emission to 10% or less of the general public, assuming some
and dispersion modelling, using meteorological data for degree of acceptance of the rural nature of their living
3 years or more. The target values range from C98, 1-hour environment.
≤ 0.5 ouE/m3 for rendering plants to C98, 1-hour ≤ 3.5
ouE/m3 for coffee roasters. The limit value for new pig production units is
complied with when, for all locations of odour
Chapter 7 provides an overview of regulatory sensitive receptors, the calculated odour exposure is
approaches taken in a number of countries, including less than an hourly average odour concentration of
Canada, Germany, the Netherlands, New Zealand, and 3.0 ouE/m3 in 98% of all hours in an average
the United States. Typical setback distances for a meteorological year.
relevant number of sows are presented in Table 9 to
allow a comparison of the proposed framework with that • Limit value for existing pig production units:
in other countries. C98, 1-hour ≤ 6.0 ouE/m3
The limit value for existing pig production units
Recently, a large-scale epidemiological study was provides a minimum level of protection against odour
conducted in the Netherlands to establish the dose-effect annoyance, aiming to limit the percentage of people
relationship between percentages of population annoyed experiencing some form of odour-induced annoyance
and calculated odour exposure. The study was carried to 10% or less, in the most tolerant tolerence section
2
(agricultural/rural) of the population. purposes.
The limit value for existing production units is Chapter 9 provides an overview of options for reducing
complied with when for all locations of odour odour emissions from pig production. The following
sensitive receptors the calculated odour exposure is options have been identified:
less than an hourly average odour concentration of
6.0 ouE/m3 in 98% of all hours in an average 1. Reduction of odour production
meteorological year. a. Reduction of the protein content in feed
b. Separation of urine and faeces, followed by
A phased plan must be made to reduce the odour treatment.
impact, with time, to the limit value for new pig c. Lowering of temperature of stored slurry
production units and, eventually, the target value. d. Collection of slurry in closed tanks,
followed by anaerobic digestion.
These criteria for odour exposure aim to provide a 2. Reduction of transfer rate from the surface of
framework that can be used to attain a general slurry
environmental quality in Ireland, while recognising that a. pH control
in some cases existing pig production units may need b. Covering the surface
some considerable period of time to achieve that target. i. Natural crusting
In some cases, the time allowed will have to take into ii. Floating biological covers (straw
account the cycle of normal replacement of assets such fibre)
as livestock housing, to allow implementation of a iii. Floating covers
structural solution, while avoiding destruction of capital iv. Liquid additives (vegetable oils)
goods. v. Air filled plastic domes (over sludge
storage lagoons)
The implementation of the proposed framework would 3. Reduction of exposed area of slurry, including
rely on using emission factors per animal. Specific storage, soiled surfaces, grids etc
measurements should not be required, with the possible a. Different housing types, which include
exception of very large units or production systems that systems such as:
are atypical. A limited programme of measurements was i. Green Label pig houses, designed for
conducted in Ireland for this report, providing indicative low ammonia emissions
emission factors for winter conditions. The results b. Frequent removal of slurry and storage in
statistically fall within the range of the wider set of closed tanks
annual mean data from the Netherlands. In the absence 4. Extraction of ventilation air with treatment to
of a sufficient dataset for Irish conditions, the presented reduce odour concentration
data from the Netherlands provide the best basis for a. Bioscrubbers
emission estimates for licensing. b. Chemical scrubbers
c. Biofilters
A flow diagram for the licensing assessment process for 5. Miscellaneous additives
odour annoyance is provided in Chapter 8. It provides a a. Feed additives
method to decide when to apply a straightforward b. Slurry additives
assessment, using standard contour lines. These standard
contours are provided in Annex E and can be used as Generally speaking, low-emission housing systems can
overlays on a map to estimate the area where exposure is achieve a reduction of odour emissions of up to 50%.
in excess of a limit or target value for odour impact, for This approach is best implemented in the course of the
a particular number of animals. In more complicated normal cycle of replacement of assets. End-of-pipe
cases, where decisions can become borderline, full treatment of ventilation air is generally speaking not an
modelling is proposed as the preferred assessment option that is economically feasible. The volume flow of
method. Detailed information is provided on how to ventilation is large in the relevant season (summer)
conduct assessments of odour impact for licensing because it is the main regulating mechanism for
3
temperature and hence is directly related to the welfare European legislation.

and growth of the animals. The odour concentrations are
relatively low from the perspective of air treatment 2) An assessment framework based on quantitative
systems, with concentrations of a few thousands ouE/m3. emissions is most likely to achieve a transparent
These factors combine to cause significant additional licensing practice that achieves a balance between
costs, in the order of €25 per pig produced. the interest of the pig producer and those who use
the surroundings as their living environment.
Chapter 9 provides detailed information on odour
abatement methods, both process integrated and end-of- 3) The proposed assessment framework identifies one
pipe methods. Indicative cost information has been environmental target for all situations. To allow for
included when available in the literature. The economics a degree of flexibility two limit values have been
of installing the technology to abate odour emissions set, one for new production unit applications and
must be assessed before the technology is imposed on one for existing facilities. The ‘space’ between the
existing pig unit operations. target and the limit values can be used in the
licensing process to tailor the conditions to the
Chapter 10 summarises two case studies that were specific requirements and opportunities that exist
conducted at pig units in Ireland for this report. In the for that licence application.
course of these case studies a limited programme of
emission measurements was conducted, with the aim of 4) The proposed framework for target and limit values
checking whether the results would fall within the range is, in general terms, compatible with the setback
of values for emission factors found in the recent Dutch distances required or advised in other countries,
study. The measurements in Ireland yielded a geometric such as Germany, the Netherlands, New Zealand.
mean value of 13.2 ouE/s per finisher, in winter
conditions, which was approximately one third lower 5) The prevailing wind direction in Ireland causes a
than the annual mean of 22.6 ouE/s per finisher from the distribution of odours that is not symmetrical. The
Dutch study. A Belgian study found a figure of 15.4 actual meteorology of the pig unit’s location and the
ouE/s for fatteners in winter conditions, very similar to position of the receptor relative to the source are,
the Irish emission. However, given the variance therefore, a greater factor than in most countries,
observed within and between farms in the Netherlands, where the wind rose is more uniform and resultant
the difference is too small to conclude that Irish contours circular. These particular circumstances
emissions are systematically lower. To draw that increase the need for specific modelling, in cases
conclusion, a larger programme of odour emission where the outcome is not clear-cut.
measurements would be required.
6) The geometric mean emission rate of 13.2 ouE/s per
Chapter 11 contains the conclusions, which are repeated finisher measured in winter conditions in Ireland for
in this summary: this study is about one third lower than the value of
22.6 ouE/s per finisher found in a larger study in the
This report is not answering a well-defined question, Netherlands.
with a concise set of conclusions. It aims to provide a
framework that will provide a balance between the 7) Given the relatively small number of samples,
economic interests of the pig producer and the collected in the Irish study, and the statistical
environmental interests of those using the vicinity in variance as derived from the larger Dutch study, the
which to live, work and play. difference in the mean outcome is too small to be
statistically significant.
On the basis of the issues explored in this report, a
number of general conclusions can be made: 8) It is, therefore, justified to use the emission factors
derived in the Netherlands for emission estimates in
1) A significant number of pig production units will Ireland, as long as emission factors specifically
require a licence, based on current National and measured in Irish conditions are not available for a
4
larger sample of study sites. the general public.
9) Options for reducing odour emissions from pig

production exist. Reductions to 50% relative to the
most common fully-slatted production unit are quite
feasible. However, the financial viability of many
retrofit methods is an issue of concern, given the
low economic returns on pig production.
10) The economics of installing technology to abate

odour emissions must be assessed before this
technology can be imposed on existing pig unit
operations. The most viable low-emission options
involve modification of pig houses, or replacement
by new low-emission design housing. In most cases,
such structural abatement can only be reasonably
achieved in the normal economic cycle of asset
replacement.
11) Retro-fitting of abatement, using air treatment

systems such as bioscrubbers, chemical scrubbers or
biofilters, can achieve significant emission
reductions of between 70% and over 95%. The main
impediment is the additional cost incurred, which
can increase the cost of a pig produced by roughly
10-20%. Market conditions in recent years,
generally speaking, do not allow such an increased
cost.
12) Good operational practice, including suitable

landscaping, tree screens and pro-active community
relations, remain a main factor in reducing
annoyance and avoiding annoyance developing into
nuisance.
13) A suitable production site for a given production

capacity will become a major asset for any pig
producer and may become a main factor in
determining the sustainability of the activity.
Producers are well advised to use the planning
process to their advantage and be pro-active in
counteracting any encroachment into the existing
setback zone by developments that may be termed
‘odour sensitive receptors’.
14) By making the assessment of the impact in the

vicinity of pig production transparent, the proposed
framework can contribute in practice to the
protection of the interests of both pig producers and
5
6
2. Scope of the study
2.1 Scope framework. It also reviews the practices, methods and

technologies that are available to reduce the odour
Livestock odours, in particular those caused by pigs, impact of pig production units.
have been a subject of study and of regulation for some
considerable time. Guidelines for planning and To illustrate the application of concepts outlined in this
licensing, aimed at maintaining adequate buffer zones report, three pilot studies were conducted to illustrate
between pig units and residents, have been introduced in practical implementation of the proposed assessment
some European countries as early as 1971. The nature of framework.
pig production has changed since then, while the
environmental impacts of pig production have become a This report will assist the Agency in formulating its
major consideration. Nutrient recycling and disposal, approach to processing the licence applications and to
ammonia emissions, greenhouse gas emissions and achieve transparent and uniform decision-making on
odours are now significant issues determining the odour issues for that purpose.
sustainability of the industry. The industry has
developed various technologies in response to these Part A of the report provides the background to the
challenges. New types of pig housing and manure assessment framework, on the basis of literature study
storage have been developed, feed technology is and direct contacts with experts and industry
evolving and even air treatment to reduce emissions of representatives.
ammonia and odour to air is under consideration.
In Part B two case studies for Irish pig production units
In Ireland, the pig production sector has seen are reported, using the methodology and approach as
considerable change over the past 12 years. The suggested in Part A of the main report. Part B includes a
production of pig meat in Ireland has doubled, while the chapter reporting a limited number of measurements
number of pig producers has been reduced. The trend that were made to assess whether emissions from pigs in
towards larger units is expected to continue. Ireland were within the range of values found in a more
extensive survey in the Netherlands. The Dutch data
IPC licensing for pig production units above a certain were used to derive emission factors for application in
size is currently being introduced, on the basis of the Ireland as proposed in this report.
EPA Act of 1992 and European requirements.
2.1.1 Study objectives
As a result, close to 200 pig units will be applying for
IPC licences in Ireland in the next few years. The The study reported here aims to achieve the following
assessment of the odour impact of these pig units will be objectives:
an important element of the licensing process, which
will be carried out by the Environmental Protection • Review and evaluate methods for odour impact
Agency. assessment and prediction.
To implement licensing, a transparent framework for • Propose criteria and standards to avoid impairment of
assessing the odour impact of livestock production units amenities adjacent to the production unit.
is required, providing consistent criteria to avoid
impairment of amenity in the vicinity of production • Identify and review options for reducing odour
units. Suitable criteria for ‘acceptable exposure’ to generation.
odours are required, in order to set a framework of
environmental quality criteria: limit and target values. • Identify and review odour abatement options suitable
This report aims to provide the basis for such a for retro-fitting to existing production units
7
2.1.2 Structure of Part A of the report. In Chapter 10 the results of three case studies are
presented. In addition to demonstrating the principles of
After providing a general overview of the characteristics odour impact assessment as proposed in this report, the
of the pig production sector in Ireland in Chapter 3, the results of a limited measurement programme are
principles of odour assessment are discussed in Chapters presented. The measurements, performed at two study
4 and 5. Aspects of formation of odourants, release to sites, were aimed at determining emission factors for
atmosphere, dispersion, exposure and detection, finishers, and to compare the results to emission factors
perception and the factors that determine whether an derived from a larger study in the Netherlands.
odour becomes an annoyance or a nuisance will be
explained. This introduction will provide basic Chapter 11 lists the main conclusions of the report.
knowledge of the way our human sense of smell works,
in the context of environmental odour annoyance. The At the end of this part A of the report, a number of
relationship between perception of odours and health annexes are included, giving detailed information that is
and well-being and the process that can lead from referred to in the main report, followed by the
perception of environmental odours to odour annoyance References.
are discussed. The main factors in the process that
determine whether annoyance will be an issue will be • Annex A Odour regulations for intensive livestock in
identified. More detailed background information is other countries gives details on national regulations,
provided in annexes. summarised in the main report.
The dose-effect relationship, between the calculated • Annex B Methods for odour assessment and units of
exposure to odour and the percentage of people measurement provides a detailed description of odour
‘annoyed’ by odours, is discussed in Chapter 6. This assessment methodology, and the units and concepts
relationship is highly relevant, as it provides the used to report on odour measurements.
underlying data for setting targets for environmental
quality, at a level that reflects the needs of a particular • Annex C Glossary contains an extensive list of terms
society. and definitions for odour related concepts.
An overview of the regulatory approach to pig odours as • Annex D References lists the literature references
an environmental issue in other countries is provided in
Chapter 7. In Chapter 8 a framework for assessing and • Annex E Contour plots for screening assessment
managing odour annoyance issues in the licensing provides standard contour plots for different pig unit
process in Ireland is proposed, specifically aimed at the sizes
pig production sector. Prevention of odour problems is
the guiding principle. The framework provides a 2.1.3 Conventions
systematic approach to achieving a defined minimum air
quality target for odour through application of best A number between square parentheses indicates a
practice in operation and design, combined with reference to the literature, e.g. [1]. These numbers refer to
adequate setback distances and possibly specific the list of references, in Annex D References. All costs
measures to reduce odour impact to an agreed set of are expressed in Euros (€).
limit and target values.
Chapter 9 describes the causes of formation of

odourants. Methods to mitigate odour impact by
reducing odourant production are reviewed, including
methods for reducing the release of odourants and
methods for reducing the impact in other ways, such as
using slurry additives or end-of-pipe air treatment, e.g.
chemical scrubbers, bioscrubbers, biofilters etc.
8
3. Introduction to the characteristics of pig production

in Ireland
This chapter provides an overview of the main fattened pigs. Each sow produces an average of 22.1
characteristics of the pig production sector in Ireland. pigs in a year. The weight of a fattened pig when
The information provided by the Teagasc Pig Advisory slaughtered is typically about 90 kg live weight. An
Service has been invaluable in preparing this chapter. average of 225 kg of feed is required for each pig
produced. Approximately 25% of the live weight is
3.1 Overview of the pig production removed in the evisceration process during slaughtering,
sector and its economics yielding a carcass of 70 kg.
Pigs are produced to be sold at a profit. That implies that The average size of Irish pig production units is
economics determine the dynamics of the pig relatively large, at 316 sows/unit (1998), compared to
production sector. The profitability also determines the Denmark (103) or the Netherlands (137). Currently
ability of the pig producers to invest to adapt to market there is a trend towards fewer, but larger units. A
conditions, including environmental requirements, such detailed overview of the size distribution of Irish pig
as odour abatement. production units is presented in Table 2.
The profitability is largely determined by feed cost and The density of pigs in Ireland, expressed as the number
the market price for pork and is subject to significant of hectares of farmed land per sow, is relatively low,
fluctuations, see Table 1. Since July 1998 the Irish pig ranging from 4 to 69 with an average of 26 hectares per
sector lost substantial sums. Currently, in 2000, the price sow. Other European countries keep much higher
for pork is rising again because of short supplies on the densities of pigs, see Table 3.
increasingly globalised market. Of the meat produced,
65% is exported to the United Kingdom. Smaller production units currently do not require a
licence. When the current licensing threshold is applied
The Irish pig production sector employs approximately (see section 8.2) an estimated total of 191 production
6000 people in pig production, slaughtering, processing, units will require a licence, (143 integrated units, 30
feed production and facilities supply. breeding units and 18 fattening units)
The investment per integrated sow, in Euro, is between
€2000 and €3200. 3.1.2 Strengths of the Irish pig industry
3.1.1 Distribution, size and type of pig production The Irish pig industry has few natural advantages and
units has managed to survive by being very competitive.
Relatively large unit size in comparison with other EU
The pig production sector in Ireland includes 657 countries means that pig producers are specialists and
commercial pig farms, with a total of approximately can employ specialist staff, maintaining a high level of
175,000 sows. In 1999 they produced 3.8 million technical expertise. Larger units have economies of
Table 1: Cost, price and profit per kilogram of pig meat produced in Ireland, 1992 to 1999.
Pig production, cost and prices, in Euros per kilogram

1992 1993 1994 1995 1996 1997 1998 1999
Finisher pig price 1.47 1.28 1.28 1.42 1.64 1.43 1.13 1.02
Feed cost 0.97 0.94 0.90 0.86 0.91 0.88 0.81 0.76
Margin over feed cost 0.51 0.34 0.38 0.57 0.72 0.56 0.33 0.27
Estimated non-feed cost for large units 0.39 0.39 0.38 0.41 0.41 0.42 0.42 0.44
Estimated profit / (loss) 0.14 -0.05 0.00 0.17 0.32 0.14 -0.09 -0.18
(Source: Teagasc Pig Advisory Service)
9
Table 2: Size distribution of Irish pig production units, Table 3: Density of pigs EU Countries and regions,
and numbers of sites to be licensed at 1994 (pigs/km2 utilised agricultural land).
current licensing limits.
Country Average Most dense Density in
Integrated breeding and finishing units density region the most
unit size units Total sows dense
region
sows
[pigs/km2] [pigs/km2]
20 to 49 24 930
Netherlands 724 N.Brabant 2338
50 to 99 48 3450
Belgium 518 Flanders 1356
100 to 299 207 38335
Denmark 396 - -
300 to 499 56 21345
Germany 139 N.Rhine-Westphalia 367
500 to 999 61 38720
Spain 61 Catalonia 375
1000+ 26 42585
France 49 Brittany 434
total 422 145365
Italy 46 Lombardy 256
To be licensed: 143 102650
UK 42 Yorkshire 181
Breeding units
Ireland 32 Cavan 220
unit size units Total sows
(Source: Eurostat, Statistical Yearbook Regions, 1997)
sows
20 to 49 31 1080 use of technical advice combined with performance
50 to 99 30 2135 monitoring and appraisal has contributed to a high level
100 to 299 40 7180 of productivity.
300 to 499 14 5345
500 to 999 11 6825 The island location is an aid to the maintenance of the
1000+ 5 6960 health status of the Irish pig herd. The introduction of
total 131 29525 liberal EU animal movement policies places a greater
To be licensed: 30 19130 responsibility on individual producers, especially
Finishing units breeders. Efficient use of information generated at post-
unit size units Total no pig slaughter veterinary inspections is increasingly required
sows finishing places to maintain the health of the Irish herd.
150 to 499 25 7450
500 to 999 27 18950 The density of pigs in Ireland is relatively low, see Table
1000 to 2999 33 55700 3. The contribution of pigs to manure output or nutrient
3000 to 4999 11 41400 input into agriculture is therefore less than in some other
5000+ 7 45500 EU countries, where this issue poses a serious restriction
total 103 169000 to further growth in production (e.g. the Netherlands). In
To be licensed: 18 142600 most cases land for manure spreading is available within
reasonable distance of pig units. Restrictions on pig
production capacity in some other EU regions of high
scale in staff training, purchasing and selling. Smaller animal density is anticipated, allowing an opportunity
units have tended to use selling groups and to a lesser for expansion of the sector in Ireland.
extent purchasing groups to achieve an adequate scale.
The progression towards integrated production has also 3.1.3 Operational characteristics
contributed to greater stability in the industry.
Most pig production units in Ireland are integrated units,
Rapid adoption of new technology has been a feature of where the entire production cycle takes place in one
the industry and explains, for example, the relatively location. The life cycle of pigs in such a unit is
high level of sow productivity in Ireland. The industry summarised in Table 4.
has tended to follow developments in Denmark and the
Netherlands where innovation has been more rapid than The housing system of choice is currently fully slatted
in the UK, which was the traditional model. Widespread pig houses, with underfloor slurry storage. The
10
Table 4: Pig life cycle in an integrated production unit
Pig life cycle Start End Weight Weight Average Optimum

start end growth rate temperature
range
day day kg kg kg/day Celsius
Dry sows, gestation -115 0 19 to 20
Farrowers 0 26 10 to 20
Piglet, birth to 6.5 kg 0 26 6.5 0.3 32 to 34
Weaner, 1st stage, 6.5 - 15kg 27 53 6.5 15 0.3
Weaner, 2 nd stage, 15 - 35kg 54 89 15 35 0.6
Finishers, 35 - 93kg 90 175 35 93 0.7 14 to 20
European directives on animal welfare will require can be delivered to the pigs separately, in a dry feeding
considerable modifications in the existing housing system, or pre-mixed in a wet feeding system. Most
systems by the end of the year 2005, which may present larger farms use wet feeding. Only a minority of farms
an opportunity to involve odour emission as one of the use food by-products, e.g. whey, yeast, beer, dough, etc.
selection criteria in selecting suitable housing systems. The feed in Ireland may be different from continental
European countries in having a slightly higher protein
The typical production parameters for a typical 100 sow content. In countries where ammonia emissions and
integrated pig production unit are presented in Table 5. nutrients are a regulatory issue, (e.g. the Netherlands),
The table shows that in an integrated unit, the odour farmers tend to limit protein content to the minimum.
emissions from fatteners are the dominant emission, at Otherwise, feeding practice in Ireland is similar to that
72% of the total. This is compatible with the fact that in other European countries.
fatteners consume between 60 and 65% of all feed in an
integrated unit. The differences in operational practice and environment
for pig production in Ireland, relative to the Netherlands
The emission of the total farm pig population, calculated are outlined below, as these are relevant to explain
per sow in a typical integrated unit is approximately 160 possible differences in emission factors used in this
ouE/s . In a breeding unit, the emission per sow is very report. Although the production methods in Ireland and
much lower, at approximately 45 ouE/s. in the Netherlands are bound to differ, the similarities
are greater than the differences as far as odour emissions
The feed used in Ireland is mainly meal and water. This are concerned. The following differences in the
Table 5: Typical characteristics for an integrated pig unit with approximately 100 sows.
Number of animals, floor areas and emissions per animal stage.
Typical parameters for approximately 100 sows in an integrated unit

Stage Animals Animals Floor Area per Emission Emission per
places area animal per animal 100 sows and
per sow progeny
no. m2 m2 ouE/s ouE/s % of emission
Sows, farrowing 23 0.22 94 4.09 18 396 3%
Sows, dry 82 0.78 130 1.59 19 1492 9%
Maiden gilts 15 0.14 21 1.40 20 323 2%
Boar 2 0.02 12 6.00 20 43 0%
Weaner stage 1 176 1.68 25 0.14 6 838 5%
Weaner stage 2 240 2.29 84 0.35 6 1371 9%
Finisher 525 5.00 368 0.70 22.5 11300 72%
15764 100%
4464 Breeding unit
11
production practices in Ireland and the Netherlands can

be listed:
1. The slaughter weight in Ireland is somewhat lower,

which implies lower feed usage per sow per year and
lower nutrient emissions.
2. In Ireland the rearing of hogs (boars) for meat is more

common than in the Netherlands. Boars are more
efficient than castrates, use less feed etc.
3. The mean unit size in Ireland is larger than in

Holland.
4. Marginally lower summer temperatures would be

expected to require lower ventilation rates, and also
result in slightly lower manure store temperatures.
5. The use of wet feeding systems in Ireland is

widespread, which reduces dust emissions.
6. Higher nutrient density feeds are used in Ireland, with

less manure produced and less fermentable residue in
the manure.
7. Raw protein content in the Netherlands is generally

lower, to reduce ammonia emissions, which are less
of an issue in Ireland.
8. The emission factors in the Netherlands were

obtained for partly slatted pig houses, while the fully
slatted systems in Ireland are more common. Fully
slatted systems will have a higher emission per pig.
Overall, it is not plausible that the impact of these

differences is significant. However, only actual
measurements on a larger scale can confirm actual
specific emission factors for Irish conditions, and their
variation from unit to unit.
12
4. Introduction to odours as an environmental stressor
Summary: This chapter aims to provide a level of memory. The nerves that connect the sensor to the brain
operational knowledge that is deemed adequate as lead directly to the hippocampus, which is the part of the
general background information for professionals brain that regulates basic functions, such as the
involved in odour annoyance management. The organisation of long-term memory and emotions. It is,
following themes are discussed in this chapter: therefore, not surprising that smells are often highly
associative and can elicit vivid memories of experiences
• Odour perception: the function of our sense of smell that occurred even in early childhood. This associative
and its evolutionary development; aspect is highly relevant to environmental odours. Once
a negative association is formed, it is very difficult to
• Characterising odours: the various attributes used to change the appreciation of that particular odour stimulus
characterise odours, and the method of measurement; in an individual. This helps to explain why an odour
problem from the past often seems to haunt site
• The mechanism that leads from the production of operators, even after the odour emissions have been
odourants from pig production units to odour significantly reduced.
nuisance complaints
When an ambient odour is detected by our senses it
4.1 Odour perception starts a chain of events. During sensory perception, the
detectability, intensity, and character of the odour
The chemical senses, for smell (olfaction) and taste stimulus are determined. This information is then
(gustation), are generally considered to be the oldest processed in the brain, in the cognitive appraisal
ones in evolutionary development. Although humans are process. At this stage the perception information input is
a relatively recent development in evolutionary terms, combined with various sources of reference information,
the function of our sense of smell is the same as for other such as the history of perception, associative
species: it helps us to evaluate our environment. In information with previous similar perception events,
simple terms of behaviour, perception of odours can lead information on the current visual, social, etc.,
to two basic behavioural responses: avoidance or behavioural status and information about the
approach. These responses can occur for example in environmental context etc. If this appraisal leads to a
judging food or water, but also in a social or sexual negative appreciation of the perceived odour, in the
context. current behavioural context, the relevance needs to be
determined, and the appropriate behaviour needs to be
The human sense of smell helps us to assess our displayed in response. This phase of the process is
environment in a very direct manner. The sensor in the characterised as ‘coping’. One type of ‘coping’
nose cavity is a direct interface between the brain and behaviour involves undertaking actions to remove the
the environment. It is a highly sophisticated sense, cause of the negative appraisal (remove the source).
which interacts with our life and behaviour on many Another type of ‘coping’ is aimed at reducing the
levels. The process of odour detection, perception and emotional impact of the negative appraisal, by
evaluation is therefore understandably complex. ‘reasoning’ that the cause is not so relevant after all and
Humans can detect and differentiate up to 3000 odours. is better ignored. Repeated ‘annoyance events’ as a
Recent research indicates that as many as 1000 genes result of ambient stressors, such as odour, over a
out of the total of 100,000 in our genome are dedicated considerable period of time, may lead to nuisance,
to our sense of smell. This significant proportion of 1% which in turn may result in complaints.
suggests that the sense of smell is of considerable
importance in evolutionary terms. Most odours can cause odour annoyance when they are
intermittently clearly detectable. Even odours that are
The sense of smell is closely related to long-term commonly not identified as unpleasant, such as coffee
13
roasting odours, cause odour annoyance in a population units used to report measurement results are described in
that is exposed to sufficiently high concentrations of some detail in chapter Annex B of this report, titled
odours intermittently, but regularly, over prolonged Methods for odour assessment and units of
periods. measurement.
4.2 Characterising odours: 4.2.1 The common traits of units used for odour
psychophysical dimensions of odour and noise assessment
perception
Our senses are equipped to respond to a wide range of
The sensory perception of odourants can be stimuli. Our ears detect faint sounds, but can also cope
characterised by four major attributes or dimensions: with very loud noises: from a whisper to a roar. We can
pick out a specific conversation in a room full of loud
• detectability; conversation. The response characteristics of our sense
of smell are similar. It is, therefore, helpful to consider
• intensity; the way in which we describe environmental noise, and
identify the relevant similarities with units used to
• hedonic tone; characterise odours.
• odour quality. The stimulus for noise is vibration of the air. The energy
of that vibration determines the strength of the stimulus.
A fifth attribute has been proposed recently[1], to This energy is measured in linear units, Watts per square
characterise the propensity of an odour to cause odour metre, (W/m2). The energy that is picked up by the
annoyance. However, no operational method for human ear is an even smaller quantity. Our eardrum is
characterisation and interpretation is available for this only 1 cm2, or 10-4 m2. Our ear, therefore, can perceive
fifth attribute: an energy level of 10-16 W, which is very little.
• annoyance potential. The loudest noise that we can perceive is close to the
pain limit, where hearing turns to hurting. The stimulus
In addition to the sensory dimensions used to describe there is 10 W/m2, or an energy uptake of 10-3 W by our
how odourants act when perceived as odours, efforts are eardrum.
ongoing to devise a more technical approach to
characterise odourants, using analysis of the chemicals As powers of ten are not the most intuitive of measures,
involved. The approach can be to look at simple key we use logarithmic measures to describe these stimuli,
substances that can be perceived as an odour, such as where the number of zeros, or the powers of ten, are
H2S or ammonia. It may involve measuring a tracer more important than the difference between two
component that is non-odorous itself, but occurs with numbers like 3000 and 5000, for example.
the odourants, e.g. methane as a tracer for landfill gas.
Finally, an attempt can be made to actually measure a The idea to describe a signal, or stimulus, in terms of a
multitude of odourants in the mixture, using advanced logarithm of the proportion (or ratio) between the actual
analytical methods such as GC-MS or ‘electronic nose’ value and the detection threshold, is attributed to
devices. The practical application of such methods is, so Alexander Graham Bell (1847-1922), the inventor of the
far, limited. The sensitivity of the analytical methods is telephone. He was interested in describing the strength
usually not nearly sufficient to approach that of the of signals, and coined the unit: bel.
human nose, and the poor capability to predict or model
the actual odour perception in humans on the basis of
measured parameters is poor.
The different dimensions used for characterising odours,

the methods for assessment of odour samples and the
14
In mathematical terms: The anchor for noise is the detection threshold for a
sound of 1000Hz in W/m2, experimentally determined
L = log(I I )
0 to be approximately 10-12 W/m2 in young people.
where I is the strength of the signal, and I0 is the smallest
detectable signal. As this measure is a bit crude, the A sound at the upper end of the sensory range of hearing
decibel is more common: (extremely loud) is 130 dB. In linear terms, the energy
of that sound is 10,000,000,000,000 times the detection
L = 10 . log(I I )
0 threshold, in linear units of W/m2.
The decibel is best known for describing noise levels.
In noise levels, the reference value I0 is the detection For odour, the range between detection thresholds and
threshold for a sound of 1000 Hz frequency, established unbearably strong smells is smaller, but still
experimentally, in sensory experiments using young considerable. Odours at the high end of the intensity
people as panel members. range (extremely strong) may contain hundreds of
thousands or even millions of ouE/m3. Therefore the
Therefore, I0 = 10-12 W/m2 at 1000 Hz, and a stimulus of range of odour intensities, in dBod, is open ended, but
10-12 W/m2 0 dB. relevant in the range of 0 to 60.
The same model can be applied to odour intensity[2]. In Although the dBod has been proposed some time ago[2],
the European standard EN13725 the threshold value, or and is also included in the draft standard EN13725, it is
zero odour decibel is defined as equivalent to an odour not commonly used. When interpreting odour
of 40 ppb n-butanol. The choice of a particular odour is measurements it is, however, useful to realise that the
not dissimilar to defining a particular frequency for odour concentration, ouE/m3 is a linear unit, just like
noise. W/m2 for noise. The principal similarity between these
units is that their relation to perceived intensity cannot
So, if 0 dBod 40 ppb n-butanol = 1 ouE/m3, then odours easily be interpreted intuitively. These linear parameters
can be expressed, just like noise, in decibels; dBod. tend to reach very large values, making clumsy numbers
in practical use.
For the reference odour, the mathematics work very
well. The stimulus of 4000 ppb = 4 ppm n-butanol can By using dB units, similar to noise, odour intensity can
be described as: be associated with a more tangible unit, directly
comparable to the decibel used in noise assessment. For
L = 10 . log(4000 40) = 10 . log(102) = 20 dBOD example, a reduction in odour concentration by a filter
with 90% efficiency amounts to a reduction with 10
For other odours, the reference may be different. For dBod while a filter performing at 99% abatement
sound a number of different reference levels are used, efficiency achieves a 20 dBod reduction in odour
hence the variety in decibels: dB(A), dB(B) and dB(C), intensity.
all with a slightly different reference value, each defined
as a spectrum of a defined set of frequencies. 4.3 The mechanism leading from pig smell
to odour nuisance
So, the strength, or intensity of both noise and smell can Pig producers have a responsibility to minimise the
be defined on the basis of a detection threshold for a impact of their activity in the vicinity of their production
particular stimulus in people. site. They have a legal obligation to avoid impairment of
amenities. Odours are probably the predominant
The anchor for the odour unit is the detection threshold, nuisance issue for pig producers, with the potential to
which has been defined using 40 ppb of reference reach well beyond the limits of the production site.
odourant n-butanol, based on tests with human subjects.
This detection threshold is described by 1 ouE/m3 which Odour nuisance can develop after long-term intermittent
is equivalent to 0 dBod. exposure to odours that cause a negative appraisal in the
individual concerned. It has to do directly with the way
15
we value our environment. It is not a straightforward

process. Our attitudes towards the source, the From odour formation to complaint
inevitability of the exposure and the aesthetic
expectations regarding our residential environment are Odourant formation process
some of the less tangible factors that are relevant to the
probability of experiencing nuisance. Once the balance Transfer to air
tips, and an environmental stressor, such as a livestock
odour, becomes a nuisance to an individual, it is very Release to atmosphere
difficult to reverse the process. What used to be a faint
odour has now become a stimulus associated with Dispersion
annoyance. Once the first complaint has been made, the
problem becomes much more serious than before for all Exposure
those affected.
Frequency of exposure
Duration of exposure
The mechanism that leads from an emission of
Intensity of exposure
odourants to the atmosphere to actual odour nuisance is
quite complex. It involves the following main factors:
Detection
and
• The characteristics of the odour that is released perception
(detectability, intensity, hedonic tone, annoyance

potential); Time of day/activity
Context
Relation to source
• Variable dilution in the atmosphere through turbulent Association with odour
dispersion (turbulence or stability of the boundary
layer, wind direction, wind speed, etc.); Appraisal
by
receptor
• Exposure of the receptors in the population (location
of residence, movement of people, time spent
Receptor characteristics
outdoors, etc.); perception of
individual health
• Context of perception (i.e. other odours, background

of odours, activity and state of mind within the Annoyance
perception context);
Receptor characteristics:
Other ambient stressors
Coping strategy
• Receptor characteristics (exposure history, noise
Attitude to 'status quo'
crowding
association with risks, activity during exposure Relation to source
dust
(economic)
episodes, psychological factors such as coping
behaviour, perceived health and perceived threats to
health). Nuisance
This process can be summarised as:

Access to complaint
formation of odourants ➛ transfer to atmosphere ➛ channel
Expected result of complaint
atmospheric dispersion ➛ exposure ➛ Access to legal instruments
population ➛ perception ➛ appraisal ➛
annoyance ➛ nuisance ➛ complaints
Complaint action
Figure 1: Pathway of odour from production to receptor
16
When we look at the underlying mechanisms, the factors

that play a role are more diverse and mutually
interactive, as is illustrated in Figure 1.
For practical purposes, such as regulatory use, the

complex relationship between nuisance (effect) and
exposure to odours (dose) can be described in a
simplified model that does not take into account all these
different factors. The dose-effect model linking
‘exposure to odours’ to ‘nuisance’ is typically described
as the relation between modelled exposure and
annoyance as measured by a standardised telephone
questionnaire or, alternatively, complaint records.
Epidemiological methods are used to describe this
relationship.
The exposure is typically quantified in terms of a

frequency of occurrence of hourly average
concentrations above a certain limit odour
concentration, e.g. 5 odour units per metre cubed
(ouE/m3) as a 98-percentile of hourly averages of odour
concentration for a year with average meteorology. In
short notation: C98, 1-hour = 5 ouE/m3. This measure of
exposure is calculated from an estimated or measured
odour emission from the source, using an atmospheric
dispersion model.
Air quality criteria for odour can be set on the basis of

combining calculated exposure with knowledge of the
dose response relationship to quantify and assess odour
impact. However, this relationship will not be the same
for every community. It is determined by factors such as
crowding, expectations of environmental quality,
economic priorities, etc. Although odour can have direct
effects on well-being, and hence on health, it is to some
degree an aesthetic factor in environmental quality.
To set environmental exposure criteria with a view to

avoiding odour nuisance is, therefore, not only a
scientific, but also a political process. The range of
political discretion is limited, however. Unlike other air
pollutants, every citizen with a functioning nose can
assess odour real-time. The appraisal is immediate and
the outcome is readily communicated to the relevant
authority in the form of complaints.
17
18
5. Impact assessment methodology
This section describes different methods for assessing applied relatively widely[3]. A specific large scale dose-
odour impact. Some of these methods start with the effect study for pig production odours has been carried
effect, by primarily studying the people involved and out there recently[4], and is an important contribution to
their behaviour. This same perspective can be used the data underlying this report (see also section 6.2). The
experimentally, using trained field panels for conducting method requires specialised expertise. Specialised odour
field observations, either for a short period of time or for survey firms with suitable experience are required to
a period of many months. apply the methodology successfully.
The most common method for impact assessment, The cost of such a survey is in the order of €15,000, for
however, is to use knowledge of the dose-effect one site. Its application in specific licensing cases is
relationship between odour exposure and annoyance to limited, as in most cases the number of people exposed
predict annoyance levels on the basis of calculated is insufficient to apply the method successfully. Direct
exposure. The starting point for calculated exposure is measurement of annoyance is a valuable method to
measurement of odour emissions at the source. determine the underlying dose-effect relationships, in
carefully selected case studies.
Measuring odour concentrations ‘at the site boundary’ is
not an effective method, even though it would fit well 5.1.2 Complaints analysis
with the legal approach to nuisance. The variability that
is introduced by weather conditions and the practical Complaint analysis is not covered by any standard
difficulty of measuring odours at very low method or recognised protocol. Usually complaints are
concentrations, ≤ 20 ouE/m3, are so far insurmountable registered by local, regional or even national authorities
methodological obstacles for measuring an interpreting or by companies who have a customer relations system
‘ambient odour concentrations’. that can be adapted for complaint registration.
5.1 Assessment of effects Complaints registration provides an insight into the

prevalence of a symptom of odour annoyance, but not of
5.1.1 Direct measurement of percentage of people the prevalence of the annoyance itself. There are many
annoyed factors at play that determine the ease or difficulty of
registering a complaint. Therefore, complaint data must
The Standardised Telephone Questionnaire (STQ, also be interpreted with some caution. Registered complaints
known by the acronym: TLO) is used to measure the are a very strong indication that odour nuisance is a
percentage of people annoyed in a sample of the reality in a specific situation. However, the absence of
population[3]. The main application is to determine dose- registered complaints does not necessarily indicate the
effect relationships, either in general or for a particular absence of nuisance. Also, once a conflict situation
site. develops over odour emissions, the registering of
complaints can become a tool in the fight, when
The TLO is typically applied to sufficiently large residents use orchestrated complaints as a political lever
samples of the exposed population, in at least four study to move the argument in their favour.
areas, with different exposure levels. By collecting at
least 100, preferably 200, TLO results for each exposure The approach in setting up complaints registration and
test area, a dose-effect relationship can be established. analysis must be determined and tailored to the purpose
of the registration.
There is no standard document with a protocol for this
technique, although a unity of application exists in
practice in the Netherlands, where the method has been
19
The minimum information that needs to be collected for 5.2 Direct assessment of odour exposure
each complaint is:
5.2.1 Direct field methods
• Location where the offensive odour occurred, within
approximately 100m. (i.e. address complete with 5.2.1.1 Field panels, short-term evaluations
house number);
Field panel measurements provide an estimate of total
• Date and time when the offensive odour was emissions from a source, including all diffuse sources.
observed;
Field panels consist of 4 to 6 trained, qualified panel
• Characterisation of offensive odour, preferably on the members selected using the same criteria as used for the
basis of a choice from standardised descriptors; odour laboratory, according to prEN13725. The field
panel makes observations on locations in the field,
• Preferably the identity of the complainant, to assess usually to determine the maximum distance of
repeated nature of complaints; detectability of the odour from a particular source. This
result, combined with the meteorological conditions
• Residential address of complainant. during the field observations, is used for ‘reverse
dispersion modelling’, which gives an estimated source
In complaint analysis each complaint should be verified emission rate as a result. Field panels can also be used to
and collated with additional information: provide information on odour intensity and/or hedonic
tone in field conditions.
• Wind direction, wind speed and stability class at the
time of complaint; A practical test procedure is described here. At any
given location the panel makes observations every 10
• Any process incidents at the time of complaint. seconds, for a duration of up to ten minutes. By
traversing the ‘plume’ at intersections at varying
The benefits of a complaints registration system can be distances, the results are gathered in the course of a
greatly improved by implementing a standard protocol number of hours[35]. The technique has been applied for
for complaint data registration and processing. a number of years, in some countries, in applied odour
Professional advice, including co-ordination with research. An unofficial guideline for carrying out these
complaint registration units of the local authority or measurements has been published in the Netherlands[6],
other organisations, is advisable. while in Germany a guideline has also been published:
VDI3940:1993[7].
A quick and adequate response to complainants is vital
in those situations where community relations can be Field panels can not only be used for evaluating
improved. This part of the complaints response process detectability of the source as a whole but also as a more
should be regarded as a fully-fledged method of ‘analytical’ instrument by teaching the panel to identify
annoyance reduction, as it can be very beneficial indeed specific smells on-site and using this perceptive
to reduce anxiety in the complainant by adequate expertise to identify individual sources downwind.
response and supply of information. Using this technique the following information is
recorded: type of smell, intensity and relative annoyance
The results of complaint registration and response potential to the overall off-site smell. This provides
should be fully analysed periodically. useful qualitative data, although they cannot lead to
decisive conclusions as they reflect an assessment by a
limited sample of the population, only briefly exposed to
these odours.
The field panel work requires certain weather conditions

and characterisation of meteorological conditions during
20
measurements (wind speed, wind direction and stability VDI3940:1993 and in the GIRL[8].
class)
The method has been applied to a pig fattening unit of
The inherent uncertainty of the method of measurement 1760 fatteners. It was found that the criteria were not
is mainly determined by the inaccuracies involved in met at distances of up to 1000 m from the pig unit[9].
characterising the turbulence in the mixing layer of the
atmosphere, and the relatively poor capabilities of The long-term field panel method is useful, in that its
models to accurately predict short-term downwind methodology and approach are easily envisaged, and
concentrations. Generally speaking, the results of understandable. The considerable resources and cost
modelling impact on the basis of source emission data involved are limiting factors in its practical application.
will give a more reliable result. Field panel data can, Concerns have been raised about the statistical basis of
however, be invaluable in providing a field check based the experimental design, when the assessment
on actual conditions, especially where sources are programme is in any way limited because of practical
complex and include diffuse sources (i.e. natural and/or cost implications.
ventilation, large area sources etc).
5.3 Assessment of odour impact by
5.2.1.2 Field panels, long-term evaluations measuring emissions at source,
followed by dispersion modelling
In 1994 an odour regulation for industrial sources on the
basis of long-term field observations was introduced in This section describes the practical methods used to
the state of Nordrheinland-Westfalen: Geruchs assess odour emissions at source. The need for carrying
Immissions Richtlinie (GIRL)[8]. This guideline is based out such measurements for actual licensing purposes
on measuring the actual frequency at which odours can will be limited, as reliable emission factors for odour
be perceived in the vicinity of the source in question, emissions for pigs at various stages of their life cycle are
over a period of 6 to 12 months. available[10]. These emission factors were obtained using
the methods described in this section.
A number of fixed observation points are determined, on
a regular grid, access allowing. A number of observers If emission factors are available, sampling can be
are assigned to the task of making observations at these avoided, and estimated emissions can then be used as
points, according to a pre-determined schedule. The input for atmospheric dispersion modelling, as described
assessor makes observations at 10-second intervals, over in section 5.3.3.
a 10-minute period. The number of observations with a
positive detection is divided by the total number of 5.3.1 Sampling
observations in the 10-minute period, and a percentage
of positive observation is calculated. If the percentage is When sampling odours, appropriate Health & Safety
above a limit value, typically 10%, the measurement at procedures must be applied. The Source Testing
that point in that 10-minute interval is considered to be Association (UK) has drafted specific guidelines for
an ‘odour hour’. The frequency of ‘odour hours’ is used environmental sampling.
as the criterion to determine if a ‘relevant nuisance’
exists at that grid point. The limit value that is applied in Sampling must be carried out in accordance with the
Germany for residential areas is 10%, while for trade CEN standard prEN13725 [11]. Samples are collected in
and industrial zones a more lenient 15% limit is applied. odour sampling bags made from a suitably odour free
material, such as Nalophane. Odour samples must be
The method requires approximately 26 measurements, analysed as quickly as possible, but no later than 30
on different days, for each point, while allowing no more hours after sample collection.
than 5 of these measurements to be done by the same
assessor. The requirements for human resources are
therefore considerable.
The methodology is described in the guideline
21
5.3.1.1 Point sources flow path section [m2]

L=
covered area [m2]
Point sources must be sampled in accordance with the
CEN standard prEN13725 [11]. Once the hood factor L is known, the specific emission
rate can be calculated from the concentration measured
If a risk of condensation of the odour sample in the bag at the exit of the hood and the flow velocity V:
exists or when concentrations are expected to be higher Esp = Chood x L x V
than the measuring range of the olfactometer to be used,
dynamic pre-dilution on site may be required. In our example the hood factor L is 0.006944, and the
flow velocity V = 0.25 m/s, which implies that at a
5.3.1.2 Area sources specific emission rate of 1 ouE/m2/s, an odour
concentration of 144 ouE/m3 is measured at the exit of
Area sources must be sampled in accordance with the the hood.
general principles of CEN standard prEN13725 [11].
To establish specific emission rates from liquid or This implies that emission rates as low as approximately
sludge surfaces a sampling hood is the preferred 0.5 ouE/m2/s can be measured without coming too close
method. to the lower detection limit of the method of the odour
concentration measurement (approximately 50 ouE/m3).
In using the sampling hood method, the choice of the
correct parameters of operation is of vital importance. 5.3.2 Odour concentration analysis
The sampling hood must use a flow velocity under the
hood of 0.2-0.3 m/s and have a headspace height of no Odour samples must be analysed in compliance with the
more than 200 mm. The path length must be sufficient to draft standard EN13725 ‘Odour concentration
allow the concentration under the hood to reach values measurement by dynamic olfactometry’, see Annex B.1.
that can be readily measured using olfactometry. for details.
Generally speaking olfactometry becomes more difficult 5.3.3 Modelling of atmospheric dispersion
at concentrations below 50 ouE/m3 because of
background odours in sample bags etc. Once the odour emission rate from the source is known
(in ouE/s) the impact in the vicinity can be estimated.
The impact of an emission is very strongly determined
by the way in which the odour is diluted in the
atmosphere, while being carried towards the receptor by
the wind. The dilution can vary considerably, depending
on the meteorological conditions: wind speed and
turbulence of the atmosphere, also called atmospheric
stability. The meteorology of a site will be a major factor
in determining the impact of a certain release of odours.
To predict the impact as well as we can, computerised
mathematical models for atmospheric dispersion models
are used.
Figure 2: Example of Lindvall sampling hood

for area sources 5.3.3.1 Characteristics of suitable atmospheric
dispersion models
As an example a Lindvall-type sampling hood of
suitable dimensions is pictured in Figure 2. The section Dispersion models are used for predicting odour
of the headspace is 160 x 160 mm with a total path exposure with a view to assessing expected annoyance.
length of 6 x 960 = 5760mm. The ‘hood constant’ L can The relationship between odour exposure and
be calculated as: annoyance has been established in a number of
22
epidemiological studies, where a particular modelling

approach was used. When using dispersion models for
Atmospheric dispersion models
odour annoyance prediction, the objective must be to
apply the models that were used to establish dose-effect
Most dispersion models are Gaussian models, which
relationships in the underlying epidemiological case
assume the concentration profile across the plume to
studies. This implies that, although better atmospheric
follow a Gaussian probability curve.
dispersion models may become available, these can only
be applied to odour problems after their results have
been validated in dose-effect studies, or by using base
data from previous dose-effect studies.
The relationship between modelled odour exposure and

actual annoyance levels has been established using
models and data with the following characteristics:
• Gaussian plume models were used;
• to represent conditions for an ‘average year’ hourly

The expansion of the plume is modelled by
meteorological data for a period of at least 3,
mathematically representing the standard deviation
preferably 5 years were used;
as a function of distance to source, wind speed, and
atmospheric stability (turbulence). By repeated
• models were used to calculate one-hour average
calculation of each receptor point in the study area,
concentrations for all hours in the meteorological
for each hour of the weather data set, a frequency
dataset;
distribution of hourly concentrations at that receptor
point can be obtained. This distribution can be
• exposure was expressed as the concentration
characterised by the concentration that is exceeded
corresponding to a certain percentile of the
only 2% of time, in terms of hourly average
distribution of hourly values, usually the 98-
concentrations. This is commonly called the 98-
percentile.
percentile. By drawing a line on the map connecting
all points with the same concentration at the 98-
A commonly used model is the Industrial Source
percentile, for example at 5 ouE·m-3, an odour
Complex (ISC) model developed by the US
contour line can be shown on a map. In the area
Environmental Protection Agency and used as a
enclosed by the contour the exposure level
regulatory tool for atmospheric emissions in many parts
5 ouE·m-3 as a 98 percentile of hourly averages will
of the world. ISC is a Gaussian dispersion model, which
be exceeded. Outside the contour the exposure will
uses input data such as wind speed, wind direction,
be less than the given criterion.
atmospheric stability and height of the mixing layer to
determine ground-level concentrations at defined
receptor points.
In establishing odour exposure with a view to assessing

the risk of odour annoyance, similar models must be
used. Models that produce results that do not closely
resemble those commonly applied to odour problems
must be validated for that application before they can be
used.
For practical purposes, other Gaussian models that are

able to predict the frequency of 1-hour average
23
concentrations can be used. Data should be obtained for the ‘most representative’
meteorological station. This may not always be the
The results are presented as contour lines for specific closest station, especially where the issue of coastal
frequencies of occurrence of odour concentration with versus inland locations is concerned. Specialist advice
hourly average values above a certain limit value (air can be obtained from the Met Office.
quality criterion).
Meteorological data from individual stations are
In this report, the COMPLEX atmospheric dispersion available from a number of suppliers (e.g. the Met
model[26,27] was used to calculate overlay contours. This Office or Trinity Consultants in Austin, Texas), at a
US-EPA model, that is based on the widely used budget cost of approximately €1200 for 3-5 years data
MPTER and ISC models, has been adapted by from a station. The data has to be formatted so that they
OdourNet to accommodate numerous sources (up to 999 are suitable for use by the software of the dispersion
sources) and to provide high percentile values (e.g. 95, model.
98 or 99.5 percentiles of 1-hour average calculated
concentration) that are used to evaluate odour impacts. For Ireland, data are available for the following
Odours, by the nature of our olfactory sense, which meteorological stations, typically for the period 1993-
responds almost immediately to a stimulus, has a very 1998, and in many instances going back to 1990 or
much shorter time frame to cause effects in receptors earlier:
than most other common forms of air pollution.
However, dispersion models are not typically designed • Belmullet Peninsula
nor validated to be used at averaging periods shorter
than 1 hour (see also section 5.3.5). • Casement
For the standard overlay contours, the hourly • Claremorris

meteorological data for Claremorris meteorological
station were used, for the years 1993-1995 (inclusive). • Connaught Airport - 1997 and later only
5.3.4 On the choice of meteorological data • Cork Airport/Corcaigh
For modelling purposes suitable meteorological data are • Dublin Airport

required, consisting of 8760 hourly observations per
year for the following parameters: • Kilkenny
1) Wind speed • Malin Head
2) Wind direction • Rosslare/Ros Lair
3) Atmospheric stability, which can be: • Shannon Airport
a) Pasquill stability class (derived from cloud cover • Valentia Observatory

observation)
It must be noted that most of these stations are coastal
b) Monin-Obukov length (derived from stations that are not suitable for characterisation of more
measurement of incoming solar energy) inland locations, even when the distance to the coastal
station may be less than the distance to an inland station.
To avoid error because of year-to-year variations, a Also, data from some stations are not continuously
minimum of 3 years continuous hourly data are registered and are therefore not suitable for dispersion
required, while a 5-year period is considered preferable. modelling for odour impact assessments.
24
A more in-depth comparative analysis of differences In other words, a criterion of C98, 1-hour ≤ 6 ouE/m3 is
between the dispersion patters of different approximately equivalent
meteorological stations with a view to odour impact to C99, 1-hour ≤ 12 ouE/m3.
assessment would be advisable in the course of
formulating a regulatory odour guideline.
5.3.5 On the choice of percentile values
In setting exposure criteria different percentiles can be

used. However, these percentiles all reflect one
distribution of values, determined by the meteorology at
the location in question. Therefore, there is a
relationship between a certain limit value at the 98-
percentile, at the 99-percentile and at the
99.5-percentile.
In choosing a suitable percentile to reflect a certain

exposure level, both fundamental and practical issues
must be considered. A fundamental issue is that for
characterising the exposure conditions that determine
odour induced annoyance, the relatively rare hours with
high exposure are more determining than the majority of
hours when the exposure is average or below average.
This is a result of the exponential relationship between
concentration and perceived intensity (see section in
Annex B.2) and the psychophysical and psychological
processes involved. A practical issue is that the
uncertainty of the prediction of the model is becoming
greater at very high percentiles. The mean of the
distribution has a smaller margin of error than the ‘tails’
of the distribution. At the high end of the percentiles, the
outcome becomes increasingly vulnerable to the small
fraction of hourly meteorological observations that are
of questionable quality, because of instrument failure,
missing values etc. The 99.5-percentile, after all, is
determined by only 44 hourly observations in a year.
In this report, the 98-percentile is used to set criteria,

because this value can be seen as a compromise: it
reflects the upper ‘tail’ of the distribution, but is based
on the top 175 hourly observations in a year. To make
this value more representative, a minimum of three years
meteorological data is used, or 525 hours in total.
The relationship between the 99-percentile and the 98-

percentile, in the conditions that prevail in Ireland, is
roughly:
C99, 1-hour ≈ 2 x C98, 1-hour
25
26
6. Dose-effect relationship
6.1 The relative properties of pig odours

compared to other environmental
odours
Evidently not all odours are the same in their ability to

cause annoyance. To account for these differences in
annoyance potential in quantitative terms is not that
simple, however. That is the reason why most
calculations used to predict the impact of odours involve
the simplification of characterising odours in terms of
detectability only, using the odour concentration, which
does not take into account the different characters of
odours. The odour concentration reduces the question of
‘how strong and unpleasant is this odour’ to a detection
threshold, and the original odour is characterised in
odour units, or multiples of the concentration at
threshold. (See section 4.2.1 for more detail on
interpretation of odour measurement results.)
This simplification is useful, as it allows calculations in

concentration terms, compatible with the general Figure 3: Relationship between odour concentration and
concepts of air quality criteria. perceived intensity, for broiler house odour (top) and the
odour of pig slurry (bottom) after application.
From Misselbrook TH, Clarkson CR, Pain BF: Relationship
It is important, however, to be aware of the limitations of Between Concentration and Intensity of Odors for Pig
Slurry and Broiler Houses. Journal of Agricultural
this simplification, and to consider the characteristics of Engineering Research 55: 163-169, 1993.
the odour at hand, relative to other odours.
The relative odour annoyance potential of pig odours, A straightforward approach to comparing the odour
relative to other odours, is relevant to the question of annoyance potential of different odours is to ask a group
assessment, in that it can assist in providing points of of people to rank a list of 20 descriptors of odours,
reference with criteria that have been established for according to like and dislike. This approach has been
other odours. used in research for generic, everyday odours[13]. More
recently, this approach has been applied to rank
Exposure guidelines based on dose-effect relationships environmentally relevant odours, using groups of people
for a number of industries were established during the who deal with odour annoyance professionally[14]. The
1990s in the Netherlands[3] and range from C98, 1-hour ≤ ranking order of a list of 20 industrial and agricultural
0.5 ouE/m3 to C98, 1-hour ≤ 3.5 ouE/m3, see also Table 7. odours was found to be remarkably consistent, when
applied to two groups of people attending an odour
The relationship between the odour concentration of pig annoyance seminar (one group in the Netherlands and
slurry and the perceived intensity has been established another in Germany). The results for the Dutch group
experimentally[12] (see Figure 3). Comparison with the are presented in Table 6.
intensity characteristic for broiler house odours shows
that the increase in perceived intensity is less steep than The ranking is strictly on order, it does not provide a
broiler odours, which are particularly pungent because comparative magnitude. The results are relevant, in that
of their high ammonia content. odours from intensive livestock operations are ranked in
27
the more unpleasant end of the list, but certainly not Table 6: Ranking of 20 generic and 20 environmental odours according
at the extreme end of dislike. Hedonic tone to like or dislike by a group of people professionally involved in odour
management, in the Netherlands, 1997
measurements for livestock odours were not
Descriptor Ranking Ranking Descriptor
identified in the review of literature for this report.
Generic odours mean mean Environmental odours
Roses 3.4 1.7 Bread Factory
Although scientifically not very relevant, it is
Coffee 4.6 4.6 Coffee Roaster
interesting to compare the ranking technique with
Orange 5.8 5.1 Chocolate Factory
odour exposure criteria that have been set for
Cinnamon 6.0 8.1 Beer Brewery
specific industries in the Netherlands. These criteria
Mowed lawn 6.4 8.3 Car Park Bldg
are only partly based on research, as they are also the
Soap 7.3 9.4 Charcoal Production
result of a consensus building process between the
Hay 7.5 9.6 Frozen Chips production
regulatory agency and the industry involved[15].
Brandy 7.8 9.8 Eel smoking
These values can be seen, however, as an expression
Raisins 7.9 9.8 Car Paint Shop
of the consensus reached in that society on the
Beer 9.3 9.8 Sugar Factory
relative odour annoyance potential of these odours.
Cork 10.5 9.8 Fragrance & Flavour Factory
In Table 7 the air quality criteria are listed, with their
Peanut Butter 11.1 11.2 Asphalt
ranks from Table 6. The ranking is generally
Cleaning Agent 12.1 12.8 Intensive Livestock Farming
reflected in the air quality criteria that were agreed.
Sauerkraut 12.8 12.9 Wastewater Treatment
Livestock odours are ranked similar to wastewater
Wet Wool 14.1 13.2 Livestock Feed Factory
treatment plants, in terms of relative dislike, see
Paint 14.4 13.2 Refinery
Table 6. For wastewater treatment odours, a range of
Vinegar 14.8 14.0 Green Fraction compositor
criteria exist in the Netherlands, ranging from 0.5 ≤
Sweat 17.2 14.1 Landfill
C98, 1-hour ≤ 3.5 ouE/m3. In the United Kingdom, a
Sour Milk 17.5 15.7 Fat & Grease Processing
limit value of C98, 1-hour ≤ 5 ouE/m3 has been accepted
Cat's Urine 19.4 17.0 Slaughter House
in a planning procedure for Newbiggin-by-the
Sea[16] as a reasonable criterion to demonstrate
absence of nuisance, as required in the UK Within the descriptor ‘livestock odours’ pig odours are
legislation. classified at the less favourable end of the spectrum of
like and dislike. In Figure 4, relative nuisance from
different types of livestock are shown, with pig odours
forming the least liked extreme, in relative terms[in:17].
60
Swine
Beef
Percentage Nuisance (%)
Layer
+ Boiler
40
20
+ +
+
+
0 +
0 1000 2000 3000
Distance (ft)
Figure 4 Relative nuisance perception for different livestock odours, expressed in relation to the
distance to the livestock housing, source: Veenhuizen, 1996 [in:17]
28
6.2 Relation between odour exposure and Table 7: Industry sector specific air quality criteria for odours,
percentage of population annoyed the Netherlands, and mean ranking score (see table 2)
Industry Limit value Target value Questionnaire

Note: This section reflects the results of a large- C98, 1-hour C98, 1-hour ranking
scale study into the relationship between exposure to
Bakeries No limit value,
pig odours and annoyance in the exposed
>>5 ouE/m3
population. The results of the study have been
Meat processing ≤ 1.5 ouE/m3 ≤ 0.8 ouE/m3
specifically interpreted to arrive at a framework of Grass dryers ≤ 2.5 ouE/m3
environmental quality criteria for application in Bakeries, pastry ≤ 5 ouE/m3 1.7
Ireland. Coffee roasters ≤ 3.5 ouE/m3 4.6
Flavours & Fragrances ≤ 3.5 ouE/m3 ≤ 2.0 ouE/m3 9.8
Wastewater treatment ≤ 0.5 ouE/m3 12.9
To establish air quality criteria for odours, with a
plant, greenfield site,
view to assuring that exposure of odours is limited to
residential dwellings in
levels that are acceptable from a public health and vicinity
well-being perspective, we need to establish a dose- Wastewater treatment ≤ 1.0 ouE/m3 12.9
effect relationship describing the relation between plant, greenfield site,
exposure to odours in the environment and the rural area or industrial
estate in vicinity
resulting degree of annoyance in the exposed
Wastewater treatment ≤ 1.5 ouE/m3 12.9
population. This epidemiological basis is
plant, existing site,
indispensable for setting environmental criteria in a residential dwellings
quantitative framework. Wastewater treatment ≤ 3.5 ouE/m3 12.9
plant, existing site,
In this section the relation between calculated rural area or industrial
estate
exposure to odours (dose) and percentage of the
Livestock feed ≤ 1 ouE/m3 13.2
population experiencing annoyance (effect) will be
production
discussed on the basis of epidemiological studies for Composting, organic ≤ 1.5 ouE/m3 ≤ 0.5 ouE/m3 14.0
a number of (bio-)industrial odour sources fraction of domestic
conducted in the late 1980’s and 1990’s [3] and a waste, greenfield site
recent extensive study aimed specifically at pig Composting, organic ≤ 3.0 ouE/m3 ≤ 1.5 ouE/m3 14.0
fraction of domestic
odours[4].
waste, existing facility
Slaughterhouses ≤ 1.5 ouE/m3 ≤ 0.55 ouE/m3 17.0
In these studies the dose is determined by
atmospheric dispersion modelling, using the odour
emission rate as measured at source, with other data In a recent paper these studies have been reviewed[3].
such as meteorology and topography, to calculate 98- The raw data of previously reported studies were re-
percentile of hourly average odour concentration analysed, this time adding a factor to reflect the relative
(C98, 1-hour). The percentage of people annoyed is ‘pleasantness’, or odour annoyance potential for each
determined by drawing a sample from the residential odour type. (See also section 4.2). The results were
addresses found in an area exposed to a certain level analysed to establish a relation between dose (odour
of exposure to odours (dose) and approaching these exposure) and effect (percentage of population
residents, typically by telephone, with a annoyed). The statistical model applied in these studies
questionnaire. The residents are not informed that shows a strong correlation for all but one of a dozen
the questionnaire is aimed at assessing odour studied cases. For the combined data for all studies, the
annoyance to reduce the probability of bias in the correlation coefficient r is 0.889. The review paper
response. The responses to defined key questions are demonstrates that the correlation coefficient improves to
used to establish whether respondents are annoyed r is 0.935 to 0.945 when an additional variable is
by odours. (see section 5.1.1). In some cases, the introduced to represent odour annoyance potential.
annoyance is classified on a scale, to differentiate
between annoyed and seriously annoyed. The odour annoyance potentials for the odours involved
29
were established in laboratory tests using assessors. The Table 8: Ranking of (bio-)industrial odours according to
odours were presented at a defined odour concentration, their perceived 'pleasantness' as determined in two types
of laboratory experiments.
e.g. 25 ouE/m3. Two methods were used, both leading to
Unpleasant
a ranking of odours according to their annoyance
Oil extraction Rendering
potential. Method a.) used paired comparison, while
Chemical plant Oil extraction
method b.) used a nine-point reference scale, with H2S at
Rendering Chemical Plant
providing a reference for scale value 2 on the unpleasant
Pig Farm Pig Farm
end and amylacetate referencing value 8 on the pleasant
Sugar factory Grass drying
end of the scale. The ranking of the odours according to
Grass drying Sugar factory
their ‘pleasantness’ is presented in Table 8 on this page.
Frozen Chip production Frozen Chip production
This table compares to Table 6.
Wire coating Pastry factory
Pastry factory Wire coating
When interpreting the results of the review paper[3] two
Cacao processing Cacao processing
issues must be kept in mind:
Tobacco processing Tobacco processing
Pleasant
• The odour units used in the review paper are odour
units as used in the Dutch NVN2820 standard. The
relation to European odour units as defined in • Odour exposure, calculated using dispersion
EN13725, as used throughout this report is: 1 ouE/m3 modelling, as the hourly concentration at the 98-
= 2 ou/m3. percentile of 1-hour averaged concentrations in a
typical meteorological year
• The annoyance is expressed as percentage of
questionnaire respondents seriously annoyed (%HA). • Percentage of a sample of the population classified as
This classification, particular to this paper, represents ‘occasionally or frequently annoyed’ on the basis of
approximately the top third of all people ‘annoyed’. their responses in a standardised questionnaire, using
interviews by telephone (see section 5.1.1 for details).
The relation between the percentage of respondents
seriously annoyed and the calculated odour exposure, The correlation between odour exposure and percentage
for all study cases combined, is: of the population ‘annoyed’ turned out to be highly
significant.
%HA = 4.775 . log(C98, 1-hour)2
This implies that at an exposure level of C98, 1-hour = 5.3 A significant conclusion was, that the percentage of
ouE/m3 the percentage of respondents seriously annoyed respondents annoyed could be predicted adequately by
by odours is 10%. As pig odours rank among the odours establishing the exposure due to the dominant source
with higher annoyance potential, this value is likely to only, defined as the individual source contributing most
be slightly lower for pig odours, although the paper does to the total C98, 1-hour odour exposure of the exposed
not give the equation for the dose-effect curve for pigs. location. Adding additional sources contributing to the
overall exposure did not provide a better prediction of
The specific dose-effect relationship for odour annoyance percentage. This finding is practically very
annoyance caused by pig odours has been established as useful, as it allows assessment of annoyance by
recently as 1999, in a study involving approximately considering the most dominant source only, when
2300 residents exposed in different degrees to odours preparing an environmental impact statement.
from pig production units[4]. As a result, the dose-effect
relation for pig odours is currently relatively well In the study, differences between areas with different
documented and can be used as a tool for predicting land use were determined, which reflects the Dutch
odour annoyance levels in a population. regulatory practice of using a qualification of ‘non-
agricultural, urban and suburban’, ‘Villages in
The dose-effect relation was determined by studying the agricultural environment’, agricultural housing’ etc. In
following variables: the study, a differentiation was made between pig
30
concentration areas, which are effectively set aside to difference between respondents in one-source and
some degree to allow elevated levels of pig production, multiple source situations was no longer apparent.
and general usage areas, outside concentration areas.
Respondents in concentration areas who are employed 3. The annoyance sensitivity of people who are directly
in the pig production trade form a specific category. involved in agriculture was found to be significantly
lower than that of the general population living in a
The dose-effect curves for percentage annoyed, as similar area. This effect is even more pronounced for
predicted from calculated exposure (C98, 1-hour) caused those living in ‘pig production concentration areas’
by the dominant source only, are shown in Figure 5. where the agricultural population displays a markedly
more tolerant attitude, leading to lower levels of
The figure shows that: annoyance at a given exposure to odours than in all
other groups.
• In the general public, exposed to pig odours from one
piggery, 10% of the respondents are annoyed at an 4. Whether people lived in a rural or more (sub-) urban
exposure level of C98, 1-hour ≈ 1.3 ouE/m3 . environment did not have a significant effect on their
annoyance sensitivity. Only people living in ‘pig
• In a selection from general public of those resident in production concentration areas’ showed lower
a pig concentration area, where pig odours form a annoyance sensitivity than in other areas, indicating a
feature of the odour context of the area, both those higher tolerance to pig odour exposure.
exposed to pig odours from one piggery and those
exposed to odours from multiple pig units, 10% of the The dose-effect relationships, differentiated for
respondents are annoyed at an exposure level of subgroups as discussed above, are shown in . The
C98, 1-hour ≈ 3.2 ouE/m3 . information as provided above can be the starting point
for a set of limit and target values for exposure,
• The most ‘pig odour tolerant’ selection from the associated with differentiated levels of protection
general public was found in the pig production against nuisance.
concentration areas, where pig odour is a feature of
the odour context in the environment, historically. For A level of 10 % of respondents annoyed has been chosen
this group, with the lowest annoyance sensitivity, the as a practical guide value, which is more than double the
10% annoyance level is reached at an exposure of C98, ‘background’ or ‘baseline’ level for odour annoyance, as
1-hour ≈ 6.3 ouE/m .
3 assessed in areas not exposed to (bio) industrial odours.
The percentage value of 10% is slightly less than the
• Those who are directly involved in agriculture were stated policy objective of the regulator in the
found to be the most ‘pig odour tolerant’ selection of Netherlands to limit the fraction of people annoyed by
the population. For this group, with the lowest odours to 12%. Although this regulatory objective lacks
nuisance sensitivity, the 10% annoyance level is a well-reasoned scientific motivation, is has been used in
reached at an exposure regulatory practice since 1985 in hundreds of licensing
of C98, 1-hour ≈ 13 ouE·m-3. cases, with overall satisfactory results, for the Dutch
regulatory environment[20].
Statistical analysis of the data yielded some remarkable
conclusions: Of course, the Agency will have to take into account
specific conditions to set criteria for Irish conditions,
1. The ‘annoyance sensitivity’ of people exposed to one reflecting a level of environmental quality and
single source was higher than for those exposed to protection compatible with Irish society. Ideally, dose-
two or more sources. effect relationships for Irish citizens in Irish conditions
should be assessed, to confirm the findings obtained
2. The annoyance percentage was best predicted by abroad.
considering the one dominant source only. When
exposure was calculated in this manner, the The guide values for establishing exposure criteria are
31
outlined below: value for ‘tolerable exposure, we have looked to this

group, with the exposure calculated on the basis of all
• The curve for One source, non-agricultural residents relevant sources of pig odour combined. For this
represents the group with the highest ‘annoyance group, 10% annoyance is associated to an exposure to
sensitivity’. This group consists of people with no pig odours of C98, 1-hour ≈ 6.3 ouE/m3.
professional involvement with the agricultural sector,
exposed to the effects of a single source, in rural or • The curve for concentration area, farmers represents
(sub-) urban areas. This sub-group can be regarded as a subgroup that can be characterised as the most
the group with the least tolerance to pig odours. tolerant in the survey, given their association with
However, it is likely to be the largest subgroup in the agriculture in pig production concentration areas.
general public. This group has therefore been This group, associated with the source of the odours,
regarded in this study as representative of the general cannot be considered to be representative for the
public. For this group, an annoyance percentage of general public. The exposure associated with 10%
10% is associated with an exposure to pig odours of annoyance in these persons directly involved in
C98, 1-hour = 1.3 ouE/m3. agriculture in concentration areas is C98, 1-hour ≈ 13
ouE/m3. This value is indicative as an absolute upper
• The curves for One source, non-agricultural exposure limit, but is as such not usable as a starting
residents, living in pig production concentration area point for an exposure guideline aimed at protecting
and Multiple source, non-agricultural residents living the general public.
in pig production concentration area coincide, when
only the one most dominant source is included in the The exposure level of C98, 1-hour ≈ 6.3 ouE/m3 associated
calculation of exposure. These groups represent a with 10% annoyance in the most tolerant selection from
selection of the general public with a lower sensitivity the general public[4], combined with the finding[3] that
to annoyance. These people live in an environment for a selection of (bio)-industry odours 10% of
where pig odours have become a feature of their respondents experiencing serious annoyance is
environment, and pig production is apparently more associated with C98, 1-hour = 5 ouE/m3, underpins a
accepted to take its place in that environment. This suggested upper limit value for odour exposure to be set
sample of the population can been regarded as not higher than C98, 1-hour ≤ 6 ouE/m3. This
representing the general public, with intermediate recommendation is based on a substantial set of data
annoyance sensitivity, showing an increased from independently conducted studies, as described
tolerance to pig odours as an existing element in the above (n = approximately 620 respondents).
status quo of their residential environment.
The value for the general public of C98, 1-hour ≈ 1.3
For this group, an annoyance level of 10% is ouE/m3 is associated with 10% annoyance in the general
associated with an exposure to pig odours public underpins a target value for odour exposure at a
of C98, 1-hour ≈ 3.2 ouE/m3. value of C98, 1-hour ≤ 1.5 ouE/m3, based on a substantial
set of data (n = approximately 1500 respondents).
• The curve for two or more sources, non-agricultural
residents living in ‘pig production concentration A framework of suggested limit and target values on the
areas’ (not shown in Figure 5) represents the sub- basis of the results of epidemiological dose-effect
group of the general public with the lowest studies as discussed above is outlined in section 8.3.
annoyance sensitivity. Although these residents had
no involvement with the agricultural sector, they live 7.1 Generic approaches to livestock odour
in areas where pig production is a significant feature management and regulation
of that area. This subgroup has been regarded to
represent the general public, with the lowest
annoyance sensitivity, showing increased tolerance to
pig odours as significant element in the status quo of
their residential environment. To arrive at a maximum
32
50%
45%
Non-concentration area
Odour nuicance by pig houses (%)
40%
35%
30%
Concentration area,
25%
non-farmers
20%
15%
Concentration area,
farmers
10%
5%
0%
0.1 1 10 100
3
C98 [ou/m ]
Figure 5 Relation between percentage of population experiencing 'annoyance' (effect) and calculated odour exposure (dose)
for one-source situations, expressed as C98, the 1-hour averaged odour concentration at the 98-percentile for a normal
meteorological year . Note: 2 Dutch ou/m3 = 1 ouE.m-3
(Source: reference [4] also presented verbally at conference with paper[37]).
33
34
7. Comparison of regulatory approaches in

different countries
guidelines were based on very limited experimental
Pig producers have a responsibility to minimise the data.
impact of their business on the environment, and to
avoid ‘loss of amenity’ and nuisance occurring in the In a limited number of European countries, specific
vicinity of their production site. Odours, as an regulations for odours from livestock were introduced
environmental stressor, tend to have a reach well beyond well before industrial odours became a regulatory issue,
the site boundary. in the early 1980s. The more quantitative approach of
measuring emissions and modelling their impact off-
In many countries, nuisance law has existed for many site, which was applied to industrial odour sources, was
years. This type of legislation is very general in its generally not applied to agricultural odours. For a
nature, often relying on appointed environmental health variety of reasons, however, there is a tendency for the
officers or law enforcement officers to make individual approach towards odour annoyance management in
judgements on the question of whether the odour agriculture and in industry to converge. This
experienced constitutes a nuisance in legal terms. development is a result of a variety of factors, such as
the enlargement of the scale of pig production units, the
To assist regulatory agencies in managing odour issues, tendency towards diversification of land use in rural
a number of countries have issued guidelines, areas and the introduction of IPC and IPPC for both
regulations or even legislation in their efforts to apply a industrial and livestock operations (over a certain
regulatory mechanism to odours from pig production minimum size).
units.
Currently, data are becoming available which allow the
The majority of these guidelines use a schedule of setting of environmental targets for odour exposure on
setback distances, based on the size of the facility the basis of actual ‘epidemiological’ studies into the
expressed as total number of animals, to define the prevalence of odour annoyance as a result of exposure to
minimum distance between pig houses from dwellings pig odours[4]. The emission factors for pigs, in different
in the vicinity, with a view to avoiding odour annoyance stages of their life cycle, have been measured in detail,
using standardised methods[5,10]. The availability of
This type of setback distance guidelines started to these data allows the design of an assessment framework
appear in the early 1970s and have been adopted in a that translates these scientific background data into
number of countries. In some cases, the setback tables objectives of a society in terms of environmental quality.
have been developed from their early, simple form into
an elaborate mechanism, taking into account different Summarising, there are three generic approaches to
types of housing, operational practices, usage of the managing the impact of pig production odours in a
surrounding area, cumulative effects of multiple pig regulatory framework:
producers, etc.
1. Nuisance Law approach: No Nuisance at the
Although this approach provides a detailed framework boundary
for assessment, the basis for the setback distances is Assessment and enforcement on the basis of the
rarely based on systematic scientific observations. The individual judgement of inspectors
first guidelines were published in the Netherlands in
1971 as a graph of setback distances versus the number 2. Setback Distances: Pragmatic semi-quantitative
of pigs[18]. These graphs were based on the limited approach based on experience
observations of a handful of Health Inspectors. This Assessment based on a schedule of distances
provided a pragmatic approach that has been applied depending on the number of animals, sometimes
effectively for many years, in spite of the fact that the more elaborate to take into account operational
35
practice, land use in the vicinity, cumulative effects sources where the exposure standard of 0.5 ouE/m3 for a
etc. Tables with setback distances have been used maximum of 2% of all hours (C98, 1-hour ≤ 0.5 ouE/m3)
typically for relatively small pig units, e.g. ≤ 500 had been in place for some years, has shown that, when
sows. this standard is met, there are no justifiable
complaints[20].
3. Air quality criteria for odour exposure:
Quantitative approach based on the dose-effect In 1995, a group of Dutch industries successfully
relationship lobbied to obtain a more lenient regulatory framework,
Assessment based on epidemiological research into which allowed differentiation between odours with a
the prevalence of annoyance caused by odours from high odour annoyance potential (e.g. rendering) and
pig production. These criteria can be implemented those with a lower annoyance potential (e.g. coffee
using the general approach common to air quality roasters, bakeries). This has resulted in a set of
issues. Impact assessment is based on emission differentiated target values of between 0.5 and 3.5
factors and atmospheric dispersion modelling. The ouE/m3 as a 98-percentile for industrial sources (see
approach in individual cases can be simplified for Table 7). A strict value of C98, 1-hour ≤ 0.5 ouE/m3 is
practical implementation similar to setback distance applied to very unpleasant odours, with a high odour
schedules, if the outcome is straightforward. annoyance potential, while more acceptable industrial
odours are regulated to more lenient values, e.g. C98, 1-
hour ≤ 3.5 ouE/m (e.g. coffee roasting).
For the framework for assessment of the impact of pig 3 [15]. Also, the
odours in Ireland, the latter option is considered the best regional and local authorities have been given more say
approach. A quantitative approach provides the best and flexibility to take account of local conditions in the
basis for judgements that are based on objective data, licensing process.
while leaving a margin of flexibility to allow more
subjective factors to be taken into account in the In the UK, the first instance of regulatory approval of
licensing procedure. this approach and an exposure standard of 5 ouE/m3 was
in a Planning Inquiry held into a proposal for a new
7.2 Background to the use of quantitative sewage treatment facility at Newbiggin-by-the-Sea in
odour exposure criteria spring 1993. The Inspector in his report accepted this
approach and stated that "he was satisfied that the
At the time of writing this report in 2001, there is no evaluation undertaken demonstrated that the proposals
clear guidance available on the assessment of impacts of would not give rise to a risk of unacceptable odour
livestock odours in Ireland. The obligation to avoid emissions beyond the boundary of the appeal site" (case
nuisance does exist, however. ref. APP/F2930/A/92 206240)[16].
Some other European countries have moved towards the 7.2.1 The reasoning underpinning the choice of
use of quantitative assessment methods for regulating odour exposure limit values varies.
odour exposure since the 1980s.
In the UK literature[36], the reasoning underlying a
Authorities in the Netherlands have used quantitative air choice of limit value is typically based on extrapolation
quality criteria for odours in licensing since 1985[19]. of findings in laboratory studies, rather than relying on
Initially, an overall air quality target was set for licensing epidemiological data. The argument is that, by
industrial sources of odour. This regulation definition, 1 ouE/m3 is the detection threshold of 50% of
differentiated between existing sources and applications a qualified panel of observers working in an odour-free
for new sources on greenfield sites. The limit value for laboratory using odour-free air as the zero reference (the
existing sources was set at C98, 1-hour ≤ 0.5 ouE/m3 , while selection criteria result in the qualified panel being more
for new sources a stricter limit at a higher percentile was sensitive to a particular odourant than the general
prescribed: C99.5, 1-hour ≤ 0.5 ouE/m3. population). The recognition threshold is generally
Data collected by survey in 1990 in The Netherlands between 1 and 5 times this concentration (1-5 ouE/m3)
from populations surrounding 200 industrial odour and the concentration at which the odour may be
36
considered to be a nuisance is between 5 and 10 ouE/m3, the prevalence of odour annoyance, as a result of
although this figure can be as low as 2 ouE/m3 for exposure to that specific odour. This dose-effect
particular offensive odours. Laboratory research in relationship has recently been established for pig odours
Ireland by Carney and Dodd found that pig odours reach in the Netherlands[4]. It should be taken into account that
a nuisance threshold at between 2.3 and 9.6 times the dose-effect relationship depends on the population
(average 4.8 times) the concentration at the detection involved. To transfer results of research in one country
threshold[21] to another country requires some interpretation, to arrive
at odour exposure criteria best suited to the conditions in
Although this argument provides a useful insight, it does that country. Environmental quality standards are set to
not tell the full story. To fully consider the relationship reflect the requirements and aspirations of a society, and
between downwind odour concentration and nuisance, hence contain a policy judgement, even if the principles
the length of time that a receptor is subject to a threshold are laid out using scientific data.
concentration must be considered. For example, a ‘faint
odour’ that can be perceived by a receptor for a few 7.3 Overview of livestock odour guidelines in
hours per year with those hours dispersed throughout the different countries
year, would be unlikely to cause a nuisance. The same
concentration perceived for the majority of the year, The guidelines for limiting the impact of pig odours in
undoubtedly would. Relationships between different countries tend to reflect the history and
concentration, time of exposure and resulting nuisance structure of the pig industry in each national setting.
(or freedom of nuisance) can only be determined by Most regulations and guidance use setback distances as
surveying populations living in the vicinity of odour the main instrument to reduce the impact of pig odours
sources, using an epidemiological approach. on people living in the vicinity. In some cases (UK) the
guidance focuses entirely on good practice, in the form
There is a need for a standardised quantitative method of a Code of Good Agricultural Practice for the
for the measurement of ‘odour annoyance potential’ to Protection of Air by the Ministry of Agriculture,
differentiate between different odours. It should allow Fisheries and Food (MAFF).
the translation of knowledge on the dose-effect
relationship, collected in an epidemiological survey, to Setback distances are in some cases suggested in general
another odour, using a more simple and effective terms, as a ‘desirable distance’ (USA), or as a minimum
laboratory method to do so. Such a method is not yet suggested distance (Ireland, currently). In other
available. A feasibility study to review the options to countries, more elaborate guidance has been developed,
develop a method for determining annoyance potential allowing a more precise determination of setback
of odours has been completed, however, giving an distances based on a number of factors, such as:
overview of the options and identifying the steps needed
to develop and validate the required method[1]. The • The number of animals on the site, specified to their
prospects for developing a standardised method for stage in the lifecycle
determining annoyance potential are promising.
• The design and operation of the facility (e.g. housing
Until a proper method for odour annoyance type, manure storage, feed composition)
measurements is available, it will be difficult to relate
specific odours to dose-effect relationship information • The use of the vicinity (e.g. neighbouring farmhouses
for other odours, on the basis of comparing the odour only, isolated dwellings, residential developments)
annoyance potentials. Until a method for assessing
odour annoyance potential becomes available, the best Examples of countries where such detailed schedules are
indicator for annoyance potential is probably the use of used to determine setback distances for licensing
hedonic tone, or a scale of like and dislike. purposes in specific cases are Germany, Austria and The
For now, the best possible basis for odour exposure Netherlands.
criteria is the use of actual epidemiological research into The approach towards reducing pig odour impact has
been described in detail in the annexes to this report for:
37
Germany, the Netherlands, the United Kingdom, New months at pre-defined locations on a grid around the
Zealand and the United States. source in question. The method has been applied on pig
units[9]. This method can be applied to determine
In this chapter, a short overview is provided for each of licensing applications.
these countries, followed by a comparison of the
different regulatory approaches, by comparing the However, in most cases, a technical guideline is applied,
setback distances for fictitious pig production sites. that provides detailed advice on the design and operation
of pig units and other livestock operations.
7.3.1 Germany
The national standard, VDI3471:1986, defines the
The law concerning air quality issues in Germany is the practice for reducing the impact of pig production units.
Bundes Immissionsschutz Gesetz (known as Setback distances are determined on the basis of a table
‘BimSchG’), or the Federal Immission Control act of to establish the number of Grossvieheinheiten (1 GV is
1990. equivalent to 500kg live weight). The number of GV is
then related to the required minimum setback distance
All odours from any commercial installation are using a graph. The setback distance graph gives
considered an annoyance, according to §3 of distances for up to 750 GV, approximately equivalent to
‘BimSchG’. For licensing and enforcement, however, 575 integrated sows in an Irish unit. A points system is
the issue is to determine whether the annoyance used to characterise the operational practice and design,
constitutes a ‘significant disturbance’, based on the which is accounted for in different curves on the setback
‘relevance of the annoyance’. However, the ‘BimSchG’ distance graph.
does not provide for criteria to determine when an
annoyance becomes a significant disturbance For a 400 sow unit, which is close to the Irish licensing
(nuisance). limit, the setback distance would be between
approximately 390 and 620m, depending on the number
Criteria on how to achieve the general principles of points.
concerning air quality are not provided for in the
BimSchG, nor in the second relevant official regulatory A more detailed description, including the distance
document, which provides technical guidance for graph, is given in Annex A.1.
specific industries,. The Technische Anleitung zur
Reinhaltung der Luft, or TA-Luft, details the technical 7.3.2 The Netherlands
measures, expected to be applied in different sectors of
industry and agriculture, including methods for Setback distances are determined on the basis of a graph
assessment. The TA-Luft is available in English, titled: relating ‘pig units’, equivalent to a fattener, to setback
Technical Instruction on Air Quality control. distances. Different curves are given, depending on the
land use in the vicinity. Different graphs have been used
The TA-Luft defines a maximum ‘odour frequency’, as since the first publication of a guideline of this type, in
an ambient air quality characteristic, but does not 1971. Currently the graphs from the 1985 guideline are
specify a method to assess this parameter. In 1994 a prescribed, providing graphs for production units of up
method was introduced by the Department of the to 2,500 pig units (equivalent to fatteners). A new
Environment of Nordrheinland Westfalen. The guideline was proposed in 1995, with an identical set of
document is available in English translation: setback distance curves. However, the description of
Determination and Evaluation of odour immissions - categories of land use to which these curves were to be
Odour exposure guideline applied was modified, effectively leading to a more
lenient policy on pig odours. However, the guideline as
This method is based on a method for long-term field revised in 1995 has been judged to be too lenient in a
panel observations, in which the fraction of ‘odour number of cases by the State Council, the appeal court
hours’ is determined by a team of assessors assigned to for planning decisions. A policy review by the Ministry
make observations at intervals over a period several of Public Planning and the Environment is now ongoing
38
(1999-2001). This review, which involves detailed facilities, and extensions or alterations to such facilities,
research into the actual annoyance experienced at need planning permission when these will be within a
different levels of exposure to pig odours, is expected to distance of 400 m from the boundary of any protected
lead to a more restrictive system based on air quality buildings (such as residential houses or schools).
criteria.
Under the Town and Country Planning (Assessment of
A more detailed description, including the distance Environmental Effects) Regulations of 1988 an
graph, is given in Annex A.2. environmental assessment is to be carried out for certain
types of major project which are likely to have
7.3.3 United Kingdom significant effects on the environment. For livestock
units this requirement is likely to apply to new pig units
The Environmental Protection Act of 1990 provides the of more than 400 sows or 5000 fatteners and new poultry
legal framework for avoiding and controlling odour units of more than 100,000 broilers or 50,000 layers.
nuisance in the United Kingdom. The Environmental
Health department of the Local Authority is responsible In planning procedures, the use of odour modelling with
for its enforcement. Under Part III, Section 79 of the application of a criterion of 5 ouE/m3 as a 98-percentile
Act, the local authority has a duty to inspect their area of hourly values has been accepted as an acceptable
and detect any statutory nuisance. Reasonably approach to demonstrate that no statutory nuisance
practicable steps are to be undertaken to investigate would arise, in a planning enquiry involving a
complaints by residents made to them. wastewater treatment plant at Newbiggin-by-the-Sea,
Northumberland, planning reference APP/F2930/A/92
Where a local authority Environmental Health 206240, UK, 1993[16].
Department is satisfied that a statutory nuisance exists,
or is likely to occur or recur, it has a duty to serve an The main document providing guidance is:
abatement notice under Part III, Section 80 of the Act
requiring: • The Air Code, Code of Good Agricultural Practice for
the protection of Air, revised 1998, Ministry of
• the abatement of the nuisance or prohibiting or Agriculture, Fisheries and Food and the Welsh Office
restricting its occurrence or recurrence; and Agriculture Departments, October 1998.
• the execution of such works and the taking of such The Code provides general guidance on the legal
other steps as may be necessary for these purposes. background, and on good practice for production., The
main part of guidance on odours is given in Part B of the
When appropriate action is not taken as a result of an Code, which contains a wealth of sound general advice,
Abatement Orders, significant fines of up to 20,000 but is remarkably limited on technical detail and
pounds can be imposed. quantitative assessment and management information.
The Code does not contain any specific recommendation
The law on statutory nuisance is far from on setback distances, other than suggesting that any pig
straightforward. A key problem is that no criteria are unit located at less than 400 metres from residences
provided to decide when occurrence of an odour should take extra care in implementing the advice given
constitutes a nuisance, and when it is acceptable. The in the Code.
system relies heavily on the individual judgement of the
Environmental Health Inspector. In practice a wide For details on the Code of Practice, see Annex A.3.
variety of licence conditions occur.
7.3.4 United States of America
Planning consents have to be granted on the basis of the
Town and Country Planning (General Permitted The regulations for odour in the United States of
Development) Order of 1995 (GPDO). New livestock
facilities, such as livestock buildings, slurry storage
39
America vary from state to state. The main legal basis on land use in the vicinity. Obviously, the New Zealand
for the regulations is the nuisance law. Various states setback distances are quite strict, reflecting the
have guidelines specifically aimed at managing odour availability of land and the high priority given to
emissions from pig units. In Annex A.4 a guideline environmental quality in that country.
issued by the EPA in Texas[17] is described in more
detail. For details on the New Zealand Code of Practice, see
Annex A.5.
In 1994, the American Society of Agricultural Engineers
indicated that a ‘desirable distance’ for siting livestock 7.3.6 Comparison of setback distances in different
facilities in general is 1600 meters from housing countries for fictitious pig units
developments and 400 to 800 meters from neighbouring
domestic dwellings. The nationally advised practice, guidelines and
regulations are compared for a number of countries, in
The EPA guidance indicates that the setback distance general terms, by applying their particular approach to
should be at least 3.6 kilometres and preferably 7.2 pig production units of a certain size. The comparison is
kilometres for ‘larger facilities’. not as detailed as the legal application in the country
might require, but serves to illustrate differences in
Details on US swine odour guidelines are provided in expectations of environmental quality for odour impact
Annex A.4. in the societies involved. The results are presented in
Table 9.
7.3.5 New Zealand
A detailed Code of Practice is applied in New Zealand,

within the general legal framework of the nuisance law.
The New Zealand Code of Practice (CoP) contains two
types of setback distances:
• Fixed setback distances, that must be observed in all

cases, regardless of the size of the production unit
• Adjustable setback distances that depend on the size

of operation and a set of correction factors for site-
specific operational characteristics.
The adjustable setback distances must be applied to pig

production units with 2000 pigs or more.
For any piggery of more than 5000 pigs, the potential to
create adverse effects needs to be determined on an
individual case basis. The size of the buffer zone for
such a piggery is determined to reflect this assessment.
Setback distances are determined on the basis of the P-

factor, which counts all pigs older than 70 days.
Breeding units with weaners only are counted using a
conversion of 1 breeding sow = 5 pigs.
For an integrated unit of approximately 400 sows (2000

P-units in New Zealand), setback distances in excess of
between 500 and 2000 meters would apply, depending
40
Table 9: Comparison of (approximate) setback distances according to

advice, guidelines and regulations in various countries.
Minimum values are marked yellow, values for
domestic dwellings green
Comparison of setback distances for pig Unit used 600 sow

units in different countries. Comparable or graph integrated
criteria indicated in one colour. unit
Germany
Graph input is the number of Grossvieh-units [GV] GV 785
(equivalent to 500 kg live weight)
Operations rating 25 points m 700
Netherlands
Graph input is the number of 'pig units' [mve] mve 3856
(equivalent to a fattener)
Category 1 (1995) - residential areas m 520
Category II - (1995) dwellings in clusters m 400
and villages
Category III - (1995) isolated clusters of m 210
dwellings in rural area
Category IV - (1995) Farm houses m 190
Ireland, currently
Minimum suggested setback distance only m >400
Ireland, proposed
Existing piggery, 600 sow integrated,limit value: m 450
minimum suggested setback
(receptor location to the SW)
Existing piggery, 600 sow integrated, limit value: m 650
maximum suggested setback distance
(receptor location to the NE)
Existing piggery, 600 sow integrated, limit value: m 600
minimum suggested setback
(receptor location to the SW)
New piggery, 600 sow integrated, limit value: m 900
maximum suggested setback distance
(receptor location to the NE)
USA
ASEA,1994 ‘desirable’ setback distances
Residential areas m 1600
Dwellings, minimum suggested m 400
Dwellings, optimum suggested m 800
EPA guidance for 'Larger facilities' m 3600
New Zealand
Residential area, urban m 2000
Place of public assembly m 1500
Rural dwelling m 500
41
42
8. Proposed framework for impact assessment

with a view to licensing
8.1 Scope of impact assessment framework framework will require careful consideration of the
transitional phase, taking into account the historical
The proposed framework for assessing the odour impact development in which dwellings were previously
of pig production units that is outlined in this chapter has allowed to coexist with pig units in situations where the
a clearly delineated scope: buffer zones were insufficient. The issue of how to
achieve such a transition is an essential aspect of policy
• The framework aims to assist the Agency in assessing implementation. However, being an issue of regulatory
the odour impact of pig production units in licensing policy, it is beyond the scope of this report.
procedures.
A consultation process involving the pig industry sector
• The assessment framework is suitable for existing pig should form part of the implementation process, to
production units and for new projects in greenfield ensure that its application in the licensing process is
locations. practicable.
• The impact of odours released from land-spreading 8.2 Legal framework

operations is outside the scope of this study.
8.2.1 European legislation
• As the issue of planning is outside the authority of
the EPA, issues and instruments of planning are As a result of implementation of European Community
outside the scope of the assessment framework, while guideline 97/11/EG, and a more recent amendment in
recognising that sound planning is the most effective Regulation SI no. 93, 1999 in the Regulations on
method of protecting the interests of the public and of Environmental Impact Assessment, an environmental
the pig producers. impact assessment is required for all pig production
units, which have
• The framework does not differentiate in assessing the
impact for a single sensitive receptor, such as a '…………installations for intensive rearing of
residential house, or larger numbers of dwellings. pigs....which would have more than 2000 places for
production pigs (over 30kg) in a finishing unit, more
• The framework is aimed at assessing the situation as than 400 places for sows in a breeding unit or more
it is described in the licence application and will not than 200 places for sows in an integrated unit'.
consider potential changes in the configuration
through building or change of use. The European Council Directive 96/61, concerning
Integrated Pollution Prevention and Control (IPPC) is
• The framework is aimed at assessing pig production relevant to pig production. According to the
units that require a licence, according to the First requirements of article 16, the Best Available
Schedule to the EPA Act, 1992, which states: Techniques (BAT) are to be identified for the European
Intensive Livestock Industry, including pig production.
6.2. The rearing of pigs in installations, whether within the
same complex or within 100 metres of that complex, where the The IPPC directive defines BAT in its article 2, sub 11:
capacity exceeds 1,000 units on Gley soils or 3,000 units on
other soils and where units have the following equivalents: "Best Available Techniques" means the most
1 pig = 1 unit effective and advanced stage in the development of
1 sow, including progeny = 10 units activities and their methods of operation which
indicate the practical suitability of particular
Implementation of a guideline based on the proposed techniques for providing in principle the basis for
43
emission limit values designed to prevent and, where 7. the commissioning dates for new and existing
it is not practicable, generally to reduce emissions installations;
and the impact on the environment as a whole.
8. the length of time needed to introduce best
"Techniques" include both the technology used and available techniques;
the way in which the installation is designed, built,
maintained, operated and decommissioned. 9. the consumption and nature of raw materials
(including water) used in the process and their
"Available" techniques mean those developed on a energy efficiency;
scale which allows implementation in the relevant
industrial sector, under economically and technically 10. the need to prevent or reduce to a minimum the
viable conditions, taking into consideration the costs overall impact of the emissions on the
and advantages, whether or not the techniques are environment and the risks to it;
used or produced inside the Member State in
question, as long as they are reasonably accessible to 11. the need to prevent accidents and to minimise the
the operator. consequences for the environment;
"Best" means most effective in achieving a high 12. the information published by the Commission
general level of protection of the environment as a pursuant to Article 16 (2) or by international
whole. organizations
When determining BAT the following considerations It should be noted that the identification of BAT at
must be taken into account, as listed in Annex 4 of the individual pig- or poultry farms is not necessarily the
directive: same for all European farms. The so-called "local
aspects" may influence the selection of BAT between
Considerations to be taken into account generally or Member States, or even between pig- or poultry farms
in specific cases when determining best available within a Member State. Relevant local aspects are for
techniques, as defined in Article 2 (11), bearing in example nuisance (odour, noise, coarse dust) and local
mind the likely costs and benefits of a measure and air or water quality.
the principles of precaution and prevention:
8.2.2 Irish legislation and guidelines
1. The use of low-waste technology;
The European requirement for environmental impact
2. the use of less hazardous substances; statements is roughly compatible with the licensing
requirements set out in the First Schedule to the EPA Act
3. the furthering of recovery and recycling of of 1992. The Act implies that licensing is required for
substances generated and used in the process any units that fall within the following description:
and of waste,where appropriate;
6.2. The rearing of pigs in installations, whether
4. comparable processes, facilities or methods of within the same complex or within 100 meters of that
operation which have been tried with success on complex, where the capacity exceeds 1,000 units on
an industrial scale; Gley soils or 3,000 units on other soils and where
units have the following equivalents:
5. technological advances and changes in scientific 1 pig = 1 unit
knowledge and understanding; 1 sow, including progeny = 10 units
6. the nature, effects and volume of the emissions Integrated Pollution Control (IPC) licensing will
concerned; become a requirement for pig production units over a
certain size. The dates at which existing units should
44
have applied for a licence, according to their size, are: batches.

- Using adequate bedding in litter based pig
housing.
>10,000 pig units 10 March 1998 - Provision of adequate manure storage
>7,000 pig units 9 June 1998 capacity.
>6,000 pig units 1 September 1999 - Stocking pig units at design level.
>5,000 pig units 4 April 2000 - Designing ventilation system to facilitate
>4,000 pig units 5 September 2000 efficient operation including maintenance.
- Filling and emptying liquid manure storage
EPA publications relevant to IPC licensing are: tanks from below the surface of the stored
manure, where feasible.
• Integrated Pollution Control Licensing - Application - Minimising the agitation of manure.
Form, Pig and Poultry sectors
3. Section 4.5:
• Integrated Pollution Control Licensing - Application Odour emissions should be contained by:
Guidance Note, Pig and Poultry sectors - Reducing uncontrolled air movement
- Filling and emptying liquid manure storage
• Integrated Pollution Control Licensing - Guide to tanks from below the surface of the stored
Implementation and Enforcement in Ireland manure, where feasible
- Transporting manure in suitably contained,
• Integrated Pollution Control Licensing - BATNEEC leak proof vehicles
guidance note for the Pig Production Sector - Minimising the agitation of manure
Minimising the generation of odours during
All pig units should comply with the existing meteorological conditions which favour the
BATNEEC guidance note for the pig production spread of odours
sector: - Landscaping pig houses using shelter belts.
• Integrated Pollution Control Licensing. BATNEEC Carcasses should be stored on site in covered
Guidance Note for the Pig Production Sector, containers and transported to a rendering facility
Revision 1 - February 1998, Environmental in covered, leak proof containers as soon as
Protection Agency, ISBN 1 8999965 36 X, Wexford, practical and at least once per week.
Ireland, 1998
4. Section 5.1: Describes the requirements for
This BATNEEC guidance is of a very general nature. compliance monitoring
The main issues raised, as far as they are relevant to the Periodic monitoring of air quality with regard to
reduction of odour impact, are: odour nuisance at the boundary of the site and
spreading areas as per licence
1. Section 4.3: Pig units should be sited a distance
of preferably not less than 400 metres from the This report outlines a wider range of methods and
nearest neighbouring dwelling and all operations techniques to reduce the impact of odours from pig
on site shall be carried out in a manner such that production units. These methods will not all necessarily
air emissions and/or odours do not result in fall within the scope of BATNEEC or BAT.
significant impairment of or significant
interference with amenities or the environment 8.3 The proposed assessment framework: general
beyond the site boundary. principles and odour exposure criteria
2. Section 4.4: The assessment framework aims to define a set of

Minimisation of odour emissions by: criteria for odour exposure to achieve a common
- Adequate cleaning of pig houses between environmental quality objective in licensing procedures.
45
The propensity to experience annoyance caused by living in the vicinity of pig production units[4], see
environmental odours, or annoyance sensitivity, is section 6.2. In these studies the percentage of the
variable in a population of residents. Due to the wide exposed population experiencing odour annoyance was
range of annoyance sensitivity of individuals, a zero- determined using a standardised telephone questionnaire
impact approach is not realistic. The aim is to set method (see section 5.2.1.1). The person was classified
environmental quality criteria for odours associated with as ‘annoyed’ if they experienced ‘occasional or frequent
an acceptable level of annoyance, sufficiently low to annoyance because of livestock housing units’.
prevent nuisance. The required level of protection is to
some degree a political choice, reflecting the In proposing the environmental quality criteria for
expectations of environmental quality of the society in Ireland, the general objective has been to reduce the
question. These expectations are not only determined by percentage of the population ‘annoyed’ by odour
human physiological characteristics, but also reflect the exposure to less than 10% of the resident population in
expectations of environmental aesthetic quality. These the vicinity. The choice of this level of annoyance to be
expectations are a function of the social, economic and ‘acceptable’ cannot be made on purely scientific
cultural outlook of a particular society at any given grounds. It is a pragmatic value, based on the experience
moment. accrued from developing an odour policy in the
Netherlands, where a policy target of 12% annoyance
The core of the framework is a set of target and limit was formulated in 1988 [23]. The experience with
values for calculated odour exposure. Using emission licensing in the Netherlands, using criteria on that basis,
factors for the animals in a given production unit, which supports the feasibility of that value, and its practical
may be adjusted for low-emission housing types and effectiveness. However, it must be clearly stated that this
other mitigating factors, the odour exposure in the practical experience, built up in one country, can only
vicinity can be calculated, using atmospheric dispersion provide a starting point for choosing an appropriate
models. environmental target for Ireland. In the end, such an
environmental target reflects the ambitions and
For straightforward cases, a simplified calculation requirements of a society, and is, therefore, a matter of
method can be used to determine whether the impact is policy, supported by scientific information.
well within the target values. If that is the case, no actual
modelling is required. If the situation is ‘borderline’, a At the selected level of annoyance, the risk of annoyance
modelling exercise is necessary to assess the impact developing to nuisance is limited, in most
specifically for the location in question. circumstances, but not excluded. Nuisance may still
occur in unfavourable situations, which can be
The calculated odour exposure is expressed as a value determined by secondary factors such as disrupted
for the one-hour average odour concentration that is not community relations, a history of ‘odour incidents’ on
exceeded during 98% of all the hours in a year with the production unit, other environmental stressors
average weather conditions. This value is the 98- (noise, dust), etc.
percentile of all calculated hourly concentrations, or C98,
1-hour in short, and is expressed as odour concentration in The ‘annoyance sensitivity’, or the propensity of a
European odour units per cubic metre (ouE/m3) This population to be annoyed by odours, at similar exposure
value is obtained using the emission rate (estimate), levels, has been found to differ. The highest annoyance
combined with the characteristics of the emissions sensitivity is found in situations where one pig
(height of emission point, exit velocity, location), in a production unit is the cause of the exposure, and where
mathematical model for atmospheric dispersion. These those exposed are not employed in the agricultural
models can take the local topography, obstacles in the sector. The lowest annoyance sensitivity was found in
wind flow and meteorological conditions into account. people who were employed in the agricultural sector,
exposed to the cumulative pig odours originating in two
The environmental quality targets are based on extensive or more pig production units, living in areas where pig
dose-effect studies that were carried out in the production is the predominant economic activity.
Netherlands in 1999, involving more than 2303 people Using dose-effect relationships that were established
46
experimentally in a large-scale study[4] as a starting to 10% or less, in the most tolerant selection of the
point, a three-tiered set of limit and target values was population.
defined for impact assessment in Ireland. These odour
exposure criteria aim to define ‘acceptable odour The limit value for existing pig production units shall
exposure’ that should not be exceeded at locations that not be exceeded in the vicinity of existing pig
can be classified as a ‘sensitive receptor’ for odours, as production units to ensure the minimum
described in section 8.4. The environmental quality environmental quality in an agricultural setting. A
criteria are: phased plan must be made to reduce the odour
impact, with time, to the limit value for new pig
• Target value: C98, 1-hour ≤ 1.5 ouE/m3 production units and, eventually, the target value.
The target value provides a general level of protection The limit value for existing production units is
against odour annoyance for the general public, complied with when for all locations of odour
aiming to limit the percentage of people experiencing sensitive receptors the calculated odour exposure is
some form of odour-induced annoyance to 10% or less than an hourly average odour concentration of
less. The target value is to be used as an 6.0 ouE/m3 in 98% of all hours in an average
environmental quality target for all situations. meteorological year.
The target value is achieved when the calculated These criteria for odour exposure aim to provide a
odour exposure for all locations of odour sensitive framework that can be used to attain a general
receptors is less than an hourly average odour environmental quality in Ireland, while recognising that
concentration of 1.5 ouE/m3 in 98% of all hours in an in some cases existing pig production units may need a
average meteorological year. considerable period of time to achieve that target. In
some cases, the time allowed will have to take into
• Limit value for new pig production units: C98, 1- account the cycle of normal replacement of assets such
hour ≤ 3.0 ouE/m
3 as housing, to allow implementation of a structural
The limit value for new pig production units provides solution.
a minimum level of protection against odour
annoyance, aiming to limit the percentage of those For those situations, a limit value is set that will reduce
experiencing some form of odour-induced annoyance the impact of odours to a level deemed acceptable for the
to 10% or less in the general public, assuming some most tolerant sections of society, with a strong affinity
degree of acceptance of the rural nature of their living with the agricultural sector.
environment.
As a general principle, however, a target value is set to
The limit value for new pig production units shall not ensure that, with time, an environmental quality is
be exceeded in the vicinity of new pig production achieved throughout the country that allows a diversified
units to ensure a minimum environmental quality. use of the countryside, not excluding or limiting
The limit value for new pig production units is recreational and residential use.
complied with when for all locations of odour
sensitive receptors the calculated odour exposure is 8.4 Definition of sensitive receptors
less than an hourly average odour concentration of
3.0 ouE/m3 in 98% of all hours in an average To make the concept of ‘impairment of amenity’
meteorological year. operational, it is useful to determine what objects or
structures, intended to be used by people, constitute a
• Limit value for existing pig production units: C98, sensitive receptor. To decide whether a location should
1-hour ≤ 6.0 ouE/m
3 be classified as an odour sensitive receptor, the
The limit value for existing pig production units following issues must be considered:
provides a minimum level of protection against odour
annoyance, aiming to limit the percentage of people • Is the facility for permanent use, throughout the year?
experiencing some form of odour-induced annoyance • Is the facility suitable as overnight accommodation?
47
• What is the type of activity associated with the

facility (work, recreation, transport, residential, • Offices.
transportation)
8.5 Emission factors
All facilities where people can stay overnight and are
permanently used are considered sensitive. In general Table 10 lists emission factors that are most suitable to
terms, the more people consider the facility as their use for odour impact assessment. The emission factors
‘territory’ and the less options they have to extract are differentiated for the category and weight of the
themselves from the situation in an episode of exposure, animal. The data are derived from a Dutch study
the more ‘sensitive’ the facility is. including conventional partly slatted pig houses[5,10,24],
similar to those most common in Ireland, as well as a
The following table lists examples of sensitive receptors: number of low-emission housing systems. The
recommended emission factors for impact studies are
• Residential dwellings, based on Dutch data as these were used to calculate
exposure in the dose-effect studies[4] discussed in
• Visitor accomodation (hotels, B&B, guesthouses), section 6.2 that underpin the air quality criteria proposed
in this report. Data from other countries are summarised
• Hospitals or nursing homes, in Table 10 and are discussed in more detail later in this
section.
• Schools,
If an applicant proposes to use lower values than those
• Churches, provided in the emission factor table, these figures must
be supported by measurement results. The
• Holiday and weekend dwellings, measurements shall be carried out according to the
standard prEN13725, and the emission factors expressed
• Campsites, and caravan parks, in ouE/s.
• Sports facilities,
Table 10: Recommended emission factors for pigs at different stages in the life cycle in European odour units per second
(ouE./s), and a summary of measured values from Netherlands, Belgium and the United Kingdom.
Category of animal Recommended Emission rate per animal Emission rate/kg
emission Netherlands Belgium UK, converted to ouE/m3
factors Calculated for animal of 85 kg
Emission per
animal annual summer winter mean Max mean max
ouE/s ouE/s ouE/s ouE/s ouE/s ouE/s ouE.kg-1.s-1 ouE.kg-1.s-1
Fatteners, conventional, fully slatte 36 128 0.43 1.50
Fatteners, conventional, partially slatted 22.5 22.4 25.4 32.7 15.4 19 47 0.22 0.55
Fatteners, restricted emitting area below slats 10 9.6
Fatteners, cooling of slurry surface below slats 11 10.8
Fatteners, flushing twice/day below slats 11 10.9
Fatteners, straw bed, scraped 20 53 0.24 0.63
Weaners, conventional, fully slatted 6 5 to 16.3 3.3 3.8 2.8
Farrowers, conventional, fully slatted 18 17.8 17.2 20.1 14.5
Dry sows, conventional 19 19.0 44.6 52.6 34.8
Dry sows, group housing with feeding station 7 6.8
Gilts 20
Boars 20
Fatteners, conventional, with air scrubber (acid) 30% 29% scrubber removal efficiency
48
The data that were obtained in the Netherlands were olfactometry used at the time was compatible with the
derived from a long-term research project, following a Dutch NVN2820 standard method, and used an n-
detailed protocol for evaluating emissions from different butanol reference value of 20 ppb/v. To convert its
types of livestock housing[10]. The sampling programme findings to European Odour Units, a conversion was
stretched over two periods of several weeks each, one in applied on that basis: 1 ouE/m3 = 2 ou/m3. The
winter conditions and one in summer conditions. measurements for this study were mainly carried out in
Preliminary results were reported in 1997 including an the winter of 1993 (November 1992 to March 1993). As
emission rate for fatteners in a conventional housing can be seen from the Belgian results in Table 12 this may
system with partly slatted floors of 22.6 ouE/s per animal have caused a bias towards lower values, by as much as
at a mean ventilation rate of 33.3 m3/hour/animal[5]. The a factor two. The report by ADAS proposes emission
full results, published more recently[10], are summarised rates ‘for inventory and planning purposes’ that are
in Table 11. Each emission factor in that table is based listed in Table 13, including conversion to European
on 20 observations. The observed variations were odour units and expressed for a typical weight of a
considerable, both between farms and temporal fattener of 85 kg. The suggested emission rate for a
variations within a farm, with coefficients of variation fattener of 18.7 ouE/s is not significantly different from
(standard deviation as a percentage of the mean) the results found in the Netherlands (22.4 ouE/s) and
typically in the order of 25 to 50%. The variation for Belgium (25.4 ouE/s), especially when considering the
weaners was considerably larger, which prompted a potential ‘winter bias’ in the UK data mentioned before.
repeat of an initial sampling run. The study indicated The limited measurements that were carried out on two
that the emission rate was significantly affected by the Irish pig production units, reported in Part B of this
ventilation rate. report, provided values between 7.3 and 20.2 ouE/s for
fatteners, measured in winter conditions. These data,
The same correlation between ventilation rate and with a geometric mean of 13.2 ouE/s and a median value
emission factor was established in recent research in of 15 ouE/s per fattener, fit well within the range of data
Belgium[35], using compatible methodology and as found in the Belgian study, for winter conditions, with
olfactometry. Initial results of the ongoing study into a geometric mean of 15.4 ouE/s. The data from the three
emission factors from pig production in Belgium were studies abroad and the data measured in Ireland for this
reported recently[35] and are summarised in Table 12. study broadly support the recommended value for the
When the emission for a production unit was corrected emission of fatteners for impact assessment in Ireland of
for the actual ventilation during the measurements, the 22.5 ouE/s.
mean emission rate remained the same, but the standard
deviation was reduced considerably: from mean 19070
ouE/hour with standard deviation 8000 ouE/hour (42%)
to mean 19840 ouE/hour with standard deviation 1500
ouE/hour (7.6%). This finding would suggest that the
temporal variations in emission rate are mainly
associated to ventilation rate, which is in turn
determined mainly by the outside temperature.
This is a relevant observation when assessing nuisance

potential, as annoyance is most likely to occur in
summer conditions, when residents are likely to be
outside in good weather. This situation would coincide
with conditions requiring relatively high ventilation
rates in the livestock unit, in turn associated with above
average emission factors.
In earlier work, emission data were collected for the
Ministry of Agriculture, Fisheries and Food in the
United Kingdom between 1987-1995[34]. The
49
Table 11: Emission factors for distinct categories of pigs held in conventional and low emission livestock housing in the
Netherlands, measured from 1996-1999. Each cell is based on 20 observations in both summer and winter.
Live stage of animal Housing type Emission Coefficient Minimum Maximum

code rate per of variation
animal
ouE/s ouE/s ouE/s
Fatteners, conventional, partially slatted FP-C 22.4 57% 8 85

Fatteners, restricted emitting area below slats FP-L1 9.6 25% 7 15
Fatteners, cooling of slurry surface below slats FP-L2 10.8 36% 6 18
Fatteners, flushing twice/day below slats FP-L3 10.9 49% 5 23
Weaners, conventional, fully slatted, first sampling run WP-C 16.3 53% 8 35
Weaners, conventional, fully slatted, second sampling run WP-C 5 79% 1 11
Weaners, restricted emitting area WP-L 4 140% 1 16
Farrowers, conventional, fully slatted FS-C 17.8 47% 7 35
Dry sows, conventional DS-C 19.0 47% 8 37
Dry sows, group housing with feeding station DS-L 6.8 69% 3 19
Source: Ogink, N.W.M., Groot Koerkamp, P.W.G., 2001[10]
Table 12: Emission factors from distinct categories of

pigs measured on one farm in Bierbeek, Belgium.
Each cell is based on 20 observations in both
summer and winter.
Pig category Annual s.d Summer s.d Winter s.d.
ouE/s ouE/s ouE/s
Fatteners, total 25.4 22.5 32.7 27.1 15.4 9.3
Fatteners, cell C5 26.4 24.8 34.4 29.9 15.4 8.5
Fatteners, cell C6 24.4 20.8 31.2 24.4 15.4 10.5
Weaners 3.3 1.8 3.8 2.1 2.8 1.4
Farrowers 17.2 11.2 20.1 13.1 14.5 8.5
Dry sows 44.6 32.2 52.6 36.2 34.8 25.2
Source: Van Langenhove, H., De Bruyn, G, 2001[32]
Table 13: Emission factors for planning purposes, as proposed by ADAS, UK, 1995,
with conversion to European odour units.
Emission Emission Weight Emission

factor factor animal animal
ou.kg-1.s-1 ouE.kg-1.s-1 kg ouE.s-1
Finishers, fully slatted, mean value 0.85 0.43 85 36.1
Finishers, fully slatted, maximum value 3.00 1.50 85 127.5
Finishers, part slatted, mean value 0.44 0.22 85 18.7
Finishers, part slatted, maximum value 1.10 0.55 85 46.8
Finishers, straw bed scraped 0.47 0.24 85 20.0
Finishers, straw bed scraped 1.25 0.63 85 53.1
Source: Peirson, S, Nicholson, R., MAFF Project WA0601 - Measurement of odour and ammonia emissions from livestock buildings Phase 1 final report.
ADAS, Cambridge, United Kingdom, March 1995.
50
After conversion of the UK results to European odour In those cases where the answer to the question: Are the
units (1 ouE.m-3 = 2 ge.m-3) the mean emission factors environmental quality targets for odour attained? is not
from the UK study are in the order of 0.43 ouE.s-1.kg-1 easily answered with an unequivocal Yes, a more
for a fattener on a fully slatted system and 0.22 ouE.s- detailed assessment route must be followed. The
1.kg-1 for a part slatted system. These values correspond detailed assessment involves making a best estimate of
with Dutch data presented in Table 11 that are equivalent odour emissions, on the basis of the number of animals
to specific emission rate for fatteners of 85 kg live present at the facility at full design capacity and a
weight on a part slatted system of 0.26 ouE.s-1.kg-1 . schedule of emission factors per animal, differentiated to
Limited measurements on an Irish pig unit, presented in the stages in the life cycle of the pig. These factors can
Part B of the report of this study, arrived at a (geometric) take low-emission practices into account, when relevant
mean specific emission rate of 0.22 ouE.s-1.kg-1 for emission factors are available or abatement efficiency
fatteners. can be reliably estimated.
As the differences between measured emissions per Only in those cases where the applicant asserts that
animal from different sources are not clearly significant, lower emission factors than those listed in this document
given the observed variation of emission rates measured are applicable, will measurements of emissions per
on different days within one pig unit and between pig animal on site be required. The statistical design of the
units, the data presented in Table 10 provide the best sampling programme must be suitable, in that case, to
estimate for emission rates currently available. yield statistically significant conclusions on differences
with the standard emission factors.
8.6 Cumulation
Contours can be calculated, using the estimated
In the Irish situation, cumulative impact of large emissions and a suitable atmospheric dispersion model,
numbers of sources affecting the same receptor site is which uses the topographical and meteorological data
relatively rare. It has been found that even in situations applicable to the specific site, and drawn on the map of
where cumulated exposure from multiple pig production the vicinity to represent the environmental quality
units occurs, the impact in terms of percentage of people targets that have been set as the basis for this assessment
annoyed can be adequately predicted on the basis of the framework. Comparing the location of odour sensitive
calculated exposure caused by the one dominant source receptors, such as residential dwellings, relative to these
only[4]. Therefore, in order to simplify the assessment, contours, will provide the licensing officer with the
only the predominant source is to be used in assessment framework for setting licensing conditions.
of impact.
The assessment process for licensing, where odour
The predominant source is defined as the source with the exposure is involved, has been outlined in Figure 6. This
largest proportional contribution to the odour impact approach to assessment has been illustrated in three case
(C98, 1-hour) at the receptor location in question. studies, see Chapter 10.
8.7 Practical application of the assessment

framework
The assessment framework provides for a simplified

assessment method for cases in which it can be
reasonably expected that the underlying environmental
quality criteria will be met. Those cases can be assessed
using a simple screening method, employing map
overlays with standard contours. The aim is to reduce
the burden of assessment where this is feasible, with
reasonable confidence.
51
Framework for Assessment of Odour Impact in EPA Licensing
Is the unit
Licence
BATNEEC/ Set conditions to achieve
application
IPC BATNEEC/IPC
submitted
compliant? compliance [c]
[a]
[b]
How large is the pig

production unit?
(number of ssows,
fatteners, gilts, boars)
[d]
Simple impact
Are
1000 assessment, using
aggravating
integrated standard contour
factors
sow overlays.
present (eg.
equivalents Select applicable
sludge
or more? odour criterion
lagoons)
[e] (limit and target value)
[f]
[g]
Full impact assessment, Sensitive

including modelling. objects within
Select applicable setback zone
odour criterion for target
(limit and target value) value?
[i] [h]
Exposure for
all sensitive
receptors below
target value?
[j]
Define impact
reduction options,
in consultation with
the applicant
[n]
Sensitive Finalise odour impact
receptors reduction plan and set
exposed in excess conditions for staged
of applicable impact reduction plan to
limit value? achieve target value
[k] [m]
No objections to
issue licence on
grounds of odour
[o]
Are odour
refuse licence on
impact reduction
grounds of odour
options feasible?
[p]
[l]
Figure 6: Odour Impact Assessment framework flowchart
52
8.8 The general, simple case: assessment by The protocol for the simple assessment using standard
screening using standard contour overlays contour overlays is summarised below:
To make a first assessment, figures with standard 1. Determine which criterion is applicable
contours for integrated units have been produced, which
are included in Annex E of this report. These contours 2. Select the appropriate overlay figure
represent a simple case, for flat terrain, for an inland
meteorological station (Claremorris). By copying these 3. Copy to the standard contour figure on an overhead
standard contours on an overhead sheet, at the scale sheet at the correct scale for the underlying
appropriate for the topographical map of the vicinity, a topographical map, on which the locations of the
quick assessment can be made, aimed at establishing sensitive receptors are marked
whether the pig production unit involved is well clear of
potentially causing a nuisance. If that is the case, the 4. Determine the size of the pig production unit of the
assessment for the application does not need to go any applicant. Choose the contour for the nearest higher
further. number of sows: this is the appropriate standard
contour
In the event that the outcome is either ‘borderline’ or
indicates that sensitive receptors are within the standard 5. Place the overlay on the topographical map, with the
contours, a more detailed assessment is to be made. centre cross in the centre of the pig production unit
The relevant figures can be found in Annex E: 6. Assess if any sensitive receptors are located within
the appropriate standard contour. If that is the case,
• Figure 16 Standard contour overlay, representing carry out a full impact assessment.
typical contours for the target value for all pig
production units of C98, 1-hour = 1.5 ouE/m3, for 7. If no sensitive receptors are located within the
integrated sow units of different sizes, scale 1:50,000. appropriate standard contour, the licence can be
approved, as far as odour impact is concerned, as long
• Figure 15 Standard contour overlay, representing as BATNEEC conditions are applied.
typical contours for the limit value for new pig
production units of C98, 1-hour = 3 ouE/m3, for The standard contours provide a reasonable indication of
integrated sow units of different sizes, scale 1:50,000. the impacted area, but specific factors, such as local
topography and meteorology and configuration of
• Figure 17 Standard contour overlay, representing emission points may have a significant impact on the
typical contours for the limit value for existing pig actual shape and location of the contours for the specific
production units of C98, 1-hour = 6 ouE/m3 , for study site. Where the final decision hinges on detailed
integrated sow units of different sizes, scale 1:10,560. location of contours, actual modelling is advisable,
particularly if the consequences of the decision are
The contours can be a-symmetrical to some degree, potentially large.
reflecting the predominant SW wind directions in
combination with wind velocity and stability of the In making impact assessments all relevant local
atmosphere in Ireland. The setback distances can, as a information should be considered and its implications
result, be dependent on the location of the sensitive for the impact of the odour exposure on the nuisance
receptor relative to the source. potential should be taken into account.
The standard contours were calculated using 8.9 Full impact assessment using atmospheric
meteorological data for Claremorris, for the years 1993 modelling
to 1995 (inclusive). The terrain was assumed to be
perfectly flat for the purpose of the calculation of To carry out a full impact assessment, a modelling study
standard contour lines for screening purposes. needs to be carried out, using a suitable atmospheric
53
dispersion model to determine the contours on the map be irrelevant at a distance of 800 m. A simplified
where a chosen exposure criterion occurs (e.g. C98, 1-hour method, dividing the area of buildings into areas
= 3 ouE/m3). (or pixels) of no more than approximately 20 m x
20 m, each represented by one point source, will
The outcome of the model depends entirely on the data provide sufficient detail for the impact
input. It is important to collect adequate data for model assessment at distances of approximately 200 m
input. This requires a balanced approach. Too little detail and more from the source.
may, therefore, lead to unreliable results, while too
much detail may require an inappropriate amount of b) Determine the emission per building, on the basis
resources to collect, perhaps wrongly suggesting an of the housing capacity and the type of animals
accuracy for the prediction that is unrealistic relative to housed in that building. Using the emission
the inherent uncertainties in the model. factors, listed in table Table 11, the emission for
that housing unit can be determined, and
The typical data required are: assigned to as many point sources as is deemed
necessary to represent the emission adequately
• Source characterisation data, describing the location, for the purpose of modelling. The ventilation
flow characteristics and emission rates of the source; flow rate, the height of the emission point, the
exit velocity and the diameter of the exit duct
• Terrain data, describing the topography of the study must be estimated based on the actual conditions
area, and characterising the ‘terrain roughness, or the and the number of animals represented by the
typical size of obstacles as far as these affect the area chosen. As odour annoyance is most likely
turbulence in the boundary layer of the atmosphere; to occur in summer, when residents are outside,
or have windows open, the exit velocity based on
• Meteorological data, for a minimum of 3 years, summer ventilation rates are arguably the most
consisting of a continuous set of hourly observations relevant to use in modelling. For establishing the
for wind direction, wind speed and ‘stability class’, a worst case, winter ventilation rates should be
parameter describing the turbulence in the boundary used to establish input for dispersion modelling.
layer. Some models accommodate different sets of
sources to be entered as input for each season, or
The collection of those data is described below. Before even per month, which is the most accurate
modelling, a decision must be made as to which approach.
exposure criterion is to be applied to the production unit
under study. Then the following data need to be 2) Characterisation of terrain in the vicinity
collected: a) Topography.
To calculate a contour for an exposure criterion,
1) Source characterisation: e.g. C98, 1-hour = 3 ouE/m3, a grid of receptor
a) Prepare an inventory of the location, dimensions points is required. The model will calculate
and height of the production units (buildings, exposure for each point on the grid, and then use
storage facilities, ventilation points, air treatment a suitable interpolation method to draw a
units, etc.) continuous contour line. Specific receptor
locations on the grid may be entered,
This inventory can be done on a detailed scale, representing the location of defined sensitive
locating individual ventilation shafts, their receptors, for detailed consideration. In setting
location, capacity, mean regulated rate, etc. up a grid, the following considerations apply.
However, this level of detail will, generally
speaking, not be required. The larger the distance A finer grid causes significantly longer runtimes
to the sensitive receptor, the less detail is for the modelling software. A coarser grid
required. At a distance of 100 m, the location of increases the risk of errors induced by the
a ventilator may have an effect, which is likely to interpolation to draw contour lines. At distances
54
up to 300 m a grid with a grid step of 50 m is times the height of these obstacles, e.g.
advisable. Between 300 m and 1000 m a grid step mature forest, low to medium density
of 100 m is recommended. Between 100 m and residential areas in villages and towns.
2000 m a grid step of 250 m is required. At larger vi) 3.0 m - Cityscape with intermittent low, single
distances a grid step of 500 m may be used. and double storey, and higher multi-storey
Slight topographical features, such as elevation buildings. Can also apply to woods with very
differences of 5 m or less, within the first 300- high trees and many, irregular openings.
400 m or 10 m or less at distances greater than
300 m are not very relevant for the model results 3) Meteorological data
in most cases, except those that are borderline A minimum of 3 years of continuous meteorological
relative to compliance with the selected criterion observations is required, giving hourly observations of
for the site. wind speed, wind direction and stability class (see also
i) If topography is not an issue, a grid of section 5.3.3).
receptor at the same elevation as the source
can be used. 8.10 Specific terms and definitions
ii) If topography is an issue, than an actual
elevation must be added to each point on the • Distance to source. The distance between the off-site
receptor grid. Elevation can be obtained from receptor point, such as the closest elevation of a
maps, preferably directly in digital format. corner of a residence, and the nearest emission point
(ventilation shaft or natural opening).
b) Terrain roughness.
Terrain roughness is a parameter with the unit
metre [m] that is used to characterise mechanical
turbulence in the lowest part of the mixing layer,
just above the earth surface. The parameter is
determined on the basis of the size of obstacles,
using the following scale:
i) 0.03 m - plain surface with only low vegetation
(grassland) with only occasional small
obstacles, e.g. airfields, plain meadows or
field after ploughing.
ii) 0.10 m - Fields with regular cover of low
crops, or grasslands with drainage ditches no
more than 20m apart. Occasional minor
obstacles can occur at distances in excess of
20 times their height, such as low windrows,
single lines of trees without leaves, individual
farm buildings.
iii) 0.25 m - Fields with intermittently high and
low crops. Larger scale obstacles (e.g. rows of
trees with leaves, low orchards) occur at
distances of more than 15 times their height.
iv) 0.50 m - Larger obstacles (e.g. larger farm
buildings, wooded areas) occur at distances of
approximately 10 times their height,
separated by shrubs, new forest with young
trees and mature orchards.
v) 1.0 m - Area regularly covered by larger
obstacles, at distances not exceeding a few
55
56
9. Methods for reduction of odour impact
9.1 What causes odourants to be produced? molecular structures that can be adsorbed into the blood
stream and used for growth and the energy need of the
The odours generated in a pig production unit originate animal. The excess nutrients and those components that
from: are not digested are excreted as urine and faeces. These
can be collected either separately, or mixed in the form
• Feed of slurry. The biological degradation process, that
started in the digestive tract, under anaerobic conditions,
• Spilled feed continues after excretion.
• Body odour of the animals There are two basic routes for biodegradation of pig
manure or slurry: anaerobic (Figure 7) or aerobic
• Urine and faeces processes. The aerobic process is faster, and produces
less odourants, than the anaerobic process. Generally
The most relevant source of odourants from pig speaking, however, urine and faeces are collected in
production is the excreta. As feed passes through the mixed form, and the resulting slurry is degraded in
digestive tract, food is transformed into smaller anaerobic conditions.
Biomass
Carbohydrate polymers, proteins, fats
Small polymers and monomers

Amino acids, sugars, glycerine,
longer fatty acids
Simple organic acids and alcohols

e.g. lactic acid, propionic acid,
butyric acid, ethanol
Figure 7: Anaerobic biogenic transformation, simplified
57
As a result of anaerobic biogenic transformation of Table 14: Compounds with low odour detection thresholds
organic matter and nutrients by mainly bacterial found in pig slurry[28]. Note: odour thresholds from
literature, typically not measured according to EN13725.
biomass, odourants are produced. The human sense of
Range of Compound
smell has evolved to be highly sensitive to these
detection
odourants. This is relevant to survival, as the ability to
threshold Cod
detect these odourants as smells is our main method for
[µg.m-3]
evaluating the chemistry of our environment and food.
Cod≤ 0.01 Methanethiol
The common word for anaerobic degradation is rotting,
2-propanethiol
and rotten food is a health threat. Our sense of smell has
2-propene-1-thiol
evolved specifically to detect by-products from rotting,
2,3-butanedione
as an immediate health warning. In particular the by-
0.01≤ Cod≤ 0.05 Phenylacetic acid
products of rotted proteins, a high-risk food, are easily
Ethanethiol
detected: substances containing sulphur (e.g. H2S and
4-methylphenol (p-cresol)
mercaptans, indicative of rotten eggs) and nitrogen (e.g.
0.05≤ Cod ≤ 0.1 Hydrogen sulphide
amines, indicative of rotten fish).
1-octene-3-one
0.1≤ Cod ≤ 0.25 Benzenethiol

Odourants in pig slurry are mainly formed as a product
2,4-decadienal
of anaerobic metabolism that occurs when all dissolved
3-methylbutanoic acid
oxygen has been depleted by bacterial respiration. The
2,6 dimethylphenol
underlying metabolism is complex, and produces a wide
3-methylphenol
range of chemical compounds, see Figure 7. Many are
2,4-nonadienal
highly odorous, such as mercaptans, organic sulphides,
Dacanal
amines, organic acids, aldehydes and ketones. A
0.25≤ Cod ≤ 0.5 Trimethylamine
secondary effect of anaerobic metabolism is the
Octanoic acid
lowering of pH due to the formation of organic acids.
Nonanal
The rate of formation of odourants, being a biological
Methylthiomethane
process, depends on a number of factors, such as the dry
Ethyldithioethane
solids content, availability of nutrients, availability of
2-phenylethanol
oxygen and, significantly, temperature.
3-methylindole (skatole)
Butanoic acid
The variety of odourants in pig odours is considerable
2-methylphenol
and some of these odourants are very smelly indeed,
2-butene-1-thiol
even at low concentrations.
2-nonenal
0.5≤ Cod ≤ 1.0 Indole

Between 100 and 200 odourants, have been identified in
Pentanoic acid
pig odours[28]. Detection thresholds can be as low as ppb
Butanal
or even ppt levels (10-9 to 10-12). At least thirty of the
identified compounds are highly odorous, having odour
detection thresholds of less than 1 µg.m-3, see Table 14. microorganisms. This conversion starts in the digestive
The most recent odour threshold for H2S, measured by system of the pig, but accelerates within hours after
dynamic olfactometry, measured according to excretion.
EN13725:1999, is 0.5 ppb, which is equivalent to 0.7
µg/m3. The basic principles for reducing odour emissions are:
9.2 Theoretical options for reducing odour 1. Reduction of odourant formation in slurry
emissions from pig production
a. Separation of urine and faeces, followed by
The odourants that are released from the operations of treatment.
pig production are the result of anaerobic metabolism of When solids are separated, the liquid fraction can
58
be treated by aeration, reducing the production of 3. Reduction of exposed area of slurry, including
highly odorous compounds. The aerated liquid storage, soiled surfaces, grids etc
can be used as flushing liquid to collect slurry
more efficiently from the pig houses. Aeration a. Different housing types, which include systems
and separation of the solid fraction also provides such as:
a benefit in reducing the odour impact from i. Green Label pig houses, designed for low
spreading. ammonia emissions,
b. Frequent removal of slurry and storage in closed
b. Lowering of temperature of stored slurry tanks.
The relation between temperature and the
anaerobic metabolic rate is exponential. A 4. Extraction of ventilation air with treatment to reduce
reduction of the slurry temperature from 20°C to odour concentration
10°C causes a reduction to less than 50% of the a. bioscrubbers,
emission rate at 20°C. b. chemical scrubbers,
c. biofilters.
c. Reduction of the protein content in feed
5. Miscellaneous additives
d. Collection of slurry in closed tanks, followed by
anaerobic digestion. a. Feed additives,
In this process the odourants that are produced b. Slurry additives.
can be destroyed by controlled incineration of the
biogas. The digested slurry is significantly less The economics of installing the technology to abate
odorous, which is a great benefit when odour emissions must be assessed before this
spreading[29]. technology can be imposed on existing pig unit
operations.
2. Reduction of transfer rate from the surface of slurry
9.3 Good operational practice
a. pH control
The pH of the slurry can be used to manipulate It is good operational practice to keep the pigs and the
the balance between soluble, ionised forms of surfaces in and around buildings clean.
odourants and less soluble, volatile forms.
Regulating the pH of slurry to low pH values, pH Pigs with manure on their skin will have a significantly
≤ 7, can effectively control the emission of increased odour emission, as the body heat of the animal
ammonia from slurry. The effect on the transfer will accelerate the release of odours significantly.
rate of other odourants is ambiguous. At lower
pH values the organic fatty acids will be released Every surface covered in manure will be a source of
more readily to the atmosphere. Active chemical odour. Reducing the exposed area of manure induces a
control of pH is not a practical tool for odour direct reduction in odour emissions.
control.
Keeping a clean operation is a matter of combining good
b. Covering the surface design with good operational practice.
i. Natural crusting,
ii. Floating biological covers (straw, fibre), Sources of odour in and around buildings include:
iii. Floating covers (plastic film, polystyrene
panels or porous stone), • Wet and manure-covered floors;
iv. Liquid additives (vegetable oils),
v. Air-filled plastic domes (over sludge storage • Dirty pigs, with manure on their skin;
lagoons).
• Spilled feed;
59
• Improper storage and disposal of dead pigs; main principles for reducing emissions to air from pig
housing are:
• Deep underfloor manure storage pits, with long
residence time; 1. Limiting the exposed area of stored manure;
• Dusty surfaces, that can capture and release 2. Frequent removal of manure by a sewage system;
odourants;
3. Cooling manure, lowering the temperature of stored
• Elevated temperatures in manure storage pits and in manure;
pig houses.
4. Faster discharge of the manure from slats, by using
9.3.1 Slurry removal triangular iron bars, which are easily cleaned;
Long storage times and large storage volumes increase 5. Frequent removal of manure by flushing or scraping.
the emissions of odourants. As a general principle, pig
manure must be removed to adequate storage pits or be A variety of systems have been developed, using these
subjected to an appropriate treatment, including land principles. In the Netherlands strong incentives have
spreading, as quickly as practicable. The current been made available to those implementing low
widespread use of deep tanks under fully slatted pig ammonia emission systems. This has resulted in a
houses is not ideal from an odour management relatively widespread application of a number of
perspective. systems. These systems are described in detail in a
document[30] that is publicly available on the Internet, at
9.3.2 Cleanliness http://www.infomil.nl/lucht/index.htm:
• Ensure that pigs remain clean; Hendriks, HJM, van de Weerdhof, AM, Dutch notes on
BAT for pig and poultry intensive livestock farming,
• Clean surfaces and slats regularly; Ministry of Housing, Public Planning and the
Environment and the Ministry of Agriculture, Nature
• Clean pens and storage units regularly, both the floors Management and Fisheries, August1999.
and the structure, using adequate means such as high
pressure water washers; The main characteristics, including extra investment
cost relative to constructing a fully slatted system, and
• Remove any stagnant water from surfaces. the suitability to retrofitting existing buildings, have
been summarised in Table 15.
9.4 Housing design
Separate detailed studies for each of these housing
9.4.1 Standard housing systems systems are available or will become available under the
so-called Green Label certification system. The detailed
The majority of existing housing in Ireland is the reports include measured emission rates for odours for
traditional fully slatted system. The main source of each housing type.
emission in these houses is the surface of the underfloor
slurry storage.
9.4.2 Low-emission housing systems
Low-emission housing systems have been developed,

mainly with the objective to reduce ammonia emissions.
Most systems will reduce odour emissions as well as
ammonia emissions, roughly in equal measures. The
60
Table 15: Low-emission housing systems, extra investment costs, operational costs and effectiveness for ammonia reduction.
Housing type Animal NH3 NH3 Extra Extra costs, Extra No animals Suitable for retrofit? Reference in draft
emission reduction investment annual, per energy housed in BAT document…
relative to, pig place use per system, NL,
fully place place 1999
slatted
[kg/place/ [%] [€] [€] [kWh/
year] year]
Pen manure channel with slanted side finishers 1.2 60% 3 0.50 0 5,000 most existing b, section 3.1.1
wall(s), partly slatted floor, concrete slats houses can be
adapted
Pen manure channel with slanted side finishers 1.0 67% 23 15.00 250,000 depending on, section 3.1.2
wall(s), partly triangular iron bar slatted existing pit design
floor
Manure surface cooling channel, concrete finishers 1.5 50% 30 5.50 14 20,000 yes b, section 3.1.3
slats
Manure surface cooling channel, finishers 1.2 60% 43 8.00 14 200,000 yes b, section 3.1.4
61
triangular iron bars
Manure channel with gutters, partly slatted finishers 1.2 60% 59 9.45 1.5 depending on b.section 3.1.5
floor with concrete slats existing pit design
Manure channel with gutters, partly slatted finishers 1.0 67% 79 12.50 1.5 50,000 depending on b,section 3.1.6
floor with triangular iron slats existing pit design
Partly slatted or convex floor with plastic weaners 0.34 43% - - 0 1,000,000 depending on b, section 4.1.1
or iron slats existing pit design
Shallow manure pit with a channel for weaners 0.26 57% 6 0.45 250,000 depending on b, section 4.1.2
spoiled drinking water existing pit design
Pen manure channel with side wall(s) on a weaners 0.17 72% 5 0.75 50,000 most existing b, section 4.1.3
slope, partly slatted floor, triangular iron bars houses can be adapted
Manure scraper under the slats weaners 0.18 70% 69 12.30 40,000 no b, section 4.1.4
Manure channel with gutters, partly slatted weaners 0.21 65% 25 4.15 75,000 depending on b, section 4.1.5
floor with triangular iron slatsexi sting pit design
Manure surface cooling channel, weaners 0.21 65% 24 4.40 150,000 yes b, section 4.1.6
(Source: Dutch notes on BAT for Pig- and Poultry intensive livestock farming, draft, August 1999
Table 15: Low-emission housing systems, extra investment costs, operational costs and effectiveness for ammonia reduction. (continued)
Housing type Animal NH3 NH3 Extra Extra costs, Extra No animals Suitable for retrofit? Reference in draft
emission reduction investment annual, per energy housed in BAT document…
relative to, pig place use per system, NL,

fully place place 1999
slatted
[kg/place/ [%] [€] [€] [kWh/
year] year]
Board on a slope under the slatted floor farrowers 5.0 40% 260 29.50 few yes b, section 5.1.1
Manure surface cooling channel farrowers 2.4 71% 302 54.25 10,000 yes b, section 5.1.2
Combination of a water- and manure farrowers 4.0 52% 60 1.00 50,000 yes b, section 5.1.3
channel
Manure pan farrowers 2.9 65% 280 45.85 10,000 yes b, section 5.1.4
62
Manure scraper farrowers 4.0 52% 785 147.20 1,000 sometimes, b,section 5.1.5
depending on
existing pit design
Flushing system with manure gutters farrowers 4.0 52% 535 86.00 8.5 500 depending on b, section 5.1.6
existing pit design
Small manure pit dry sows 2.4 43% 18 5.80 1,500 depending on b, section 6.1.1
existing pit design

Flushing gutters dry sows 2.5 40% 162 57.90 1,000 depending on b, section 6.1.2
existing pit design
Manure surface cooling channel dry sows 2.2 48% 113 20.35 8.5 3,000 yes b, section 6.1.3
(Source: Dutch notes on BAT for Pig- and Poultry intensive livestock farming, draft, August 1999
9.5 Optimisation of ventilation and 9.5.2 Optimisation of atmospheric dilution

atmospheric dilution
9.5.2.1 Increase of emission height
9.5.1 Ventilation in livestock housing
Increasing the height at which the odorous emissions are
The ventilation rate in a pig production unit is released into the atmosphere can be a relatively
determined by the needs of the animals. Adequate economic and effective method to reduce the impact.
ventilation is vital, to perform the following functions:
For a mechanically ventilated pig house with a number
• Regulate temperature of ventilators, raising the emission point to 4 or 5 metre
Pigs produce a significant quantity of energy, in the above the roof will limit the ‘building wake effect’ and
form of heat. The excess heat is removed from the pig hence enhance dilution downwind. This is especially
houses with the ventilation air, so that the temperature noticeable at relatively short distances from the source
remains within the optimum temperature range, (up to 300 m). To optimise dispersion of odours in the
which varies for different stages of the pig life cycle. atmosphere, apex ventilators, especially those with
Too high temperatures cause growth rates to slow appropriate ducting aimed at lifting the emission point
down, while low temperatures increase the risk of above the roof, are preferable over side wall ventilators
disease and cause a lower feed conversion efficiency.
More significant benefits can be obtained when the
• Remove excess carbon dioxide (and ammonia etc.) ventilation flow is ducted to a central stack. By
Due to the high density of biomass in pig houses, increasing the stack height to between 10 m and 25 m, a
ventilation needs to be carefully regulated to remove marked benefit is achieved in terms of downwind
the CO2 produced by the respiration of the animals. dilution. The extent of this beneficial effect should be
CO2 levels are kept below 3%. determined for each situation on the basis of dispersion
modelling. The point of maximum concentration at
Because of the importance of ventilation rates for pig ground level will move away from the source when
welfare and productivity, ventilation rates cannot be raising the height of the emission point, which is
easily modified. generally a benefit, but in some cases may cause an
increase of exposure for specific receptors.
Ventilation rates vary with the seasons. In summer, the
main determining factor is temperature control, while in In Figure 8 the effect of raising the emission height is
winter the CO2 concentration may become the main illustrated, for a typical emission for an integrated sow
determining parameter for the ventilation rate. unit of 670 sows, with:
The ventilation rates in spring, summer and autumn are • Natural ventilation or ventilation at roof level (4m);
the most relevant for the odour impact, as these are the
seasons where exposure in the vicinity of the facility is • Mechanical ventilation, exit velocity 5 m/s, at 4 m
most likely to cause annoyance. As odour annoyance is above roof level (8m);
most likely to occur in summer, when residents are
outside, or have windows open, the exit velocity based • Mechanical ventilation, emission through stack at 15
on summer ventilation rates are arguably the most m, exit velocity 5 m/s;
relevant to use in modelling. For establishing the worst
case, winter ventilation rates should be used to establish • Mechanical ventilation, emission through stack at 15
input for dispersion modelling. Some models would m, exit velocity 15 m/s.
accommodate inputting different sets of sources for each
season, or even per month, which is the most accurate 9.5.2.2 Increase of emission velocity
approach.
Optimising the vertical velocity of the emission helps
‘plume rise’ to occur due to the momentum of the
63
ventilation flow. This increases the effective emission impact will have an effect on the appreciation of the
height, which benefits the dilution downwind. Optimum odour. When the source is not visible, or obscured by
exit velocities are in the range of 10-15 m/s. pleasant vegetation, the odour may be appreciated as
less negative than the same odour emanating from a
The effect of optimising the way in which emissions are secretive production unit behind a forbidding fence, with
released into the atmosphere is illustrated in Figure 8 signs saying DANGER - KEEP OUT.
below.
The visual impact of a site is the starting point of
9.5.2.3 Vegetation and landscaping community relations and should be considered as such.
The use of barriers and earth screens and/or vegetation 9.6 Feed
is strictly speaking not an odour abatement technology.
The adsorption of airborne contaminants to foliage is A high protein diet increases the availability of nitrogen
generally very low to the point of insignificance, with and sulphur in the manure. These substances are the
the possible exception of ammonia. precursors to very odorous substances when the
anaerobic digestion of that manure occurs. From the
The effect on dispersion can be either favourable or odour reduction perspective, it is, therefore, advisable to
unfavourable. In the wake of obstructions in the flow, an reduce crude protein levels, while providing the
area of uniform mixing occurs that can either reduce or essential amino acids in adequate amounts to ensure
increase concentrations at receptor level, depending on optimum growth.
the actual site configuration. As a rule the benefits are
not clear, and this should not be relied upon as an odour The benefits in odour emissions per animal are relatively
abatement method. poorly documented so far, but indications are that a
reduction in odour emission is not likely to be greater
As odour annoyance is ultimately a cognitive than 50%, and more likely to be in the order of 25-
psychological process, however, landscaping and visual 30%[31].
Figure 8 Odour contours for for a fictitious pig unit with a) natural ventilation at roof height, b) mechanical ventilation 4m
above roof apex, exit velocity 5 m/s, c) all emissions through stack of 15m height, exits velocity 5 m/s, d)
as c but velocity 15 m/s. Meteorological data for Claremorris.
64
9.7 Additives character of the odour becomes less offensive.
9.7.1 Feed additives • Counteractants are agents that interfere with

odourant molecules with the aim of reducing the
Many feed additives have been introduced to the market odour intensity of the mixture, as well as making the
in recent years, and a number of producers claim character of the odour more acceptable. The
benefits in terms of reduced emissions of odourants as a underlying process is not specified, but a form of
result. encapsulation on a molecular level is implied.
Additives include: Masking agents may have a public relations benefit in

the short-term, communicating that some short-term
1. Fats and oils. By reducing feed dust generation, the action has been undertaken in acute conflict situations.
release of dust-bound odourants can be reduced. In the long-term they may be counterproductive,
however, as the masking odour becomes associated with
2. Absorbing additives, e.g. calcium bentonite, zeolite the cause of the annoyance. As the intensity of the
and activated charcoal. These substances are added masking odour is higher than that of the original odour,
with the objective to absorb odourants to counteract it will increase the magnitude of the problem, rather than
their release from the manure. reduce it.
3. Plant extracts. A natural extract from the yucca plant, Masking agents and counteractants are used fairly
sarsaponin, has been reported to reduce ammonia frequently to provide immediate relief.
emissions. No significant effect on odourant releases
has been confirmed, however. There is a lack of quantitative data to quantify potential
benefits of these techniques. A 1993 study for
4. Enzymes. wastewater treatment odour abatement by the Water
Research Centre in the UK did not demonstrate a
5. Microbial formulae. significant difference between the effectiveness of a
counteractant and water in reducing odours. Most
9.7.2 Slurry additives performance evaluations are anecdotal.
Slurry additives are available in a wide variety of 9.8 Extraction and treatment of ventilation air
products. The main generic types are reviewed in the
following sections. Once the air can be ducted to a central point for
treatment there are a number of options for reducing the
9.7.3 Odour counteractants and masking agents odour concentration in the exhaust air. A number of
options have been listed here:
Masking agents and counteractants are gas phase
treatment methods, in which an odour treatment agent is • Chemical scrubbing;
mixed directly with the foul airflow, usually by
atomising a liquid using sprays. This may be done in • Biological scrubbing;
ducts but also after release of the odourants into the
atmosphere, using open-air sprays. • Biofiltration, potentially combined with pre-
treatment for H2S using a catalytic iron filter;
• Masking agents are odourants with a relatively
pleasant odour, that are used to mix with the foul air • Biofiltration on fixed medium substrate covered with
to produce a more acceptable mixed odour or even a biofilm, such as a lava-rock filter.
‘drown out’ or overpower the foul odour with the
masking odour. The resultant odour is inherently These techniques all have the potential to reduce the
more intense than the original odour, but arguably the odour concentration in the ventilation air significantly,
65
with an efficiency of up to 90-95%. The techniques are abatement efficiency of 70-90%.

well known, proven technologies in applications in other
processes, e.g. wastewater treatment. The treated In a wet scrubber, foul air is vigorously mixed with a
ventilation air could be released from a raised stack, to scrubbing liquid. Typically the airflow is counter-current
further reduce the odour impact by achieving (upwards) but crosscurrent is an option. Most systems
atmospheric dilution. are packed with random plastic media. Unpacked
scrubbers rely on fine droplets (mist scrubbers) and
The main disadvantage of this approach is cost. The usually require demisting. Scrubbers are essentially a
quantities of ventilation air are considerable, and chemical processing system and proper chemical
determined by the ventilation needs of the animals. For engineering is required for successful application.
this study a ventilation rate of between 50 and 100
m3/hour per fattening pig was used for estimating Wet scrubbing relies on mass transfer of (odorous)
emission volume flow rates. compounds from the gas phase to the liquid phase.
Usually chemical scrubbing is applied, where chemicals
The cost and feasible abatement efficiency for the are added to the scrubbing liquid that can react to
different systems that can be applied are summarised in transform odorous substances to ionised forms or
Table 18. In reviewing the cost of abatement systems, decompose them to less odorous compounds by
the cost of ducting to collect the air and duct it to the oxidation. Alkaline or acid scrubbers rely on the
treatment unit must be taken into account. As a rough formation of salts. Oxidising scrubbers (e.g. chlorite,
budget estimate, ducting amounts to an investment of potassium permanganate, peroxide) will oxidise
between €0.80 and €1.60 per m3/hour ventilation dissolved pollutants. Ozone can be used to react with
capacity. pollutants both in the gas and the liquid phase. Catalysts
can be used to make the chemical reactions in the
9.8.1 Chemical scrubbers scrubbing liquid more effective and reduce requirements
for chemicals.
Chemical scrubbers are used as an end-of-pipe
treatment, before the ventilation air is released into the Single-stage acid scrubbers have been reviewed as a
atmosphere. Chemical treatment of air in scrubbers is a method for ammonia emissions reduction for the pig
proven technology, which can achieve an odour production sector in in the Netherlands[30], and were
Table 16: Budget cost (€) for different end-of-pipe treatment methods, per unit ventilation capacity.
Treatment method Cost per m3/hr ventilation capacity Remarks
Investment [€] Operation [€] Total [€]
From To From To From To
Single stage chemical 0.41 0.86 0.14 0.28 0.23 0.45 Designed for pig houses
scrubber (Cost data from the
Netherlands[30] )
Bioscrubber 0.41 0.86 0.17 0.33 0.25 0.50 Designed for pig houses
(Cost data from the
Netherlands[30] )
Biofilter 2.05 2.55 4.60 - Industrial applications
Bioscrubber 3.55 6.33 3.27 15.92 6.82 22.25 Industrial applications
Two stage chemical
scrubbing 1.87 1.87 10.82 27.40 12.68 29.27 Industrial applications
Dry scrubbing 3.70 4.03 7.73 9.27 industrial applications
66
Table 17: Indicative budget costs (€) of ammonia (and odour) abatement in ventilation air, using single-stage chemical
scrubbers. (Netherlands, 1997) [30]
Cost per animal capacity (fatteners)

Abatement system When data are not stated in the reference, the table cell is left blank Source Remarks
Capital investment [€] Life Capita Operational cost [€] Total
cost [€] annual
[€]
Abatement Ducting Total Years Total Energy Other Total
unit etc
Single-stage 42 43.18 85.00 10 8.50 3.89 10.11 14.00 22.50 a Optimal
capacity
chemical scrubber of a unit is for
2000 fatteners
or equivalent
Single-stage 43 14.00 b fatteners
chemical scrubber
Single-stage 9 3.00 b weaners
chemical scrubber
Single-stage 84 28.00 b farrower
chemical scrubber
Single-stage 63 25.05 b farrower
chemical scrubber
found capable of reducing NH3 emissions with 9.8.2 Bioscrubbers

efficiency up to 90%. The efficiency for odour
removal[10] was found to be lower, at an average of 29%. Biological treatment of air in scrubbers is a proven
technology, which can achieve an odour abatement
Diluted sulphuric acid is the most used scrubbing liquid efficiency of 70-80 %.
in this system. Hydrochloric acid solutions may also be
used. Bioscrubbers have been reviewed as a method for
ammonia reduction in the Netherlands, and were found
A total commercial production capacity of capable of reducing NH3 emissions from 3.0-0.8
approximately 2000 gestating sow places, 40,000 kg/animal capacity/year for finishers.
weaner places and 100,000 finisher places have been
equipped with this system in commercial; application in The technology is reasonably simple, and suitable for
the Netherlands (1999)[30]. application in an agricultural environment. No
chemicals are required, that might otherwise imply an
The cost for single-stage chemical scrubbers is additional environmental and Health & Safety risk
summarised in Table 17. Please note that these single- factor.
stage scrubbers were aimed at ammonia removal only.
The cost of a multi-stage scrubbing system that is The cost of bioscrubbers, as designed and applied
effective at removing odours may be significantly more specifically to pig production, seems to be relatively low
costly, see Table 16 (see Table 18), especially when compared to odour
abatement systems used for industrial applications.
67
Table 18: Indicative budget costs (€) of odour abatement in ventilation air using bioscrubbers (Netherlands, 1997)
Cost per animal capacity (fatteners)

Abatement system When data are not stated in the reference, the table cell is left blank Source Remarks
Capital investment [€] Life Capita Operational cost [€] Total
cost [€] annual
[€]
Abatement Ducting Total Years Total Energy Other Total
unit etc
Bioscrubber 49 34.09 83.18 10 8.32 2.50 14.23 16.73 25.05 a finishers
Bioscrubber 49 10 16.70 b finishers
Bioscrubber 10 10 3.35 weaners
Bioscrubber 111 10 32.75 farrower
Bioscrubber 111 10 16.70 dry sows
Practical application has been identified in the behind establishing a specialist, well-adapted population
Netherlands, for a capacity of in total approximately in the filter to utilise the compounds in the airflow to
1000 gestating sows, 20,000 weaners and 100,000 their metabolic optimum.
finisher places (1999)[30].
Biofilters are typically used for treating medium to high
9.8.3 Biofilters volumes of odorous air, from 5000 m3/hr to large flows
of several hundreds of thousands m3/hr, at low to
Biofilters are used widely for odour treatment, achieving medium odour concentrations
abatement efficiencies from 70 to over 95%. (5,000-100,000 ouE/m3).
In a biofilter, a solid porous medium (e.g. compost) acts Biofilters are most effective at temperatures between
as a carrier matrix for a biomass of micro-organisms 15°C and 50°C. Air at higher temperatures cannot be
(e.g. bacteria, actinomycetes and fungi). The treated. Cooling of warmer emissions may be required.
biologically active layer is moist, and the filter is kept at The optimum temperature is between 25°C and 35°C.
a high humidity. When an odorous airflow is passed As the treatment rate depends on biological metabolism,
through the filter bed, odorous compounds (gases, the rate of treatment doubles with every 10 degrees,
aerosols) that are soluble in water will be transferred to between 0°C and 40°C.
the water in the biologically active layer. The biomass
will use these compounds for their aerobic metabolism, The foul air needs to have a high relative humidity to
leading to biological oxidation of odourants to usually avoid drying out of the filter bed. Humidification of the
less odorous compounds. incoming flow may be necessary. At high flows even a
relatively small deficit in humidity may cause
Biofilters can be open or closed, can use upward or top- considerable mass loss of water in the biofilter bed.
down flow and can contain any of a variety of media, When the flow temperature increases in the filter, which
such as mature stable compost, peat/heather, coconut is after all metabolically active and will produce some
shells, seaweed, tree bark, woodchips etc. excess heat, the drying effect of the treated air may be
exacerbated
The medium layer is typically 1 m thick to avoid an
unacceptable pressure drop. At airflows of typically 50- The foul air needs to be relatively free of particles, as
150 m3/m2/hour this implies that biofilters may require these may clog the medium. The foul air should not
considerable areas of space. contain substances at concentrations that can have a
toxic impact on the biomass.
Some suppliers market specific cultures of
microorganisms. Typically, however, natural selection is Pre-treatment of the foul airflow may combine the
68
removal of particles and the cooling and humidification down odourants to less odorous oxidised forms.
of the flow before entering the biological treatment,
using counter flow sprays. The working principle is that odourants in the foul air
are mixed with air containing the treatment agent. In the
The design of the humidification or irrigation system of case of UV treatment, ultraviolet radiation is used to
the filter is crucial to its effective use. Irrigation can have ionise the oxygen in the foul air itself. The radicals that
multiple functions: are formed, including ozone (O3) will react with the
odourants, oxidising these to less odorous compounds.
• Moisture control The process depends mainly on turbulence (mixing),
The primary function of irrigation is to keep the oxidant concentration and residence time. Residence
moisture content of the filter at the required level. The time in the mixing chamber should be at least > 1
optimum moisture content may vary for different second, preferably as long as several seconds.
media.
UV systems seem to be more effective when used to
• Removal of metabolic products and pH control break down odourants with small molecular weight,
In beds with, for example, a high H2S loading, the pH such as H2S, but less effective on larger molecules.
in the filter tends to come down to very low values. Although manufacturers claim that the technology is
The effectiveness of the filter is reduced, as a result. suitable for large volume flows, residence time in the
The optimum pH is between pH=6 and pH=8. mixing chamber is difficult to maintain at the required
Irrigation can help to remove acids to the drain. Some length of one or more seconds.
materials are more suitable to be irrigated in this
manner (e.g. tree bark is suitable, while peat will tend Performance is claimed to be >90% for H2S. It is known
to become soggy and lose its structure). that efficiency is reduced for odourants with increasing
molecular mass. Little information is available for such
• Nutrient supply odourants, but indicative data would point to abatement
If the loading of the filter is very low additional efficiencies around 50%.
nutrients may be required to maintain sufficient
biomass. Surface water can be considered for this • Strengths
purpose. - Low capital cost;
- Suitable for on/off operation.
Irrigation is best provided by an array of nozzles, with
an even distribution over the surface of the bed. Dry • Weaknesses
patches should be avoided, as these may cause - Odour abatement efficiency for odourants with
irreversible loss of effectiveness locally. higher molecular mass remains questionable.
Biofilters typically have a distinctive residual odour, 9.9 Slurry storage

which will not be far below 100-300 ouE/m3. However,
this residual odour will in most cases resemble the odour Slurry storage can be a highly significant source in terms
of the soil, which is an earthy odour generally not of odour annoyance potential. Under anaerobic
recognised as annoying, as its character resembles that conditions, high concentrations of odourants can be
of odours naturally emitted from soil. formed in slurry, which can be released in highly
concentrated ‘puffs’ when slurry is being handled.
9.8.4 Ozone treatment of ventilation air Turbulence, resulting from ‘stirring’ and pumping, can
increase the emissions from the surface by an order of
Treatment of air using ozone is applied in feed magnitude (factor 10) compared to a still surface.
production as an end-of-pipe odour treatment method.
Practical application in pig production is rare. On the other hand, natural crust formation can reduce
the emissions from the surface significantly.
Ozone is a highly reactive oxidising gas that can break
69
In assessing the relevance of slurry storage for odour emission reductions of 70-80% have been reported.
annoyance potential it helps to realise that odour
concentrations over slurry, or in headspaces, can reach 9.9.2 Enclosed slurry storage
tens or even hundreds of thousands of ouE/m3, whereas
the odour concentration in pig house ventilation air There are a number of options for enclosed slurry
rarely exceeds 5,000 ouE/m3. Small volumes of very storage. This has a direct advantage in stopping almost
strong odours can, under unfavourable circumstances, all emissions to atmosphere. In rigid tanks, however,
travel quite far and cause a strong odour intensity to be great care should be taken that highly odorous air from
perceived by persons downwind. As this perception can the headspace is not released in a ‘puff’ when loading or
be of high intensity, this can trigger annoyance and unloading. The headspace from tanker and tank can be
exacerbate nuisance. attached, to form a closed system, or some form of
odour control can be used. Enclosed tanks inherently
It is therefore a significant area of interest to assess form a potential Health & Safety risk, when incidental
slurry storage and handling for transport as an area of access is required, as they are likely to contain lethal
opportunity for minimising odour impacts. concentrations of gases like H2S.
9.9.1 Odour emission reduction in open slurry Flexible solutions are becoming more popular, using
storage methods of cover that avoid creation of headspace. An
example of a covered storage without headspace are foil
In many cases, slurry is stored in open tanks. In Ireland, basins in an earth enclosure, with a floating foil cover.
storage underneath the fully slatted floors is common. Floats support the cover, and an extraction system for
Extra storage in the open air is becoming more of an escaping digestion gas is provided in the design. They
issue lately. This is typically achieved in open to tanks or are made out of reinforced plastic (PVC) foil of 1 mm
lagoons, built from foil, glass coated metal or concrete. thick (see Figure 9). Stirring of the slurry is achieved
through pumping slurry through a specially designed
To reduce emissions from open, uncovered tanks, the fixed tubing system. These fully enclosed foil basins
following factors and approaches are relevant: have an economic lifespan of at least ten years.
Hundreds if not thousands of these systems have been
• Reduce the surface area to volume ratio (deeper installed in the Netherlands, at commercial pig units. No
tanks, rather than a larger area); precise cost data are available, but the supplier indicates
that the investment cost is close to half the cost of a
• Minimise turbulence, by careful design of tubing. All concrete storage tank of the same capacity. The foil liner
filling should occur under the liquid surface to avoid will be viable for an economic life of 10-14 years.
turbulence;
A relatively new approach is the use of large ‘slurry
• If possible, formation of a natural crust should be cushions’, derived from military storage systems for fuel
encouraged; and water.
• Treatment of sludge by mechanical separation or Existing open top tanks can be covered by add-on
digestion or aeration can drastically reduce content of covers, typically made out of glass reinforced plastics.
odourants, and hence the odour emission; From Canada the use of inflatable domes has been
reported, although few details are available as to the
• Temporary cover such as straw, in a 10-20 cm layer, practical application of such systems.
has been suggested as an effective method for
reducing emissions from the surface;
• Reduce area of liquid exposed to air, using temporary

or floating covers (e.g. polystyrene floating panels).
These can reduce emissions significantly. Ammonia
70
71
Figure 9 Fully enclosed slurry storage in earth shape with 1mm reinforced PVC lining
(drawing courtesy of www.steenbergen.org)
9.9.3 Operational aspects of slurry storage
The following operational practices can be applied to

reduce odour emissions from slurry storage:
• When the storage is not in use, clean the surfaces

carefully. Even a few inches of anaerobic sludge at
the bottom can be a significant source of odour
emission;
• Avoid turbulence and splashing, as these can cause

emissions from the surface to be increased by an
order of magnitude;
• Always use pipes and tubes that extend to below the

liquid surface when filling;
• Promote and preserve natural crust formation;
• Keep storage periods as brief as possible;
• Choose your moment for any sludge handling

operation, so that the wind direction is favourable,
winds moderate to high and the atmosphere turbulent.
72
10. Two case studies
In the course of this project two case studies were cases dispersion modelling was used to assess the odour
conducted for pig units in the Irish context. The case impact. Abatement options were considered to reduce
studies included odour emission measurements. For the impact.
both locations dispersion modelling was used to assess
the odour impact. Abatement options were considered to The objectives of these case studies were:
reduce the impact.
• To illustrate the approach as outlined the main
The case studies were conducted at two sites, and the Agency study Odour Impacts and Odour Emission
characteristics are summarised below: Control Measures for Intensive Agriculture;
• A large integrated pig unit, with approximately 1000 • To obtain a limited set of emission measurements for
sows in fully slatted pig houses; Irish conditions, to assess whether these are
significantly different from the distribution of results
• A medium sized integrated pig unit, with found in the much larger data set from the
approximately 590 sows in fully slatted pig houses; Netherlands[5, 10].
The case studies for the two sites listed above are It should be noted that the case studies were not intended
reported in Part B of this report, titled: Case studies to provide an overview representative for most Irish pig
assessing the odour emissions and impact of two pig units. The scope of the case studies was too limited to do
production units in the Irish Situation The so. Similarly, the scale of the emission measurements
measurements of odour emissions carried out at these was not sufficient to yield specific emission factors
two sites are also included in Part B. generally representative for Irish conditions.
For full details of the case studies, including a detailed The case studies prompted the following conclusions:
report on odour emission measurements, please refer to
Part B of this report, that is structured as a complete Conclusions on the results of finisher emission rate
document. The essence of this report is provided below measurements in Ireland
in an executive summary.
The geometric mean emission rate of 13.2 ouE/s per
10.1 Executive summary Part B: Case studies finisher measured in Ireland in winter conditions for this
assessing the odour emissions and impact study is about one third lower than the value of 22.6
of two pig production units in the Irish ouE/s per finisher found in a larger study in the
Situation Netherlands[10].
The Environmental Protection Agency has initiated a Given the relatively small number of samples, collected
study into Odour Impacts and Odour Emission Control in the Irish study, and the statistical variance as derived
Measures for Intensive Agriculture, with the objective to from the larger Dutch study, the difference in the mean
assist the Agency in formulating its approach for outcome is too small to be statistically significant.
processing the license applications with a view to Therefore, it is justified to use the emission factors
achieving transparent and uniform decision-making. derived in the Netherlands for emission estimates in
Ireland, until emission factors specifically measured in
In the course of this project, three case studies were Irish conditions are become available for a larger sample
conducted to assess the odour impact of pig production of study sites.
units in the Irish context. At two of these locations odour
emission measurements were conducted. For all three
73
Odour impact study Farm A
Farm A is a large integrated unit, containing over 17 000

animals. It is therefore no surprise that total emissions
are high, and the odour footprint relatively large.
However, given the locality of the farm - its distance
from residential units and its rural context, there seems
to be no urgency as no complaints have been registered.
The only concerns that were identified during the site
visits were the uncovered slurry store and carcass skips.
It is the opinion of Odournet UK that these sources may
become a significant emitter of odours during the warm
summer months. However, having undertaken sampling
during a cool spring day, there is no quantitative data to
support this.
The modelling shows that in the current situation a

limited number of (five) dwellings may be affected by
odour impacts in excess of the limit value. It is therefore
necessary to seriously consider the need to reduce
emissions and the options to do so. In the short-term
measures to reduce emissions from sludge storage
should be considered. In the longer term, replacement of
housing assets could reduce the number of dwellings
exposed to odour impacts in excess of the limit value,
conceivably to zero. The target value will be difficult to
attain for farm A. The farm can be made sustainable at
current stock levels, from the perspective of odour
impact, provided that the community recognises and
accepts the rural context in the vicinity.
Odour impact study Farm B
Farm B is a relatively small-scale operation. Under

current circumstances, the odour footprint does not
include any domestic dwellings within the limit value
contour. This farm does not require the installation of
any abatement options if current stock numbers and
good practice are maintained.
In the long-term, a reduction of odour emissions should

be considered when normal renewal of housing assets
becomes an issue. This could achieve attainment of the
target value for all dwellings in the vicinity, or create
room for some growth of stock when the agricultural
context of the area is recognised and accepted as the
status quo.
74
11. Conclusions
This report does not aim to provide one answer to a factor in determining the setback distance required in
well-defined question, with a concise set of conclusions. a particular direction. These particular circumstances
Rather, it aims to provide the information, required by increase the need for specific modelling, in cases
the Environmental Protection Agency, for defining a where the outcome is not clear-cut.
framework for answering the complex questions that
need to be addressed to achieve a systematic and 6) The geometric mean emission rate of 13.2 ouE/s per
transparent IPC licensing policy. Such a policy will finisher measured for winter conditions in Ireland for
ultimately be required if a reasonable and durable this study is about one third lower than the annual
balance is to be achieved between the economic interests mean value of 22.6 ouE/s per finisher found in a
of the pig producer and the environmental interests of larger study in the Netherlands.
those using the vicinity to live, work and play.
7) Given the relatively small number of samples,
On the basis of the issues explored in the full text, a collected in the Irish study, and the statistical
number of general conclusions can be made to highlight variance as derived from the larger Dutch study, the
the essence of the report: difference in the mean outcome is too small to be
statistically significant.
1) A significant number of pig production units will
require a licence, based on current National and 8) It is therefore justified to use the emission factors
European legislation derived in the Netherlands for emission estimates in
Ireland, until emission factors specifically measured
2) An assessment framework based on quantitative in Irish conditions are available for a larger sample
emissions is the most likely to achieve a transparent of study sites.
licensing practice that achieves a balance between
the interests of the pig producer and those who use 9) The options for reducing odour emissions from pig
the surroundings as their living environment. production do exist. Reductions to 50% relative to
the most common fully slatted production unit are
3) The proposed assessment framework identifies one quite feasible. However, the financial viability of
environmental target for all situations. To allow for a many retrofit methods is an issue of concern, given
degree of flexibility two limit values have been set, the low economic returns on pig production.
for new production unit applications and for existing
facilities. The ‘space’ between the target and the 10)The economics of installing the technology to abate
limit values can be used in the licensing process to odour emissions must be assessed before this
tailor the conditions to the specific requirements and technology can be imposed on existing pig unit
opportunities that exist for that licence application. operations.
4) The proposed framework for target and limit values 11)The most viable low-emission options involve
is, in general terms, compatible with the setback modification of pig houses, or replacement by new
distances required or advised in other countries, such low emission design housing. Such structural
as Germany, Netherlands, New Zealand. abatement can only be reasonably achieved in the
normal economic cycle of asset replacement, in most
5) The prevailing wind direction in Ireland causes a cases
distribution of odours that is not entirely
symmetrical. The actual meteorology of the location 12)Retro fitting of abatement systems, using air
of the pig unit in question and the actual position of treatment systems such as bioscrubbers, chemical
the receptor relative to the source, are therefore a scrubbers or biofilters, can achieve significant
75
emission reductions of between 70% and over 95%.

The main impediment is the additional cost incurred,
which can increase the cost of a pig produced by
roughly 10-20%. Market conditions in recent years,
generally speaking, do not allow such an increased
cost.
13)Good operational practice, including suitable

landscaping, tree screens and pro-active community
relations, remain a main factor in reducing
annoyance and avoiding annoyance developing into
nuisance.
14)A suitable production site for a given production

capacity will become a major asset for any pig
producer, which may become a main factor in
determining the sustainability of the activity.
Producers are well advised to use the planning
process to their advantage and be pro-active in
counteracting any encroachment into the existing
setback zone by any developments that may be
termed an ‘odour sensitive receptor’.
15)By making transparent the assessment of the impact

of pig production on the vicinity, the proposed
framework can contribute in practice to the
protection of the interests of both pig producers and
the general public.
76
Annex A. Odour regulations for intensive livestock

in other countries
A.1 Germany Table 19: Conversion table from pig life stages to
German GV units, according to VDI3471
A.1.1 General regulatory framework and relevant Conversion of pig numbers to GV, VD13471
guideline documents Animal stage
Dry sow, boar 0.3 GV
The law concerning air quality issues in Germany is the Wet sow with progeny < 4 weeks 0.4 GV
Bundes Immissionsschutzgesetz (known as Wet sow with progeny > 4 weeks 0.5 GV
‘BimSchG’), or the Federal Immission Control act of Gilts 0.15 GV
1990, which is available in English from the Ministry for Weaners ≤ 15 kg 0.01 GV
Environment, Nature Conservation and Nuclear Weaners >15 kg and < 25 kg 0.02 GV
Safety. Fatteners (batch finishing) ≤ 45 kg 0.06 GV
Fatteners (batch finishing) > 45 kg 0.15 GV

All odours from any commercial installation are Fatteners (continuous finishing) 0.12 GV
considered an annoyance, according to §3 of from 25 to 105 kg
‘BimSchG’. For licensing and enforcement, however,
the issue is to determine whether the annoyance at assessing ambient air quality for odours in the vicinity
constitutes a ‘significant disturbance’, on the basis of the of an existing source. The document is available in
‘relevance of the annoyance’. However, the ‘BimSchG’ English translation:
does not provide for criteria to determine when an
annoyance becomes a significant disturbance • Determination and Evaluation of odour immissions -
(nuisance). Odour exposure guideline (Feststellung und
Beurteilung von Geruchsimissionen -
The second relevant official regulatory document, aimed Geruchsimmissionsrichtlinie), Länderausschuß für
at providing technical guidance for specific industries on Imissionsschutz, LAI-Schriftenreihe No. 5, Berlin
how to achieve the general principles concerning air 1994.
quality in the ‘BimSchG’ also fails to provide
operational annoyance criteria. The Technische This method describes a method for long-term field
Anleitung zur Reinhaltung der Luft, or TA-Luft, details panel observations, in which the fraction of ‘odour
the technical measures, expected to be applied in hours’ is determined by a team of assessors on pre-
different sectors of industry and agriculture, including defined locations on a grid around the source in
methods for assessment. The TA-Luft is available in question. The method has been applied on pig units[9].
English: This method can be applied apply to determine licensing
applications.
• Technical Instruction on Air Quality control (Erste In most licensing cases, however, technical guidelines
Allgemeine Verwaltungsvorschrift zum are applied, that provide detailed advice on the design
Bundesimmissionsschutzgesetz), Federal Ministry for and operation of pig units and other livestock
the Environment, Nature Conservation and Nuclear operations:
Safety, Bonn, 1986 (GBBl. P.95)
• VDI3471:1986 Emission Control. Livestock
The TA-Luft defines a maximum ‘odour frequency’, as management - Pigs
an ambient air quality characteristic, but does not
specify a method to assess this parameter. In 1994 the • VDI3471:1986 Emission Control. Livestock
Department of the Environment of Nordrheinland management - Hens
Westfalen introduced a method for this purpose, aimed
77
• VDI3473:1994 Part 1 (draft) Emission Control. Table 20: Correction factors and their points value for
Livestock farming - Cattle. Odourants. use with the setback distance graph, VDI3471
Point system for setback distance graph, VDI3471

In addition to technical guidance on the design and Criteria Points
operation of pig units, the technical standard, Waste removal and storage
VDI3471:1986 contains a graph providing setback Solid manure removal
distances, for pig units of different sizes. This graph is ‘Tiefstall’ 60
presented in Figure 10. In determining setback Mechanical manure removal to storage 50
distances, operational methods and design of the pig unit enclosed by walls on three sides
are taken into account, using a system of assigning Mechanical manure removal to transport 40
points. A correction on the standard setback distance can vehicle
be applied on the basis of the total number of points. Mechanical manure removal to open air 20
manure heap
If a pig unit complies with the VDI3471 standard, and is Liquid manure removal
located so that the setback distances are respected, that Slatted floors, >45% 10
is in most cases sufficient ground for the local authority Slatted floors, <45% 5
to grant a licence. In those cases where the distance to Mechanical removal 0
residences is less than 100m, or in cases where the Slurry storage
setback distances cannot be attained fully, expert advice Storage tank fully enclosed 50
is sought to determine the application, using detailed Storage with cover 30
assessments, typically based on atmospheric dispersion Storage with full natural crust formed 30
modelling. Storage without cover 0
Underfloor storage in the pig house 30
A.1.2 The distance graph in VDI3471 Ventilation
Summer ventilation rate, according to
The capacity of a pig unit is expressed in DIN18910
‘Grossvieheinheiten’ (GV), that are equivalent to 500 kg Temperature difference ≤ 2 K 10
live weight. The number of GV in a pig unit is calculated Temperature difference ≤ 3 K 5
using the conversion factors in Table 19. Temperature difference > 3 K 0
Ventilation exit duct
For a typical Irish integrated unit, capacity for one Vertical, height ≥ 1.5 m above roof apex 15
integrated sow and progeny, including finishers, would Vertical, height < 1.5 m above roof apex 5
be approximately equivalent to 1.3 GV. Horizontal side vents 0
Vertical exit velocity at summer ventilation rate
Once the number of GV units has been determined, a Velocity ≥ 12 m/s 25
point system is applied to take design and operational 10 ≤ velocity < 12 m/s 20
practice into account. The point system is summarised in 7 ≤ velocity < 10 m/s 10
Table 20. Velocity < 7 m/s 0
Miscellaneous
For a 400-sow unit, which is close to the lower limit for Special feeds, dry waste food 0
licensing in Ireland, the setback distance would be Kitchen wastes with weak odour up to -10
between approximately 390 and 620m, depending on the Wastes with a strong odour Up to -25
number of points. Location Up to plus or
minus 20
A.2 Netherlands Slurry storage capacity
≥ 6 months 10
A.2.1 Guideline documents ≥ 5 months 5
≥ 4 months 0
The pig production sector in the Netherlands is
78
Figure 10: Setback distances graph, for different numbers of 'Livestock Units' (GV), wit
correction based on points for operational practice and design of the pig unit. Germany,
VDI3471. One sow on an integrated unit in Ireland is approximately equivalent to 1.3 GV.
considerable in size, relative to both the size of the and 1996.

population and the surface area of the country. Annual
production is approximately 30 million pigs, which The successive guideline documents are:
amounts to 2 pigs per head of the population, There are
1.4 million places for sows and 7.4 million places for • Brochure Livestock Rearing and Nuisance Law
finishers in Dutch pig houses (1998). It is therefore not (1976)[25]
surprising that odour impact of pig production is a major
environmental issue, given the high density and • Guidance note on the application of the Nuisance
proximity of both residents and pigs. Very recently, in Law on livestock production units (1984)
2000, a considerable budget of more than 200 million
Euro was made available by the Dutch government to • Brochure on Livestock Production and Nuisance Law
buy out existing pig production units to reduce the (1985)
capacity of the sector and its environmental impact.
• Assessment of accumulation by intensive livestock
The first guideline on how to take account of production, Publication Series Air no. 46, Ministry of
environmental odour aspects for licensing as a result of Public Planning and the Environment (1985)
application of the existing the Nuisance Law was issued
in 1971, and revised several times in later years: 1984
79
• Guideline Livestock Production and Odour • 1 mve is equivalent to 1.4 fatteners held in a ‘Green
Annoyance (1996) Label’ low-emission housing system
The guideline of 1996 is currently used. The main The ‘distance graph’ provides four lines, differentiated
instrument for managing odour impact through licensing for the category of the land use in the vicinity of the pig
has been retained in all these successive guidelines, in unit.
the form of a graph relating the required setback • Category I provides a higher degree of protection
distances to the number of animals in the pig production and is applied to non-agricultural, residential areas,
unit, see Figure 11. The size of a pig production facility hospitals, recreational accommodation etc.
is expressed in ‘mestvarkeneenheden’ or [mve], which
translates to ‘fattener units’. In many translated • Category II is characterised by more disperse
publications the less correct and confusing term ‘pig residential use in villages or hamlets, where the living
units’ has been used. One [mve] represents the emission environment has a rural character.
of odours of one fattener, held in a traditional housing
system (i.e. partly slatted) over one year. • Category III is for more or less isolated residences
or clusters of residences in an otherwise rural
There is a table for conversion of various animals and environment.
life stages to [mve]. In the 1996 guideline, the following
numbers of animals are equivalent to 1 mve: • Category IV is the most lenient level of protection
and is used for environments with farmhouses only.
• 1 mve is equivalent to 11 weaners
The ‘distance graph’ has remained largely the same over
• 1 mve is equivalent to 1.5 wet sows the years. In 1996 the lines were extended for higher
numbers of pigs. The main drive behind the 1996
• 1 mve is equivalent to 3.0 dry sows revision was to find ways to allow further expansion of
pig units that had reached the limits of their expansion
• 1 mve is equivalent to 1.0 fatteners on a conventional, possibilities. Although the ‘distance graph’ remained
partly slatted system (and equivalent to 22.6 ouE/s ) largely unchanged, the interpretation of the categories
Figure 11 Setback distances for pig production units in the Netherlands, 500 to 2500 pig units
(equivalent to fatteners), Brochure Hinderwet en Veehouderij, 1996
80
and the method of measuring distances between practicable steps are to be undertaken to investigate
buildings and residences was adapted to allow some complaints by residents made to them.
degree of further expansion, while maintaining the
objective that no ‘severe odour nuisance’ should occur. Statutory Nuisances are defined in Section 79 of the Act.
Another modification concerned the conversion for The relevant passages, relating to odour, are:
various animals and life stages to [mve] units, which
now included values for low emission housing, the • …any dust, steam, smell or other effluvia arising on
certified ‘Green Label’ housing systems. Using the industrial, trade or business premises and being
expanded table for these ‘Green Label’ systems allowed prejudicial to health or a nuisance;
existing pig units to expand, providing that they used
these low emission housing systems. . • fumes or gases so as to be prejudicial to health or a
nuisance;
In 1997 and 1998 a number of rulings of the Council of
State, the highest appeal court for planning cases, have • any animal kept in such a place or manner as to be
challenged parts of the revisions of 1996. Particularly prejudicial to health or a nuisance; and
the modifications in the application of the categories,
that implied that a number of categories of residences • any other matter declared by any enactment to be a
were moved to a less protected category, was not statutory nuisance.
accepted. The Council of Stated judged the motivation
for the revisions insufficient. As a result, the Ministry of Where a local authority Environmental Health
Public Planning and the Environment has started a Department is satisfied that a statutory nuisance exists,
number of research projects, to establish the scientific or is likely to occur or recur, it has a duty to serve an
basis for the relation between exposure to pig odours abatement notice under Part III, Section 80 of the Act
and actual levels of annoyance. A second project requiring:
involves the measurement of emission factors for
different life stages of pigs. The results will be used as • the abatement of the nuisance or prohibiting or
the starting point for a major policy review, named restricting its occurrence or recurrence; and
Revision of policy instruments for rural odour policy,
with the Dutch acronym VIAS. The review is currently • the execution of such works and the taking of such
ongoing and is expected to lead to introduction of a fully other steps as may be necessary for these purposes.
revised guideline in autumn 2000 or early in 2001.
A person served with an abatement notice may appeal to
It is expected that the ‘distance graph’ will continue to the Magistrates Court within 21 days of being served
be applied, but the use of four categories of land use may with the notice. In the event the person does not comply
be simplified and adapted to reflect the results of the with the Abatement Notice fines up to GBP 20,000 may
dose-effect studies. In addition, a revision of the be imposed. In such cases, it is a defence to show that
conversion factors to [mve][ for different life stages is you have used the best practicable means (BPM) to
expected to be revised, on the basis of recently measured prevent or counteract the nuisance. BPM is defined to
values. have regard among other things to local conditions and
circumstances, to the current state of technical
A.3 United Kingdom knowledge and to the financial implications.
The Environmental Protection Act of 1990 provides the The law on statutory nuisance is far from
legal framework for avoiding and controlling odour straightforward. A key problem is that no criteria are
nuisance in the United Kingdom. The Environmental provided to decide when occurrence of an odour
Health department of the Local Authority is responsible constitutes a nuisance, and when it is acceptable. The
for its enforcement. Under Part III, Section 79 of the system relies heavily on the individual judgement of the
Act, the local authority has a duty to inspect their area Environmental Health Inspector. In practice a wide
and detect any statutory nuisance. Reasonably variety of licence conditions occur.
81
Planning consents have to be granted on the basis of the A.4 United States
Town and Country Planning (General Permitted
Development) Order of 1995 (GPDO). New livestock Swine CAFO odours: Guidance for Environmental
facilities, such as livestock buildings, slurry storage Impact Assessment, US EPA, Region 6, Dallas, Texas,
facilities, and extensions or alterations to such facilities, contract no. 68-D3-0142, 1997
need planning permission when these will be within a
distance of 400m from the boundary of any protected A.4.1 Setback distances:
buildings (such as residential houses or schools).
The American Society of Agricultural Engineers
Under the Town and Country Planning (Assessment of (ASEA,1994) indicates that a ‘desirable distance’ for
Environmental Effects) Regulations of 1988 an siting livestock facilities in general is 1600 m from
environmental assessment is to be carried out for certain housing developments and 400-800 m from
types of major project which are likely to have neighbouring domestic dwellings.
significant effects on the environment. For livestock
units this requirement is likely to apply to new pig units The EPA guidance indicates that the setback distance
of more than 400 sows or 5000 fatteners and new poultry should be at least 3.6 km and preferably 7.2 km for
units of more than 100,000 broilers or 50,000 layers. ‘larger facilities’.
For planning procedures, the use of odour modelling A.5 New Zealand
with application of a criterion of 5 ouE/m3 as a 98-
percentile of hourly values has been accepted as an A Code of Practice for pig production and siting is
acceptable approach to demonstrate that no statutory available on the Internet,
nuisance would arise, in a planning enquiry involving a at www-aghort.massey.ac.nz/centres/mrc/extension/cop
wastewater treatment plant at Newbiggin-by-the-Sea,
Northumberland, planning reference APP/F2930/A/92 The New Zealand Code of Practice (CoP) contains two
206240, UK, 1993[16]. types of setback distances:
The main document providing guidance is: • Fixed setback distances, that must be observed in all
cases, regardless of the size of the production unit
• The Air Code, Code of Good Agricultural Practice for
the protection of Air, revised 1998, Ministry of • Adjustable setback distances that depend on the size
Agriculture, Fisheries and Food and the Welsh Office of operation and a set of correction factors for
Agriculture Departments, October 1998. operational characteristics.
The Code provides general guidance on the legal The adjustable setback distances must be applied to pig
background, and on good practice for production. The production units with 2000 pigs or more.
main part of guidance on odours is given in Part B of the
Code, which contains a wealth of sound general advice, For any piggery having more than 5000 pigs, the
but is remarkably limited on technical detail and potential to create adverse effects will have to be
quantitative assessment and management information. determined on an individual case basis. The size of the
The Code does not contain any specific recommendation buffer zone for such a piggery will reflect this.
on setback distances, other than suggesting that any pig
unit located at less than 400 metres from residences One pig is counted for the P-factor when it is older than
should take extra care in implementing the advice given 70 days. Breeding units with weaners only are counted
in the Code. using a conversion of 1 breeding sow = 5 pigs.
82
The fixed setback distances are listed below: There is a system to take into account good operational
practice. Based on a number of operational
Table 21: Fixed setback distances for intensive pig characteristics, see Table 23, a piggery reduction factor
production units, New Zealand.
can be calculated, the percentage by which the standard
Feature Separation distance setback distance required can be reduced, up to 40%
(meters) maximum:
Residential building on same site 50
Milking shed and yard 45 Piggery reduction factor = 100 (1-[AxBxCxDxExFxG])

Slaughterhouse 50
Reservoir for domestic water supply 800 Specific setback distance calculations are included for
Well for domestic supply 30 situations where local topography or meteorology may
Water course 20 have a significant influence in the odour exposure.
Public highway 50
Property boundary 20
(with adjoining paddocks)
The flexible setback distances are differentiated,

depending on the land use of the area surrounding the
pig unit. Three classes of land use are distinguished:
• Zone 1A - Piggery reference point to a residential

zone in an urban area
• Zone 1B - Piggery reference point to a place of public

assembly
• Zone 2 - Piggery reference point to a rural dwelling

not on the same property
Depending on the zone the adjustable setback distance is

determined on the basis of a table, relating the P factor
to adjustable setback distance, see Table 22.
Table 22: P values and zone distances -

standard piggery, New Zealand
P value Zone 1A Zone 1B Zone 2

(No. of pigs) meters meters meters
Up to 2000 2000 1500 500
2500 2500 1875 625
3000 3000 2250 750
3500 3500 2675 875
4000 4000 3000 1000
4500 4500 3375 1125
5000 5000 3750 1250
83
Table 23: Piggery reduction factors, New Zeland
Designator Detail Adjustment

factor
A Emmissions to air from buildings
Weather by natural airflow or fan driven1
(1) Bio filtration of exhaust air 0.80
(2) Ridge-ventilators plus trees (more than10 m high) 0.90
surrounding the piggery compound
(3) Ridge-ventilators only 0.95
(4) Ridge & side-ventilators 0.90
(5) Side ventilators only 1.00
B Effluent collection system within all pig buildings

Faeces, urine and other biological material removed from the confines of the building
(1) Less than 12 hours old 0.75
(2) As compost formed in sawdust bed penning system 0.75
(3) While essentially aerobic but not greater than 30 hours old 0.90
(4) Greater than 30 hours, but less than 4 days old 1.00
(5) Greater than 4 days old 1.20
C Effluent collection system outside all pig buildings (but within the piggery compound)
(1) Closed pipes (pig buildings to aerobic holding thank/pump) 0.95
(2) Open channels (pig buildings to aerobic holding tank/pump) 1.00
D Effluent treatment system(within the piggery compound)

(1) Anaerobic ponds(s) (inc. all inlet pipes/channels) 1.00
(2) Facultative ponds(s) (inc. all inlet pipes/channels) 0.95
(3) Aerobic pond(s) 0.60
(4) Aerated pond(s) (aerobic surface layer over entire pond) 0.75
(5) Composting using aeration or regular turning 0.80
E Noise
(1) Maintaining noise recommendations (see part 4) 0.95
F Power supplies for ventilation, water supply, effluent handling and pumping
(1) Reliable power supply (loss of supply for not more than an aggregate of
2 hours month) 0.95
(2) Standby power supply for each 25% reduction in full load standby capacity 0.80 plus
0.05 per 25%
84
Table 23: Piggery reduction factors, New Zeland (continued)
Designator Detail Adjustment

factor
G Management
Stock under surveillance:
24 hours/day 0.90
12-23 hours/day 0.95
6-11 hours/day 1.00
1-5 hours/day 1.10
less than 1 hour/day 1.20
Stock under surveillance shall mean that a person, qualified or competent to have charge of stock and deal with routine or
emergency conditions is monitoring the fuctions of that piggery.
1 For ventilation requirements see:

British Standards Institution. (1990). BS 5502: 1990 Building & Structures for Agriculture. Part 42: Code of practice for design and
construction of pig buildings. London. BSI
The Code of Practice - Pig Farming calls for a high standard at all piggeries, which is achieved by good management control of
odour generating procedures. A piggery scoring a negative adjustment factor (i.e. an increase) in variable buffer zone distance
would not meet the requirements of this Code of Practice.
85
86
Annex B. Methods for odour assessment and

units of measurement
This annex describes the concepts used for CEN EN 13725:1999, Air quality - Determination of
characterising odours, the methods used for assessing odour concentration by dynamic olfactometry,
odours and the units of measurement used. CEN/TC264/WG2 ‘Odours’, 1999
B.1 Detectability The odour concentration of a gaseous sample of

odourants is determined by presenting a panel of
Detectability (or odour threshold) refers to the minimum selected and screened human subjects (Figure 12) with
concentration of odourant stimulus necessary for that sample, in varying dilutions with neutral gas, in
detection in some specified percentage of the test order to determine the dilution factor at the 50%
population. The odour threshold is determined by detection threshold (D50). The odour concentration of
diluting the odour to the point where 50% of the test the examined sample is then expressed as multiples of
population or panel cannot detect the odour any more. one European Odour Unit per cubic metre [ouE/m3] at
The original odour concentration of an odour sample can standard conditions.
be characterised by the number of dilutions to reach this
detection threshold. At the detection threshold the odour
concentration is 1 odour unit per metre cubed (ouE/m3).
Threshold values are not fixed physiological facts or
physical constants but statistically represent the best
estimate value from a group of individual responses.
Odour concentration is the most common attribute used
to characterise odours. It provides the most common
measure to characterise the magnitude of stimulus for
determining the other attributes of an odour (the
horizontal axis).
A European CEN standard method for measuring odour

concentration is available[11], as described in the
following sections. The European odour unit (ouE) is
pegged to a well-defined reference material, through the
exclusive use of trained assessors selected for their Figure 12: Dynamic olfactometry with human assessors,
to measure odour concentration according to EN13725,
specific sensitivity to the reference odour of n-butanol. photo courtesy of Odournet UK Ltd.
In this manner the ouE has been made traceable:
1 ouE/m3 ∫ 40 ppb/v n-butanol. B.1.2 The unit of measurement
B.1.1 Measurement of odour concentration using The odour unit is a difficult unit to define, because it
olfactometry relates a physiological effect to the stimulus that caused
it. The stimulus, in this case, can be a multitude of
Odour measurement is aimed at characterising substances. The way in which the response of our sense
environmental odours, relevant to human beings. As no of smell is reduced to a single value of a parameter
methods exist at present that simulates and predict the amounts to a gross simplification of the rich spectrum of
responses of our sense of smell satisfactorily, the human sensory information that is actually perceived by the
nose is the most suitable ‘sensor’. Objective methods brain. Such a simplification may be useful, however, in
have been developed to establish odour concentration, describing potential effects. The reduction of a very
using human assessors. A draft European CEN standard complex set of physiological processes to a simple
applies to odour concentration measurement: parameter is methodologically very similar to
87
expressing the effects of toxic substances on an to that elicited by one European Reference Odour Mass
organism as the LD50, indicating the dose that causes a (EROM), evaporated in 1 cubic metre of neutral gas at
lethal effect in 50% of a well-defined test population. standard conditions
The complex physiological response is regarded as the
unifying reaction that can be caused by a wide range of One EROM, evaporated into 1 cubic metre of neutral gas
substances, at an equally wide range of dosages. In at standard conditions, is equivalent to the D50
general terms, this approach can be used to describe the physiological response (detection threshold), assessed
potential of a certain amount of a substance to cause a by an odour panel in conformity with this standard, and
physiological effect, by expressing the dose as a has, by definition, a concentration of 1 ouE/m3. There is
multiple of the dose that would cause an effect in 50% one relationship between the ouE for the reference
of a population. The definition and use of the unit are odourant and that for any mixture of odourants. This
highly analogous to that of the odour unit. In odour relationship is defined only at the D50 physiological
research, the D50 could be described as the 50% of a response level (detection threshold), where:
population that can detect a sensory stimulus. In the past
odour researchers have not used populations of standard 1 EROM (for n-butanol, CAS 71-36-3) ∫ 1 ouE
test subjects, and have only related the physiological for the mixture of odourants.
response to the number of dilutions of the dose of a
sample to be measured. That practice implies a This linkage is the basis of traceability of odour units for
fundamental inability to compare the dosage of the any mixture of odourants to that of the reference
samples through other means than the population itself. odourant. It effectively expresses odour concentrations
This can only be justified if the researcher is convinced in terms of ‘n-butanol mass equivalents’.
that the samples of the population are sufficiently large
to compensate for biological variability within this The odour concentration is expressed as a multiple of
population. This assumption, however, cannot be one ouE in a cubic metre of neutral gas. The odour
fulfilled in the practice of odour measurement. The concentration can only be assessed at a presented
small sample from the population (4-8 subjects, more or concentration of 1 ouE/m3. The odour concentration, in
less randomly chosen) is far too limited a sample to be ouE/m3, can be used in the same manner as mass
representative, knowing the variability of sensitivity concentrations (kg/m3).
within the population. This practice does not comply
with statistical requirements as used in toxicological Note: When using odour concentrations one should be
experimental design, as the sample size from the aware that the relationship between the odour intensity
population required to be representative (hundreds) is and the odour concentrations is not linear, and may be a
far larger than the regular number of panel members different relationship for different (mixtures of)
used in olfactometry for environmental applications. odourants.
The solution is to standardise the test subjects used to B.1.3 Odour concentration measurement using
assess the sensory response. Reproducible results can be quantitative olfactometry
obtained by selecting panel members with a known
sensitivity to an accepted reference material (now n- The odour concentration of a gaseous sample of
butanol CAS-nr [71-36-3]). The assumption made is that odourants is determined by presenting a panel of
the sensitivity for the reference odourant will be a selected and screened human subjects with that sample,
predictor for sensitivity to other substances. The dose of varying the concentration by diluting with neutral gas, in
other substances and mixtures is then expressed in order to determine the dilution factor at the 50%
multiples of the dose that would elicit a physiological detection threshold (Z50 ZITE,pan ).
reaction equivalent to that of the reference. In practical
terms: The European odour unit [ouE] is that amount of At that dilution factor, the odour concentration is 1
odourant(s) that, when evaporated into 1 cubic metre of ouE/m3 by definition. The odour concentration of the
neutral gas at standard conditions, elicits a physiological examined sample is then expressed as a multiple (equal
response from a panel (detection threshold) equivalent to the dilution factor at Z50) of 1 European odour unit per
88
Table 24: Confidence interval for replicated measurements using dynamic olfactometry
n Sr . Sr lower limit m upper limit Lower limit m upper limit

2,0. 10 2,0 √n
√n log(ouE/m3) log(ouE/m3) ouE/m3 ouE/m3
1 0.3443 2.2093 2.6557≤ 3.0 ≤3.3443 453≤ 1000 ≤2209

2 0.2434 1.7516 2.7566≤ 3.0 ≤3.2434 571≤ 1000 ≤1752
3 0.1988 1.5804 2.8012≤ 3.0 ≤3.1988 633≤ 1000 ≤1580
4 0.1721 1.4864 2.8279≤ 3.0 ≤3.1721 673≤ 1000 ≤1486
5 0.1540 1.4255 2.8460≤ 3.0 ≤3.1540 702≤ 1000 ≤1425
6 0.1405 1.3821 2.8595≤ 3.0 ≤3.1405 724≤ 1000 ≤1382
7 0.1301 1.3493 2.8699≤ 3.0 ≤3.1301 741≤ 1000 ≤1349
8 0.1217 1.3235 2.8783≤ 3.0 ≤3.1217 756≤ 1000 ≤1323
9 0.1148 1.3024 2.8852≤ 3.0 ≤3.1148 768≤ 1000 ≤1302
10 0.1089 1.2849 2.8911≤ 3.0 ≤3.1089 778≤ 1000 ≤1285
cubic metre (ouE/m3) at standard conditions for Table 25: Confidence interval for determining the
olfactometry (Room temperature (293 K), normal abatement efficiencyusing dynamic olfactometry,
for different numbers of replicates
atmospheric pressure (101,3 kPa) on a wet basis). m
n Lower limit 1-10 D Upper limit
nod
confidence = confidence
The measurement must be carried out with a selected
interval interval
panel. The method should comply fully with the CEN
standard EN13725. nod
1 90 69.3% ≤90.0% ≤96.7%
2 90 77.9% ≤90.0% ≤95.5%

The performance of odour concentration measurements
3 90 80.9% ≤90.0% ≤94.8%
has been defined in the performance criteria of the
4 90 82.5% ≤90.0% ≤94.3%
standard. These imply that for one single measurement
5 90 83.5% ≤90.0% ≤93.9%
result, the 95% confidence interval will be:
6 90 84.2% ≤90.0% ≤93.7%
7 90 84.7% ≤90.0% ≤93.5%

The geometric repeatability of the sensory calibration
with n-butanol is r’ ≤ 3, complying with the EN13725. 8 90 85.1% ≤90.0% ≤93.3%
9 90 85.5% ≤90.0% ≤93.1%
The confidence limits for a value x for two
10 90 85.7% ≤90.0% ≤93.0%
measurements (k=2) is:
x . 2.09-1 ≤ x ≤ x . 2.09
In other words, if the real concentration is 1000, the B.2 Intensity

result of analysis will in 95% of cases lie in the interval
between 478 and 2090 ouE/m3. Analysing more than one Intensity is the second dimension of the sensory
replicate of a sample can reduce the uncertainty. Table perception of odourants, which refers to the perceived
24 shows the 95% confidence interval for replicated strength or magnitude of the odour sensation. Intensity
measurements, for the repeatability that is required in increases as a function of concentration. The relation
the EN13725 standard. between perceived intensity and the logarithm of odour
concentration is linear.
For the assessment of the efficiency of an odour
abatement unit, the repeatability is an important Odour intensity refers only to the magnitude (strength)
consideration. Again assuming the repeatability required of the perception of an odour. Intensity has a second
in the CEN standard EN13725, Table 25 gives meaning, in that it can refer to the magnitude of the
confidence intervals for the filter efficiency in relation to stimulus causing the perception.
the number of samples taken both before and after the
abatement unit n, when the actual efficiency is 90%. The relationship between perceived intensity I and the
89
stimulus may be described as a theoretically derived 1 no odour

logarithmic function according to Fechner: 2 very faint odour
3 faint odour
S = kw . log I I (A.1) 4 distinct odour
where ˚ 5 strong odour
S perceived intensity of sensation 6 very strong odour
(theoretically determined) 7 overwhelming odour
I physical intensity (odour
concentration) The values for I are then plotted against the logarithm of
I threshold concentration the odour concentration or the dilution factor. The
˚
kw Weber-Fechner coefficient regression line characterises the relation between
perceived intensity and odour concentration, see Figure
or as a power function according to Stevens: 3. The point where the regression line intersects with the
horizontal axis is equivalent to the detection threshold.
S=k.In (A.2)
where By comparing the slope of the regression line for
S perceived intensity of sensation different odours they can be characterised. Some odours
(empirically determined) cause a rapid increase in perceived intensity with
I physical intensity (odour increasing concentration (such as NH3). Other odours
concentration) cause only a slow rise of perceived intensity, such as
n Stevens’ exponent commercial toilet air fresheners that are designed to be
k a constant perceived at a similar intensity, regardless of dilution.
Which one of these two descriptions applies depends on B.3 Odour quality (descriptive)
the method used. To date no theory has been able to
derive the psychophysical relationship from knowledge Odour quality is the third dimension of odour. It is
about the absolute odour threshold of various expressed in descriptors, i.e. words that describe what
substances. the substance smells like. This is a qualitative attribute
that is expressed in words, such as fruity.
The method for measuring intensity is derived from the
following standard documents: B.4 Hedonic tone
• VDI 3882:1997, part 1, Determination of Odour Hedonic Tone is the fourth dimension of odour. This is a
Intensity, Düsseldorf, Germany. category judgement of the relative like (pleasantness) or
dislike (unpleasantness) of the odour. The method for
The principle of measurement is the presentation of the measuring intensity is derived from the following
odour to human assessors in an odour panel, at varying standard documents:
degrees of dilution, hence varying perceived intensity.
VDI 3882:1997, part 2; Determination of Hedonic Tone,
The members of a panel of assessors are asked to Düsseldorf, Germany
indicate perceived intensity at each presentation as a
value for the perceived intensity I on the seven-point The principle of measurement is presentation of the
intensity scale: odour to human assessors in an odour panel, at varying
degrees of dilution; hence varying perceived intensity
and hedonic tone.
The members of a panel of assessors are asked to

indicate perceived hedonic tone at each presentation as a
value from the nine-point hedonic tone scale:
90
+4 very pleasant between the dimensions of odour, which may help to

+3 pleasant define the potential of that odour to cause annoyance,
+2 moderately pleasant has yet to be clarified. Once a method is available it will
+1 mildly pleasant help in the definition of differentiated air quality
0 neutral odour / no odour standards for specific odours.
-1 mildly unpleasant
-2 moderately unpleasant B.6 Characterisation of odours using
-3 unpleasant chemical analysis
-4 offensive
The characterisation of odours on the basis of
For each concentration level, the mean of the values for measurement of concentrations of chemicals is only
H of all panel members is calculated, and plotted against possible in those situations where one chemical is
the odour concentration in ouE/m3. A fictitious example dominant in terms of odour perception. This is only
of the plotted result is presented in Figure 13 on this rarely the case.
page. Using a suitable curve-fitting procedure a line can
be fitted through the points obtained in the experiment.
Using this interpolation, characteristic values can be
derived from the plot, such as the odour concentration at
H = -2.
B.5 Annoyance potential
Annoyance potential is a proposed attribute to quantify

the propensity of an odour to cause annoyance within a
population when exposed to this odour intermittently,
over a long period of time. Annoyance potential is likely
to be a function of both odour quality and hedonic tone
in addition to perceived intensity (and odour
concentration). The exact nature of the interaction
Figure 13: Hedonic tone as a function of odour concentration
91
92
Annex C. Glossary
accuracy: Closeness of agreement between test result being detected under the conditions of the test.
and the accepted reference value.
detection threshold (for an environmental sample): The
(sensory) adaptation: Temporary modification of the dilution factor at which the sample has a probability of
sensitivity of a sense organ due to continued and/or 0.5 of being detected under the conditions of the test.
repeated stimulation. [ISO 5492:1992]
diffuse sources: Sources with defined dimensions
amenity: The quality of being pleasing or agreeable in (mostly surface sources) which do not have a defined
situation, prospect, disposition etc. [Websters waste air flow, such as waste dumps, lagoons, fields after
Encyclopedic Unabridged Dictionary] manure spreading, non-aerated compost piles.
anosmia: Lack of sensitivity to olfactory stimuli. [ISO dilution factor: The dilution factor is the ratio between
5492:1992] flow or volume after dilution and the flow or volume of
the odorous gas. [AFNOR X 43-104E, see bibliography,
assessor: Somebody who participates in odour testing. Appendix J]
Best Available Technique (BAT) The most effective and dynamic olfactometer: A dynamic olfactometer
advanced stage in the development of activities and their delivers a flow of mixtures of odorous and neutral gas
methods of operation which indicate the practical with known dilution factors to a common outlet.
suitability of particular techniques for providing in [AFNOR X 43-101E, modified, see bibliography,
principle the basis for emission limit values designed to Appendix J]
prevent and, where it is not practicable, generally to
reduce emissions and the impact on the environment as dynamic olfactometry: Olfactometry using a dynamic
a whole. olfactometer
• "Techniques" include both the technology used and emission factor: The emission per unit product. (e.g. for
the way in which the installation is designed, built, wastewater treatment works expressed in this report the
maintained, operated and decommissioned. emission rate in ouE/s per kg BOD, in screened sewage)
• "Available" techniques mean those developed on a European odour unit, ouE/m3 : That amount of
scale which allows implementation in the relevant odourant(s) that, when evaporated into 1 cubic metre of
industrial sector, under economically and technically neutral gas at standard conditions, elicits a physiological
viable conditions, taking into consideration the costs response from a panel (detection threshold) equivalent
and advantages, whether or not the techniques are to that elicited by one European Reference Odour Mass
used or produced inside the Member State in (EROM), evaporated in one cubic metre of neutral gas at
question, as long as they are reasonably accessible to standard conditions
the operator.
European Reference Odour Mass , EROM : The
• "Best" means most effective in achieving a high accepted reference value for the European odour unit,
general level of protection of the environment as a equal to a defined mass of a certified reference material.
whole. [IPPC directive 96/61, 1996, article 2, sub One EROM is equivalent to 123 µg n-butanol (CAS 71-
11] 36-3). Evaporated in 1 cubic metre of neutral gas this
produces a concentration of 0.040 µmol/mol.
detection threshold, (for a reference material): The
odourant concentration which has a probability of 0.5 of
93
forced choice method: For this standard the following odour: Organoleptic attribute perceptible by the
definition applies: An olfactometric method in which olfactory organ on sniffing certain volatile substances.
assessors are forced to make a choice out of two or more [ISO 5492]
air flows, one of which is the diluted sample, even if no
difference is observed. odour abatement efficiency: The reduction of the
odour concentration or the odourant flow rate due to an
fugitive sources: Elusive or difficult to identify sources abatement technique, expressed as a fraction (or
releasing undefined quantities of odourants e.g. valve percentage) of the odour concentration in the odourant
and flange leakage, passive ventilation apertures etc. flow rate of the untreated gas stream.
hedonic tone: attribute of an odour, indicating like or odour concentration: The number of European odour
dislike units in a cubic metre of gas at standard conditions.
Note: The odour concentration is not a linear measure for the intensity
individual threshold estimate, ITE : The detection of an odour. Steven’s Law describes the a-linear relation between
threshold applying to an individual estimated on the odour stimulus and its perceived intensity. When using odour
basis of one dilution series. concentrations in dispersion modelling, the issue is complicated by the
effects of the averaging time of the dispersion model, further
lower detection limit, LDL : Lowest value of the air complicating the use of the odour concentration as a direct measure for
quality characteristic which, with 95% probability, can dose. To define a ‘no nuisance level’, the entire method of dosage
be distinguished from a zero sample [ISO 6879]. evaluation, including the dispersion model, will yield a ‘dose’. The
relation between this ‘dose’ and its effect (odour annoyance) should be
neutral gas : Air or nitrogen that is treated in such a way validated in practical situations to be a useful predictive tool for
that it is as odourless as possible and that does, occurrence of odour nuisance.
according to panel members, not interfere with the
odour under investigation. odour detection: To become aware of the sensation
Safety Warning: Nitrogen is only used to predilute the resulting from adequate stimulation of the olfactory
sample itself. For the olfactometer the neutral gas used system.
to dilute the sample and present a reference shall be air.
odour panel: See panel.
objective method: Any method in which the effects of
personal opinions are minimised. [ISO 5492] odour sensitive receptor: The closest fixed building or
installation where odour annoyance may occur, such as
odourant— A substance which stimulates a human residential homes, school, hospital, overnight facility for
olfactory system so that an odour is perceived. holidays etc.
odourant flow rate: The odourant flow rate is the odour threshold: See panel threshold.
quantity of odorous substances passing through a
defined area at each time unit. It is the product of the odour unit: See European Odour Unit
odour concentration cod and the outlet velocity v and the
outlet area A or the product of the odour concentration olfactometer: Apparatus in which a sample of odorous
cod and the pertinent volume flow rate V&. Its unit is gas is diluted with neutral gas in a defined ratio and
ouE/h (or ouE/min or ouE/s, respectively.) presented to assessors.
Note: The odourant (emission) flow rate is the quantity equivalent to
the emission mass or volume flow rate, for example in dispersion olfactometry: Measurement of the response of
models. assessors to olfactory stimuli. [ISO 5492]
odorous gas: Gas that contains odourants. olfactory: Pertaining to the sense of smell. [ISO 5492]
94
olfactory receptor: Specific part of the olfactory reproducibility: Precision under reproducibility
system which responds to an odourant. [after ISO 5492] conditions. [ISO 5725-part 1]
olfactory stimulus: That which can excite an olfactory reproducibility conditions: Conditions where test
receptor. [ISO 5492, modified] results are obtained with the same method on identical
panel: A group of panel members. test material in different laboratories with different
operators using different equipment. [ISO 5725-part 1]
panel member: An assessor who is qualified to judge
samples of odorous gas, using dynamic olfactometry sensitive receptor: see odour sensitive receptor:
within the scope of this standard.
sensory fatigue: Form of adaptation in which a decrease
panel selection: Procedure to determine which in sensitivity occurs. [ISO 5492]
assessors are qualified as panel members.
to smell: To detect or to attempt to detect an odourant.
panel threshold: Detection threshold applying to a
panel. specific emission rate: The emission rate per unit of
area of liquid or solid
perception: Awareness of the effects of single or
multiple sensory stimuli. [ISO 5492] standard conditions for olfactometry: At room
temperature (293 K), normal atmospheric pressure
population (detection) threshold: Detection threshold (101.3 kPa) on a wet basis [as in ISO 10780].
applying to the general population, if this population is Note: This applies both to olfactometric measurements and volume
not specified. flow rates of emissions.
proficiency testing: The system for objectively testing static olfactometer: A static olfactometer dilutes by
laboratory results by an external agency. mixing two known volumes of gas, odorous and
odourless, respectively. The rate of dilution is calculated
quality: The totality of features and characteristics of a from the volumes. [AFNOR X 43-101E, see
product or service that bear on its ability to satisfy stated bibliography, Appendix J]
or implied needs. [ISO 6879]
step factor: The factor by which each dilution factor in
quality assurance: All those planned and systematic a dilution series differs from adjacent dilutions.
actions necessary to provide adequate confidence that a
product, process or service will satisfy given subjective method: Any method in which the personal
requirements for quality. [ISO 6879] opinions are taken into consideration. [ISO 5492]
recognition threshold: The odour concentration which substance: Species of matter of definite chemical
has a probability of 0.5 of being recognised under the composition.
conditions of the test (definition not applied in this
standard). test result: The value of a characteristic obtained by
completely carrying out a specific measurement, once.
repeatability: Precision under repeatability conditions.
[ISO 5725-part 1] volatile organic compound: organic substance that will
readily evaporate from a liquid into gas phase.
repeatability conditions: Conditions where
independent test results are obtained with the same
method on identical test material in the same laboratory
by the same operator using the same equipment within
short intervals of time. [ISO 5725-part 1]
95
96
Annex D. References
1 Van Harreveld, APh, Schakel, A., Valk, CJ, Vreeburg, S,, Feasibility study for the development of a
standardised method for assessment of odour nuisance and odour annoyance potential, Project PD 207
Action programme Standardisation and Validation of Environmental Assessment Methods (English
Summary, Project Research Amsterdam BV, Amsterdam, 1999. ISBN90-805641-1-7 Electronically
available at www.odournet.com).
2 Oberthür, R., Ein logarithmischer Maßstab mit dB-Einheiten zur Darstellung des Geruchsstoffpegels
(translation: a logarithmic measure of odour intensity, using decibel units), Staub - Reinhalt. luft 50, pp.
175--183, 1990
3 Miedema, H.M.E., Walpot, J.I., Steunenberg, C.F., Exposure-annoyance relationships for odours from
stationary sources, Atmospheric Environment 34 (2000) 2927-2936, 2000.
4 anon. (2001A), Geurhinderonderzoek Stallen Intensieve Veehouderij (English: Odour annoyance caused
by intensive livestock housing), in print, Publication number 010164/h/04-01 22317/209, Ministry of
Housing, Public Planning and the Environment, The Hague, The Netherlands, 2001.
5 Ogink, N.W.M., Klarenbeek, J.V., Evaluation of a standard sampling method for determination of odour
emission from animal housing systems and calibration of the Dutch pig odour unit into standardised odour
units, In: Proceedings of the Int. Symposium on Ammonia and Odour Control from Animal Production
Facilities, NVTL, Rosmalen, Netherlands, 1997.
6 Anzion CJM, Dragt A, Kuijk van AHJ, Post JG, Document meten en rekenen geur (Document on
measuring and calculating odours), Publikatiereeks Lucht 115, Ministry for Housing, Public Planning and
the Environment (VROM), Den Haag, Netherlands, 1994
7 VDI 3940 Determination of Odorants in Ambient Air by Field Inspections, May 1991
8 Determination and Evaluation of Odour Emissions - Directive on Odour (Feststellung und Beurteilung von
Geruchsimmissionen (Geruchsimmissions-Richtlinie) mit Begründung und Auslegungshinweisen),
Laenderausschuss fuer Immissionsschutz, LAI Schriftenreihe No.5, Berlin, 1994
9 Steinheider, B., Both, R., Winneke G., Die Erfassung der Geruchsbelästigung durch Tierstallimmissionen
bei Anwohnern (The assessment of odour impact on people living in the vicinity of livestock houses),
Gefahrstoffe-Reinhaltung der Luft 58 , 1998.
10 Ogink, N.W.M., Groot Koerkamp, P.W.G., Comparison of odour emissions from animal housing systemes
with low ammonia emission., in: Proceedings of the 1st IWA International Conference on Odour and
VOCs: Measurement, Regulation and Control Techniques, University of New South Wales, Sydney,
Australia, March 2001, ISBN 0 7 334 1769 8-
11 EN13725 ‘Odour concentration measurement by dynamic olfactometry’, CEN, 1999.
12 Misselbrook TH, Clarkson CR, Pain BF: Relationship Between Concentration and Intensity of Odors for
Pig Slurry and Broiler Houses. Journal of Agricultural Engineering Research 55: 163-169, 1993.
97
13 Dravnieks A.., T. Masurat, R.A.Lamm, Hedonics of Odors and Odor Descriptors, Journal of the Air
Pollution Control Association, Vol.34 no.7, 1984.
14 Valk, CJ, Anzion, C.J.M, van Harreveld, A.Ph., Van Belois, H, Geurhinder: de kwaliteit van de
besluitvorming moet verbeteren (Odour annoyance: the quality of decisions in environmental
management must improve), Journal: Lucht, 3, september 1998, ISSN0925-9953
15 NeR Nederlandse Emissie Richtlijn, Hindersystematiek Geur. (Netherlands Emissions Guideline.

Framework for odour nuisance), InfoMil, 1996
16 Department of the Environment, Northern Reg. Office, Town and Country Planning Act - Section 78.
Appeal by Northumbrian Water Ltd: Land adjacent to Spital Burn, Newbiggin-by-the-Sea,
Northumberland, planning reference APP/F2930/A/92 206240, UK, 1993
17 US EPA / Lee Wilson & Associates, Swine Cafo Odors: Guidance for Environmental Impact Assessment,
US EPA, 1996
18 van Harreveld, A. Ph. , Klarenbeek, J.V., On the regulations, measurement and abatement of odours
emanating from livestock housing in the Netherlands, In: New Knowledge in Livestock Odour,
Proceedings of the International Livestock Odour Conference 1995, pp.16-21, Iowa, USA
19 Van Harreveld, A.Ph., Introduction and Related Practical Aspects of Odour Regulations in the
Netherlands: in: Derenzo and Gnyp, TR18, Recent Developments and Current Practices in Odour
Regulations, Controls and Technology, AWMA, 1991, ISBN 0-923204-0
20 Inventarisatie Epos; Eindrapport van de eerste fase van het Evaluatie Project Ontwerp
Stankconcentratienormen (EPOS90) (translation: Inventory for Project Evaluation of Provisional Odour
Concentration Standards), Ministry of Public Planning and the Environment, 1990.
21 Carney, P.G., Dodd, V.A., The Measurement of Agricultural Malodours, J. Agric. Engng. Res. (1989) 43,
197-209.
22 VDI3471:1986 Emission control - Livestock Management - Pigs (Emissionsminderung - Tierhaltung -

Schweine), VDI Handbuch Reinhaltung der Luft, Band 3, Beuth Verlag, Jun 1986
23 NMP 1 Nationaal Milieubeleidsplan I,(National Environmental Plan), Ministry of Housing, Public Planning
and the Environment, Den Haag, Netherlands, 1988.
24 Beoordelingsrichtlijn in het kader van Groen Label Stallen (Guideline for the assessment of Green Label
livestock housing systems), ed. 1996, Ministry of Housing, Public Planning and the Environment and the
Ministry of Agriculture, The Hague, Netherlands, 1996.
25 Brochure Veehouderij en Hinderwet (Brochure Livestock rearing and nuisance law), Staatsuitgeverij, Den
Haag, The Netherlands, 1976
26 EPA, Guideline on air quality models (revised), july 1986, EPA, EPA-450/2-78-027R.
27 EPA, User's guide to MPTER: a Multiple Point Gaussian dispersion algorithm with optional TERrain
Adjustment, April 1980, EPA, EPA-600/8-80-016
98
28 O’Neill, D.H., Phillips, V.R., A review of the control of odour nuisance from livestock buildings: Part 3,
properties of the odorous substances which have been identified in livestock wastes or in the air around
them, J.agric.Engng.Res. (1992), 53, 23-50
29 Van Harreveld, A.P., De geuremissie tijdens en na het verspreiden van varkensmengmest (translation:
Odour emissions after land spreading of pig slurry), Research report 37, Institute for agricultural and
Environmental Engineering IMAG-DLO, Wageningen, the Netherlands, 1981.
30 Hendriks, HJM, van de Weerdhof, AM, Dutch notes on BAT for pig- and poultry intensive livestock
farming, Ministry of Housing, Public Planning and the Environment and the Ministry of Agriculture, Nature
Management and Fisheries, August 1999. Electronically available at http://www.infomil.nl/lucht/index.htm
31 Saskatchewan Pig Odour code of practice, Prairie Swine Centre Inc., Lee Whittington, (306) 373-9922.
Electronic enquiries at [email protected],
32 EPA, Guideline on air quality models (revised), July 1986, EPA, EPA-450/2-78-027R.
33 EPA, User's guide to MPTER: a Multiple Point Gaussian dispersion algorithm with optional TERrain
34 Peirson, S, Nicholson, R., MAFF Project WA0601 - Measurement of odour and ammonia emissions from
livestock buildings - Phase 1 final report. , ADAS, Cambridge, United Kingdom, March 1995.
35 Van Langenhove, H., De Bruyn, G., Development of a procedure to determine odour emissions from
animal farming for regulatory purposes in Flanders., in: Proceedings of the 1st IWA International
Conference on Odour and VOCs: Measurement, Regulation and Control Techniques, University of New
South Wales, Sydney, Australia, March 2001, ISBN 0 7 334 1769 8
36 Valentin F.H.H, North, A.A. (eds.), Odour Control - A Concise Guide, Warren Spring Laboratory, UK, 1980
37 Bongers, M.E., Van Harreveld, A.P., Jones, N., Recent developments in research supporting pig odour
policy reviews in the Netherlands and in Ireland, In: Proceedings of the 1st IWA International Conference
on Odour and VOC’s: Measurement, Regulation and Control Techniques, University of New South Wales,
Sydney, March 25-28, 2001, ed. J. Jiang, International Water Association, ISBN 0 7 334 1769 8.
99
100
Annex E. Contour plots for screening assessment
Plots with standard contours are provided as a means to

carry out a quick screening assessment. The contours are
provided at scale: 1:50.000 for the limit value for new
pig units and the target value, and at scale 1:10,560 for
the limit value for existing pig units.
To calculate these contour lines, the COMPLEX

atmospheric dispersion model was used to calculate
overlay contours. This US-EPA model, which is based
on the widely used MPTER and ISC models, has been
adapted by OdourNet to accommodate numerous
sources (up to 999 sources) and to provide percentile
values that are used to evaluate odour impacts that by
their nature have very much shorter time frames to cause Figure 14: Wind rose for meteorological station
Claremorris, hourly observations,
effects in receptors than most other common forms of air period 1993 to 1995 (inclusive).
pollution.
For a detailed description of this model the following

references are available:
[32] EPA, Guideline on air quality models (revised),

July 1986, EPA, EPA-450/2-78-027R.
[33] EPA, User's guide to MPTER: a Multiple Point

Gaussian dispersion algorithm with optional TERrain
For the standard overlay contours, the hourly

meteorological data for a Claremorris meteorological
station were used, for the years 1993 to 1995 (inclusive).
The wind rose for the data-set used is presented
graphically in Figure 14.
101
E.1 Standard contours overlay, scale 1:50,000, for the limit value for new pig production units,
Figure 15: Standard contour overlay, representing typical contours for the limit value for new pig production units of C98, 1-
hour = 3 ouE/m3 , for integrated sow units of different sizes, scale 1:50,000.
[note: Please ensure that reproduction has not distorted the image. This can be achieved by measuring the box at the bottom right of
the image and ensuring that it is 1cm x 1cm]
102
E.2 Standard contours overlay, scale 1:50,000, for the target value for all pig production units
Figure 16: Standard contour overlay, representing typical contours for the target value for all pig production units of C98, 1-
hour = 1.5 ouE/m3 , for integrated sow units of different sizes, scale 1:50,000.
103
E.3 Standard contours overlay, scale 1:10,560, for the limit value for existing pig production units.
Figure 17 Standard contour overlay, representing typical contours for the limit value for existing pig production units of C98,
1-hour = 6 ouE/m3 , for integrated sow units of different sizes, scale 1:10,560.
104
ODOUR IMPACTS AND ODOUR EMISSION
CONTROL MEASURES
FOR INTENSIVE AGRICULTURE
Part B Case studies assessing the odour emissions and impact of
two pig production units in the Irish Situation
© Environmental Protection Agency 2001
All or part of this publication may be reproduced without further permission, provided the
source is acknowledged.
Although every effort has been made to ensure the accuracy of the material contained in this
publication, complete accuracy cannot be guaranteed. Neither the Environmental Protection
Agency nor the author(s) accept any responsibility whatsoever for loss or damage occasioned
or claimed to have been occasioned, in part or in full, as a consequence of any person acting,
or refraining from acting, as a result of a matter contained in this publication.
Project information
This report was commissioned by:
Environmental Protection Agency
Johnstown Castle Estate, Wexford, Ireland
Contact: Dr. Vera Power / Dr. Tom Stafford
and was issued by:

OdourNet UK Ltd
5 St Margaret’s Street, Bradford-on-Avon, Bath BA15 1DA
Contacts: Mr. Ton van Harreveld, Mr. Matt Stoaling
Copyright notice
© OdourNet UK Ltd, 2000.
Sources of information used
The following sources of information were among those used in preparing this report:
• Scale site plans provided by farm representatives
• Information collected during site survey by Odournet consultants
• The draft report of the Dutch IMAG data
Acknowledgements
The authors wish to thank the owners and operators of farms A and B, the officers involved
and all others who provided the information on which this report is based,
for their cooperation.
Contents
1. SCOPE OF STUDY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. DETAILED DESCRIPTION OF APPROACH AND METHODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4. DISCUSSION OF THE RESULTS OF ATMOSPHERIC DISPERSION MODELLING

OF THE EXISTING SITUATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5. OPTIONS FOR REDUCING THE ODOUR IMPACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
7. ANNEX: SUMMARY OF LARGE-SCALE SURVEY TO ESTABLISH PIG ODOUR

EMISSION FACTORS IN THE NETHERLANDS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
8. ANNEX: REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
9. ANNEX: FIGURES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Executive Summary
The Environmental Protection Agency has initiated a in the Irish study, and the statistical variance as derived
study into Odour Impacts and Odour Emission Control from the larger Dutch study, the difference in the mean
Measures for Intensive Agriculture, with the objective to outcome is too small to be statistically significant.
assist the Agency in formulating its approach for Therefore, it is justified to use the emission factors
processing the license applications with a view to derived in the Netherlands for emission estimates in
achieving transparent and uniform decision-making. Ireland, as long as emission factors specifically
measured in Irish conditions are not available for a
In the course of this project two case studies were larger sample of study sites.
conducted to assess the odour impact of pig production
units in the Irish context. At these locations odour Odour impact study Farm A
emission measurements were conducted and dispersion
modelling was used to assess the odour impact. Farm A is a large integrated unit, containing over 17000
Abatement options were considered to reduce the animals. It is therefore no surprise that total emissions
impact. are high, and the odour footprint relatively large.
However, given the locality of the farm - its distance
The objectives of these case studies was: from residential units and its rural context, there seems
• To illustrate the approach as outlined the main The only concerns resulting from the site visits were the
Agency study Odour Impacts and Odour Emission uncovered slurry store and carcass skips. It is the
Control Measures for Intensive Agriculture opinion of Odournet UK that these sources may become
a significant emitter of odours during the warmth of
• To obtain a limited set of emission measurements for summer. However, having undertaken sampling during a
Irish conditions, to assess whether these are cool spring day, there is no quantitative data to support
significantly different from the distribution of results this.
found in the much larger data set from the
Netherlands[5, 10]. The modelling shows that in the current situation a
limited number of (ten) dwellings may be affected by
The case studies were not intended to provide a odour impacts in excess of the limit value. It is therefore
representative overview representative for most Irish pig necessary to seriously consider options to reduce
units. The scope of the case studies and is too limited to emissions. In the short-term measures to reduce
do so. Similarly, the scale of the emission measurements emissions from sludge storage should be considered. In
was not sufficient to yield specific emission factors for the longer term, replacement of housing assets could
Irish conditions. reduce the number of dwellings exposed to odour
impacts in excess of the limit value, conceivably to zero.
The case studies prompted the following conclusions: The target value will be difficult to attain for farm A. The
farm can be made sustainable at current stock levels,
Conclusions on the results of finisher emission rate from the perspective of odour impact, provided that the
measurements in Ireland community recognises and accepts the rural context in
the vicinity.
The emission rate of 13.2 ouE/s per finisher measured in
Ireland for this study in winter conditions is about one Odour impact study Farm B
third lower than the annual mean value of 22.6 ouE/s per
finisher found in a larger study in the Netherlands Farm B is a relatively small-scale operation. Under
current circumstances, four properties fall within the
Given the relatively small number of samples, collected odour footprint and therefore may be affected by odour
1
impacts in excess of the limit value. This farm would

require to consider the implementation of abatement
options if current stock numbers are maintained.
In the long term, a reduction of odour emissions should

be considered when normal renewal of pig housing
assets becomes an issue. This could achieve attainment
of the limit value for all dwellings in the vicinity, and
possibly create room for some growth of stock when the
agricultural context of the area is recognised and
accepted as the status quo.
2
1. Scope of study
1.1 Scope would be preferable to have emission factors specifically

obtained for Irish conditions. However, if reliable
The Environmental Protection Agency has initiated a specific emission factors for Irish production practice
study into Odour Impacts and Odour Emission Control are required, a much larger programme of odour
Measures for Intensive Agriculture, with the objective to emission measurements is required.
assist the Agency in formulating its approach for
processing the license applications with a view to The case studies were conducted at three sites, and the
achieving transparent and uniform decision-making. characteristics summarised below:
In the course of this project two case studies were • Case A: a large integrated pig unit, with
conducted to assess the odour impact of pig production approximately 1000 sows in fully slatted pig houses
units in the Irish context. At these locations odour
emission measurements were conducted and dispersion • Case B: a medium sized integrated pig unit, with
modelling was used to assess the odour impact. approximately 590 sows in fully slatted pig houses
Abatement options were considered to reduce the
impact. The case study reports were prepared on the basis of
information provided by the client, the Environmental
The objectives of these case studies was: Protection Agency and observations made during a site
visit by the authors to the two farms in question. Odour
• To illustrate the approach as outlined the main samples were taken at representative locations and
Agency study Odour Impacts and Odour Emission analysed at the Odournet odour laboratory in Bradford-
Control Measures for Intensive Agriculture (Part A) on-Avon.
• To obtain a limited set of emission measurements for 1.1.1 Study objectives

Irish conditions, to assess whether these are
significantly different from the distribution of results The study aimed to achieve the following objectives:
found in the much larger data set from the
Netherlands[5, 10]. 1 Assess the local situation during two site visits;
The case studies were not intended to provide a 2 Determine appropriate sampling locations and
representative overview representative for most Irish pig numbers;
units.
3 Collect samples and analyse in the Odournet UK
Similarly, the scale of the emission measurements was laboratory.
not sufficient to yield specific emission factors for Irish
conditions. The aim was limited to assessing whether the 4 Compare these the measured emissions in this limited
results found in this limited study can be considered to sample relative to the distribution of results of the
fall within the distribution of the data found in the much Dutch data. These have been used as the basis of
larger study in the Netherlands. emission estimates in the general background study.
The operational practice in pig production in Ireland is 5 Estimate the potential impact of the pig production
different in a number of aspects from practices in the units on any nearby residential properties;
Netherlands, as are the environmental conditions, see
also section 7. These differences are not likely to cause 6 Identify financially viable options for reducing the
significantly different odour emissions. Of course it odour impact of the production units.
3
4
2. Detailed description of approach and methodology
2.1 Approach regular intervals during the collection of each sample.

The velocity within each extract duct from the room
This study can be regarded as having two distinct aims: being sampled was measured using a hot wire
anemometer. A number of the ducts were fitted with
1. Limited measurement of emission rates to assess large manually adjustable dampers, these resulted in
whether results in Ireland lie within the range of data uneven airflow through the ducts and made precise
from a larger study in the Netherlands measurement difficult to achieve.
2. Detailed odour emission estimate and odour • Other Parameters - Ambient temperature and wind
dispersion model study for each of the two farms speed/direction were recorded at each site.
The results of these distinct parts of the study will be 2.2.2 Odour sampling and analysis
discussed in separate sections in the Results section
Odour concentration was determined according to the
2.2 Limited measurement of emission rates EN13725 Odour concentration measurement by
for pigs dynamic olfactometry[11] and results expressed in odour
units (ouE·m-3).
2.2.1 Objectives and approach to sampling
2.3 Detailed odour impact assessment for
The purpose of the sampling was principally to collect the two farms
data for comparison with odour emissions rates
measured at similar pig units in the Netherlands. As the Initially a model of the sources was constructed on a
contribution of fatteners to the total odour emissions is geographic information system (GIS), using:
the largest, during the rearing process, odour sampling
was principally concentrated within fattening houses. • Site plans,
In addition a limited number of samples was collected to • Waypoints logged onto a portable Global Positioning
determine the emission rate from external uncovered System (GPS)
sludge storage facilities.
• Digitised Ordnance Survey 1:50,000 maps (provided
The sampling took the following relevant factors into by the Irish EPA)
account:
Then the emissions for each building on the farms were
• Sample timing - One day of sampling was carried estimated on the basis of the number of animals and
out at each of the two farms. At each farm samples their size and age, using emission factors that were
were taken from several groups of fatteners. The established on the basis of extensive measurements in
groups varied from one another in age and the Netherlands[5, 10].
consequently average live weight of each group.
Using the estimated emission rates, a source
• Time of the day - Due to the necessity to collect all characterisation model was constructed and used as
number of samples within one day, and transport the input for an atmospheric dispersion model. Combined
samples for analysis the following day, sampling was with additional input data, such as hourly
carried out between 07.30am and 3.00pm. meteorological data for a nearby station, and digitised
terrain data (topography), the model was constructed
• Ventilation - The ventilation rate was measured at and applied to establish a first estimate of the
5
environmental impact with regard to odorous emissions. elevation of the sources considered.
As in the other studies, meteorological data were used
for the station located at Claremorris, covering the COMPLEX-1 is a modification of the MPTER model
period 1993 to 1995 (inclusive). that incorporates the plume impaction algorithm of the
VALLEY model. As a result COMPLEX-1 can also
The COMPLEX-1 atmospheric dispersion model was calculate ambient concentrations in situations where
used, as described in the next section. The model MPTER is not applicable, i.e. at receptors situated at a
produces a probability of exceeding a certain hourly level above stack height.
average concentration, at locations where people may be
exposed. The COMPLEX-1 model calculates and summarises the
individual contribution to the overall emission pattern of
When an annoyance criterion is applied to these results, each source, for each hour, at each point in the receptor
a contour can be plotted which contains the area in grid. This summarised data can then be plotted as
which odour annoyance may occur. contours of equal odour exposure on a topographic map.
A discussion on how the results of odour dispersion 2.3.2 Methodology: Construction of the model of
modelling should be interpreted is included in section emissions and atmospheric dispersion
2.3.3 below.
The following scenarios are modelled within this study
2.3.1 The dispersion model used: Complex-I for Farm A and Farm B:
COMPLEX-1 is an air pollution dispersion model, • Scenario 1: Represents the actual existing situation at
which can be used for estimating air pollutant the farms, with mechanical ventilation. Modelling
concentrations in complex terrain. It was originally will encompass ventilation rates based upon those
developed in the United States by the Environmental measured on the day of survey - which will not be
Protection Agency (EPA). equivalent to those under different climatic
conditions. Emissions will be through the roof vents
Complex terrain models are normally applied to at a height of approximately 4m.
stationary sources of gaseous pollutants (like SO2) and
particulates. The model is based on the so-called • Scenario 2: The situation will be as scenario 1, but
Gaussian plume model and it can be used to calculate assuming an abatement of 50% (utilising best
one-hour to twenty-four-hour averaged concentrations at practice techniques).
specific receptor points in a complex topography. The
receptor points may be situated below as well as above The COMPLEX-1 dispersion model requires 3 distinct
stack height. types of input data to run:
COMPLEX-1 is based on the MPTER model[27] and • Data concerning the location, physical dimensions,
incorporates the plume impaction algorithm of frequency of activity and odour emission of the
VALLEY. With COMPLEX-1 computations can be source.
made for up to 250 point sources and 180 receptors
located within 50 kilometres of the sources[26]. • Local meteorological data.
MPTER is most applicable for source-receptor distances • Local topographic data.

less then 10 kilometres. It can be used in rural as well as
in urban areas, with flat or rolling terrain where a single 2.3.2.1 Odour source data
wind vector for each hour adequately approximates the
flow over the source and receptor sites. The Terrain Odournet UK have identified and included all relevant
adjustment option of MPTER is limited to receptors sources in the odour emission model. Emission rates for
whose elevation is no higher than the lowest stack top each building were estimated on the basis of the number
6
of animals housed and their life stage, using emission concentration of 1.5 ouE·m-3 in 98% of all hours in an
factors as proposed in the main EPA odour impact study average meteorological year.
document, that were derived from a large scale Dutch
study[5, 10], summarised in an Annex, section 7 of this • Limit value for new pig production units: C98, 1-
report. These data were used to construct a model of the -3
hour < 3.0 ouE·m
source emissions providing the input data for odour The limit value for new pig production units provides
dispersion modelling. a minimum level of protection against odour
annoyance, aiming to limit the percentage of those
2.3.2.2 Meteorological data experiencing some form of annoyance to 10% or less
in a the general public, assuming some degree of
The COMPLEX-1 model requires hourly averaged acceptance of the rural nature of their living
values for wind speed, wind direction and height of the environment.
mixing layer. The mixing layer height enables the
atmospheric stability to be classified into one of six The limit value for new pig production units shall not
`Pasquill' categories, ranging from very stable to very be exceeded in the vicinity of new pig production
unstable. Unstable conditions are the most favourable units to ensure a minimum environmental quality.
for dispersion of odours in the atmosphere.
The limit value for new pig production units is
Hourly meteorological data for the years 1993 to 1995 complied with when for all locations of odour
(inclusive) recorded at Claremorris were used for the sensitive receptors the calculated odour exposure is
modelling exercise. less than an hourly average odour concentration of
3.0 ouE·m-3 in 98% of all hours in an average
2.3.2.3 Topographic data meteorological year.
We often refer to this as receptor data as it describes the • Limit value for existing pig production units: C98,
location of receptors potentially exposed to odours. It -3
1-hour < 6.0 ouE·m
consists of an X,Y,Z grid with an additional parameter H
referring to the height of the receptor above ground (i.e. The limit value for new pig production units provides
nose height). a minimum level of protection against odour
annoyance, aiming to limit the percentage of people
2.3.3 Interpretation of odour modelling output data experiencing some form of annoyance to 10% or less,
in the most tolerant selection of the population.
The results of the modelling are presented in the form of
contours or isopleths (lines connecting equal frequency The limit value for existing pig production units shall
of occurrence) for: 1.5, 3 and 6 ouE/m3 as a 98 not be exceeded in the vicinity of existing pig
percentile. These contours describe the area associated production units to ensure the minimum
with the following proposed air quality criteria: environmental quality in an agricultural setting. A
phased plan must be made to reduce the odour
• Target value: C98, 1-hour < 1.5 ouE·m-3 impact, with time, to the target value.
The target value provides a general level of protection
against odour annoyance for the general public, The limit value for new existing production units is
aiming to limit the percentage of people experiencing complied with when for all locations of odour
some form of annoyance to 10% or less. The target sensitive receptors the calculated odour exposure is
value shall be used as an environmental quality target less than an hourly average odour concentration of
for all situations. 6.0 ouE·m-3 in 98% of all hours in an average
meteorological year.
The target value is achieved when the calculated
odour exposure for all locations of odour sensitive These criteria are proposed in the main study prepared
receptors is less than an hourly average odour for the EPA, and are underpinned by actual dose effect
7
relationships established on the basis of the percentage

of residents annoyed, as determined by telephone
survey[3].
The modelling results cannot be assumed to be an exact

mirror of reality. The model will be less effective in
predicting the actual exposure in a single hour,
especially at distances to the source of less than 100 m.
The models are more effective in predicting the
probability of exposure levels over a large number of
hours, such as the four-year data set of hourly
observations used for this study.
The contours represent the area where the maximum

hourly average ground level concentration will be
greater than 1.5, 3 or 6 ouE·m3 for more than 2 % of the
hours in the year.
The predicted ground level concentrations are above

background concentrations and only relate to odour
originating from the works.
2.3.4 Model of source: odour emission estimation
2.3.4.1 Assumptions used in determining odour

emission rate
The following assumptions have been used to model the

emissions for all scenarios.
• The odour concentration is homogeneous within the

building.
• The numbers of pigs as provided by the farm

representatives are accurate, and the animals are
evenly distributed amongst the buildings as
appropriate.
• The situations found during the site survey are

representative of normal operation
• The movement of animals has not been modelled. For

instance, transfer between buildings and delivery /
dispatch.
8
3. Results
3.1 Farm A Table 1: Estimated numbers of animals for Farm A
Life stage Number of animals

3.1.1 Description of the study area: Farm A
Weaners 6210
Farrowers 90
Farm A is a large integrated unit with over 1000
Dry sows 1072
breeding sows. The unit is not fully integrated, as a
Fattening pigs 10408
proportion of weaners and fatteners are moved of site for
Gilts 172
finishing elsewhere. The site manager provided a tour of
the facilities. The site has eight separate fattening houses
of traditional slatted floor type, with mechanical normally necessary to agitate or stir any of these tanks
ventilation through roof fans. Each house is divided into prior to removal of slurry. During the visit the open
three rooms, which are in turn subdivided into 12 pens lagoons were mostly full.
containing an average 18 pigs each. In each room three
fans draw air from one wall, fresh air is introduced via a 3.1.2 Site visit observations Farm A
perforated ceiling. All animals are fed liquid feed via a
central distribution system. A site visit was carried out on Monday 17th April 2000
by Matthew Houseman (Odournet UK Ltd.) and
Sows, gilts and farrowers are contained within separate Margarethe Bongers (PRA bv). The sampling program
slatted floor buildings. A number of these buildings have was undertaken on the 18th of April 2000 by Matthew
slurry tanks beneath the floors, whilst slurry from the Houseman and Matthew Stoaling (Odournet UK Ltd.).
dry sow house and gilt houses is stored in open top, During the visit to Farm A on the 17th April odour was
concrete lined lagoons. All these buildings were faintly detectable for a distance of approximately 50
ventilated by centrally mounted roof fans. metres across fields at that time downwind of the site.
The operation was run to a high standard, and no clear Sampling was carried out at Farm A on the 18th April on
operational flaws were observed. There is no history of a clear and bright day, with temperatures averaging
registered odour complaints. 15ºC. There was a gentle breeze from the north west (up
to 2.5 m/s). Operations at the site were normal with a
Slurries are tankered off site for land spreading. Tankers tanker intermittently removing slurry from the tank
draw slurry directly from either the above ground beneath one of the fattening houses.
lagoons or the under floor storage tanks. It is not
Image 1 - A fattening house at Farm A - the primary

unit type at this facility
9
3.2 Farm B Table 2: Estimated numbers of animals for Farm B
Life stage Number of animals

3.2.1 Description of Farm B
Weaners 2964
Suckling sows with piglets 114
Farm B is an integrated production unit with
Dry sows 479
approximately 590 breeding sows. All production takes
Fattening pigs 2192
place within one site.
Gilts 101
Boars 18
Weaners are raised in groups of 160 in slatted floor
rooms with a single centrally mounted ventilator in each
room. Fattening then takes place in rooms of twelve was undertaken on the 19th of April 2000 by Matthew
pens, with approximately 250 pigs per room. The Houseman and Matthew Stoaling (Odournet UK Ltd.).
fattening rooms are also mechanically ventilated with No odour was detectable off site from Farm B on the
two centrally mounted fans in each room. Final finishers 17th April.
of approximate 90 kg live weight are kept in smaller
groups of approximately 36. These groups are in smaller Sampling was carried out at Farm B on the 19th April as
stalls with sidewall mounted mechanical ventilation. All heavy rain crossed Southern Ireland, with outside
pigs at Farm B are fed with a dry blended feed. temperatures falling to 8˚C at the end of the sampling.
There was a strong wind from the south west (up to 6.5
The operation was run in a professional manner, and the m/s).
site was kept clean and tidy. There is no history of
registered odour complaints. Operations at the site were normal with no sludge
removal operations taking place.
Slurries are usually stored in underground tanks beneath
the slatted floors. Tankers take slurry from land 3.3 Odour emission measurements
application directly from these tanks. An additional
circular, open topped storage tank is available. However In order to simplify the practice of producing a detailed
due to the low density of pig units in the vicinity of Farm odour study for farms in Ireland, the use of emission
B finding land for spreading is not usually difficult and estimates based on emission factors per animal has been
hence the tank is rarely in use. proposed in this document. The most recent and detailed
study regarding pig emission factors was identified, to
3.2.2 Site visit observations Farm B form the basis for these emission factors. The emission
A site visit was carried out on Monday 17th April, 2000 estimates were derived from data collected in a study
by Matthew Houseman (Odournet UK Ltd.) and commissioned by the Ministry of Public Planning and
Margarethe Bongers (PRA bv). The sampling program the Environment in the Netherlands, and carried out by
Image 2 - The fattening houses at Farm B. The centrally mounted

extraction vents can be clearly observed.
10
the Institute for Agricultural Engineering IMAG-DLO in Three samples were taken at each point. The surface of
Wageningen[5]. the lagoon was sampled twice, and one sample of
ambient air was collected. Samples in the pig units were
A detailed account of the approach and methodologies typically taken within the central passageway of each
used in the IMAG-DLO study is available in annex 7 of pig house at a height of 1.5 metres above the slatted
this report. In general, the methods employed when floor. Each sample took approximately 30 minutes to
sampling from the two Irish farms were as close to those collect. Samples were collected between 08.00 and
used in the Dutch project as possible. However, due the approximately 16.00 hours.
variable nature of the construction of pig housing, and
conditions, some assumptions were made. These are No sampling took place during the feeding of pigs.
also outlined in annex 7. During sampling the air flow velocity of each fan within
the particular room was measured. At least five
It was decided to conduct the measurements in Ireland measurements were taken per fan. During the fan
for one housing type only, so the statistical uncertainty measurements the percentage of fan capacity employed
could be limited within the given scope of the sampling was recorded from the display on the ventilation control
programme. As the primary sources of odour at an units.
integrated unit are the fattening houses, all samples were
taken within these buildings. The results of the 3.3.1.2 Sampling at site B
measurements in Ireland were compared with the Dutch
data for all housing types. In order for the Irish samples Sampling was carried out at Site B on the 19th April as
to be comparable to those collected in the Dutch study, heavy rain crossed the area, with outside temperatures
the protocols used in the Dutch study were followed falling to 8˚C at the end of the sampling. There was a
wherever practicable. strong wind from the south west (up to 6.5 m/s).
3.3.1 Description of sampled pig units and Operations at the site were normal with no sludge
conditions during sampling removal operations taking place.
All fatteners sampled within the two Irish case study Odour sampling was principally concentrated within
farms were housed within slatted floor buildings fattening houses. Three samples were taken at each
equipped with mechanical ventilation. The number of point. Samples were typically taken within the central
pigs per room varied from 36 to 260. passageway of each pig house at a height of 1.5 metres
above the slatted floor. Each sample took approximately
• Floor area and animal places - Floor area per pig 30 minutes to collect. Sampling took place between
varied depending upon animal weight, but was in line 07.30 and approximately 15.50 hours.
with Irish Animal Welfare Regulations (Note: these
are currently only applicable to new pig units). No sampling took place during the feeding of pigs.
During sampling the air flow velocity of each fan within
• Feed - Pigs were fed a liquid feed at Farm A and a dry the particular room was measured. At least five
feed at Farm B. measurements were taken per fan. During the fan
measurements the % of fan capacity employed was
3.3.1.1 Sampling at Farm A recorded from the display on the ventilation control
units.
Sampling was carried out at Farm A on the 18th April on
a clear and bright day, with temperatures averaging 3.3.1.3 Number of samples
15˚C. There was a gentle breeze from the north west (up
to 2.5 m/s). Operations at the site were normal with a Twelve samples were taken from fattening pigs housing
tanker intermittently removing slurry from the tank at each farm. This represented triplicate samples from
beneath one of the fattening houses. four groups of fatteners at each farm.
11
In addition an ambient sample and two samples from the the basis of triplicate samples, with a value of 2.8 ouE/s
surface of a slurry lagoon were taken at Farm A and per animal place.
three samples from a weaner house at Farm B. Therefore
of the 31 samples taken, 24 were directly aimed at the In addition to the measurements in the pig houses, a
comparison with emission rates as determined in the sample was collected of ambient air on the lee side of
Netherlands. Farm A. The odour concentration was below the
detection limit of 40 ouE/m3 as expected.
3.3.2 Results of emission rate measurements
Two samples were collected using a Lindvall hood for
The results of the measurements of odour emissions, sampling area sources. The flow velocity under the hood
carried out on site A and site B, are presented in Table 3. was 0.39 m/s. The specific emission rate calculated on
The emission factors per animal (in ouE/s per animal the basis of the geometric mean of the results of 2813
place) were calculated by multiplying odour ouE/m3 is 31 ouE/m2/s.
concentration (ouE/m3) with ventilation rate (m3/s) per
animal place, to obtain an emission rate in ouE/s per 3.3.3 Discussion of emission measurement results
animal.
The median emission rate measured in this limited study
For each sampling location, the geometric mean of the in Ireland, of 13.2 ouE/s, is approximately one third
replicate samples was calculated. lower than the value of 22.6 ouE/s per fattening pig that
was the result of the study in the Netherlands[5,10] (see
The overall emission for a finisher was calculated as the summary of that research in annex 7).
median value of all measured emission factors for
finishers of all weights, giving an emission rate of 15 The difference is considerable, but not sufficiently large
ouE/s per animal place at an average ventilation rate of to conclude that the observations obtained in Ireland are
50 m3/h per animal. The geometric mean was 13.2 ouE/s significantly different. The observed difference would
per animal fall within the statistical variability between farms and
within farms that was found in the study in the
For weaners, only one emission rate was obtained, on Netherlands[5].
Table 3: Emission rates for finishers and weaners, calculated from measured odour concentration and ventilation
rates per finisher measured at site A and B, Ireland, April 2000.
Stage Odour concentration ventilation No of fans % of Odour No. of Emission
rate per room max. fan emission rate per
capacity animals rate animal
sample a sample b sample c Geomean

ouE/m3 m3/s ouE/s ouE/s
Site A
Fatteners 35 - 40 kg 1499 1914 262 909 3.9 3 45% 3546 216 16.4
Fatteners 70 kg 1483 1394 679 1120 3.9 3 44% 4367 216 20.2
Fatteners 90 kg 764 n/a 379 538 3.0 3 37% 1587 216 7.3
Fatteners 50 kg 1073 n/a 1345 1201 2.8 3 34% 3364 216 15.6
Site B
Fatteners 70 kg 961 1450 1508 1281 3.0 2 21% 3868 262 14.8
Fatteners 40 kg 2243 1641 1505 1769 2.7 2 18% 4777 256 18.7
Fatteners 60 kg 862 908 731 830 3.2 2 24% 2640 239 11.0
Fatteners 90 kg 403 289 1498 559 0.5 1 n/a 279 36 7.8
Weaners 40 kg 779 549 400 555 0.8 1 43% 455 164 2.8
12
Table 4: Results of odour analysis of measurements of ambient air and of samples collected from
a Lindvall hood over sludge storage at Farm A.
Analysis Sample ID Date of Odour Panel Valid Remarks
datafile analysis concentration size ITE’s
[ouE/m3]
00041913 Ambient air 19/04/2000 <40 -- -- Below lower detection limit,
not valid under CEN prEN 13725
00041915 Slurry Lagoon 19/04/2000 3465 5 8 Geomean = 2813 ouE/m3
00041916 Slurry Lagoon 19/04/2000 2283 5 10
In the Dutch study, a total of 80 samples were collected The measured specific emission rate of 31 ouE/m2/s is
over a longer period of time. The geometric mean of relatively low for a sludge basin. Sludge emission rates
odour emission from the fattening pig system amounted can vary considerably, depending on the conditions that
to 22.6 ouE/s per finisher place at a mean ventilation rate determine the microbial degradation of the sludge, such
of 33.3 m3/h per animal[5,10]. The estimated variance as oxygen availability and temperature.
components in that study, expressed as variation
coefficients (or the standard deviation as a percentage of OdourNet would expect the summer emissions from the
the mean value), for fattening pigs were: sludge to be an order of magnitude higher.
• Between farm variation: 22% Emission rates are significantly increased when the
sludge is turbulent, through handling or stirring.
• Within farm variation: 34% Turbulence can cause emissions to increase by an order
of magnitude again compared to a still surface.
• Variation between duplicate samples: 16%
3.4 Odour emission scenarios
The within farm variance is the largest component and is
about two times larger than the between farm variance. 3.4.1 Odour emission scenarios Farm A
Sampling in Ireland was carried out on two days, under The emissions from Farms A were estimated using the
weather conditions that cannot be described as summer data resulting from the Dutch larger scale study[5,10], see
conditions. The ventilation per animal in Ireland was Table 5 overleaf. These estimations were used as the
relatively high, compared to the ventilation rate basis for the atmospheric dispersion modelling
observed in the Netherlands. Both factors will tend to scenarios.
lead to lower emission rates.
3.4.2 Odour emission scenarios Farm B
The odour concentration measured in the ventilation air
is higher than the values measured in Ireland in 1989 by The emissions from Farm B were estimated using
Carney and Dodd[21], which ranged between 20 and 40 emission factors as proposed in the larger EPA study on
dilutions to threshold. However, it must be noted that the the basis of the data from a larger scale Dutch study[5, 10],
olfactometry utilized in that research did not comply see Table 6 overleaf. These estimates were used as the
with any of the national or international standards that basis for the atmospheric dispersion modelling
have been introduced since that time. Increases of an scenarios.
order of magnitude in the value of odour thresholds
measured by laboratories since the introduction of odour
units traceable to reference odours have been reported in
other instances[18].
13
Table 5: Estimated numbers of animals and their odour emissions from Farm A
Estimated odour emission rate from Farm A

Number of Emission rate in Emission rate, Emission rate Percent of total
animals ouE/animal/second [ouE/s] [106 ouE/hour]
Weaners 6,210
Farrowers 90
Dry Sows 1,072
Gilts 172
Subtotal excluding fatteners 62,649 226 21%
Fattening pigs 10,408 22.6 235,221 847 79%
Total 17,952 297,870 1,073 100%
Table 6: Estimated numbers of animals and their odour emissions from Farm B
Estimated odour emission rate from Farm B

Number of Emission rate in Emission rate, Emission rate Percent of total
animals ouE/animal/second [ouE/s] [106 ouE/hour]
Weaners 2,964
Suckling sows with piglets 114
Dry Sows 497
Gilts 101
Boars 18
Subtotal excluding fatteners 31,676 114 39%
Fattening pigs 2,192 22.6 49,539 178 61%
Total 5,886 81,215 292 100%
14
4. Discussion of the results of Atmospheric

Dispersion modelling of the existing situation
Note: For guidance on interpretation of the odour The area in which the impact is such that the target value
contours, refer to section 2.3.3. of C98, 1-hour = 1.5 ouE/m3 is exceeded does include a
significant number of approximately 100 dwellings,
The results of the model run for the two scenarios are even for Scenario 1.
discussed below. The contour lines indicate the area in
which the limit value values of target value C98, 1-hour = The overall implication is that for Farm A reductions are
1.5 ouE/m3, the limit value for new pig units in feasible that would enable continuation of operations at
Greenfield sites C98, 1-hour = 3ouE/m3 and the limit value the current site, when exposure in excess of the target
for existing pig units of value is deemed acceptable. There is, however, a need to
C98, 1-hour = 6 ouE/m3 will be exceeded. reduce emissions to bring exposure within the limit
value. The options for an increase of production levels at
• In reality, the case studies all concerned existing pig the site of farm A are not easily attained, without more
units, so that the limit value should be applied. The drastic odour control measures that are not generally
target value should be regarded as a long-term goal. considered financially viable at current market
conditions.
The C98, 1-hour indicates that the limit concentration is
exceeded during 2% of the 8760 hours in an ‘average 4.1.2 The situation at Farm B
year’. This equals approximately 175 hours.
The results of the existing situation at Farm B, modelled
4.1.1 The situation at Farm A in Scenario 1, are presented in a figure, annex 9.3. The
contour for the limit value C98, 1-hour = 6 ouE/m3 extends
The results of the model run for the existing situation, for relatively modest distances, of less than 500 m from
Scenario 1, are presented in a figure, annex 9.1. The the farm buildings. OdourNet does not have access to a
contour for the limit value C98, 1-hour = 6 ouE/m3 extends detailed survey of the dwellings in the vicinity, nor their
for considerable distances, of over 1 km from the farm use. From the map it would appear, however, that four
buildings. Given the location of the farm, the number of dwellings are located within the limit value contour.
dwellings affected is limited. OdourNet does not have
access to a detailed survey of the dwellings in the The area in which the impact is such that the target value
vicinity, nor their use. From the map it would appear that of C98, 1-hour = 1.5 ouE/m3 is exceeded does include a
ten dwellings are within the limit value contour. These significant number of dwellings (nine).
may be farmhouses, in which case it can be argued that
these dwellings can expect a more lenient level of Although in the short term no odour emission reduction
protection, in the absence of complaints. To achieve a would be required, a reduction should be the aim when
more acceptable situation, however, the emissions from replacing assets, by using specifically low-emission
Farm A should be reduced in the long term. housing systems. These can achieve a 50% reduction of
emissions. The situation that would arise as a result of
When a 50% reduction of emissions is achieved, the such a reduction is shown on the map in Annex 9.4. In
odour impact of Farm A is significantly reduced. The this situation, the contour representing the target value
number of dwellings indicated on the map that lie within (C98, 1-hour = 1.5 ouE/m3) is reduced to such a degree that
the limit value contour is reduced to two, see the map in four dwellings are affected
Annex 9.2. These dwellings are quite close to the limit
value of the contour. It is likely that optimisation of The overall implications for Farm B are limited, in that
ventilation, in combination with a 50% odour emissions no immediate actions are required. The odour impact in
reduction, would reduce the odour footprint so far as to the current situation is reasonably acceptable, given its
not include any dwellings. rural context. To adhere to the proposed limit value
15
limited abatement options can be considered, e.g.

optimising the release of ventilation air by increasing
height of release and vertical velocity. Continued
attention to avoiding incidental releases caused by
actions such as slurry loading is essential, especially
when weather conditions are unfavourable, given the
short distance to the nearest dwellings. In the long term
low-emission housing is an option to be considered
when replacing current housing assets, with a view to
attaining the target value and, potentially, creating
options for growth of stock numbers in this location.
16
5. Options for reducing the odour impact
The two pig production units surveyed in this study are carcass skip was not housed in an appropriate building
similar in their design, and abatement options discussed or covered (see image 3, below). A simple and cheap
here will apply to both, generally speaking. cover, and regular emptying, would result in emissions
from this source being dramatically reduced.
There is one difference in that Farm A has a sludge
storage and transfer unit, with a rectangular open buffer The risk of causing annoyance may be reduced by
tank and a storage tank. This source will be discussed carefully planning actions on the farms and related to
separately. land spreading. Odournet UK recommends:
To reduce the odour impact from the pig production at • Suspend sludge transfer operations in unfavourable
current capacity levels there are fundamentally three conditions (low winds in summer conditions in the
options: direction of the closest residences)
1. Reduce the emissions by modification of production • Avoid land spreading in the vicinity to reduce odour
method and type of pig housing units impact associated with the farm operations,
especially in dry weather or summer conditions
2. Reduce the emissions by collection of ventilation air
followed by treatment in an odour abatement unit 5.2 Modification of production method
3. Reduce emissions by good operational practice - A modification that could be implemented in the short
especially storage of slurries at Farm A term at Farm A, is an overhaul of slurry storage practice.
Images 4 and 5 show the current situation. There are a
4. Reduce emissions by reducing production capacity number of options available for sludge storage that
reduces emissions.
5.1 Operational practice
Modifications may include:
Although farms are already aware of operational
requirements for reducing odour emissions through • Collection of slurry in closed tanks, followed by
appropriate cleaning, covering of skips, etc., there were anaerobic digestion.
still many adaptations that could be made to the
operational practice that would have a significant effect In this process the odourants that are produced can be
on the odour emitted. For instance, at Farm A, the destroyed by controlled incineration of the biogas.
Image 3 - An uncovered carcass skip is a source with a high potential to cause

annoyance in summer conditions.
17
The digested slurry is significantly less odorous,

which is a great benefit when spreading
• Covering the surface

i. Natural crusting
ii. Floating biological covers (straw, fibre)
iii. Floating covers (rigid or custom designed foil
covers)
iv. Liquid additives (vegetable oils)
Image 4 - The slurry is open to the ambient air

These methods are covered in detail in the associated and is frequently disturbed
report ‘Part A Odour annoyance assessment and criteria
for intensive livestock production in Ireland’
A longer-term solution to reduce emissions would be

modification of the production method. Replacing the
current fully slatted system with low-emission pig
housing, such as the so-called ‘green label’ buildings
that haven been introduced in the Netherlands. Using
these housing systems can potentially reduce emissions
to 50% of the current levels[24, 30]. Such a replacement
would be feasible only when planned within the normal
cycle of asset replacement. Image 5 - The main slurry storage tank covers an
area of over 1000m2
5.3 Collection of ventilation air and odour

control treatment
To be able to apply odour abatement techniques to the

current installations, the ventilation air that is now
released via roof ventilators would need to be collected
and ducted to a central point for treatment in an odour
control unit for each building. The scale of each farm
would deem the ducting of all air to a central unit as
financially inhibitive.
Image 6 - Roof ventilation at Farm A

Due to the fluctuating nature of the ventilation rates, as
determined by the temperature in the houses, each
exhaust would be required to retain regulating
mechanism. Ducting would require a significant capital
investment. However this would be less than the
installation of multiple abatement units.
Once the air can be ducted to a central point on each

building for treatment there are a number of options for
reducing the odour concentration in the exhaust air.
These options include:

Image 7 - Roof ventilation at Farm B
1) Release of all extracted air through an elevated stack.
18
This can achieve significant reduction of ground level

concentrations through enhanced atmospheric
dilution
2) Odour abatement through treatment by:

a) Chemical scrubbing
b) Biological scrubbing
c) Biofiltration, potentially combined with pre-
treatment for H2S using a catalytic iron filter
d) Biofiltration on fixed medium substrate covered
with a biofilm, such as a lava-rock filter.
These treatment techniques are described in detail in the

associated reports. These techniques all have the
potential to reduce the odour concentration in the
ventilation air significantly, with a reported efficiency of
up to 90 to 95%.
After treatment the treated ventilation air could be

released from a raised stack, to further reduce the odour
impact by achieving atmospheric dilution. Increasing
the velocity of the exit air will also increase this effect.
This approach is a costly but effective solution. The
quantities of ventilation air are considerable, and
determined by the ventilation needs of the animals.
Assuming a ventilation rate of 0.5 m3/kg pig/hour the
total estimated ventilation flow rate for Farm A would be
524,000 m3/hour.
Treatment options are not considered in detail in this

report, as the associated cost is prohibitive under current
market conditions.
19
20
6. Conclusions
6.1.1 Conclusions on the results of finisher farm can be made sustainable at current stock levels,
emission rate measurements in Ireland from the perspective of odour impact, provided that the
community recognises and accepts the rural context in
The geometric mean emission rate of 13.2 ouE/s per the vicinity.
finisher measured in Ireland for this study is about one
third lower than the value of 22.6 ouE/s per finisher 6.1.3 Odour impact study Farm B
found in a larger study in the Netherlands.
Farm B is a relatively small-scale operation. Under
Given the relatively small number of samples, collected current circumstances, the odour footprint includes four
in the Irish study, and the statistical variance as derived domestic dwellings within the limit value contour. This
from the larger Dutch study, the difference in the mean farm may require the installation of abatement options.
outcome is too small to be statistically significant. However, this would not be critical if current stock
numbers and good practice are maintained. Continued
Therefore, it is justified to use the emission factors attention to avoiding incidental releases caused by
derived in the Netherlands for emission estimates in actions such as slurry loading is essential, especially
Ireland, as long as emission factors specifically when weather conditions are unfavourable, given the
measured in Irish conditions are not available for a short distance to the nearest dwellings.
larger sample of study sites.
In the long term, a reduction of odour emissions should
6.1.2 Odour impact study Farm A be considered when normal renewal of housing assets
becomes an issue. This could achieve attainment of the
Farm A is a large integrated unit, containing over 17000 limit value for all dwellings in the vicinity and,
animals. It is therefore no surprise that total emissions potentially, create room for some growth of stock when
are high, and the odour footprint relatively large. the agricultural context of the area is recognised and
However, given the locality of the farm - it’s distance accepted as the status quo.
from residential units and it’s rural context, there seems
The only concerns resulting from the site visits were the
uncovered slurry store and carcass skips. It is the
opinion of Odournet UK that these sources may become
a significant emitter of odours during the warmth of
summer. However, having undertaken sampling during a
cool spring day, there is no quantitative data to support
this.
The modelling shows that in the current situation a

limited number of (ten) dwellings may be affected by
odour impacts in excess of the limit value. It is therefore
necessary to seriously consider options to reduce
emissions. In the short-term measures to reduce
emissions from sludge storage should be considered. In
the longer term, replacement of housing assets could
reduce the number of dwellings exposed to odour
impacts in excess of the limit value, conceivably to zero.
The target value will be difficult to attain for farm A. The
21
22
7. Annex: Summary of large-scale survey to

establish pig odour emission factors in the
Netherlands.
7.1 Background operations), Ministry of Agriculture, Nature
Management and Fisheries, The Netherlands, 1996;
The regulatory system in the Netherlands to control
odour nuisance from agricultural sources, uses the • [24] Beoordelingsrichtlijn in het kader van Groen
‘mestvarkeneenheid’ or MVE (literally ‘fattening pig Label Stallen (Guideline for the assessment of Green
unit’) as the basic emission unit for livestock emissions. Label livestock housing systems), ed. 1996, Ministry
One MVE represents the odour emission from one of Housing, Public Planning and the Environment
fattening pig place in a traditional housing system with and the Ministry of Agriculture, The Hague,
a fully slatted floor and slurry storage tanks beneath the Netherlands, 1996.
slats.
7.2 Study sites selection and methodology
In the Netherlands, fattening pigs are defined as boars
from about 25 kg till 7 months (final live weight Four pig units were randomly selected by IMAG-DLO,
between 90 and 120 kg) or sows from about 25 kg till each including substantial numbers of fattening pigs. To
their first cover. be suitable for the study the following features were
required at each site:
Odour measurements were carried out by the Institute of
Agricultural and Environmental Engineering (IMAG- • Housing system - The housing system should consist
DLO) in a project funded by the Dutch government to of mechanically ventilated compartments with pens
establish an emission factor for the Dutch pig unit into equipped with partially slatted floors and slurry
standardised odour units[5]. The measurements are based storage underneath.
on an official Dutch measurement protocol
(‘Meetprotocol voor geuremissies uit stallen’). For • Floor area and animal places - Each pig in the final
aspects of methodology the protocol refers to ‘Guideline weight range should have at least 0.70 m2 floor area
for assessment of low ammonia emission housing of which a maximum 0.40 m2 may be slatted. Animal
systems’. places varied between sites from 80 to 130 pigs per
compartment. The floor area is determined by the
The housing conditions, sampling method and results housing conditions required to meet animal welfare
are described below. The design of the sampling regulations.
programme was derived from the guidance in the
references listed below (the numbers between • Feed - Pigs were fed standard feeds (starter feed and
parentheses refer to the main list of references in this finishing feed) with protein and energy contents
report): within prescribed ranges that are similar to those
prescribed for ammonia measurements[24].
• [5] Ogink, N.W.M., Klarenbeek, J.V., ‘Evaluation of
a standard sampling method for determination of • Sample timing -Two successive fattening rounds
odour emission from animal housings and calibration were sampled, of which one round had to start in the
of the Dutch pig unit into standardised odour units’, second quarter of the year. In each round five odour
Institute of Agricultural and Environmental samples were taken evenly distributed over the
Engineering (IMAG-DLO); fattening period.
• Werkgroep emissiefactoren, Meetprotocol voor At each production site the odour emission of a single
geuremissies uit stallen (translated:Measurement compartment was measured. As one group of pigs is
protocol for odour emission from livestock kept in one compartment during their fattening stage,
23
the same pigs are sampled at five different stages. standard method NVN2820:1995 and results
expressed in Dutch odour units (ge/m3). Please note
• Time of the day - Samples were collected during that 1 ouE/m3 = 2 ge/m3. This is a fixed proportion,
two-hour periods between 10:00 AM and 12:00 AM caused by selecting assessors with an average
only. These sampling periods were timed to coincide detection threshold of 20 ppb/v n-butanol for
with relatively high odour emissions during the daily NVN2820, while EN13725 selects using 40 ppb/v n-
cycle. Two reasons support this choice. Firstly whilst butanol.
emissions are known to be strongest around midday,
variations in emission for the rest of a 24-hour day are • Ventilation - During sampling the ventilation rate
not yet fully quantified. Second, strongest emissions was measured continuously by a fan wheel
will contribute most to odour nuisance. anemometer in the ventilation shaft.
• Number of samples - Samples were collected at each • Other parameters - Temperature and humidity were
of the four study sites, in the course of five sampling measured in the compartment and outdoors by
sessions at each site, at different stage life cycle combined sensors.
stages of a group of fatteners. As two groups of
fatteners were followed sampling took place on ten 7.2.1 Results
discrete days per site. Each sample was taken in
duplicate, giving a total of 20 samples per site, or a Odour emissions (ouE/s per animal place) were
total of 80 samples for the four farms within the calculated by multiplying odour concentration in
study. ouE·m-3 with ventilation rate (m3/s) per animal place.
The required number of samples was defined by the For sampling day the geometric mean of the duplicate
statistical accuracy (uncertainty) per sample and by samples was calculated. The average per facility was
possible variation in emission during the year calculated as the geometric mean of the results from
(seasonal differences). The timing of sampling was each sampling day.
designed to follow the guidance provided in
‘Beoordelingsrichtlijn Groen Label Stallen’, in order The geometric mean of odour emission from the
to derive a representative yearly average. fattening pig system amounted to 22.6 ouE/s per animal
place at a ventilation rate of 33.3 m3/h per animal.
• Sampling method - The samples were collected
using the lung method. This requires odour sample The estimated variance components, expressed as mean
bags to be placed within airtight containers. A percentage deviations, for fattening pigs were:
vacuum is then applied to the airtight container,
surrounding the sample bag. As the bag expands to Between farm variation : 22%
fill the vacuum it draws odorous air into the bag via a Within farm variation : 34%
tube from the site of interest. Variation between duplicate samples : 16%
The evacuation rate was volume proportional to the The within farm variance is the largest component and is
ventilation rate of the fan (controlled by a critical about two times larger than the between farm variance.
orifice). The inlet of the sampling tube was located in
the ventilation shaft before the fan. Both the sampling
tubes and the containers were heated if necessary to
avoid condensation. The odour bags remained in the
container until analysis in the odour lab within 30
hours after collection.
• Odour Analysis - Odour concentration were

determined according to the Dutch olfactometric
24
8. Annex: References
3 Miedema, H.M.E., Walpot, J.I., Steunenberg, C.F., Exposure-annoyance relationships for odours from stationary
sources, Atmospheric Environment 34 (2000) 2927-2936,, 2000.
5 Ogink, N.W.M., Klarenbeek, J.V., Evaluation of a standard sampling method for determination of odour emission
from animal housing systems and calibration of the Dutch pig odour unit into standardised odour units, In:
Proceedings of the Int. Symposium on Ammonia and Odour Control from Animal Production Facilities, NVTL,
Rosmalen, Netherlands, 1997.
10 Ogink, NWM, Lens, PN, Odour Emissions from livestock operations (original title: Geuremissie uit de veehouderij.
Overzichtsrapportage van geurmetingen in de varkenshouderij, pluimveehouderij en rundveehouderij), Institute for
agricultural and Environmental Engineering IMAG-DLO, Netherlands, in press, publication March 2001
11 EN13725 ‘Odour concentration measurement by dynamic olfactometry’, CEN, 1999.
18 van Harreveld, A. Ph. , Klarenbeek, J.V., On the regulations, measurement and abatement of odours emanating from
livestock housing in the Netherlands, In: New knowledge in Livestock Odor, Proceedings of the International
Livestock Odor Conference 1995, pp.16-21, Iowa, USA
20 Inventarisatie Epos; Eindrapport van de eerste fase van het Evaluatie Project Ontwerp Stankconcentratienormen
(EPOS90) (translation: Inventory for Project Evaluation of Provisional Odour Concentration Standards), Ministry
of Public Planning and the Environment, 1990.
21 Carney, P.G., Dodd, V.A., The Measurement of Agricultural Malodours, J. Agric. Engng. Res. (1989) 43, 197-209.
24 Beoordelingsrichtlijn in het kader van Groen Label Stallen (Guideline for the assessment of Green Label livestock
housing systems), ed. 1996, Ministry of Housing, Public Planning and the Environment and the Ministry of
Agriculture, The Hague, Netherlands, 1996.
26 EPA, Guideline on air quality models (revised), july 1986, EPA, EPA-450/2-78-027R.
27 EPA, User's guide to MPTER: a Multiple Point Gaussian dispersion algorithm with optional TERrain Adjustment,
April 1980, EPA, EPA-600/8-80-016
30 Hendriks, HJM, van de Weerdhof, AM, Dutch notes on BAT for pig- and poultry intensive livestock farming,
Ministry of Housing, Public Planning and the Environment and the Ministry of Agriculture, Nature Management and
Fisheries, August 1999. Electronically available at http://www.infomil.nl/lucht/index.htm
25
26
9. Annex: Figures
9.1 Scenario 1, Existing situation impact of farm A, mechanical ventilation, with vents modelled as
point sources (residences referred to in text marked with stars)
Farm A, Scenario 1, C98 ouE/m3
27
9.2 Scenario 2, Farm A, mechanical ventilation, with vents modelled as point sources, after
reduction of emissions by 50% after abatement (residences referred to in text marked
with stars)
Farm A, Scenario 2, C98 ouE/m3
28
9.3 Scenario 1, Existing situation impact of farm B, mechanical ventilation, with vents modelled as
point sources (residences referred to in text marked with stars)
Farm B, Scenario 1, C98 ouE/m3
29
9.4 Scenario 2, Farm B, mechanical ventilation, with vents modelled as point sources, after
reduction of emissions by 50% after abatement (residences referred to in text marked
with stars)
Farm B, Scenario 2, C98 ouE/m3
30

EPA (2001) Odour Impacts and Odour Emission Control Measures For Intensive Agriculture

Uploaded by

Copyright:

Available Formats

EPA (2001) Odour Impacts and Odour Emission Control Measures For Intensive Agriculture

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EPA (2001) Odour Impacts and Odour Emission Control Measures For Intensive Agriculture

Uploaded by

Copyright:

Available Formats

E N V I R O N M E N TA L R E S E A R C H

R&D REPORT SERIES No. 14

Odour Impacts and Odour Emission Control

and was issued by:

Funding and Copyright

Management and organisation and consultation

ANNEX A. ODOUR REGULATIONS FOR INTENSIVE LIVESTOCK IN OTHER COUNTRIES . . . . . . . . 77

ANNEX B. METHODS FOR ODOUR ASSESSMENT AND UNITS OF MEASUREMENT. . . . . . . . . . . . . 87

ANNEX E. CONTOUR PLOTS FOR SCREENING ASSESSMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

(agricultural/rural) of the population. purposes.

temperature and hence is directly related to the welfare European legislation.

larger sample of study sites. the general public.

9) Options for reducing odour emissions from pig

10) The economics of installing technology to abate

11) Retro-fitting of abatement, using air treatment

12) Good operational practice, including suitable

13) A suitable production site for a given production

14) By making the assessment of the impact in the

2. Scope of the study

2.1 Scope framework. It also reviews the practices, methods and

Chapter 9 describes the causes of formation of

3. Introduction to the characteristics of pig production

Pig production, cost and prices, in Euros per kilogram

(Source: Teagasc Pig Advisory Service)

Table 4: Pig life cycle in an integrated production unit

Pig life cycle Start End Weight Weight Average Optimum

(Source: Teagasc Pig Advisory Service)

Typical parameters for approximately 100 sows in an integrated unit

production practices in Ireland and the Netherlands can

1. The slaughter weight in Ireland is somewhat lower,

2. In Ireland the rearing of hogs (boars) for meat is more

3. The mean unit size in Ireland is larger than in

4. Marginally lower summer temperatures would be

5. The use of wet feeding systems in Ireland is

6. Higher nutrient density feeds are used in Ireland, with

7. Raw protein content in the Netherlands is generally

8. The emission factors in the Netherlands were

Overall, it is not plausible that the impact of these

4. Introduction to odours as an environmental stressor

The different dimensions used for characterising odours,

we value our environment. It is not a straightforward

(detectability, intensity, hedonic tone, annoyance

• Context of perception (i.e. other odours, background

This process can be summarised as:

Figure 1: Pathway of odour from production to receptor

When we look at the underlying mechanisms, the factors

For practical purposes, such as regulatory use, the

The exposure is typically quantified in terms of a

Air quality criteria for odour can be set on the basis of

To set environmental exposure criteria with a view to

5. Impact assessment methodology

5.1 Assessment of effects Complaints registration provides an insight into the

The field panel work requires certain weather conditions

5.3.1.1 Point sources flow path section [m2]

Figure 2: Example of Lindvall sampling hood

epidemiological studies, where a particular modelling

The relationship between modelled odour exposure and

• Gaussian plume models were used;

• to represent conditions for an ‘average year’ hourly