EC104/2191AR02[0] AWN Consulting Limited

APPENDIX I

Air Quality Standards & Guidelines

The current ambient air quality standards for nitrogen dioxide, sulphur dioxide, particulates and carbon monoxide are the Air Quality Standards Regulations 2002 (S.I. No. 271 of 2002). These are based on Directives 1999130lEC and 2000/69/EC which are daughter legislation under EU Directive 96162lEC. The Air Quality Standards Regulations 2002 have set limit values which came into operation on 17th June 2002 and are detailed in Table 2.2. The Air Quality Standards Regulations 2002 also detail margins of tolerance, which are trigger levels for certain types of action in the period leading to the attainment date. The margin of tolerance varies from 30% for 24-hour limit value for PMlo, 40% for the hourly and annual limit values for NO2, 26% for hourly SO2 limit values and 60% for the 8-hour CO limit value. The margins of tolerance commenced from June 2002, and started to reduce from 1 January 2003 with further reductions every 12 months thereafter by equal annual percentages to reach 0% by the attainment date.

EU Council Directive 96162lEC on ambient air quality and assessment has recently been adopted into Irish Legislation (S.I. No. 33 of 1999). The act has designated the Environmental Protection Agency (EPA) as the competent authority responsible for the implementation of the Directive and for assessing ambient air quality in the State.

Page 18

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

EC/04/2191AR02[0] AWN Consulting Limited

APPENDIX I I

Description of the AERMOD Model

The AERMOD dispersion model has been recently developed in part by the U.S. Environmental Protection Agency USEP PA)'^). The model is a steady-state Gaussian model used to assess pollutant concentrations associated with industrial sources. The model is an enhancement on the Industrial Source Complex-Short Term 3 (ISCST3) model which has been widely used for emissions from industrial sources. The Proposed Determination 2000 Federal Register Part II (Guidelines on Air Quality Models) has proposed that AERMOD (earlier version of AERMOD without the PRIME algorithm) become the preferred model for a refined analysis from industrial sources, in all terraind3). A ruling by the USEPA on this proposal is due shortly.

Improvements over the ISCST3 model include the treatment of the vertical distribution of concentration within the plume. ISCST3 assumes a Gaussian distribution in both the horizontal and vertical direction under all weather conditions. AERMOD with PRIME, however, treats the vertical distribution as non-Gaussian under convective (unstable) conditions while maintaining a Gaussian distribution in both the horizontal and vertical direction during stable conditions. This treatment reflects the fact that the plume is skewed upwards under convective conditions due to the greater intensity of turbulence above the plume than below. The result is a more accurate portrayal of actual conditions using the AERMOD model. AERMOD also enhances the turbulence of night-time urban boundary layers thus simulating the influence of the urban heat island.

In contrast to ISCST3, AERMOD is widely applicable in all types of terrain. Differentiation of the simple versus complex terrain is unnecessary with AERMOD. In complex terrain, AERMOD employs the dividing-streamline concept in a simplified simulation of the effects of plume-terrain interactions. In the dividing-streamline concept, flow below this height remains horizontal, and flow above this height tends to riseup and over terrain. Extensive validation studies have found that AERMOD (precursor to AERMOD with PRIME) performs better than ISCST3 for many applications and as well or better than CTDMPLUS for several complex terrain data sets@).

Due to the proximity to surrounding buildings, the PRIME (Plume ' ~ i s e Model Enhancements) building downwash algorithm has been incorporated into the model to determine the influence (wake effects) of these buildings on dispersion in each direction considered. The PRIME algorithm takes into account the position of the stack relative to the building in calculating building downwash. In the absence of the building, the plume from the stack will rise due to momentum and/or buoyancy forces. Wind streamlines act on the plume leads to the bending over of the plume as it disperses. However, due to the presence of the building, wind streamlines are disrupted leading to a lowering of the plume centreline.

When there are multiple buildings, the building tier leading to the largest cavity height is used to determine building downwash. The cavity height calculation is an empirical formula based on building height, the length scale (which is a factor of building height & width) and the cavity length (which is based on building width, length and height). As the direction of the wind will lead to the identification of differing dominant tiers, calculations are carried out in intervals of 10 degrees.

In PRIME, the nature of the wind streamline disruption as it passes over the dominant building tier is a function of the exact dimensions of the building and the

Page 19

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

EC/04/2191AR02[0] AWN Consulting Limited

angle at which the wind approaches the building. Once the streamline encounters the zone of influence of the building, two forces act on the plume. Firstly, the disruption caused by the building leads to increased turbulence and enhances horizontal and vertical dispersion. Secondly, the streamline descends in the lee of the building due to the reduced pressure and drags the plume (or part of) nearer to the ground, leading to higher ground level concentrations. The model calculates the descent of the plume as a function of the building shape and, using a numerical plume rise model, calculates the change in the plume centreline location with distance downwind.

The immediate zone in the lee of the building is termed the cavity or near wake and is characterised by high intensity turbulence and an area of uniform low pressure. Plume mass captured by the cavity region is re-emitted to the far wake as a ground- level volume source. The volume source is located at the base of the lee wall of the building, but is only evaluated near the end of the near wake and beyond. In this region, the disruption caused by the building downwash gradually fades with distance to ambient values downwind of the building.

e AERMOD has made substantial improvements in the area of plume growth rates in comparison to ISCST~('). ISCST3 approximates turbulence using six Pasquill- Gifford-Turner Stability Classes and bases the resulting dispersion curves upon surface release experiments. This treatment, however, cannot explicitly account for turbulence in the formulation. AERMOD is based on the more realistic modern planetary boundary layer (PBL) theory which allows turbulence to vary with height. This use of turbulence-based plume growth with height leads to a substantial advancement over the lSCST3 treatment.

Improvements have also been made in relation to mixing height''). The treatment of mixing height by ISCST3 is based on a single morning upper air sounding each day. AERMOD, however, calculates mixing height on an hourly basis based on the ,

morning upper air sounding and the surface energy balance, accounting for the solar radiation, cloud cover, reflectivity of the ground and the latent heat due to evaporation

'

from the ground cover. This more advanced formulation provides a more realistic sequence of the diurnal mixing height changes.

AERMOD also contains improved algorithms for dealing with low wind speed (near calm) conditions. As a result, AERMOD can produce model estimates for conditions when the wind speed may be less than 1 mls, but still greater than the instrument

- threshold.

Page 20

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

EC/04/2191 AR02[0] AWN Consulting Limited

APPENDIX Ill

METEOROLOGICAL DATA - AERMET PRO

AERMOD incorporates a meteorological pre-processor AERMET PRO'^'^). AERMET PRO allows AERMOD to account for changes in the plume behaviour with height. AERMET PRO calculates hourly boundary layer parameters for use by AERMOD, including friction velocity, Monin-Obukhov length, convective velocity scale, convective (CBL) and stable boundary layer (SBL) height and surface heat flux. AERMOD uses this information to calculate concentrations in a manner that accounts for changes in dispersion rate with height, allows for a non-Gaussian plume in convective conditions, and accounts for a dispersion rate that is a continuous function of meteorology.

The AERMET PRO meteorological pre-processor requires the input of surface characteristics, including surface roughness (h), Bowen Ratio and albedo by sector and season, as well as hourly observations of wind speed, wind direction, cloud cover, and temperature. A morning sounding from a representative upper air station, latitude, longitude, time zone, and wind speed threshold are also required.

Two files are produced by AERMET PRO for input to the AERMOD dispersion model. The surface file contains observed and calculated- surface variables, one record per hour. The profile file contains the observations made at each level of a meteorological tower, if available, or the one-level observations taken from other representative data, one record level per hour.

From the surface characteristics (i.e. surface roughness, albedo and amount of moisture available (Bowen Ratio)) AERMET PRO calculates several boundary layer parameters that are important in the evolution of the boundary layer, which, in turn, influences the dispersion of pollutants. These parameters include the surface friction velocity, which is a measure of the vertical transport of horizontal momentum; the sensible heat flux, which is the vertical transport of heat tolfrom the surface; the Monin-Obukhov length which is a stability parameter relating the surface friction velocity to the sensible heat flux; the daytime mixed layer height; the nocturnal surface layer height and the convective velocity scale which combines the daytime mixed layer height and the sensible heat flux. These parameters all depend on the underlying surface.

The values of albedo, Bowen Ratio and surface roughness depend on land-use type (e.g., urban, cultivated land etc) and vary with seasons and wind direction. The assessment of appropriate land-use type was carried out to a distance of 3km from the source location in line with USEPA recommendation^(^). In relation to wind direction, a minimum sector arc of 30 degrees is recommended. In the current model, the surface characteristics for the site were assessed and two sectors identified with distinctly varying land use characteristics.

Surface rouqhness

Surface roughness length is the height above the ground at which the wind speed goes to zero. Surface roughness length is defined by the individual elements on the landscape such as trees and buildings. In order to determine surface roughness length, the USEPA recommends that a representative length be defined for each sector, based on an area-weighted average of the land use within the sector, by using the eight land use categories outlined by the USEPA. The area-weighted

Page 21

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

EC/04/2191AR02[0] AWN Consulting Limited

surface roughness length derived from the land use classification within a radius of 3km from the site is shown in Table A l .

I Sector I Area Weighted Land Use Classification I Spring I Summer I Autumn I inter(') I

Table41 Surface Roughness based on an area-weighted average of the land use within a 3 km radius of Glanbia, Edenderry.

245 - 160 160 - 245

Noon-time Albedo is the fraction of the incoming solar radiation that is reflected from the ground when the sun is directly overhead. Albedo is used in calculating the hourly net heat balance at the surface for calculating hourly values of Monin-Obuklov length. The area-weighted albedo derived from the land use classification within a radius of 3km from the site is shown in Table &2.

I Sector I Area Weighted Land Use Classification I Spring I Summer I Autumn I winter''' I

(1) Winter defined as periods when surfaces covered permanently by snow whereas autumn is defined as eriods 8 when freezing conditions are common, deciduous trees are leafless and no snow is present (Iqbal (1983))'~'. ' ). Thus for the current location autumn more accurately defines "winter" conditions at Glanbia, Edenderry.

1.0 (grassland) 0.5 (grassland) + 0.5 (urban)

(1) Winter defined as periods when surfaces covered permanently by snow whereas autumn is defined as eriods 8 when freezing conditions are common, deciduous trees are leafless and no snow is present (Iqbal (1983))'~' ' '. Thus for the current location autumn more accurately defines "winter" conditions at Glanbia, Edenderry.

0.050 0.525

245 - 160 160 - 245

Table A2 Albedo based on an area-weighted average of the land use within a 3 km radius of Glanbia, Edenderry

BOWEN RATIO

0.100 0.550

1.0 (grassland) 0.5 (grassland) + 0.5 (urban)

The Bowen ratio is a measure of the amount of moisture at the. surface of the earth. The ' presence of moisture affects the heat balance resulting from evaporative cooling which, in turn, affects the Moniri-Obukhov length which is used in the formulation of the boundary layer. The area-weighted Bowen ratio derived .from the land use classification within a radius of 3km from the site is shown in Table A3.

8

0.010 1 0.010 0.505 1 0.505

I Sector I Area Weighted Land Use Classification I Spring I Summer I Autumn I winter"' I

0.180 0.160

when freezing conditions are common, deciduous trees i r e leafless'anb no snow is present (Iqbal (1983))'~'. "'. Thus for the current location autumn more accurately defines "winter" conditions at Glanbia, Edenderry.

0.180 0.170

245 - 160 1 I .O (grassland) 160 - 245 1 0.5 (grassland) + 0.5 (urban)

Table A3 Bowen ratio based on an area-weighted average of the land use within a 3 km radius of Glanbia, Edenderry.

Page 22

0.200 0.190

(1) Winter defined as periods when surfaces covered ~ermanentlv bv snow whereas autumn is defined as oeriods

0.400 0.700

0.200 0.190

0.800 1.400

1 .OOO 1.500

1 .OOO 1.500

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

EC/04/2191AR02[0] AWN Consulting Limited

APPENDIX IV

Ambient N02/NOx Ratio

NO2 has been modelled following the approach outlined by the USE PA(^,'^-'^) for assessing the impact of NOx from point sources. The approach involves assessing the air quality impact through a three tiered screening technique. The initial analysis, termed the Tier 1 approach, assumes a worst-case scenario that there is total conversion of NOx to NO2. The guidance indicates that if this worst-case assumption leads to an exceedence of the appropriate limit value, the user should proceed to the next Tier. Tier 2 is appropriate for estimating the annual average NO2 concentration, though not for estimating the maximum one-hour limit value. The Tier 2 approach indicates that the annual average concentration should be derived from an empirically derived N02/NOx ratio. The guidance suggests that the N02/NOx ratio should be based on data representative of area wide quasi-equilibrium conditions.

In the current assessment, a representative ratio has been developed in the absence of both site-specific nitrogen oxides and ozone monitoring data. Evidence from monitoring stations in Dublin and Mullingar (see Table A4) indicates a lower ratio than the 0.75 default value recommended by the USEPA. However, as a worst-case the USEPA default value of 0.75 has been used in the current assessment.

Table A4 Nitrogen Oxides Results For Dublin and Mullingar Monitoring Stations (EPA Monitoring Reports 1998,2003)

In relation to the maximum one-hour value, the Tier 3 approach is recommended by the USE PA'^). The Tier 3 approach involves the application of a detailed screening method on a case-by-case basis. The suggested methodologies include the ozone- limiting method or a site-specific N02/NOx ratio. The site-specific method requires ambient monitors to be sited to obtain the NO2 and NOx concentrations under quasi- equilibrium conditions. Thus, a literature study was used to derive a conservative N02/NOx ratio at the location of the maximum concentration.

Station

Rathmines, Dublin

College Street

Mullingar

Evidence from monitoring stations in Ireland (see Table A4) indicates a ratio over a wide range of 99.9*c!ile concentrations *of-0.16-0.35. Thus, empirical evidence suggests that a conservative estimate of this site-specific ratio would be 0.35. Therefore, as a worst-case, a ratio of 0.35 for N02/NOx has been used in the current assessment for the 99.8th%ile of one-hour maximum concentrations.

Page 23

Year

1997

1997

2001

Annual NO2

16

83

17

Annual Ratio

NOzlNOx

0.26

0.30

0.57

99.gth%ile I -hr NO2

6 1

255

67

Annual NO

46

194

13

99.gth%ile Ratio

NOdNOx

0.20

0.16

0.35

99.gth %ile I-hr NO

24 1

1294

122

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

AWN Consulting Limited

APPENDIX V

Table A5 Summary of Emission Data Used in Dispersion Model

Summary of Emission Data

Page 24

Stack

Reference

Boiler

(1) 1020 rngl~rn3 SO2 at the stack equates to a LFO sulphur content of 0.6%.

Temp.

(K)

527

Exit

Velocity

(ms")

23.6

Stack cross-

section area

(m2)

0.246

Maximum Operation

~0~ (as ~ 0 ~ )

(mglm3)

650

soz (mglm3)

1020(~)

PM

(mglm3)

125

CO

(mglm3)

150

NOx (as NO2)

(mglm3)

1.96

SO2 (mglm3)

3.08")

PM

(mglm3)

0.38

CO

(mglm3)

0.45

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

For the Attention of: Mr Gerard Cadogan, Glanbia Meats Carrick Road Edenderry .Co. Offaly

BORD NA MONA ENVIRONMENTAL LIMITED

Ref: ECS0390lR3 - Dec 03

A DISPERSION MODELLING ASSESSMENT OF ODOROUS

EMISSIONS FROM THE GLANBIA MEATS FACILITY AT EDENDERRY,

Co. OFFALY

Prepared by: Mr John Conway Air Quality Section Head

Reviewed by: Mr Sean Creedon Environmental Scientist

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glanbia Meats Report No. ECS0390LR3

Executive Summarv

Bord na Mbna Technical Services was commissioned by Glanbia Meats to carry out an odour

impact assessment at the Glanbia facility, Edenderry, Co. Offaly. The purpose of the

assessment was to determine the potential of the facility for the generation of odour impact on

the surrounding vicinity and provide adequate abatement measures if required. Potential odour

sources were identified from a site visit and were agreed by the Environmental Protection

Agency. These odour sources were used to construct the bases of the modelling assessments. Odour emission rates/fluxes were calculated f?om site specific olfactometric data.

Using the odour emission rates calculated, a dispersion modelling investigation was undertaken

in accordance with recommended odour annoyance criterion for other odorous industries. The following modelling scenarios are contained within this report:

1. Odour emission contribution of the facility to odour plume dispersal for the current scenario.

2. Covering the Dewatered Sludge Holding and Balance Tanks and extracting through an abatement unit and assessing the odour emission contribution of the facility to odour plume dispersal.

3. Future scenario to include the addition of new tanks for proposed expansion including the above abatement option and assessing the odour emission contribution of the facility to odour plume dispersal.

In keeping with the Irish EPA recommendations for other industries, those residences that have complained should be located outside the 5 3 o u ~ l m ~ isopleth for the 98th percentile in one year as determined by atmospheric dispersion modelling software. For the purposes of this dispersion modelling assessment, the <30udm3 as a 98th

. percentile has been adopted as the odour annoyance criterion.

It was concluded that:

> In accordance with an odour annoyance criterion of 5 3 oudm3 for the 98" percentile,

an odour impact criterion may be perceived at the selected sensitive locations included in the model as observed in Figure 2.

> The Dewatered Sludge holding Tank was most significant on-site odour source based

on odour emission rates and subjective on-site assessment

> The most favourable solution for the reduction of odour emission is to cover the

Dewatered Sludge Holding and Balance Tank and extract the headspace air through an abatement unit.

Bord na M6na Technical Services T n m z n n i l ? /Id

Page 2

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Gianbia Meats Report No. ECS0390Bt3

: a k Under the modelled scenarios, the installation of the abatement unit to cover, extract

and treat the headspace air for the above sources i.e. Dewatered Sludge Holding Tank

and the Balance Tank, will not exceed the proposed odour annoyance criterion of 5 3

ouE/m3 as a 9gth percentile. This is illustrated in Figure 3.

The future expansion of the Glanbia facility to include the upgrade of the WwTP is not predicted to result in an odour impact at any sensitive receptor for the odour annoyance criterion of 13 o u ~ l m ~ as a 9gth percentile. This is illustrated in Figure 4.

The expansion of the WwTP at the Glanbia facility may have a marginally greater impact on the receiving environment. However, the model shows that this is not significant and no further mitigation measures are required.

a Respectively Submitted

~ r . John Conway- Air Quality Section Head

Mr Sean Creedon Environmental Consultant

Bord na Mdna Technical Sewices January, 04

Page 3

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glan bia Meats Report No. ECS0390BZ3

TABLE OF CONTENTS

1.0 INTRODUCTION

2.0 SCOPE

3.0 OLFACTOMETRY

4.0 SAMPLING AND ANALYTICAL METHODOLOGY

4.1 Odour Sampling

4.2 Dynamic Olfactometry

4.3 Sampling Conditions

4.4 Quality Control 4.5 Commitment to Quality

5.0 DISPERSION MODELLING DESCRIPTION

Introduction Industrial Source Complex 3 Odour Annoyance Criteria Terrain Description Sources Odour Emission Rate Calculation Meteorological Data Site Map

6.0 DISPERSION MODELLING ASSESSMENT

6.1 Introduction

6.2 Source, Input Data 6.3 Dispersion Modelling Results

7.0 DISCUSSION OF RESULTS

8.0 CONCLUSION

APPENDIX I Questionnaire to Residents APPENDIX II Residents Feedback from Questionnaire APPENDIX m Abatement Specification Details

Bord na Mdna Technical Services r --------- n l

Page 4

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glanbia Meats Report No. ECS0390LR3

INTRODUCTION

Bord na M6na, Technical Services was commissioned in July 2003 by Glanbia Meats in Edenderry, Co. Offaly to conduct an odour emission monitoring survey. Subsequent

dispersion modelling assessments of the current and future scenarios were carried out.

The project has been undertaken in response to a request by the Environmental

Protection Agency (EPA) arising from complaints lodged by residents in the vicinity of

the Glanbia Meats facility. The EPA requested that all complainants be contacted by

letter to ascertain from past experience the day(s) on which the odorous emissions were most likely to be greatest. The responses received indicated that Friday was considered

the worst day of the week for odour generation. This was concurrent with Gambia's

opinion. This day was then selected to carry out the odour monitoring survey. This

questionnaire is presented in Appendix I along with the responses from complainants

(Appendix II).

The dispersion modelling assessment allows for the estimation of the predicted odour

impact area for the odour annoyance criteria presented in Table 5.3 on the surrounding

vicinity for each scenario.

The impact assessments are presented in the form of concentration contours/isopleths

produced using US EPA approved and recommended Irish EPA dispersion modelling

techniques (Industrial Source Complex Short Term 3 - ISCST3). Concentration

contours are plotted on Ordnance Survey Maps of the locality indicating 98fi percentile

odour concentrations (using a worst case year of hourly meteorological data).

Bord na Mdna Technical Services Page 5 January, 04

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glanbia Meats Report No. ECS0390LU3

2.0 SCOPE

2.1 Scope of Project

The scope of the project was agreed as follows:

Identi@ on-site odour sources and submit the proposed sources to be sampled, to

the EPA for approval. From the on-site odour sources identified, carry out a sampling survey to

determine odour concentrations.

Determine the current impact of odorous emissions using dispersion modelling

techniques from sources monitored o Following analysis of the predicted results for the current scenario, make

recommendations, if necessary, on the sources that require abatement If required, undertake dispersion modelling for this scenario (current scenario

with abatement) to determine the impact on the receiving environment for compliance with proposed odour annoyance criterion

Undertake dispersion modelling for the proposed expansion of the WwTP

(hture scenario) to determine the impact on the receiving environment for compliance with proposed odour annoyance criterion. This may include

modelling the future scenario with or without abatement.

Bord na M6na Technical Services Tnmrnnwrr LlA

Page 6

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glanbia Meats Report No. ECS039OLU3

@ 3.0 OLFACTOMETRY

Olfactometry using the human sense of smell is the most valid means of measuring

odour and is currently the most commonly used method of measuring the concentration

of odour in air. Olfactometry is carried out using an instrument called an olfactometer

(Dravnieks et al., 1986, Atlas of Odor Character Profiles. ASTM Committee on sensory

evaluation of materials and products, ASTM data series. Baltimore, MD, USA). Three

types of dynamic dilution olfactometers exist:

Yes/No Olfactometer

Forced Choice Olfactometer

Triangular Forced Choice Olfactometer

In the dynamic dilution olfactometer, the odour is first diluted and is then presented to no less than four screened panellists (EN13725, 2001). Panellists are previously screened to ensure they have a normal sense of smell. According to the prEN13725

standard this screening must be performed using a certified reference gas n-butanol. This screening is applied to eliminate anosmia (low sensitivity) and supernoses @ugh sensitivity). The odour analysis has to be undertaken in a low odour environment such as an air conditioned odour fiee laboratory. Analysis should always be performed

preferably within 24 hours of sampling.

The odour concentration of a gaseous sample of odorant is determined by presenting a

panel of selected screened human panellists with a sample of odorous air and varying

the concentration by diluting with odourless gas (nitrogen) in order to determine the

dilution factor at the 50% threshold. The Z50 value (threshold concentration) is

expressed in odour units (ou5/m3).

Although odour concentration is an odourless number, by analogy, it is expressed as a

concentration in odour units per cubic metre (oudm3), a term which simplifies the calculation of odour emission rate. The European odour unit is the amount of odorant(s) that, when evaporated into one cubic metre of neutral gas (nitrogen), at standard

conditions elicits a physiological response from a panel (detection threshold) equivalent to that elicited by one European Reference Odour Mass (EROM) evaporated in one cubic metre of neutral gas at standard conditions. One EROM is the mass of a substance (n-butanol) that will elicit the Z50 physiological response assessed by an odour panel in accordance with this standard. n-butanol is one such reference standard and is equivalent to 123pg of n-butanol evaporated in one cubic metre of neutral gas at standard conditions (293K and 101.3kPa on a wet basis) (EN13725,2001)

Bord na Mdna Technical Sewices January, 04

Page 7

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glanbia Meats Report No. ECS0390lR3

4.0 SAMPLING AND ANALYTICAL METHODOLOGIES

4.1 Odour Sampling

4.1.1 Point and Volume Sources

Samples of gas of approximately 60 litres are collected via PTFE tubing into

~ a l o ~ h a n e @ gas sampling bags by means of the "lung principle" method. Using this

method, the sample bag is housed in a sealed car buoy that is evacuated using a vacuum

pump. The volume of air removed from the car buoy is replaced by sample gas entering

the bag through the PTFE tube whose inlet is placed in the odour stream, thus avoiding

contamination of sample by pumps or meters. Sampling was carried out in accordance with the CEN standard EN 13725 entitled 'Air Quality - Determination of Odour

Concentration by Dynamic Olfactometry'.

4.1.2 Area Sources

Samples from the locations without outward flow are taken using a Lindvall box. This

device consists of a stainless steel rectangular box that is open at one side and is used to

cover an area of 0.333 m2 of the emitting surface. Using a fan and an activated carbon

filter, odour-£tee air is passed through the box to simulate wind movements across the

surface of the wastewater. Analysis of the samples collected at the outlet of the box in

conjunction with the box dimensions and wind speed generated allows calculation of the

odour emission rate per unit area £tom the surface.

4.2 Dynamic Olfactometry

The samples are analysed by Dynamic Olfactometry. The instrument used was an

Olfactomat-e Olfactometer (Project Research Amsterdam) and the analytical procedures

were in accordance with the CEN Standard EN 13725 using a trained panel of 4

assessors. The odour concentration of the sample is expressed in odour units per cubic

metre of gas (oudm3). These values, sometimes referred to as "dilutions to threshold" are equivalent to the number of times the sample gas required dilution with odour free air to reach the panels odour threshold (i.e. the concentration at which there is a 50%

probability of the panellists detecting the odour). The results are expressed in o u ~ / m ~ .

4.3 Sampling Conditions

The weather was characterised as sunny, warm, light winds and a temperature of 15°C on 1 5 ~ ~ August 2003. On 22nd August 2003, the weather was warm, mild, heavy drizzle in the morning and a temperature of 13 OC.

Bord na M6na Technical Services T,,u~,~Y.~ nA

Page 8

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glanbia Meats Report No. ECS0390Dt3 --

a 4.4 Control Chain of Custody

As part of the Quality System in place at Bord na M6na, Environmental Limited,

measures are taken to ensure controlled chain of custody. An outline of the chain of

custody is given overleaf.

BORD NA M ~ N A ENVIRONMENTAL LIMITED

CONTROLLED CHAIN OF CUSTODY L

and packaging Transport Transport to Sample Receiving of samples at Bord na of were Document laboratory by Reception M6na Environmental Laboratory

carried out by Bord na Fom Bord na M6na Form complex by: Mbna Technical Team: Technical Team. Laboratory Manager

John Conway, Sean (Secure. laboratory complex access

Creedon, Kieran Gordon to authorised personnel only)

Storage of all samples for 1 month

period after report issue.

Supervised Disposal

Bord na Mdna Technical Services January, 04

-

Page 9

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glanbia Meats Report No. ECS0390LR3

4.5 Commitment to Quality

4.5.1 ILAB Accreditation

Bord na Mona Technical Services Analytical Laboratories is ILAB accredited by the

National Accreditation Board (NAB) since 1997. It has always been the policy of the

laboratories to achieve and maintain a high standard of quality consistent with client's

requirements in all aspects of the work carried out within the laboratory.

Bord na Mona Technical Services Analytical Laboratories successfblly transferred to

the new standard of accreditation IS0 17025 on the 16" of November 200 1.

4.5.2 Accredited to IS0 17025

This new standard contains all of the requirements that testing laboratories have to

meet if they wish to demonstrate that they operate a quality system, are technically

competent, and are able to generate technically valid results. IS0 17025 incorporates

all those requirements of IS0 9001 and IS09002 that are relevant to the scope of

testing services that are covered by the laboratory's quality system. Thus a laboratory

that complies with IS0 17025 will therefore also operate in accordance with IS0 9001

or IS0 9002.

4.5.3 Interlaboratory Proficiency Schemes

To ensure the accuracy of the analytical testing we participate in several external

proficiency schemes. The ongoing competence of the laboratory and its staff is

assessed by participation in various inter-laboratory proficiency testing schemes, such

as Aquacheck and the EPA scheme organised for environmental laboratories throughout Ireland.

4.5.4 EPA Quality Control Register

Bord na Mona Technical Services Analytical Laboratories performance in the EPA

intercalibration scheme has insured its listing on the EPA's register of Quality

Controlled Laboratories. Both accredited and non-accredited test methods are

assessed by these schemes.

-

Bord na Mdna Technical Services Page 10

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:23

Glan bia Meats Report No. ECS0390A?3

DISPERSION MODELLING DESCRIPTION

1 5.1 Introduction

Any material discharged into the atmosphere is carried along by the wind and diluted by

wind turbulence which is always present in the atmosphere. This process has the effect

of producing a plume of polluted air that is roughly cone shaped with the apex towards

the source and can be mathematically described by the Gaussian equation. Atmospheric dispersion modelling has been applied to the assessment and control of odour for many

years, originally using Gaussian form ISCST 3 and more recently utilising advanced

boundary layer physics models such as ADMS and AERMOD. Once the odour

emission rate from the source is known (ouE/s), the impact on the vicinity can be

estimated. These models can effectively be used in three different ways. Firstly, to

assess the dispersion of odour and to correlate with complaints. Secondly, in a "reverse"

mode, to estimate the maximum odour emissions which can be permitted from a site in order to prevent odour complaints occurring. Thirdly, to determine which process is

contributing greatest to the odour impact and estimate the amount of required abatement to reduce this impact to within acceptable levels. In this latter mode, models have been

employed for imposing emission limits on industrial processes, odour control systems

and intensive agricultural processes.

5.2 Industrial Source Complex 3 (ISC 3)

The Industrial Source Complex Short-Term 3 (ISCST3) model is a refined air dispersion

model used to predict pollutant concentrations fiom a wide range of sources that may be

present at typical industrial facilities. This model is recommended by the US

Environmental Protection Agency (EPA) guideline on Air Quality Modelling for applications to refinery-like soukes i d other industrial sources. It was also recently

recommended by the Irish EPA to model the potential odour impact from intensive agriculture, mushroom composting and tannery facilities. The basis of the model is the straight-line, steady-state Gaussian plume equation, which is used with some modifications to model simple point source emissions from stacks that experience the effects of aerodynamic downwash due to nearby buildings, isolated vents, multiple vents, storage piles, conveyor belts, and the like. Essentially, emission sources are categorised into four basic types of sources, i.e., point sources, volume sources, area sources, and open pit sources.

The ISCST3 Model accepts hourly meteorological data records to define the conditions for plume rise, transport, diffusion, and deposition. Depending on the location of the

facility the appropriate meteorological data is chosen. The model also takes into account

the local terrain surrounding the facility. The model estimates the concentration or

deposition value for each source and receptor combination for each hour of input

meteorology, and calculates user-selected short term averages. Since most air quality

Bord na Mbna Technical Services January, 04

Page 11

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:24

Glan bia Meats Report No. ECS0390Dt3

standards are stipulated as averages or percentiles the ISCST3 model allows further

analysis of the results for comparison with these standards.

Percentile analysis for odour emissions are calculated for the maximum 1-hour

averages using the ISCST3-PERCENT post-processing utility. This utility determines

the maximum concentration of a pollutant fiom all receptors at a specific percentile,

for a specific averaging period. Employing the percentile facilitates the omission of

unusual short term meteorological events that may cause elevated pollutant

concentrations and hence a more accurate representation of the likely average pollutant

concentrations over an averaging period.

5.3 Odour Annoyance Criteria

Commonly used odour annoyance criteria for Ireland are illustrated in Table 5.3.

Generally, odour concentration should be below a maximum of 6oudm3 for the 9~~

percentile of the time in one standard meteorological year (i.e. 2% or 175 hours per annum) in order to prevent complaints arising. In Ireland, an odour concentration of 26 (existing facility), S 3 (new facility) and 11.5 (Target value) oudm3 for the 98" percentile has been implemented for intensive agricultural facilities, mushroom compost facilities and the tanning industry in order to limit complaints. In practice, if

complaints exist around a facility, an odour concentration of 5 3 oudm3 for the 9gth percentile may be implemented to eliminate these complaints.

An odour threshold of 1 ouE/m3 is the level at which an odour is detectable by 50% of the screened panellists. According to research on wastewater treatment works, the odour recognition threshold is approximately 3-5 times this concentration and is liable

to cause offence (3-5 oudm3) (Keddie A. W. C., 1980, Chapter 11 - Dispersion of

Odours - A Concise Guide, Valentine F.H.H., and North A.A., Warren Springs

Laboratory, Warren Springs, UK). An odour impact criterion of 5 5 ouE/m3 was

Table 5.3: Odour Annoyance Criteria for Dispersion Modelling

Bord na Mdna Technical Services T n n l ~ n m , Ad

Page 12

Application

Limit Value for new facilities

Limit Value for Existing Facilities

Restrictive Criterion for Landfill Odour at this site

Used as a planning guideline for the Upgrade of WwTP

Concentration Limit (ou~ lm~) Ireland (pig, mushroom compost and

tanning industry) ISC ST and Complex 1 Model

53.0

56.0

Brogborough Landfill (UK)

r3.0

UK (WwlT, ADMS Model)

9 . 0

Percentile Value (%)

98

98

98

98

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:24

Glanbia Meats Report No. ECS0390LR3

adopted for a WwTP in the UK and was accepted in the planning application for this facility to limit odour impact.

As complaints from local people have been received, it is considered prudent to limit

the chance of odour impact and apply an odour impact criterion of <30udm3. Furthermore, in accordance with the odour annoyance criterion presented in Table 5.3 and in keeping with the Irish EPA recommendations for other industries, those residences that have complained should be located outside the <30udm3 isopleth for the 98th percentile in one year as determined by atmospheric dispersion modelling

software.

5.4 Terrain Description

The terrain surrounding the Glanbia Meats facility was complex, so a topography file was created for the facility and the surrounding area to investigate the effects of complex terrain on the modelling scenarios.

Odour concentrations were calculated at each 50 metre x y Cartesian grid receptor location that is predicted to exceed for 2% (175 hours) of an average of a worst case meteorological year. The total grid size was 1.8 km x 2.4 km (ie 37 x 49 Cartesian grid

locations).

The elevations of the receptor locations 'were obtained fi-om a 1 : 10000 scale map of the

area obtained fi-om the Ordnance Survey office. Elevations were taken fi-om map

contours and bench marks throughout the area of the receptor grid. Base elevations of

all of the proposed buildings include in the model were supplied by the client.

a . The site layout was obtained fi-om a *.dwg drawing, provided by E. G. Pettit. Using

this tagged drawing, the boundary, significant buildings and location of the odour

sources were inputted into the model.

5.5 Sources

A site layout plan provided by the E. G Pettit & Co. as a *.dwg file and was imported into the model program and used as a template for the source locations and the boundary of the facility.

5.5.1 Point Sources

A point source is a source that releases effluent pollutants &om a limited opening, such as a stack or vent. The ISCST3 model for stacks uses the steady-state Gaussian plume

equation for a continuous elevated source.

Bord na Mdna Technical Services January, 04

Page 13

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:24

Glanbia Meats Report No. ECS0390423

When one or more buildings in the vicinity of a point source interrupt wind flow, an area of turbulence known as a building wake is created. Pollutants emitted fkom a

relatively low level (e.g. a roof vent or a short stack) can be caught in this turbulence, affecting their dispersion. This phenomenon is called building downwash. In order to

conduct an extensive analysis of downwash effects of all the point sources the

dimensions of all significant buildings and structures on site were obtained fiom the

site layout drawing supplied by the client and inputted into the model. The downwash

effects are determined using the building profile input program (BPP). The height of

each relevant building, as selected by Bord na M6na Technical Services was supplied

by E. G. Pettit.

5.5.2 Area Sources

Tanks are typically modelled as area sources. In order to take a representative air

sample fiom these sources a portable wind tunnel sampling device known as a

Lindvall box is used. The principal of the wind tunnel system is that controlled 'odour

free7 air (filtered through an activated carbon device) flows over the water surface

body absorbing any odours fiom the surface. The odour emission rate is defined as

the quantity of odour emitted per m2 of surface area per unit of time.

5.5.3 Volume Sources

The openings associated with the sludge dewatering building and the openings to the

lairage building were modelled as volume sources. Due to practical constraints it is

impossible to get a representative sample either by covering the sample area or by

using the Lindvall box. Consequently, samples were taken of the ambient air over the

sampling area. In such cases the emitted airflow is dependent upon meteorological

conditions, wind speed, wind direction etc. As a result, average wind speeds were

used to calculate the number of air changes over the samplearea

5.6 Odour Emission Rate Calculation

The measurement of the strength of a sample of odorous air is, however, only part of the problem of quantifying odour. Just as pollution fiom a stack is best quantified by

a mass emission rate, the rate of production of an odour is best quantified by the odour emission rate. For a chimney or ventilation stack, this is equal to the odour threshold concentration (ouE/m3) of the discharge air multiplied by the flow rate (m3/s). It is equal to the volume of air contaminated every second to the threshold odour limit (OUE/S). The odour emission rate can be used in conjunction with dispersion modelling to predict the approximate radius of impact or complaint.

Bord na Mdna Technical Sewices .Ta~uarv- 04

Page 14

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:24

Glanbia Meats Report No. ECS0390A3

a Area source mass emission rateslfluxes are calculated as either oudm2/s or O U ~ S

depending if they are being represented as discrete point sources or area sources in the atmospheric dispersion model. The point and volume sources are calculated as ouds.

5.7 Meteorological Data

Three years of hourly sequential meteorological data (Claremoms 1993 to 1995 inclusive as recommended by the EPA) was used for the operation of ISC ST3. This

allowed for the determination of the predicted worst case overall impact of odour emissions £tom the Glanbia facility on the surrounding environment.

5.8 Site Map

Figure 1 illustrates the facility and the selected sensitive receptors (marked with yellow crosses) superimposed on an Ordnance Survey base map is illustrated on the following page. The closest residential premises in any direction are deemed sensitive receptors for the purposes of this modelling assessment. As a consequence, once the odour annoyance criterion of G o u ~ / m ~ as a 98th percentile is not exceeded at the five selected sensitive receptors, this annoyance criterion will not be exceeded at any of the

additional receptors.

Bord na Mdna Technical Services Page 15 January, 04

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:24

Glahdia Meat8 RepoH No. ECS03yL/R3

Figw 1: Site Map with locations of the Selected sen sit.^ ..crrr.'rrs Note 1: The base map may not be exactly to scale i.e. l:6inch

Bard na Mdna Tecl~nical Services Page 16

For

insp

ectio

n pur

pose

s only

.

Conse

nt of

copy

right

owne

r req

uired

for a

ny ot

her u

se.

EPA Export 25-07-2013:18:08:24