EirGrid Evidence Based Environmental

Studies Study 8: Noise

Literature review and evidence based field study on the
noise effects of high voltage transmission development

May 2016
This document has been prepared by EirGrid plc with the assistance of RPS Group.

The evidence contained in this study has been provided by RPS Group.

Reproduction of this study is authorised provided the source is acknowledged.

May 2016
Evidence Based Studies Noise – Main Report

SUMMARY

This is an evidence-based study carried out by experts in noise. The research looks
at the actual noise effects of the construction and presence of high voltage
transmission infrastructure in Ireland. Such projects include overhead lines and
substations. Underground cables have been excluded from the study as they do not
create any significant noise. This document is intended for use by professionals
working in the area if environmental impact assessment for transmission lines. It is
also intended to inform best practice in future planning of this infrastructure.

The purpose of this study has been to:

• Review literature on the noise effects of transmission infrastructure.

• Report on and discuss noise surveys of existing overhead lines and

substation transmission infrastructure of different voltages across the country
– 110 kV, 220 kV and 400 kV.

• Present conclusions and recommendations in relation to noise for the future

design and siting of transmission infrastructure projects.

In addition to the survey work completed as part of this study, EirGrid separately
commissioned a long term noise survey at the existing Dunstown-Moneypoint 400kV
overhead transmission line in Cloney, Co. Kildare. This work was completed by
AECOM and supplements the work completed as part of this study, feeding into the
overall discussion, conclusions and recommendations.

There are a number of ways in which noise can be generated from electricity
infrastructure. Generally, these fall within four categories of noise:

• audible noise associated with “Corona Noise” from high voltage transmission
lines – generally heard as crackling and hissing;

• audible noise associated with dirty, damaged or cracked insulators;

• audible noise associated substation equipment;

i
Evidence Based Studies Noise – Main Report

• audible noise associated with wind blowing through electricity infrastructure –

this is called “Aeolian Noise”.

The literature review confirms that the level of noise impact likely from electricity
transmission lines increases with the increase of the voltage strength of the line.
Much of the literature indicates that “Corona Noise” only becomes a significant issue
from 350-500 kilovolts (kV) and above. This would suggest that significant “Corona
Noise” impacts may not be likely for 110 kV and 220 kV transmission lines and that
the potential for more significant impacts may only relate to 400 kV lines.

Noise surveys (measuring noise over a period of time) were carried out at locations
along existing 110 kV, 220 kV and 400 kV overhead lines in Ireland, and at
substations of the same voltages. The approach was to survey when the line was in
operation (“on”), and when it was switched out (“off”), say for routine maintenance,
and to compare the noise recorded at these two different survey times.

It wasn’t possible to switch off any substation, so the overall noise environment of the
operating substation was surveyed. In addition, it was decided not to carry out noise
surveys of any substations under construction. This would be similar to any
construction project, rather than specific to a transmission substation development.

The results of the surveys were compared with the information contained in the
literature, and with information from the separate AECOM survey.

The results from the 110 kV and 220 kV overhead line surveys present strong
evidence that these lines are not likely to result in a significant noise impacts in their
vicinity. On this basis, the planning of 110 kV and 220 kV lines should not be
significantly constrained on the basis of potential noise issues.

The noise study on the 400 kV overhead line (OHL) provided evidence which showed
that these lines do produce significant corona noise effects under certain conditions
(especially at night under humid or wet conditions). This evidence is consistent with
the literature review which shows that corona effects start to become significant in
noise impact terms at voltages in the range of 350-500 kV. There is potential for
noise impacts from such corona effects on properties located very close to such
infrastructure in quiet rural locations. On this basis, in the design and siting of new

ii
Evidence Based Studies Noise – Main Report

400 kV overhead lines it may be appropriate to seek a separation distance of 200m

between any property and a 400 kV tower, and 100m between any property and the
OHL.

Noise measurement surveys completed at 110 kV, 220 kV and 400 kV substations
recorded steady state noise levels in the vicinity of the boundaries of these
substations. To avoid any noise impacts at sensitive receptors, it is recommended
that in the design and siting of new substations: a minimum distance of 5m is
maintained between a 110 kV substation and the land boundary of any noise
sensitive receptor. A distance of 20m is to be maintained between a 220 kV
substation and the land boundary of any sensitive receptor. A minimum distance of
150m is to be maintained between a 400 kV substation and the land boundary of any
noise sensitive receptor.

iii
Evidence Based Studies Noise – Main Report

TABLE OF CONTENTS

1 INTRODUCTION ........................................................................................................................ 1
1.1 THE SCOPE OF THIS PROJECT ............................................................................................................ 1
1.2 THE SCOPE OF THIS STUDY........................................................................................................... 2
1.3 THE TRANSMISSION NETWORK AND NOISE ..................................................................................... 3
2 LITERATURE REVIEW.............................................................................................................. 6
2.1 INTRODUCTION ............................................................................................................................. 6
2.1.1 Scope and Aims ................................................................................................................. 6
2.1.2 Noise Fundamentals .......................................................................................................... 6
2.1.3 Audible Noise from Corona Discharge .............................................................................. 7
2.1.4 Insulator Noise ................................................................................................................... 8
2.1.5 Substation Noise ................................................................................................................ 8
2.1.6 Aeolian Noise ..................................................................................................................... 8
2.2 AUDIBLE NOISE FROM OVERHEAD TRANSMISSION LINES ................................................................. 9
2.2.1 Introduction ........................................................................................................................ 9
2.2.2 Measurement of Audible Noise from Transmission Lines ................................................. 9
2.2.3 Factors Influencing Corona Discharge ............................................................................ 10
2.2.4 The Effect of Rainfall on Corona Discharge .................................................................... 12
2.2.5 Character of Audible Noise from Corona Discharge ........................................................ 13
2.2.6 Methods for Calculating Audible Noise of High Voltage Transmission Lines .................. 17
2.2.7 Mitigation Measures for Corona Discharge Audible Noise .............................................. 17
2.2.8 Discussion ....................................................................................................................... 18
2.3 AUDIBLE NOISE FROM SUBSTATIONS ............................................................................................ 19
2.3.1 Introduction ...................................................................................................................... 19
2.3.2 Source of Noise in Substations ........................................................................................ 19
2.3.3 Noise from Substations .................................................................................................... 21
2.3.4 The Effect of Impulsive Substation Noise on Wireless Communications Systems .......... 22
2.3.5 Mitigation Measures for Substations ................................................................................ 22
2.3.6 Discussion ....................................................................................................................... 23
2.4 AEOLIAN NOISE FROM ELECTRICAL INFRASTRUCTURE ................................................................... 24
2.4.1 Introduction ...................................................................................................................... 24
2.4.2 General Mechanism of Aeolian Noise ............................................................................. 24
2.4.3 Aeolian Noise Characteristics .......................................................................................... 25
2.4.4 Mitigation Measures ......................................................................................................... 26
2.4.5 Discussion ....................................................................................................................... 27
2.5 HEALTH EFFECTS OF NOISE ........................................................................................................ 28

iv
Evidence Based Studies Noise – Main Report

3 METHODOLOGY ..................................................................................................................... 30
3.1 SITE SELECTION ......................................................................................................................... 30
3.1.1 Site Selection Criteria ...................................................................................................... 30
3.1.2 Outline of Sites Selected ................................................................................................. 31
3.1.3 Limitations to Site Selection Process ............................................................................... 39
3.2 NOISE SURVEY METHODOLOGY ................................................................................................... 40
3.2.1 Survey Equipment - Overhead Lines ............................................................................... 40
3.2.2 Survey Methodology - Overhead Lines ........................................................................... 40
3.2.3 Problems Encountered During Surveys - Overhead Lines .............................................. 41
3.2.4 Survey Equipment - Substations ..................................................................................... 41
3.2.5 Survey Methodology - Substations .................................................................................. 42
3.3 METHODOLOGY FOR ANALYSIS OF DATA ...................................................................................... 42
3.3.1 Overhead Lines ............................................................................................................... 42
3.3.2 Substations ...................................................................................................................... 44
4 SURVEY RESULTS ................................................................................................................. 45
4.1 SURVEY RESULTS FOR 110KV OVERHEAD LINES .......................................................................... 45
4.1.1 Cathleen Falls (C’Fall) Golagh Tee - Letterkenny 110kV OHL (1) .................................. 45
4.1.2 Cathleen Falls (C’Fall) Golagh Tee - Letterkenny 110kV OHL (2) .................................. 48
4.1.3 Cathleen Falls (C’Fall) - Srananagh 110kV OHL .............................................................. 49
4.1.4 Golagh Tee - Letterkenny 110kV OHL ............................................................................ 52
4.2 SURVEY RESULTS FOR 220KV OVERHEAD LINES .......................................................................... 52
4.2.1 Dunstown-Maynooth 220kV OHL - Location 1 Betaghstown ........................................... 52
4.2.2 Dunstown-Maynooth 220kV OHL - Location 2 Currabell ................................................. 55
4.2.3 Dunstown-Maynooth 220kV OHL - Location 3 Thomastown ........................................... 57
4.3 SURVEY RESULTS FOR 400KV OVERHEAD LINES ........................................................................... 60
4.3.1 Oldstreet-Woodland 400kV OHL - Ardrums Great .......................................................... 60
4.4 SURVEY RESULTS FOR SUBSTATIONS ........................................................................................... 63
4.4.1 Dunfirth 110kV Substation ............................................................................................... 63
4.4.2 Gorman 220kV Substation ............................................................................................... 64
4.4.3 Woodland 400kV Substation ........................................................................................... 66
5 AECOM SURVEY ON 400KV OVERHEAD LINE ................................................................... 69
5.1 INTRODUCTION ........................................................................................................................... 69
5.2 METHODOLOGY .......................................................................................................................... 69
5.3 MEASURED DATA ....................................................................................................................... 69
6 DISCUSSION AND CONCLUSION ......................................................................................... 74
6.1 DISCUSSION ON SURVEY RESULTS FROM 110KV OVERHEAD LINES ............................................... 74
6.2 DISCUSSION ON SURVEY RESULTS FROM 220KV OVERHEAD LINES ............................................... 75
6.3 DISCUSSION ON SURVEY RESULTS FROM 400KV OVERHEAD LINES ............................................... 76
6.4 DISCUSSION ON AECOM 400KV SURVEY .................................................................................... 76

v
Evidence Based Studies Noise – Main Report

6.5 DISCUSSION ON SURVEY RESULTS FROM SUBSTATIONS ................................................................ 77

6.6 CONCLUSION .............................................................................................................................. 78
7 RECOMMENDATIONS ............................................................................................................ 80
8 REFERENCES ......................................................................................................................... 82

GLOSSARY OF ACOUSTIC TERMINOLOGY ..................................................................................... 85

APPENDIX A OVERVIEW OF ELECTRICITY INFRASTRUCTURE

APPENDIX B 400KV TRANSMISSION LINE MONITORING NOISE REPORT PREPARED BY
AECOM CONSULTANTS, APRIL 2015

LIST OF FIGURES
Figure 1.1: EirGrid and SONI’s Transmission System Map (January 2015)
Figure 2.1: Typical Lateral Profile of Noise from a 400kV Transmission Line
Figure 2.2: Schematic Representation of Corona Discharge
Figure 2.3: Lateral Profile of A-weighted Tonal Component (100Hz)
Figure 2.4: Typical Frequency Spectrum of Transmission Line AN Wet with 40, 50 and 60Hz
Supply Voltage
Figure 2.5: Discrete Frequency Pattern of a Wet Overhead Transmission Line
Figure 2.6: Discrete Frequency Pattern of a Dry Overhead Transmission Line
Figure 2.7: Typical Noise Spectrum of a Transformer
Figure 2.8: Schematic View of Noise Generation and Transmission in a Power Transformer
Figure 2.9: Photo of Typical Transformer Core
Figure 2.10: Generation Mechanism of Aeolian Noise
2
Figure 2.11: Aeolian Noise Characteristics of an 8-Bundle LN-ACSR 960 mm Conductor
Figure 2.12: The Spiral Rod Method
Figure 2.13: Relationship Between Protrusion Height and Aeolian Noise Level in a single
Conductor System
Figure 3.1: Noise Monitoring Location on Golagh Tee - Letterkenny 110kV OHL
Figure 3.2: Noise Monitoring Location on C'Fall Golagh Tee - Letterkenny 110kV OHL (1)
Figure 3.3: Noise Monitoring Location on C'Fall Golagh Tee - Letterkenny 110kV OHL (2)
Figure 3.4: Noise Monitoring Location on C'Fall - Srananagh 110kV OHL
Figure 3.5: Noise Monitoring Location on Dunstown - Maynooth (Betaghstown) 220kV OHL
Figure 3.6: Noise Monitoring Location on Dunstown - Maynooth (Currabell) 220kV OHL
Figure 3.7: Noise Monitoring Location on Dunstown - Maynooth (Thomastown) 220kV OHL
Figure 3.8: Noise Monitoring Location on Oldstreet - Woodland (Ardrum Great) 400kV OHL
Figure 3.9: Noise Monitoring Locations at Dunfirth 110kV Substation
Figure 3.10: Noise Monitoring Locations at Gorman 220kV Substation
Figure 3.11: Noise Monitoring Locations at Woodland 400kV Substation
Figure 4.1: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night
(C'Fall Golagh Tee - Letterkenny 110kV OHL)

vi
Evidence Based Studies Noise – Main Report

Figure 4.2: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night
(C'Fall - Srananagh 110kV OHL)
Figure 4.3: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night at
Location 1 (Betaghstown) on the Dunstown-Maynooth 220kV OHL
Figure 4.4: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night at
Location 2 (Currabell) on the Dunstown-Maynooth 220kV OHL
Figure 4.5: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night at
Location 3 (Thomastown) on the Dunstown-Maynooth 220kV OHL
Figure 4.6: Comparison of All Background Noise Levels (LA90) for 'On' and 'Off' Surveys on the
Oldstreet-Woodland 400kV OHL (Rain & Wind Affected Data Included)
Figure 4.7: Comparison of All LA5 Noise Levels for 'On' and 'Off' Surveys on the Oldstreet-
Woodland 400kV OHL (Rain & Wind Affected Data Included)
Figure 4.8: Typical Spectral Profile of Noise Levels at Dunfirth 110kV Substation
Figure 4.9: Typical Spectral Profile of Noise Levels at Gorman 220kV Substation
Figure 4.10: Typical Spectral Profile of Noise Levels at Woodland 400kV Substation
Figure 5.1: Typical Dry 24-hour Variation in LAeq, 10min Level for All Locations
Figure 5.2: Distribution of LAeq, 1min Levels at the Control Location (Dry Conditions)
Figure 5.3: Distribution of LAeq, 1min Levels at the Mid-Span Location (Dry Conditions)
Figure 5.4: Distribution of LAeq, 1min Levels at the Tower Location (Dry Conditions)
Figure 5.5: Measured Noise Levels Against Rainfall Rate for All Sites
Figure 5.6: 1/24-Octave Band Analysis of Corona Noise at Mid-Span Locations

LIST OF TABLES
Table 2.1: Relationship between Rainfall Intensity and Measured AN in KEPCO Surveys
Table 2.2: Summary of Health Effects and Noise Threshold Limits
Table 4.1: Comparison of Averaged Data from 'On' and 'Off' Surveys at C’Fall Golagh Tee -
Letterkenny 110kV OHL (1)
Table 4.2: Comparison of Averaged Data from 'On' and 'Off' Surveys at C’Fall Golagh Tee -
Letterkenny 110kV OHL (2)
Table 4.3: Comparison of Averaged Data from 'On' and 'Off' Surveys at C’Fall - Srananagh
110kV OHL
Table 4.4: Comparison of Averaged Data from 'On' and 'Off' Surveys at Location1 (Betaghstown)
on the Dunstown-Maynooth 220kV OHL
Table 4.5: Comparison of Averaged Data from 'On' and 'Off' Surveys at Location 2 (Currabell) on
the Dunstown-Maynooth 220kV OHL
Table 4.6: Comparison of Averaged Data from 'On' and 'Off' Surveys at Location 3 (Thomastown)
on the Dunstown-Maynooth 220kV OHL
Table 4.7: Comparison of Averaged Data from 'On' and 'Off' Surveys at Ardrums Great on the
Oldstreet-Woodland 400kV OHL
Table 4.8: Noise Monitoring Data for Dunfirth 110kV Substation
Table 4.9: Noise Monitoring Data for Gorman 220kV Substation
Table 4.10: Noise Monitoring Data for Woodland 400kV Substation
Table 5.1: Summary of Measured Noise Levels Over 10-Week Period

vii
Evidence Based Studies Noise – Main Report

1 INTRODUCTION

1.1 THE SCOPE OF THIS PROJECT

In April 2012, EirGrid published the Grid25 Implementation Programme 2011-2016, and its associated
Strategic Environmental Assessment (SEA).

The SEA identified a number of Environmental Mitigation Measures expected to prevent, reduce and,
as fully as possible, offset any significant adverse impacts on the environment of implementing the
Implementation Programme.

Environmental Mitigation Measure (EMM) 3 concerns Preparation of Evidence-Based Environmental

Guidelines. These are intended to be a series of authoritative studies which look at the actual effects
of the construction and existence of transmission infrastructure in Ireland. The studies would provide
benchmarks to facilitate the robust preparation of projects with an evidence-based understanding of
likely environmental impact.

Three types of studies are envisaged under EMM3:-

• Environmental Benchmarking Studies: to determine the actual effect, in respect of a

number of environmental topics, of the construction and existence of transmission projects in
a representative range of Irish environmental conditions – typical, non-standard, and worst-
case. The studies, while authoritative, are conceived as an ongoing body of work that can be
continuously updated to take account of new information and/or developments in
understanding arising from practice and research;

• Evidence-based Environmental Design Guidelines: deriving from the factual basis and
evidence contained in the initial benchmarking studies, these will provide practical guidance to
practitioners and consultants in the planning and design of transmission infrastructure from the
perspective of a particular environmental topic. These might comprise new guidelines, or the
updating of existing guidelines;

• Guidelines on EIA for Transmission Projects in Ireland: Accompanying, or incorporated

into the design guidelines, these are intended to provide an agreed and authoritative format
for the preparation of Environmental Impact Assessment (EIA) for transmission projects in
Ireland, in respect of particular environmental topics.

This study is one of the Environmental Benchmarking Studies – to determine the actual effect of the
construction and existence of transmission infrastructure in Ireland on its environment.

1
Evidence Based Studies Noise – Main Report

1.2 THE SCOPE OF THIS STUDY

The key aim of this study is to examine and benchmark the actual noise effects of the construction and
operation of high-voltage transmission projects in Ireland.

A review of available documentation and literature provides the basis for an assessment of noise in
the context of the development of the transmission network.

Consideration was given to assessing the noise impact associated with the construction of electrical
infrastructural projects; however this was scoped out, as it was considered that such noise would be
consistent with any construction project, rather than specific to transmission infrastructure. As such,
this study concentrates on the noise impacts associated with existing operational infrastructure.

Two separate survey strategies have been adopted for the purposes of investigating noise from
overhead lines and noise from substations. For overhead lines, long term noise surveys have been
completed adjacent to various infrastructural types (i.e. 110 kV, 200 kV and 400 kV) during time
periods when the infrastructure is live (referred to as 'On' surveys in this report) and during periods of
outage when the infrastructure is subject to maintenance work (referred to as 'Off' surveys in this
report). By making direct comparison over a large sample period (e.g. 160 hours) and using an array
of relevant noise parameters (i.e. LAeq, LA50, LA5, LA90), the effect of corona, insulator or any other noise
sources associated with overhead lines should become apparent as increased noise levels in the 'On'
surveys.

In order to reduce the variability of factors influencing measured noise levels to enable more
meaningful comparisons to be made, hundreds of datasets were created as part of the analytical
process to isolate measured noise levels in different time periods (e.g. weekday, night, quiet day etc)
and under different weather conditions (e.g. rain, no rain, wind below 10mph, wind below 5mph, wind
= 0mph etc.). This was done for each noise parameter at each location for 'On' and 'Off' surveys.

In the case of the substation surveys, long term noise surveying was not adopted as it was not
possible to complete surveys around the outage programme. While individual items of plant within a
substation were subject to outage, an entire substation was not, and hence a direct comparison
between 'On' and 'Off' conditions was not feasible.

The substation surveys were completed with a view to determining the typical noise levels associated
with substations of each type (e.g. 110kV, 220kV and 400kV). It was difficult to incorporate sporadic
effects such as corona into such a survey strategy, as the onset of these conditions cannot readily be
predicted or planned for. Without an option for complete outage at substation sites, it is considered
that a long term survey strategy could not isolate and illustrate corona effects from substations.

On the basis of the practical considerations discussed above, random and sporadic effects such as
corona noise could not be fully explored for substation sites as part of the evidence based studies.

2
Evidence Based Studies Noise – Main Report

Underground electrical cables did not form a part of the current study as they are not considered to be
a noise source. The soil covering any underground cable acts as a substantial insulator absorbing any
potential noise energy produced by underground cables and preventing any significant emission of
noise energy above ground.

In addition to the survey work completed as part of this study, EirGrid separately commissioned a long
term noise survey at the existing Dunstown-Moneypoint 400kV overhead transmission line in Cloney,
Co. Kildare. This work was completed by AECOM and supplements the work completed as part of
this study, feeding into the overall discussion, conclusions and recommendations.

1.3 THE TRANSMISSION NETWORK AND NOISE

Electricity supply is an essential service in Ireland’s economy. The transmission system is a meshed
network of 400kV, 220kV and 110kV high voltage circuits, with 156 high voltage substations (see
1
Figure 1.1). The transmission system therefore plays a vital role in the supply of electricity . The
development of the transmission network is the responsibility of EirGrid, the Transmission System
2
Operator (TSO), under statutory instrument SI445/2000 .

Grid development requires a careful balance between meeting the technical requirement for a project,
the costs of that project, and the environmental impact of that project.

ESB, as the Transmission Asset Owner (TAO), is charged with constructing the transmission assets
as specified by the TSO. ESB also has the role of Distribution System Operator (DSO) with which the
TSO coordinates planning and development requirements. An overview of the primary types of
transmission infrastructure, including an outline of construction methodology is set out in Appendix A.

EirGrid is committed to ensuring that transmission infrastructure development is undertaken in an

environmentally sensitive manner, including in terms of impact from noise. The significance of any
adverse effects of transmission infrastructure development depends on the location and scale of the
proposed infrastructure and the potential for mitigation measures. This is why transmission
infrastructure development is undertaken and constantly reviewed by suitably qualified experts as the
design of a scheme progresses.

1
Transmission Development Plan 2008-2012 EirGrid

2
SI445/2000, entitled European Communities (Internal Market in Electricity Regulations, 2000)

3
Evidence Based Studies Noise – Main Report

Figure 1.1 EirGrid and SONI’s Transmission System Map (January 2015)

There are a number of ways in which noise can be generated from electricity infrastructure. Generally,
these fall within four categories of noise, namely:

• audible noise associated with Corona Discharge from high voltage transmission lines;

• audible noise associated with dirty, damaged or cracked insulators;

• audible noise associated substation equipment such as transformers, quadrature boosters and
mechanically switched capacitors;

• audible noise associated with wind blowing through electricity infrastructure - Aeolian Noise.

4
Evidence Based Studies Noise – Main Report

1.4 STUDY LAYOUT

Chapter 2 sets out a literature review relating to the noise effects of transmission infrastructure
development. At the outset, it is clear that Aeolian Noise is not as common as Corona Noise, as the
conditions that give rise to it (high winds and a very specific angle of incidence) do not occur regularly.
Furthermore, such conditions that give rise to Aeolian Noise also assist in masking it (Cigre, 2009). In
addition, the literature evidence does not provide any standard method for measuring and analysing
Aeolian Noise from existing infrastructure - any work in this area has been by means of wind tunnel
experiments (Akagi et al, 1998). On account of the practical difficulties associated with measuring
Aeolian Noise and the fact that Aeolian Noise has generally not been perceived of as presenting a
very significant noise impact at sensitive receptors (e.g. Cigre, 2009, DECC, 2011), it does not form a
significant part of the evidence based studies included in this report.

Chapter 3 sets out the methodology undertaken for site survey of noise. As noted above, two separate
survey strategies have been adopted as part of the evidence based surveys for noise for the purposes
of investigating noise from overhead lines and noise from substations. Chapter 4 sets out the survey
results for 110 kV, 220 kV and 400 kV overhead lines (underground cables having been scoped out of
this study), and existing operation substations (substations under construction also having been
scoped out of the study).

Chapter 5 discusses the separate AECOM survey work conducted at a location along the existing
400kV Dunstown-Moneypoint overhead transmission line in Cloney, Co. Kildare (also refer to
Appendix B). Finally Chapter 6 sets out a discussion and conclusions in respect to the actual noise
effects of transmission infrastructure development, and Chapter 7 sets out recommendations.

5
Evidence Based Studies Noise – Main Report

2 LITERATURE REVIEW

2.1 INTRODUCTION

2.1.1 Scope and Aims

There are a number of types of noise associated with high voltage electricity transmission
infrastructure and these have been summarised in Chapter 1. The aim of this review is to examine
literature relating to the topic of noise in the context of electricity infrastructure, and to summarise the
evidence base under a series of headings that cover the different aspects of this topic.

There is a difference in the type and character of noise that is generated by different types of
transmission infrastructure (overhead lines, underground cables and substations). This is further
complicated by the strength of the voltage passing through the infrastructure. As noted in Chapter 1,
underground cables are not identified as being a significant source of noise, as the ground covering
the cable acts as an insulator absorbing a large proportion of the noise energy released prior to it
propagating above ground. As such, it is not considered further in this literature review.

On account of the difference in the noise associated with overhead lines and substations, the noise
associated with these two categories of infrastructure is discussed under separate headings (Section
2.2 and 2.3 below). These headings are supplemented with evidence under a number of other
headings that relate to the overall topic of noise, including measurement of noise associated with
electrical infrastructure and mitigation measures for noise associated with electrical infrastructure.

2.1.2 Noise Fundamentals

Sound is a sensation detected by the ear as a result of pressure variations set up in the air by a
vibrating source. Noise is essentially unwanted sound or sound that is not desired by the recipient.
Any sound that has the potential to cause disturbance, discomfort or psychological stress to a subject
exposed to it, or any sound that could cause actual physiological harm to a subject exposed to it, or
physical damage to any structure exposed to it, is known as noise (Environment Agency, 2004).

On account of the human ear's sensitivity to a wide range of fluctuations in pressure levels, sound is
typically measured in terms of a logarithmic ratio of sound pressures. These values are expressed as
sound pressure levels (SPL) in decibels (dB). In terms of sound pressure levels, audible sound ranges
from 0dB (the threshold of hearing) to approximately 120dB (the threshold of pain). A doubling/halving
of pressure equates to a 3dB increase/decrease in decibel level, which under normal circumstances is
the smallest change in noise level that is noticeable to the human ear. A 10dB increase/decrease in
sound level normally equates to a subjective doubling/halving of noise (EPA, 2012).

6
Evidence Based Studies Noise – Main Report

For a point source such as the fan of a chimney stack, the sound energy spreads out spherically, so
that the SPL is the same for all points at the same distance from the source. The SPL decreases from
a point source by 6dB per doubling of distance until ground and air attenuation noticeably affect the
level. Items of plant within a substation are point sources. An overhead line is a line source, whereby
the sound level spreads out cylindrically so the SPL is the same at all points at the same distance from
the line. The SPL decreases from a line source by 3dB per doubling of distance until ground and air
attenuation noticeably affect the level.

The frequency of a particular sound wave is the rate at which the sound wave oscillates. Frequency is
measured in Hertz (Hz). Human hearing is not uniform across the frequency range, being less
sensitive at very low and very high frequencies. In order for measured sound to relate more closely to
the human hearing, a weighting (i.e. A-weighting) is commonly applied (EPA, 2012). Examples of
some common sounds and their decibel ratings are listed below (Scottish Government, 2011):

• Unsilenced pneumatic drill (at 7m distance) - 95dB(A);

• Heavy diesel lorry (40km/h at 7m distance) - 83dB(A);

• Modern twin-engine jet (at take-off at 152m distance) - 81dB(A);

• Passenger car (60km/h at 7m distance) - 70dB(A)

• Office environment - 60dB(A);

• Ordinary conversation - 50dB(A);

• Quiet bedroom - 35dB(A).

2.1.3 Audible Noise from Corona Discharge

Audible noise associated with high voltage transmission lines usually occurs when the conductor
surface electric stress exceeds the inception level for corona discharge activity, resulting in the release
of acoustic energy which radiates into the air as sound. Conductors are designed to operate below
the inception level for corona discharge; however surface contamination or accidental damage to the
conductor can cause local enhancement of electrical stress, leading to the discharge activity and
subsequent generation of noise (DECC, 2011).

Corona discharges are most often associated with rainy conditions, where water droplets collecting on
the surface of the conductor initiate the discharge activity. The audible noise levels experienced are
directly related to the level of rainfall - the higher the rainfall the higher the noise level. Fog will also
give rise to corona noise, however the noise levels are generally lower than those experienced during
rainfall. The build up of contamination or surface grease on conductors can give rise to or exacerbate
corona discharge noise.

7
Evidence Based Studies Noise – Main Report

Corona discharges are perceivable as a broadband crackling and hissing noise in the frequency range
of one to several kilohertz (Straumann and Fan, 2009).

2.1.4 Insulator Noise

Insulator noise is similar to corona noise with the one exception that it is not dependent on weather
conditions. Insulator noise is caused by dirty, damaged or cracked insulators and is generally a
feature of older ceramic or glass insulators. New polymer insulators have been found to minimise this
type of noise emission from high voltage transmission systems. No specific literature was identified
that explores any aspects of insulator noise and hence there is no specific section on insulator noise in
this report. Much of the discussion on line transmission corona may be related to insulator noise also.

2.1.5 Substation Noise

Audible noise can be generated by substation equipment such as transformers, quadrature boosters
and mechanically switched capacitors. Transformers typically generate a low frequency humming
noise, the extent of which depends on the transformer type and the level of nose attenuation at the
substation (DECC, 2011).

Transformer hum is the predominant noise generated at electricity substations and is associated with
magnetic and electrical forces within the core of an electrical transformer. These forces generate
vibrations in the core laminations within the transformer which generates noise. Typically, the noise
level does not vary with transformer load as the core is magnetically saturated and cannot produce
variations in the amplitude of the noise.

On account of the typically tonal nature of noise from substation plant, spectral analysis is commonly
used to determine if a particular hum is due to noise at a distinct frequency.

Generally, modern transformers are manufactured with a specified and guaranteed emission level.
Improvements in the manufacture of transformers have reduced the associated level of noise emission
and hence modern transformers are typically quieter than equivalent capacity older transformers.

2.1.6 Aeolian Noise

The effect of wind blowing through electricity plant and its supporting structures can result in two types
of noise - a general broadband turbulence noise and Aeolian noise. The general turbulence noise is
no different for electricity infrastructure than it is for a wide range of physical structures in the
environment and therefore this type of noise is not considered a nuisance. Aeolian noise,
characterised by a series of tones and whistles that vary in frequency with the wind speed, is caused
by vortex shedding (regular air fluctuations) across the surface of the item of infrastructure. This form
of noise is independent of the infrastructure being energised or not.

8
Evidence Based Studies Noise – Main Report

Aeolian noise may become problematic at wind speeds higher than 10m/s; however it is often masked
by the noise of the rain or by the wind itself (Cigré, 1999). Aeolian noise effects can be reduced by a
range of infrastructure design techniques such as reducing the number of sub-conductors, increasing
the spacing between sub-conductors, spiral wire wrapping and using composite insulators instead of
glass or porcelain ones.

In general, Aeolian noise is not as noticeable as corona noise as the conditions that give rise to it (i.e.
high winds at a specific angle of incidence) also help to mask its effects. It is also not as common as
corona noise, as the conditions required to give rise to Aeolian noise do not occur regularly.

2.2 AUDIBLE NOISE FROM OVERHEAD TRANSMISSION LINES

2.2.1 Introduction

Corona from overhead transmission lines (OHL) refers to the discharge phenomenon produced by the
ionisation of the air around the conductors, when the conductor surface potential gradient exceeds a
critical value, namely, the corona onset gradient (Zhang et al, 2009). The corona characteristics of DC
OHL are different from those of AC OHL mainly on account of the magnitude of environmental space
charge arising from ionisation in the space between the two conductors of the lines and between each
conductor and the ground.

It is generally recognised that audible noise from AC OHL is a concern in foul weather only, principally
in rain. In the case of DC OHL, it is generally recognised that if there is any concern with audible
noise, it is during fair weather (Task Force of the Corona and Field Effects Subcommittee, 1982) The
cause of strong DC corona in fair weather is pollution on the conductors due to airborne particles and
insects, which collect much more in the case of DC by being continuously attracted to the
corresponding pole (Gela et al, 1993). As the Irish National Grid uses AC OHL, all of the background
literature evidence outlined in this report relates to AC OHL.

There is no single standard guidance document that details the background, the requirements for
measuring, assessing or predicting audible noise or that outlines mitigation measures for addressing
audible noise on OHL. The consensus of authoritative evidence on the subject is derived from
dedicated sub-sections within regional policy guidance documents on electricity infrastructure (DECC,
2011, Cigré, 2009) and international research papers

2.2.2 Measurement of Audible Noise from Overhead Transmission Lines

Corona noise is usually measured at the edge of the right away, or under the OHL as appropriate
(Cigré, 2009). It can also be measured along a lateral profile. As one moves away from an OHL, the
audible noise decreases approximately in inverse proportion to the square root of the distance, i.e. by
approximately 3dB as the distance from the power line doubles (Cigré, 2009).

9
Evidence Based Studies Noise – Main Report

An example of a lateral profile of noise, averaged for the meteorological variation over one year, is
illustrated in Figure 2.1. This profile was measured at 1.5m above ground, under a 400kV OHL (double
circuit vertical configuration) with twin bundles of conductors of diameter 38mm and with distance
between sub conductor of 40cm (Cigré, 2009).

Figure 2.1: Typical Lateral Profile of Noise from a 400kV Transmission Line (Cigré, 2009)

Many noise measurements show evidence of higher noise levels near towers (especially near angle
towers) than at mid-span on the transmission line. These higher noise levels are due to the presence
of insulators and fittings on the towers, where corona discharges can occur when these items of
equipment are dirty, wet or damaged (Cigré, 2009).

Corona discharge occurs on all types of OHL, but it becomes more noticeable at higher voltage
(approximately 350kV and higher). Under fair weather conditions, audible noise from corona is minor
and rarely noticed (Wordpress, 2011). On account of the fact that audible noise from corona discharge
is more of an issue at higher voltage, much of the evidence base relates to measurement data from
AC OHL at higher voltages (e.g. Zhang et al, 2009; Al-Faraj et al, 1997; Straumann and Fan, 2009;
Task Force of the Radio Noise an Corona Subcommittee of the Transmission and Distribution
Committee, 1975). There is literature that explores the measurement of audible noise from corona
discharge at voltage levels between approximately 170 and 350kV (e.g. Chartier et al, 1995; Muhr et
al, 2004), which is relevant to the OHL transmission network in Ireland (110kV, 220kV and 400kV).

In a review of previous measurement studies, Al-Faraj et al (1997) presented data from surveys
completed by the Korean Electric Power Corporation (KEPCO) on the audible noise performance of a
6-rail conductor on a 765kV test line. The survey included experimentation with a number of different
conductor bundle arrangements to determine which set up ensured the audible noise criteria of

10
Evidence Based Studies Noise – Main Report

50dB(A) for environmental protection of residential amenity was met. A 6-rail bundle conductor set up
achieved an audible noise of 48.8dB(A) in rain at 15m from the outermost phase, compared with a
measurement in fair weather of 42.1dB(A). The relationship between audible noise and voltage was
also explored, demonstrating that an increase in voltage from 630kV to 800kV resulted in an increase
in noise levels of 5.2dB(A).

Chartier et al (1995) completed long term measurements on an operating Puget Power 230kV line in
USA in order to confirm the accuracy of the Bonneville Power Administration (BPA) audible noise
calculation method. While the research team experienced partial contamination of the measurement
data due to background noise at various stages of the survey, they were able to eliminate
contaminated data and analyse the remaining data. Based on the analysis of the good data, the L50
levels during stable rain were 45.5 and 41dB(A) at 15m and 32.2m respectively.

The literature research detailed above provides good background information for the purposes of
assessing the measured noise levels from the EirGrid study. Of particular interest to this study is the
fact that the literature indicates that corona noise only becomes a noticeable phenomenon at
approximately 350kV and above. The measured noise levels in the Chartier et al (1995) study appear
to indicate a measureable difference in noise levels form a 230kV line; however these measurements
are in the low 40s dB(A). The noise profile drawing shown in Figure 2.1 shows similarly low noise
levels peaking in the mid 40s dB(A) for a 400kV overhead transmission line. Such low noise levels
could easily be masked by other extraneous noise sources in a study area.

2.2.3 Factors Influencing Corona Discharge

The amount of corona produced by an OHL is a function of the voltage of the line, the diameter of the
conductors, the locations of the conductors, the locations of the conductors in relation to each other,
the elevation of the line above sea level, the condition of the conductors and hardware, and the local
weather conditions. Power flow does not affect the amount of corona produced by an OHL
(Wordpress, 2011).

Irregularities (such as nicks and scrapes on the conductor surface or sharp edges on suspension
hardware) concentrate the electric field at these locations and thus increase the electric field gradient
and the resulting corona at these locations. Similarly, foreign objects on the conductor surface (e.g.
dust, insects etc.) can cause irregularities on the surface, thus giving rise to corona (Wordpress, 2011)

Corona also increases at higher elevations where the density of the atmosphere is less than at sea
level. An increase in 1,000 feet of elevation will result in an increase in audible noise of approximately
1dB(A). Raindrops, snow, fog, frost and condensation accumulated on the conductor surface are also
sources of surface irregularities that can increase corona, therefore during wet weather the corona
effects are greater (Cigré, 2009; Wordpress, 2011). On account of the significant effect of rainfall on
corona discharge, it is discussed in greater detail in Section 2.2.4.

11
Evidence Based Studies Noise – Main Report

Corona discharge has an inverse relationship with air density and with humidity at power frequencies
(i.e. the frequency of oscillation of alternating current in an electric power grid). Corona will generally
be greater on new conductors and will decrease to a steady-state value over a period of approximately
one year in service (Wordpress, 2011). The trend towards compacting lines has a negative effect on
audible noise, therefore there is a balance to be found between the positive and negative effects of
compaction (Chartier et al, 1995: Cigre, 2009).

2.2.4 The Effect of Rainfall on Corona Discharge

Water droplets are elongated in the electrical field due to their dielectric behaviour. With the
deformation of a water droplet, the electric field strength is increased again. Increase in the elongation
of the water droplet to a certain extent leads to the formation of what is referred to as a Taylor-cone
(Straumann and Fan, 2009). The subsequent instability created by the formation of the Taylor-cone
results in a water jet being ejected from the tip of the water drop. With the excessive field strength in
this process, corona discharges of various types may set in.

In the event of a corona-discharge occurring, two clear regions can be observed: the ionisation region
and the drift region. Figure 2.2 illustrates the location of these two regions in a schematic
representation of the corona discharge.

Figure 2.2: Schematic Representation of Corona Discharge (Straumann and Fan, 2009)

12
Evidence Based Studies Noise – Main Report

Negative ions are formed in the drift region, which drift together with the positive ions in the electric
field E. Starting from the conductor, ions can drift over distances of several decimeters during a half-
wave. Within the drift zone, there is a build-up of collisions between the ions and the neutral gas. The
movements and collisions of ions with neutral gas within the drift zone creates energies and forces
that can give rise to the sound emissions typical of corona discharge.

The effect of rainfall on the measured audible noise can be clearly illustrated by the data retrieved
from the KEPCO surveys discussed in Section 2.2.2 of this report. Table 2.1 presents the correlation
between the rainfall intensity and the measured AN (Al-Faraj et al, 1997).

Table 2.1: Relationship between Rainfall Intensity and Measured AN in KEPCO Surveys (Al-
Faraj, 1997)

Rainfall Intensity
0.1 0.5 0.9 2.3 6.6 31.9
(mm/hour)

Audible Noise [dB(A)] 41.3 43.4 44.2 44.9 48.1 52.0

The study concluded that at higher rainfall intensities, the nuisance effect of corona noise for residents
living near the line will become less as the noise from corona discharge is masked by the noise
created by the falling rain. Measurement data from the study indicates that audible noise reaches a
saturation point at a rain intensity of approximately 30mm/hr.

Measurement of audible noise from a Puget Power 230kV compact line between Sedro Woolley and
March Point substations in the USA demonstrated a relatively small difference in the measured L50
between heavy rain conditions and wet conductor conditions, a conclusion which contradicted some of
the prediction models in standard use (e.g. EPRI, HVTRC) (Chartier et al, 1995).

2.2.5 Character of Audible Noise from Corona Discharge

Audible noise from corona discharge has two different character components: broadband noise which
is subjectively observed as a frying, cracking or hissing noise; and tonal noise which contains pure
tone components at a number frequencies (generally 120Hz and multiples). The pure tone component
can be discerned as a noticeable 'hum' which is superimposed on the broadband noise. The
broadband noise is caused by a random sequence of pulses produced by partial discharge in the area
at the surface of the conductor (Al-Faraj et al, 1997).

While much of the literature evidence on the topic of audible noise from corona discharge relates to
broadband noise, some research has been undertaken on the tonal aspects of audible noise from
corona discharge. Straumann and Fan (2009) completed a theoretical comparison of broadband noise
and tonal components from corona audible noise. The basis for their work was the fact that a
significant tonal component from corona audible noise (i.e. the 2f component, which is twice the main
frequency) had not been adequately investigated in previous studies which concentrated on

13
Evidence Based Studies Noise – Main Report

broadband noise. Standard assessment criteria throughout the world use the A-Weighting noise level
for assessing broadband noise as it reflects the sensitivities of the human ear. However, the use of
the A-weighting scale results in a much reducing representation of the 2f component in the overall
noise level (i.e. it is reduced by 19dB at 100Hz due to the use of A-Weighting).

Despite its reduced representation in the A-weighting, the 2f emission has the potential to present a
significant problem as low frequency noise is less likely to be attenuated by building structures and
tonal noise is in general perceived of as being annoying. This latter fact is highlighted by the fact that
most national regulations add an additional noise penalty where tonal noise is involved (e.g. 5dB
penalty in Irish NG4 and UK BS4142:1997 guidance documents) (EPA, 2012; BS4142:1997).

Figure 2.3 shows the lateral profile near the ground of the A-weighted tonal component (100Hz) for the
line design investigated in the study with new conductor cables and under moderate rain (Straumann
and Fan, 2009). Although the effect of the ions on the gas is small, it is still high enough to evoke a
significant tonal emission. The levels included in Figure 2.3 were calculated for new cables, a
substantial reduction in this cables is expected with the ageing of the conductors.

Figure 2.3 Lateral Profile of A-weighted Tonal Component (100Hz) (Straumann and Fan, 2009)

Muhr et al (2004) used a narrowband analysis method to analyse the acoustic spectrum of dry and wet
conductors as opposed to the usual octave and one-third octave analyses. Energised overhead lines
with high AC voltages show a typical acoustic pattern, whereby a double frequency of the supply
voltage is noticeable as a 'hum' which is supplemented by broadband noise at the upper frequencies.

14
Evidence Based Studies Noise – Main Report

The 'hum' effect changes if the line is affected by rain, snow or fog. Figure 2.4 displays a typical
frequency spectrum of a transmission line audible noise wet with 40, 50 and 60Hz supply voltage.

Figure 2.4: Typical Frequency Spectrum of Transmission Line AN Wet with 40, 50 and 60Hz
Supply Voltage (Muhr et al, 2004)

Figure 2.5 shows the discrete frequency spectrum of a wet overhead transmission line at a voltage
from 100kV to 320kV in 20kV steps, the frequency of the supply voltage being 50Hz. The figure
illustrates that the even numbered harmonics show a higher sound pressure level than non even
numbered harmonics if this spectrum is compared to the equivalent spectrum under dry conditions
(see Figure 2.6).

15
Evidence Based Studies Noise – Main Report

Figure 2.5: Discrete Frequency Pattern of a Wet Overhead Transmission Line (Muhr et al, 2004)

Figure 2.6: Discrete Frequency Pattern of a Dry Overhead Transmission Line (Muhr et al, 2004)

16
Evidence Based Studies Noise – Main Report

2.2.6 Methods for Calculating Audible Noise of High Voltage Transmission Lines

In 1982, subsequent to the emergence of audible noise as a critical design consideration for high
voltage transmission lines in USA, the Task Force of the Corona and Field Effects Subcommittee (a
subcommittee of the IEEE Transmission and Distribution Committee) published a paper comparing the
different methods for calculating audible noise from AC and DC lines. A total of nine methods for AC
lines and four methods for DC lines were applied to 4 different bipolar line geometries.

It was not the intention of the paper to recommend a particular calculation method over any other but
to provide an objective comparison between methods available at that time. The general conclusion of
the paper was that most AC calculation methods give acceptable results for transmission line voltages
up to 765kV and a number of conductors per phase of 4 or less. For higher voltages and greater
numbers of sub-conductors, fewer methods appear reliable.

More recent use of calculation methods and comparison with measured noise levels has consolidated
the view that there is generally good agreement achieved with the standard AC methods. For
example, Chartier et al (1995) compared the Electrical Power Research Institute (EPRI) and
Bonneville Power administration (BPA) methods and found that predicted L50 audible noise levels
using both methods, were in good agreement with the measured audible noise levels.

2.2.7 Mitigation Measures for Corona Discharge Audible Noise

There are many ways in which corona discharge can be reduced or avoided. One way is by
minimising the voltage stress and electrical field gradient, which can be achieved by maximising the
distance between conductors that have large voltage differentials, using conductors with large radii
and avoiding parts that have sharp points or sharp edges (Wordpress, 2011).

Corona inception voltage can in some instances be increased by using a surface treatment such as
semiconductor layer, high voltage putty or corona dope. Usage of a good homogenous insulator, such
as a prepared silicone and epoxy potting material, can work well also.

When one is limited to using air as your insulator, geometry is the key factor. In such instances, steps
should be taken to reduce or eliminate unwanted voltage transients, which can initiate corona. Using
multiple conductors per phase helps reduce resistance and hence corona loss. The use of corona
rings help to distribute charge across a wider area, thereby reducing the electric field and the resulting
corona discharge.

For the 345-400kV voltage level, bundles of two to three conductors are generally considered
sufficient. For voltages exceeding this level, the number of sub-conductors increase and levels of
765kV may require bundles of up to 6 sub-conductors to meet the noise design requirements (Cigré,
2009).

17
Evidence Based Studies Noise – Main Report

Different types of corona noise countermeasures are being tested in Japan on very high tension
overhead lines. One of the most promising methods was the artificial ageing of the conductors by
sandblasting and coating with a thin porous boehmite film. Measurements have shown that in the
case of new conductors, corona noise levels equal or are even lower than those of equivalent
conductors whose surface has become adequately saturable (Cigré, 2009).

Straumann and Weber (2010) have more recently demonstrated that hydrophilic coating of new
conductors can cause substantial reduction of audible noise. Hydrophilic coating of well aged
conductors was found to be superfluous, as ageing of conductors results in a comparable effect to the
hydrophilic coating process. The paper concluded that hydrophilic coating produced the optimal
hydrophilicity, but did not explore if hydrophilic coating produced a superior noise reduction in
comparison to sandblasting.

2.2.8 Discussion

From reviewing the literature, it is clear that the level of impact likely from electricity transmission lines
increases with the increase of the voltage strength of the line. Much of the literature (e.g. Zhang et al,
2009; Al-Faraj et al, 1997; Straumann and Fan, 2009; Task Force of the Radio Noise an Corona
Subcommittee of the Transmission and Distribution Committee, 1975) indicates that corona noise only
becomes a significant issue from 350-500kV and above. In terms of this study, this would suggest that
significant corona noise impacts may not be likely for 110kV and 220kV transmission lines and that the
potential for more significant corona noise impacts may only relate to the 400kV lines. The field
surveys set out in Chapter 4 and 5 of this study address this in some detail.

As outlined in Section 2.2.2, many measurement surveys of transmission lines have used the lateral
profile method of survey (e.g. Figure 2.1, Cigré, 2009). This was not the approach used in the field
surveys for this study, which used a single monitoring location under the transmission lines but with
surveys completed both with the line energised and during outage. The lateral profile is a method that
can be used when the line is energised and there is no prospect of an outage for the line.

By completing the lateral profile, with distance the noise levels drop to the ambient noise level on
either side of the line. As there was the option to complete surveys during outage for this study, this
methodology gave an attractive opportunity to make direct comparisons of noise levels under the line
during periods of energised and outage at the same location. The predominant use of the L50 and L5
parameters for assessing corona noise was noted in the literature and are incorporated into the
analysis of the field survey contained in the subsequent sections of this study.

18
Evidence Based Studies Noise – Main Report

2.3 AUDIBLE NOISE FROM SUBSTATIONS

2.3.1 Introduction

Audible noise effects can arise from a number of substation equipment such as transformers,
quadrature boosters and mechanically switched capacitors (Cigré, 2009). Transformer acoustic noise
is a hum characterised by special spikes at harmonics of the fundamental frequency (100Hz/120Hz)
which is twice the line supply frequency. The transformer's low frequency tonal noise components are
the most likely source of annoyance to nearby residents from substation noise (Masti et al, 2004).
Figure 2.7 illustrates a typical noise spectrum of a standard transformer (Chang et al, 2009).

Figure 2.7: Typical Noise Spectrum of a Transformer (Chang et al, 2009)

2.3.2 Source of Noise in Substations

2.3.2.1 Transformers

The primary source of acoustic noise generation in a transformer is the periodic mechanical
deformation of the transformer core and the winding coils, under the influence of fluctuating
electromagnetic flux associated with these parts (Masti et al, 2004). The physical phenomena
associated with this tonal noise generation can be classified as follows:

• The material of a transformer core exhibits magnetostrictive properties. The vibration of the core
is due to its magnetostrictive strain varying at twice the frequency of the alternating magnetic flux.
The frequencies of the magnetic flux is equal to the power system supply frequency and its
harmonics.

• When there are residual gaps between laminations of the core, the periodic magneto-motive force
may cause the core laminations to strike against each other and produce noise. Also, the periodic
mutual forces between the current-carrying coil windings can induce vibrations if there are any
loose turns of the coil.

19
Evidence Based Studies Noise – Main Report

During a transformer's operation, the vibrations from its core and windings get transmitted to the
transformer tank surface through transmission through the air surrounding the core. Additional
vibrations can be transferred through the structure at the points where the core mount is attached to
the tank. The vibrating tank also radiates noise into the exterior air. Taken as a whole, there is a
complex set of vibro-acoustic interactions that occur during the standard operation of the transformer
and this is presented in a schematic format in Figure 2.8.

Figure 2.8: Schematic View of Noise Generation and Transmission in a Power Transformer
(Masti et al, 2004)

Figure 2.9 is photograph of a typical transformer core (400kVA). The yokes and the legs of the core
have stepped cross-sectional areas formed by a stacked arrangement of thin laminations. Although
clamped, there is scope for relative in-place motion over the remaining interface areas. Laminations
do not always have good matching flat surfaces and therefore residual gaps between the laminations
can occur. Magneto-motive forces acting across these air gaps can set up relative transfer motions
between the laminations.

20
Evidence Based Studies Noise – Main Report

Figure 2.9: Photo of Typical Transformer Core (Masti et al, 2004)

2.3.2.2 Other Sources of Noise Within Substations

There is very limited discussion in the literature relating to other specific noise sources from
substations other than transformers. One reference has indicated that audible noise effects can arise
from quadrature boosters and mechanically switched capacitors, however this is not elaborated upon
further (Cigre, 2009). Alternatively, Medeiros and Kroeff (1998) state that there is strong evidence that
the only relevant noise sources are the power transformers and associated cooling systems. They
state that other sources exist such as infrequent events (e.g. switching operations) of low intensity
events (e.g. corona), but these are masked by the transformer noise.

2.3.3 Noise from Substations

There is limited literature available which illustrate direct measurements of audible noise (broadband
or narrowband) from substations. Chang et al (2009) completed broadband noise measurements of
low-noise power transformer design at manufacture stage and in-situ in a location in Austria. The
design noise testing which was completed in a noise-isolated room (i.e. background 32dB) and
resulted in a noise level of less than 48dB(A) with the transformer operating with inductance of 1.3(T).
The exact distance between transformer and measuring point is not stipulated, however it is assumed
to be short range. The value of the outdoor measurements completed in-situ are of unknown value as
there appears to be significant contamination from other noise sources (especially road noise) in the
measured results.

Medeiros and Kroeff (1998) explore aspects of substation noise, including the main considerations for
the acoustic modelling of substation noise. Sample models of a typical substation are illustrated,
showing predicted noise levels of over 60dB approximately 20 metres from the substation. No
substation noise measurements are included in the paper.

21
Evidence Based Studies Noise – Main Report

Masti et al (2004) completed detailed experimentation on the core of the substation transformer using
various lamination settings and clamping pressures and the effects the alternative set ups would have
on the fundamental frequency. The ultimate objective of the paper was to present the optimum
arrangement of lamination and clamp settings to produce the lowest noise output from the transformer.
At this stage of development in the project, no noise monitoring had been completed.

2.3.4 Effect of Impulsive Substation Noise on Wireless Communications Systems

A number of papers have reviewed the character of substation impulsive noise and the possibility that
impulsive noise from substations has the potential to degrade the performance and reliability of
wireless communications systems (Bhatti et al, 2009; Shan et al, 2009a; Shan et al, 2009b; Shan et al,
2009c). Typical wireless transceiver designs are based on assumptions that noise is additive, white
and Gaussian. These transceivers perform fine in normal environments but their applicability in noise
intensive electricity substation environment is not fully understood (Bhatti et al, 2009).

Bhatti et al (2009) demonstrated that narrowband impulsive noise is more benign than additive white
Gaussian noise in both moderately and highly impulsive environments (i.e. environments where
impulsive noise events are common, e.g. electrical substations) for low signal to noise ratio (SNR)
values but for high SNR values, it substantially degrades the bit error rate (BER) of both IEEE 802.11b
and IEEE 802.11a WLAN receivers.

Shan et al (2009c) developed a system for the measurement of impulsive noise covering a total
frequency range from 716MHz to 5GHz. The system was deployed in a 400kV electricity transmission
substation for the characterisation of the substation noise environment at frequencies higher than has
previously been attempted. The paper includes an experiment to determine the degrading effect of
substation noise on the performance of ZigBee technology, concluding that there was no significant
adverse impact on the technology.

2.3.5 Mitigation Measures for Substations

As outlined in Section 2.3.3, Masti et al (2004) experimented with various lamination and clamping set
ups to derive the optimal set up of the transformer core. They illustrated how the dynamic interaction
of the core structure with the surrounding fluid and the core mounting can have significant impacts on
the overall generation of noise from the transformer. This paper clearly demonstrated the significant
potential for noise reduction at source with transformers.

Belardo et al (2006) examined a number of different noise control measures for power transformers.
The first part of the paper describes the methods and results of using dynamic vibration absorbers
(DA) for the purposes of noise reduction. The principle is to transfer the vibration from the transformer
panel to the DA, which has minimal acoustic coupling with the air. The results indicated that DA can
reduce the vibration and the sound radiation of the transformer tank panels, but also identifies that
there is potential for further improvements as part of further investigations. These include:

22
Evidence Based Studies Noise – Main Report

• optimising the size and fixing point of the DA on the panel by taking into consideration the
coincidence of the wavelength of the panel flexural vibrations and the wavelength of sound in air;

• cancelling the damping of the dynamic absorber;

• fine tuning the natural frequency of the dynamic absorber.

The second part of this paper examined the use of acoustic active cancelling (effectively an 'anti-noise'
wave) for transformer noise. The experiments concluded that active noise control is a viable solution
for noise related problems generated by large power transformers, especially for control of the first
harmonics contained in the noise. Active Noise Cancelling (ANC) works by introducing a cancelling
wave through an appropriate array of secondary sources, which are interconnected through an
electronic system using a specific processing algorithm for the particular cancellation scheme.
Reductions in sound pressure levels up to approximately 20dB may be achieved as long as the phase
errors remain considerably small.

Medeiros and Borges (2005) examined ANC with a view to determining the efficiency of a system by
means of the attenuation coefficient. A typical application of such a system is the control of the noise
emitted by electrical transformers in substations. The various analyses presented in the paper
illustrate that an incorrect evaluation of the relationship between local and global attenuation
coefficients can lead to an improper estimation of the performance of a system. This work provides a
valuable supplement to work on active noise control by highlighting potential stumbling blocks in the
design of active noise control systems.

2.3.6 Discussion

The literature review of substation noise has provided a valuable backdrop to the sources of noise
from substations and the significant potential for noise reduction at source in substations. Mitigation
measures at source should be the first consideration in terms of trying to reduce the potential noise
impact associated with any new substation. There is a limited amount of measurement data relating to
substation noise, the field surveys of this study provide a valuable source of new data in this regard.

Masti et al (2004) indicated that the transformer's low frequency tonal noise components are the most
likely source of annoyance to nearby residents from substation noise. The substation noise surveys
completed as part of this study included spectral analysis to determine if there were particular tonal
features at specific frequency ranges.

23
Evidence Based Studies Noise – Main Report

2.4 AEOLIAN NOISE FROM ELECTRICAL INFRASTRUCTURE

2.4.1 Introduction

Aeolian noise may occur when wind blows over conductors and insulators or through lattice towers or
hollow components like arcing horns. Aeolian noise is independent of the line being energised or not.
The noise caused by wind blowing over the electrical infrastructure results from the shedding of air
vortices. This noise may become noticeable when wind speeds approach and exceed 10m/s and may
become very prominent as wind speeds rise further. The noise however, in many cases will be
masked by the noise from rain, if any, or by the wind itself (Cigré, 2009).

2.4.2 General Mechanism of Aeolian Noise

In order to discuss the mechanism of Aeolian noise generation, an example of a conductor is

presented in this section. Figure 2.10 shows a model of fluid flow around a cylindrical conductor. A
boundary layer, which is significantly affected by viscosity, exists on the surface of the cylinder. The
fluid flow is separated from the surface of the cylinder in a region where pressure increases within the
boundary layer, thus creating eddies rearward of the cylinder (Akagi et al, 1998).

Figure 2.10: Generation Mechanism of Aeolian Noise (Akagi et al, 1998)

The separation of the fluid flow creates areas of lift and areas of drag in the vicinity of the cylinder and
it is the interaction between these areas that creates the Aeolian noise. In order to reduce Aeolian
noise, it is the lift that must be reduced. The boundary layer flow on the surface ranges from laminar to
turbulent depending on the roughness of the surface. The laminar boundary-layer flow on a smooth
surface with less surface roughness has a wide range of pressure fluctuation at which separation of
flow occurs and the generated Aeolian noise level increases. On the other hand, when turbulence is

24
Evidence Based Studies Noise – Main Report

enhanced by a rough surface, the point of flow separation moves rearward of the conductor and the
region of varying separation point becomes narrow, thus decreasing the Aeolian noise levels.

2.4.3 Aeolian Noise Characteristics

Akagi et al (1998) conducted wind tunnel experiments on various sized conductors to determine the
extent and characteristics of the Aeolian noise. The spectral analysis of Aeolian noise generated from
2
a 4-bundle conductor (ACSR 410mm ) system produced a predominant spectrum in a low frequency
range from 50Hz to 250Hz. Figure 2.11 shows a spectral analysis of Aeolian noise from an 8-bundle
2
conductor system (LN-ACSR 960mm ) at a 20m/s wind velocity. The graph shows the significant
reduction in noise level at the predominant frequency (i.e. more than 10dB) when compared with a
2
standard conductor system, the ACSR 810 mm . The graph also illustrates that this reduction using
2
the 8-bundle conductor is similar to the reduction achieved by the standard ACSR 810 mm using
spiral rods (See Section 2.4.4) (Akagi et al, 1998).

2
Figure 2.11: Aeolian Noise Characteristics of an 8-Bundle LN-ACSR 960 mm Conductor (Akagi
et al, 1998)

25
Evidence Based Studies Noise – Main Report

2.4.4 Mitigation Measures

Section 2.4.3 has highlighted the significant attenuation that can be achieved by using alternative
sized conductor bundles and by using spiral rods instead of the standard straight rods with the
conductors (Figure 2.11). Figure 2.12 shows a photograph of the spiral road method, an effective
countermeasure to reducing Aeolian noise. In this method, aluminium wires are wound around the
conductor. The outer diameter and the pitch of spiral rods have been optimised by experiments in a
low-noise wind tunnel. As a result of the use of the spiral rod method, Aeolian noise can generally be
decreased in the predominant frequency band range by 10dB or more.

Figure 2.12: The Spiral Rod Method (Akagi et al, 1998)

Aeolian noise generated from a conductor is a phenomenon in which pressure fluctuation in a domain
(i.e. where the air separates from the conductor surface) propagates as noise. Therefore, the pressure
fluctuations can be reduced by changing the boundary-layer flow on the conductor surface to a
turbulent state. The conductor surface has a certain roughness, so to create increased turbulent flow
over the surface, coarser roughness (protrusions) can be added to the surface.

Akagi et al (1998) completed various experiments measuring Aeolian noise on conductor systems with
various heights and angles of protrusions. Figure 2.13 shows the behaviour of Aeolian noise
characteristics by applying a protrusion height of 1.5mm and a protrusion angle of 20-120 degrees,
giving a 15dB noise reduction for this particular single conductor system. In the case of multi-bundle
conductor transmission lines, Aeolian noise can be problematic as the effect of turbulence on the
windward side conductors may influence the surrounding state of the leeward-side conductors.
Experiments completed on the two conductor system model concluded that the optimum protrusion
height was 2.5mm and protrusion angle approximately 45 degrees for such a system.

26
Evidence Based Studies Noise – Main Report

Figure 2.13: Relationship Between Protrusion Height and Aeolian Noise Level in a single
Conductor System (Akagi et al, 1998)

Aeolian noise around insulators may occur also at high wind speeds for certain insulator configurations
and designs. The occurrence of this type of noise is difficult to anticipate but it can be reduced by
using composite insulators instead of glass or porcelain insulators, by changing the rib profile or the
type of insulators in the string or by introducing a number of different profile insulators in the string
(Cigre, 2009)

2.4.5 Discussion

Aeolian noise is very different to the other noise sources associated with electricity infrastructure, most
notably because it does not relate to whether the infrastructure is energised or not. As the conditions
required to generate Aeolian noise are very specific in terms of requiring high wind speeds at very
specific angles of incidence, it is not as common as the other sources of noise associated with
electricity infrastructure. The conditions that give rise to Aeolian noise (i.e. high wind speeds) will in
most instances also mask the Aeolian noise. For the reasons stated above, Aeolian noise has
generally not been perceived of as presenting a very significant noise impact at sensitive receptors
(e.g. Cigré, 2009, DECC, 2011).

In terms of measurement of Aeolian noise, any significant measurement data and analysis in the
literature relates to wind tunnel experiments (e.g. Akagi et al, 1998). Attempting to measure Aeolian
noise along existing electricity infrastructure is fraught with difficulty on account of the not being in a

27
Evidence Based Studies Noise – Main Report

position to anticipate its onset. The literature evidence would suggest that there is no standard or
practical method for measuring and analysing Aeolian noise from existing infrastructure. On the basis
of this, the measurement and analysis of Aeolian noise has not formed a significant consideration in
the field surveys of this study.

The review of literature on Aeolian noise does demonstrate that there is significant potential for the
reduction of Aeolian noise by applying a number of different mitigating measures as detailed in Section
2.4.4 of this report. The construction of any new electrical infrastructure on the EirGrid network should
be done with full consideration of the mitigation measures available for the reduction of Aeolian noise.

2.5 HEALTH EFFECTS OF NOISE

In 1999, the World Health Organisation (WHO) published a report on the effects of environmental
noise on human communities (WHO, 1999). This report outlined the research and consolidated views
with regard to the health effects that environmental noise was known or suspected to cause on human
communities. The most significant health effects were identified as hearing impairment, interference
with speech communications, sleep disturbance, cardiovascular and physiological effects, mental
health effects, effects on performance and effects on residential behaviour and annoyance.

The most prominent noise sources specifically referred to were industrial, transportation, construction,
building services, domestic and leisure activity noise. No specific reference was made to noise from
electricity plant or equipment.

The WHO 1999 Guidelines set threshold limits for moderate and serious annoyance (i.e. 50dB and
55dB LAeq 16hrs) and a range of internal room noise thresholds based around the requirements for
speech intelligibility and sleep disturbance for various room types. While this study has illustrated the
potential for noise impacts from electrical infrastructure under various conditions, such noise sources
have had little literary review in terms of health effects, as noise from electrical plant and equipment is
generally considerably less significant than noise from other sources (e.g. industry, road, rail, airports).

In 2009, the WHO published an additional report aimed specifically at examining the health effects
from night-time environmental noise (WHO, 2009). No specific discussion of noise from electrical plant
or equipment was included in this report; however noise threshold limits for night-time were included
which illustrated where environmental noise was known to cause effects. Table 2 includes a summary
of effects and threshold noise levels for effects where sufficient evidence is available. These threshold
limits apply to all environmental noise sources, including noise from electrical plant and equipment.

While the literature base shows very limited reference to the health effects associated specifically with
electrical plant and equipment, it is clear from the threshold limits outlined in Table 2.2 below that there
is the potential for noise from electricity infrastructure to exceed these threshold limits under various
scenarios, and hence contribute to the health effects discussed in the WHO Night Noise Guidelines.

28
Evidence Based Studies Noise – Main Report

Table 2.2: Summary of Health Effects and Noise Threshold Limits (WHO, 2009)

Effect Indicator Threshold (dB)

EEG Awakening LAmax, inside 35

Motility, Onset of motility LAmax, inside 32

Biological Effects
Changes in duration of various stages of
sleep, in sleep structure and LAmax, inside 35
fragmentation of sleep

Waking up in the night and/or too early

LAmax, inside 42
Sleep Quality in the morning

Increase average motility when sleeping Lnight, outside 42

Self-reported sleep disturbance Lnight, outside 42

Well-being
Use of somnifacient drugs and sedatives Lnight, outside 40

Medical Conditions Environmental insomnia Lnight, outside 42

The WHO published a further report in 2011 outlining in detail the synthesised reviews of evidence on
the relationship between environmental noise and specific health effects (WHO, 2011). As with the
2009 report, the aim of this report was to supplement the 1999 guidelines and incorporate all of the
most recent developments in the research of the health effects of environmental noise. Noise from
electrical plant and equipment was not discussed directly, but this source of noise is assumed under
the heading of environmental noise.

In addition to the broad ranging coverage of the health effects from environmental noise in the WHO
reports, additional reports and research papers have been published covering various aspects of the
health effects associated with environmental noise. One example is a UK government sponsored
review document into low frequency noise and its effects (Leventhall, 2003). This document includes
a section on the health effects associated with low frequency noise exposure for humans which can be
directly related to the various low frequency noise impacts from electrical plant and equipment as
outlined in this literature review under the headings for corona, substation and Aeolian noise.

Meyer et al (1989) includes a review of published data in relation to the potential for physical agents
such as noise and electromagnetic fields to provide an adverse impact on reproductive outcomes.
The paper presents published data on possible biological mechanisms, considerations for exposure
assessments and suggestions for further epidemiological research.

29
Evidence Based Studies Noise – Main Report

3 METHODOLOGY

3.1 SITE SELECTION

3.1.1 Site Selection Criteria

The general rationale for site selection in respect of this noise study was that representative noise data
could be measured for each of the principal types of infrastructure associated with the Irish electricity
transmission network. The four principal types of infrastructure considered in the studies were
substations, 110kV OHL, 220kV OHL and 400kV OHL.

There is a 275kV double circuit OHL between Louth and Tandragee (Northern Ireland) substations.
This was not included in the site selection process as there are no plans to use this infrastructure type
in the future on the Irish electricity network. Again as previously noted, underground cables were not
considered in the site selection process as they are not a source of significant noise emissions.

The initial concept for the site selection process was to select a relatively large number of sample
locations (e.g.100) over a wide range of different land use types (e.g. urban, suburban, agricultural,
forest, open, wetland, coastal etc.). Consideration would also be given to topography (i.e. upland,
lowland etc.), the sensitivity of the landscape and the presence of specific ecological designations
(e.g. Special Areas of Conservation, Special Protection Areas, National parks, Nature Reserves etc.).
The site selection process would also build in the effect of distance between the relevant electricity
infrastructure and the respective sensitive location/land use.

As the site selection process developed, the key emerging issue was how noise monitoring surveys
could isolate and record noise from electricity infrastructure and ensure that the noise measurements
were not contaminated with any number of other noise sources in the respective study area. After
further consideration of this issue, it was clear that only a survey strategy that was based around the
EirGrid outage programme could achieve an effective means of recording the noise levels associated
with the live infrastructure.

Based on this new approach, the focus in the site selection process shifted from choosing a large
number of sample locations in varied environments to selecting a smaller number of locations where
more extensive surveys could be arranged around the outage schedule. By measuring noise levels in
close proximity to the relevant infrastructure for a sufficient period of time with the infrastructure live,
and then again during a period of outage, the variation in noise levels during different periods of the
day and different weather conditions could be established for both scenarios. Analysis of the datasets
would reveal how the noise levels compared for both scenarios at the same time of day under the
same weather conditions.

30
Evidence Based Studies Noise – Main Report

By focusing surveys around the outage programme, it would be possible to determine if certain types
of electrical infrastructure produced an audible noise that could be directly observed as an increase in
the measured noise levels and to what extent the noise levels increased (i.e. what decibel level). This
strategy would produce very clear evidence on whether various types of infrastructure did or did not
result in elevated noise levels. The previous strategy of using a large number of sample points in
various locations was flawed in that there was no means of determining whether the measured noise
levels were or were not linked to the electrical infrastructure.

Having established the strategy to be adopted for the surveys, EirGrid’s outage programme was
reviewed. A record was made of all OHL (110kV, 220kV and 400kV) that were scheduled for outage
for a period of time that would enable a sufficient sample of noise levels to recorded (i.e.
approximately one week). This list of OHL was reviewed with a view to determining which of these
were most suitable for the purposes of the noise surveys. OHL that had sections of line that were a
significant distance from other noise sources (e.g. roads, industries, running water etc.) were sought.

The number of sites used for each type of infrastructure as part of the study reflects the relative
prevalence of each type of infrastructure in the transmission network. This was also reflected in the
number of each type of infrastructure that was included in the outage programme. For this reason,
more surveys were completed on the 110kV overhead line network, followed by the 220kV network
and lastly the 400kV network. For the purposes of reference in this document, all surveys completed
on overhead lines while infrastructure has been live are referred to as 'On' surveys, while all surveys
completed on overhead lines during periods of outage are referred to as 'Off' surveys.

While the outage programme presented a direct means of determining what noise emission were
associated with OHL, this strategy was not practical in terms of surveying substations. While various
items of plant in substations were scheduled for outage, no overall existing substation was scheduled
for outage at one time. Leaving aside potential corona noise, substations typically produce a steady
state noise level (or hum) which could be recorded by short-term noise measurements. For these
reasons, the noise surveys completed for substations involved short-term measurements in the vicinity
of the boundaries of the substation sites.

3.1.2 Outline of Sites Selected

On the 110kV overhead line network, surveys were completed during live periods and during outage
periods at the following locations:

• Golagh Tee - Letterkenny 110kV overhead line;

• Cathleen’s Fall (C’Fall) Golagh Tee - Letterkenny 110kV overhead line (1)

• Cathleen’s Fall (C’Fall) Golagh Tee - Letterkenny 110kV overhead line (2)

• Cathleen’s Fall (C’Fall) - Srananagh - 110kV overhead line.

31
Evidence Based Studies Noise – Main Report

The locations of the four monitoring points for the OHL listed above are illustrated in Figures 3.1 to 3.4.

Figure 3.1: Noise Monitoring Location on Golagh Tee - Letterkenny 110kV OHL

32
Evidence Based Studies Noise – Main Report

Figure 3.2: Noise Monitoring Location on C’Fall Golagh Tee - Letterkenny 110kV OHL (1)

Figure 3.3: Noise Monitoring Location on C’Fall Golagh Tee - Letterkenny 110kV OHL (2)

33
Evidence Based Studies Noise – Main Report

Figure 3.4: Noise Monitoring Location on C’Fall - Srananagh 110kV OHL

On the 220kV overhead line network, surveys were completed during live periods and during outage
periods at the following locations:

• Dunstown - Maynooth 220kV overhead line (Betaghstown);

• Dunstown - Maynooth 220kV overhead line (Currabell);

• Dunstown - Maynooth 220kV overhead line (Thomastown);

The locations of the three monitoring points on the Dunstown - Maynooth 220kV overhead line are
illustrated in Figures 3.5 to 3.7.

34
Evidence Based Studies Noise – Main Report

Figure 3.5: Noise Monitoring Location on Dunstown - Maynooth (Betaghstown) 220kV OHL

Figure 3.6: Noise Monitoring Location on Dunstown - Maynooth (Currabell) 220kV OHL

35
Evidence Based Studies Noise – Main Report

Figure 3.7: Noise Monitoring Location on Dunstown - Maynooth (Thomastown) 220kV OHL

On the 400kV overhead line network, surveys were completed during live periods and during outage
periods at the following location:

• Oldstreet - Woodland 400kV OHL (Ardrums Great).

The location of the monitoring point on the Oldstreet - Woodland 400kV overhead line is illustrated in
Figure 3.8.

36
Evidence Based Studies Noise – Main Report

Figure 3.8: Noise Monitoring Location on Oldstreet - Woodland (Ardrum Great) 400kV OHL

Noise measurement surveys were completed at the following substation sites:

• Dunfirth 110kv Substation;

• Gorman 220kV Substation;

• Woodland 400kV Substation.

The locations of the monitoring points at the substations listed above are illustrated in Figures 3.9 to
3.11.

37
Evidence Based Studies Noise – Main Report

Figure 3.9: Noise Monitoring Locations at Dunfirth 110kV Substation

Figure 3.10: Noise Monitoring Locations at Gorman 220kV Substation

38
Evidence Based Studies Noise – Main Report

Figure 3.11: Noise Monitoring Locations at Woodland 400kV Substation

3.1.3 Limitations to Site Selection Process

On account of the survey strategy being worked around the outage programme, the most obvious
limitation to the site selection process was the availability of suitable sites that were scheduled for a
period of outage during the relevant survey window.

As the majority of the OHL transmission network consists of 110kV lines, there was a significant
number of 110kV lines included in the outage programme. It was a relatively simple task to identify
suitable locations along the 110kV network for undertaking site surveys.

The number of 220kV lines that were included in the outage programme was more limited and
therefore the process of identifying suitable survey locations were more constrained. On account of
the significant timeframe which the Dunstown-Maynooth 220kV line was scheduled for outage,
multiple locations along this line were selected.

The number of opportunities for completing surveys during periods of outage on the 400kV network
was very limited and this was reflected in the fact that only one survey could be completed on the
400kV network.

The choice of substation sites was not limited to the outage programme, and therefore there were
more options available in determining which substation sites would be used for the surveys.

39
Evidence Based Studies Noise – Main Report

3.2 NOISE SURVEY METHODOLOGY

3.2.1 Survey Equipment - Overhead Lines

All noise monitoring surveys were completed using outdoor unattended noise monitoring kits. All of
the noise monitoring kits used conformed to the requirements for integrating averaging sound level
meters (Type 1) as specified in BS EN 60804. All noise monitoring equipment was accurately
calibrated before use. All measurements were made at a height of 1.2 - 1.5m above ground level and
all meters were placed directly in the path of the overhead line.

Noise measurements at all of the monitoring locations were recorded using the following noise
monitoring equipment:

1. Rion NL-32 Class 1 Sound Level Analyser; Outdoor kit enhanced NL-31/32; Rion WS-03SO1
Windscreen head assembly (inc WS-03051); Rion EC-04 2m Extension Cable (7 Pin) and Rion NC-74
Class 1 Acoustic Calibrator.

2. Brüel & Kjær Type 2250 (BZ7224 Version 1.4.) with Brüel & Kjær Type 4189 0.25” microphone,
mounted on an UA1404 weatherproof environmental kit. The monitoring kit was powered by 2 no.
fused 12 volt valve regulated lead acid batteries and housed in a sealed iM2700 Stormcase.

Weather data was recorded at each site using a Davis Vantage Pro2 Met Station with Davis Weather
Envoy logger and Meteovue GPRS Modem. The met station was equipped to measure wind speed,
wind direction, rainfall, temperature and humidity. All measurement periods for the weather station
were synchronised with the measurement periods for the noise monitoring equipment.

The met station was configured to log data to a server via GPRS so as weather conditions could be
observed remotely in real time. On a number of occasions however, the GPRS modem failed to
connect and transmit this data to the server. It is not clear as to whether this was an intermittent fault
with the modem, the server or the local mobile network. As highlighted in Section 3.2.3 and 4.3.1
some of the weather data for the 400kV Oldstreet- Woodland survey was lost on account of this.

3.2.2 Survey Methodology - Overhead Lines

Section 3.1 outlines the site selection strategy and lists the sites where noise monitoring was
completed. Section 3.2.1 gives details of the noise monitoring equipment and the weather recoding
equipment used for the surveys. The general strategy for each location was to obtain approximately
one week of data when the infrastructure was live and one week of data during the period of outage.
There was some variation on this during the surveys as conditions arose which altered the length of
surveys in particular instances.

At each location, the noise meter and the weather station were synchronised to record in short logging
periods of either 15 minutes or 30 minutes. During each logging period, a range of noise parameters

40
Evidence Based Studies Noise – Main Report

were recorded including LAeq, LAmax, LAmin, LA5, LA10, LA50 and LA90. The parameters that were extracted
for the purposes of analysis are listed below with a brief explanation regarding the relevance of the
parameter:

• LAeq The continuous equivalent A-weighted sound pressure level. This is an “average” of the
sound pressure level for the relevant measurement period

• LA5 This is the A-weighted sound level that is exceeded for 5% of the sample period and is
commonly used in the literature relating to noise from electrical infrastructure.

• LA50 This is the A-weighted sound level that is exceeded for 50% of the sample period and is
commonly used in the literature relating to noise from electrical infrastructure.

• LA90 This is the A-weighted sound level that is exceeded for 90% of the sample period and is
commonly used as a reference for recording the 'background' noise level.

Parameters such as LAmax, LAmin and LA10 were not used in the analysis of the datasets as they did not
offer any insight in terms of establishing what noise levels are associated with electrical infrastructure.

3.2.3 Problems Encountered During Surveys - Overhead Lines

During the 'On' survey at Cathleen’s Fall Golagh Tee - Letterkenny 110KV (2), livestock were
introduced into the field that contained the noise and weather monitoring equipment. The dataset that
was retrieved for this survey showed a substantial difference to the measured noise levels for the 'Off'
survey completed at the same site. These substantial differences were not observed in the other
110kV or 220kV survey results, and therefore the only conclusion to be drawn was that the dataset
was subject to significant contamination from the animals in the field during the 'On' survey. This
dataset was therefore not used for the purposes of drawing conclusions based around the survey
findings in the noise study.

During the 'Off' survey on the Oldstreet - Woodland 400kV OHL, a weather station malfunction meant
that a portion of weather data for this survey was not recorded. The loss of this data meant that the
analysis of the two datasets for this line was completed with more emphasis on the entire dataset and
the use of graphical representations to illustrate analytical conclusions. Further discussion of this data
deficiency is included in Section 4.3.1 of this report.

3.2.4 Survey Equipment - Substations

Noise measurements at all of the substation monitoring locations were recorded using a Brüel & Kjær
Type 2250 (BZ7223 Version 1.5) with Brüel & Kjær Type 4189 0.25” microphone mounted on an UA-
0801 Lightweight tripod.

41
Evidence Based Studies Noise – Main Report

3.2.5 Survey Methodology - Substations

Completing long-term noise surveys at substations was not a practical consideration for this study as
the outage programme did not allow for an entire substation to be scheduled for outage at one time.
The outage programme only allowed for individual items within a substation to be scheduled for
outage, therefore there was no means of getting direct comparable data for complete substations in
'On' and 'Off' modes.

The primary purpose of the substation noise surveys was to record the steady state noise level (or
'hum') associated with substations of varying voltage (i.e. 110kV, 220kV or 400kV). By completing
noise measurements at varying distances from the substation boundaries, the noise levels associated
with the substation could be determined.

The same parameters as were used for the overhead line surveys were recorded for the substation
surveys, namely LAeq, LA5, LA50 and LA90. In addition to this, spectral analysis was completed at all of
the substation sites between frequencies 12.5Hz and 20kHz.

3.3 METHODOLOGY FOR ANALYSIS OF DATA

3.3.1 Overhead Lines

Sections 3.1 and 3.2 outlined the site selection process and the survey methodology used for the
evidence based noise studies. Based on the strategy used for the surveys, large datasets (>100 hours
in most instances) were recorded at each measurement location for the two principal scenarios (i.e.
with infrastructure live and during outage).

The main aim of the analysis process was to determine if there was any clear difference between the
data for both scenarios which would provide evidence that the live infrastructure is producing noise
emissions that are detectable. In order to do this, it was imperative to remove all other noise sources
from the datasets to ensure that a like for like comparison could be made. For ease of reference in
this report, the survey with the infrastructure live will commonly be referred to as the 'On' survey while
the survey completed during the period of outage will be referred to as the 'Off' survey.

The first part of this process of removing other noise sources from consideration was covered in the
site selection process. As far as reasonably practicable, quiet survey locations were selected on the
basis of their distance from all obvious noise sources in the study area (e.g. roads, industry, running
water etc.). It was understood as part of this process that there was no guarantee in avoiding certain
potential noise sources (e.g. animal noises, agricultural plant etc.). However, the data analysis did
include a mechanism for extracting random noisy periods from the dataset, as further elaborated upon
in the text below.

42
Evidence Based Studies Noise – Main Report

The second part of this process was to refine all of the datasets so that like for like comparisons could
be procured For every dataset, all of the recorded noise parameters (LAeq, LA50, LA5, LA90) were
synchronised in 15 or 30 minute logging periods with the weather data. Each dataset was divided up
into four different datasets based on different time periods as follows:

• All data for the survey period;

• Data during weekdays only (i.e. Monday to Friday, 08:00 - 18:00 hours);

• Data during night time only (i.e. every day, 23:00 - 07:00 hours);

• Data during quiet daytime periods only (i.e. Monday to Friday 18:00 - 23:00 hours, Saturday 13:00
- 23:00 hours, Sunday/Bank Holidays 07:00 - 23:00 hours).

For each survey location and survey strategy (i.e. infrastructure live or outage), separate datasets
were created of synchronised noise and weather data for each noise parameter (i.e. LAeq, LA50, LA5,
LA90) for each of the time periods listed above (i.e. 16 datasets, e.g. LAeq - All, LAeq - Weekday, LAeq -
Night, LAeq - Quiet Day etc.). Each of these datasets was put through a series of further refinements
as described in the bullet points below:

• Deletion of all data during periods of rain (including the data periods that immediately preceded
and followed periods of rain);

• Deletion of wind affected data in a graded manner (e.g. deletion of all data with wind speeds
greater than 3mph, 5mph, 10mph etc.). There was some variation on how this was applied to
different datasets based on the extent to which the dataset was impacted by the wind;

• Deletion of all data where wind speeds were greater than 0 mph;

• Deletion of 'outlier' data, which is defined as any measurement period where the measured noise
level was 10dB(A) or greater above the measurement period which directly preceded it and the
measurement period that followed it. Removing these outliers enables random sources of noise
(e.g. animal noises, agricultural plant etc.) to be removed from the dataset which prevents these
from skewing the overall trend for the survey period. The removal of outlier data was not required
for every dataset.

Based on the refined analysis completed on the datasets as described above, a series of greater than
60 datasets were created for each survey location and for each survey strategy period. For each of
these datasets, the average noise was calculated for the respective noise parameter. In the case of
the LA5, LA50 and LA90 parameters, the arithmetic average was calculated. In the case of the LAeq
parameter, the logarithmic average was calculated. The calculated averages derived from each
dataset were used for purposes of comparison between each survey strategy (i.e. infrastructure live
and outage) at each location.

43
Evidence Based Studies Noise – Main Report

The refinement process that was applied to the analysis of the datasets resulted in certain datasets
being reduced to a low sample rate. Where the refined dataset has fallen below 2.5 hours worth of
data, this has been highlighted in Section 4 of this report. Interpretation of survey results included a
consideration of the sample sizes.

The analytical process described above was completed for all of the survey locations selected and
described in Section 3.1.2 of this report.

3.3.2 Substations

As short-term measurements were taken at the substation sites, the extent of data available for
analysis is substantially lower for the substation surveys and hence there is no requirement for the
extensive set of analyses that was conducted for the overhead lines. The analysis of this data was by
means of review of the small datasets to determine the noise level associated with each substation
type and to determine if there were any particular tonal features distinctly visible in the frequency
analysis.

44
Evidence Based Studies Noise – Main Report

4 SURVEY RESULTS

4.1 SURVEY RESULTS FOR 110KV OVERHEAD LINES

4.1.1 Cathleen’s Fall (C’Fall) Golagh Tee - Letterkenny 110kV OHL (1)

Table 4.1 below includes a summary of all averaged data from the analysis of the synchronised noise
and weather data at Cathleen’s Fall (C’Fall) Golagh Tee - Letterkenny 110kV (1). For each noise
parameter, the averaged noise levels for the survey with the infrastructure live (i.e. 'On' in the table)
has been placed beside the averaged noise levels for the survey completed during the period of
outage (i.e. 'Off' in the table). This enables direct comparison to be made between 'On' and 'Off'
surveys for each of the time periods and weather conditions.

The 'On' survey produced a dataset of approximately 166 hours for the purposes of analysis, while the
'Off' survey produced a dataset of approximately 118 hours. For each broad dataset included in the
table below (e.g. all noise, weekday, night and quiet day), the dataset has been put through a series a
refinements to delete perceived extraneous source noisy periods during the survey period.

Each subsequent entry includes the refinement made in the previous step of the data analysis.
Therefore, the '0mph' entry includes all of the deletions to the dataset from the '- Rain Data', '<3mph
wind speed' and 'outlier' steps.

In the table below, some of the entries measured noise levels show no change when the dataset is
refined to omit perceived noisy periods during the survey (e.g. during rain and windy periods). This is
the case as the original dataset did not contain these conditions.

45
Evidence Based Studies Noise – Main Report

Table 4.1: Comparison of Averaged Data from 'On' and 'Off' Surveys at C’Fall Golagh Tee -
Letterkenny 110kV OHL (1)

Averaged Noise Levels dB(A)

Dataset LA5 LA50 LA90 LAeq
On Off On Off On Off On Off
All Noise Data 39.9 46.7 34.3 40.0 32.0 37.3 41.5 51.3
- Rain Data 39.6 45.9 33.6 39.6 31.4 36.8 40.7 51.2
< 3mph wind speed 39.1 38.9 33.6 34.2 31.5 32.4 40.5 38.2
Outliers 39.0 38.6 33.6 34.2 31.4 32.4 38.4 38.2
0mph 35.2 39.4 30.8 34.8 29.3 33.0 34.0 37.2

Weekday 43.1 51.4 36.1 43.8 33.1 40.6 43.9 53.2

- Rain Data 41.1 51.9 36.1 44.2 33.1 40.6 43.6 53.2
< 3mph wind speed 43.1 42.8 36.1 37.4 33.1 35.0 43.3 39.6
Outliers 43.1 42.4 36.1 37.4 33.1 35.0 40.7 39.6
0mph 37.9* 42.9* 36.1* 38.9* 33.1* 37.2* 39.3* 39.9*

Night 34.5 43.5 30.2 37.7 28.5 35.1 33.6 50.5

- Rain Data 34.4 43.3 30.1 37.6 28.5 35.0 33.1 50.6
< 3mph wind speed 33.4 35.9 29.8 31.9 28.3 30.3 32.4 34.6
Outliers 33.2 35.9 29.8 31.9 28.3 30.3 32.4 34.6
0mph 31.6 36.5 29.3 32.0 28.2 30.3 30.9 33.7
Rain (<5m/s Wind) [2 Hours 40* 47* 32* 40* 29* 37* 37* 43*
Dataset for 'On' & 'Off']

Quiet Day 42.3 45.8 37.0 39.6 34.7 36.9 42.7 50.3
- Rain Data 44.0 45.5 35.0 39.2 33.3 32.6 41.3 51.3
< 3mph wind speed 39.9 41.1 34.9 35.8 33.3 33.7 41.4 40.0
Outliers 39.9 41.1 34.9 35.8 33.3 33.7 40.8 38.4
0mph 33.8* 41.3* 30.5* 37.3* 29.4* 34.6* 31.3* 38.5*
Rain (<5m/s Wind) 50 49 42 42 37 40 45 46
* Small dataset, i.e. less than 2.5 hours of data

The comparison of the averaged noise levels for all parameters included in the table above
demonstrate that measured noise levels during the 'On' survey are less than those measured for the
'Off' survey for all analysis scenarios investigated.

When all of the noise data is compared between the 'On' and 'Off' surveys before any refinement of
the data was undertaken, there is a significant difference between the averaged noise levels for all of
the parameters. All parameters were recorded as 5-10d(A) greater during the 'Off' surveys. The
process of deleting rain and wind affected data reduces the difference between the noise parameters
for both datasets generally into the 2-4dB(A) range.

The data in Table 4.1 provides strong evidence that live 110kV overhead lines do not produce
significant noise levels that contribute to increasing the background (i.e. LA90) or ambient (LAeq) noise
level in the vicinity of the lines. The fact that the 'Off' survey is producing marginally higher measured
noise levels than the 'On' survey is simply a factor of the natural variation in the noise from other
sources in the study area during the respective measurement periods.

46
Evidence Based Studies Noise – Main Report

Figure 4.1 below includes a typical cross-section of data points for the background noise level (i.e.
LA90) for both 'On' and 'Off' surveys during the night-time period only. This dataset has been refined to
exclude all rain and wind affected measurements. The y axis has been altered to give a simple
numerical value to represent the passage of time as opposed to having a date and time. This has
been done for ease of reference and because two datasets are being represented which have different
measurement dates and times.

In the figure below, the background noise levels are quite similar with the 'Off' survey measuring
marginally higher noise levels. The night-time period is the quietest period of the day and therefore, if
the live infrastructure was producing any significant noise, it would be most noticeable as an elevated
background noise level during the 'On' survey at night. The figure below clearly illustrates that there is
no elevated background noise levels at night for the 'On' survey.

Figure 4.1: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night
(C’Fall Golagh Tee - Letterkenny 110kV OHL)

Figure 4.1 illustrates that during the quietest periods when all other rain and wind affected data has
been removed, the steady state background noise level in the vicinity of the overhead line is not higher
when the infrastructure is live as when compared with a period of outage.

Presenting the data in this way does not account for potential random and sporadic corona discharge
events during the survey period. The LA90 parameter omits the top 10% of noisy activities during a
measurement period (i.e. hence it is a record of the background noise level) and therefore omits any
likely occasional events such noisy corona discharges.

47
Evidence Based Studies Noise – Main Report

The LA5 parameter records the top 5% of noisy activity within a given measurement period. One would
expect that significant increases in random corona activity when comparing the 'On' survey with the
'Off' survey would result in a significant increase in the LA5 parameter for the 'On' survey as compared
with the off survey. Table 4.1 illustrates that for all weather conditions and during all time periods, the
LA5 parameter is similar or even lower during the 'On' survey as compared with the 'Off' survey. This
indicates that either random corona discharges were not frequent enough throughout the 'On' survey
to register a noticeable change in the recorded noise level or that if they were relatively frequent, they
were not of sufficient noise energy as to register a change in the overall LA5 measurements. Either
way, the data does not give any evidence that random corona discharge is a significant noise impact
from 110kV overhead lines.

As random corona discharge is often linked to wet weather conditions, a further exercise was
undertaken to refine the analysis to compare the measurements for all parameters during the 'On' and
'Off' surveys for periods of rain only. There was a reasonable sample size of rain affected data for
both 'On' and 'Off' surveys during the quiet daytime and night-time periods, with average measured
noise levels for all parameters being either similar or marginally less during the 'On' survey as
compared with the 'Off' survey. The higher measured levels during the 'Off' survey may have been a
factor of marginally higher average wind speeds for some of the data.

Overall, there was no evidence from these measurements during rainy conditions of significant
recordable corona effects from the 110kV overhead line.

The weather-adjusted noise levels included in Table 4.1 are below the daytime WHO threshold limits
for serious annoyance (55dB LAeq) and moderate annoyance (i.e. 50dB LAeq) for outdoor living areas
and the night-time free-field threshold limit of 42dB (LAeq) for preventing negative effects on sleep.

A review of this data in the context of all 110kV survey data is included in Section 6.1 of this study.

4.1.2 Cathleen’s Fall (C’Fall) Golagh Tee - Letterkenny 110kV OHL (2)

Table 4.2 below includes a summary of all averaged data from the analysis of the synchronised noise
and weather data at C’Fall Golagh Tee - Letterkenny 110kV (2).

For each noise parameter, the averaged noise levels for the survey with the infrastructure live (i.e. 'On'
in the table) has been placed beside the averaged noise levels for the survey complete during the
period of outage (i.e. 'Off' in the table). This enables direct comparison to be made between 'On' and
'Off' surveys for each of the time periods and weather conditions.

The 'On' survey produced a dataset of approximately 167 hours for the purposes of analysis, while the
'Off' survey produced a dataset of approximately 166 hours.

48
Evidence Based Studies Noise – Main Report

Table 4.2: Comparison of Averaged Data from 'On' and 'Off' Surveys at C’Fall Golagh Tee -
Letterkenny 110kV OHL (2)

Averaged Noise Levels dB(A)

Dataset LA5 LA50 LA90 LAeq
On Off On Off On Off On Off
All Noise Data 58.1 52.0 50.4 42.0 45.9 36.5 62.2 54.0
- Rain Data 55.8 50.6 48.1 40.6 43.9 35.3 59.5 41.2
< 3mph wind speed 55.8 41.7 48.1 30.0 43.9 27.0 59.5 41.2
0mph 55.8 42.1 48.1 29.6 43.9 27.3 59.5 41.2

Weekday 61.4 52.8 52.9 42.04 47.7 36.1 63.6 50.1

- Rain Data 58.4 51.9 49.5 41.7 44.6 35.8 60.1 43.7
< 3mph wind speed 58.4 47.9 49.5 32.7 44.6 27.8 60.1 43.7
0mph 58.4 49.9* 49.5 32.9* 44.6 30.1* 60.1 46.5*

Night 56.6 48.6 50.3 40.4 46.3 30.1 61.5 56.0

- Rain Data 54.9 45.2 48.8 37.1 45.1 35.6 59.3 35.2
< 3mph wind speed 54.9 35.3 48.8 28.1 45.1 32.9 59.3 35.2
0mph 54.9 35.1 48.8 28.3 45.1 25.9 59.3 35.8

Quiet Day 55.7 55.0 46.8 44.2 43.0 38.4 59.9 54.4
- Rain Data 54.4 54.8 45.6 44.0 43.0 38.2 57.2 41.7
< 3mph wind speed 54.4 44.8 45.6 30.3 43.0 27.1 57.2 41.7
0mph 54.4 45.3* 45.6 28.8* 43.0 25.6* 57.2 41.8*
* Small dataset, i.e. less than 2.5 hours of data

The comparison of the averaged noise levels for all parameters included in the table above
demonstrate that measured noise levels during the 'On' survey are significantly more than those
measured for the 'Off' survey for all analysis scenarios investigated.

In the course of collecting the noise monitoring equipment after the 'On' survey, it became apparent
that the field in which the noise monitoring equipment was placed in was occupied by sheep (no
livestock were in the field when the monitoring equipment was initially placed there). The presence of
the sheep in close vicinity to the noise meter would indicate that the sheep were a source of significant
contamination to the noise readings recorded during the 'On' survey. The fact that none of the other
noise monitoring surveys on 110kV lines produced results where the 'On' survey results were
significantly greater than the 'Off' survey results would indicate that this survey may have been subject
of contamination from a noise source other than the electrical overhead line.

A review of this data in the context of all 110kV survey data is included in Section 6.1 of this study.

4.1.3 Cathleen’s Fall (C’Fall) - Srananagh 110kV OHL

Table 4.3 below includes a summary of all averaged data from the analysis of the synchronised noise
and weather data at C’Fall - Srananagh 110kV OHL. For each noise parameter, the averaged noise
levels for the survey with the infrastructure live (i.e. 'On' in the table) has been placed beside the
averaged noise levels for the survey complete during the period of outage (i.e. 'Off' in the table). This

49
Evidence Based Studies Noise – Main Report

enables direct comparison to be made between 'On' and 'Off' surveys for each of the time periods and
weather conditions.

The 'On' survey produced a dataset of approximately 166 hours for the purposes of analysis, while the
'Off' survey produced a dataset of approximately 164 hours.

Table 4.3: Comparison of Averaged Data from 'On' and 'Off' Surveys at C’Fall - Srananagh
110kV OHL

Averaged Noise Levels dB(A)

Dataset LA5 LA50 LA90 LAeq
On Off On Off On Off On Off
All Noise Data 44.1 37.8 35.1 29.7 31.8 27.2 49.0 38.0
- Rain Data 43.9 37.1 35.0 28.8 31.7 26.5 48.8 36.8
< 3mph wind speed 40.6 37.1 31.5 28.6 28.7 26.3 41.2 36.3
Outliers 40.4 37.0 31.5 28.6 28.7 26.3 39.4 35.9
0mph 36.8 34.2 28.0 27.2 25.8 25.3 35.8 34.3

Weekday 49.5 42.4 39.1 32.6 35.0 29.5 52.7 39.8

- Rain Data 49.5 42.4 38.8 31.8 34.7 28.7 52.5 39.0
< 3mph wind speed 45.1 41.5 33.6 31.2 30.0 28.1 43.5 38.0
Outliers 44.6 41.5 33.6 31.2 30.0 28.1 42.3 37.8
0mph 41.7 41.9 29.4 32.0 25.3 28.7 39.3 38.2

Night 38.7 31.7 31.6 26.7 29.2 25.2 42.1 35.0

- Rain Data 38.7 31.7 31.6 25.7 29.2 24.3 42.1 29.6
< 3mph wind speed 37.3 30.5 30.0 25.7 27.9 24.3 40.5 29.6
Outliers 36.9 30.5 29.7 25.7 27.9 24.3 37.2 29.6
0mph 33.8 29.6 26.9 25.1 25.4 23.9 32.8 28.5

Quiet Day 44.5 39.6 35.3 29.6 31.9 26.6 46.7 38.2
- Rain Data 44.5 39.6 35.3 29.0 31.9 26.2 46.7 37.8
< 3mph wind speed 41.0 39.1 31.1 29.0 28.7 26.2 38.0 37.8
Outliers 40.9 38.8 31.1 29.0 28.7 26.2 38.0 37.5
0mph 37.6 37.1 28.1 27.4 26.1 24.8 34.9 35.5

The comparison of the averaged noise levels for all parameters included in the table above
demonstrate that measured noise levels during the 'On' survey are more than those measured for the
'Off' survey prior to considering wind and rain affected data.

After considering rain and wind affected data, there are only small differences in the recorded data
when comparing the datasets for 'On' and 'Off' survey periods.

The data in Table 4.3 provides strong evidence that live 110kV OHL do not produce significant noise
levels that contribute to increasing the background (i.e. LA90) or ambient (LAeq) noise level in the vicinity
of the lines. There are marginal differences between the measured noise parameters (i.e. LAeq, LA90,
LA5, LA50) for 'On' and 'Off' surveys but these differences are small and fall within the range of natural
variation that would be expected in a low noise environment.

50
Evidence Based Studies Noise – Main Report

Figure 4.2 below includes a typical cross-section of data points for the background noise level (i.e.
LA90) for both 'On' and 'Off' surveys during the night-time period only. This dataset has been refined to
exclude all rain and wind affected measurements. The y axis has been altered to give a simple
numerical value to represent the passage of time as opposed to having a date and time. This has
been done for ease of reference and because two datasets are being represented which have different
measurement dates and times.

In the figure below, the background noise levels are quite similar. As in the case of the illustration
included in section 4.1.1, if the live infrastructure was producing any significant noise, it would be most
noticeable as an elevated background noise level during the 'On' survey at night. The figure below
clearly illustrates that there is no elevated background noise levels at night for the 'On' survey.

Figure 4.2: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night
(C’Fall - Srananagh 110kV OHL)

The LA5 parameter records the top 5% of noisy activity within a given measurement period. One would
expect that significant increases in corona activity when comparing the 'On' survey with the 'Off' survey
would result in a significant increase in the LA5 parameter for the 'On' survey as compared with the off
survey. Table 4.3 illustrates that for all weather conditions and during all time periods, the LA5
parameter is higher during the 'On' survey as compared with the 'Off' survey.

This may initially indicate that the difference in LA5 measurements may be linked to higher corona
discharge events during the 'On' survey. However, upon viewing all of the data in Table 4.3, it is clear
that all of the datasets for all of the parameters are higher during the 'On' survey when compared with
the 'Off' survey before any of the data is refined to exclude rain or wind affected data. When the wind

51
Evidence Based Studies Noise – Main Report

data is removed, all of the parameters, including the LA5, show closer correlation between the 'On' and
'Off' surveys. This would indicate that it was the higher wind speeds during the 'On' survey that was
giving higher noise levels for all parameters (including LA5) and not corona discharge.

As corona noise is commonly linked with wet weather conditions, an analysis exercise was undertaken
to isolate all of the rain affected LA5 measurement data and compare the 'On' and 'Off' surveys for the
various times of day. Unfortunately, there wasn't enough rain affected data in each daytime period for
each survey (i.e. 'On' and 'Off') to make a viable comparison.

The weather-adjusted noise levels included in Table 4.3 are below the daytime WHO threshold limits
for serious annoyance (55dB LAeq) and moderate annoyance (50dB LAeq) for outdoor living areas and
the night-time free-field threshold limit of 42dB (LAeq) for preventing negative effects on sleep.

A review of this data in the context of all 110kV survey data is included in Section 6.1 of this study.

4.1.4 Golagh Tee - Letterkenny 110kV OHL

Damage was caused to the weather station during the 'Off' survey for this location, and therefore it
was not possible to analyse the effect of weather conditions on noise levels during this survey. As
three full 'On' and 'Off' surveys had been completed at other locations adjacent to existing 110kV
overhead lines (as discussed above), and on account of the difficulty in analysing this data without the
weather records, a decision was made not to analyse the datasets for this location.

4.2 SURVEY RESULTS FOR 220KV OVERHEAD LINES

4.2.1 Dunstown-Maynooth 220kV OHL - Location 1 Betaghstown

Table 4.4 below includes a summary of all averaged data from the analysis of the synchronised noise
and weather data at Location 1 (Betaghstown) on the Dunstown-Maynooth 220kV OHL.

For each noise parameter, the averaged noise levels for the survey with the infrastructure live (i.e. 'On'
in the table) has been placed beside the averaged noise levels for the survey complete during the
period of outage (i.e. 'Off' in the table). This enables direct comparison to be made between 'On' and
'Off' surveys for each of the time periods and weather conditions.

The 'On' survey produced a dataset of approximately 143 hours for the purposes of analysis, while the
'Off' survey produced a dataset of approximately 102 hours.

52
Evidence Based Studies Noise – Main Report

Table 4.4: Comparison of Averaged Data from 'On' and 'Off' Surveys at Location 1
(Betaghstown) on the Dunstown-Maynooth 220kV OHL

Averaged Noise Levels dB(A)

Dataset LA5 LA50 LA90 LAeq
On Off On Off On Off On Off
All Noise Data 46.5 44.2 41.3 39.7 38.3 37.6 45.3 46.0
- Rain Data 46.5 44.2 41.0 39.7 38.0 37.7 45.0 46.0
< 10mph wind speed 45.2 43.9 38.7 39.0 36.7 37.5 44.5 46.0
< 5mph wind speed 44.4 48.5 38.7 39.0 36.0 37.0 44.0 46.0

Weekday 49.6 45.1 44.5 41.0 41.3 38.0 47.1 46.0

- Rain Data 49.5 45.8 44.3 41.0 41.3 38.0 47.0 46.0
< 10mph wind speed 49.1 45.9 44.0 41.0 40.6 38.0 46.9 46.0
< 5mph wind speed 50.4 45.8 45.7 41.0 42.4 38.0 47.5 46.0

Night 42.2 41.4 36.5 38.0 34.6 36.7 41.3 46.0

- Rain Data 41.9 41.3 36.2 37.9 34.6 36.7 40.8 46.0
< 10mph wind speed 40.2 41.2 34.2 37.6 32.3 36.4 38.6 46.0
< 5mph wind speed 39.6 41.0 33.5 37.0 31.8 35.6 37.8 46.0

Quiet Day 47.7 45.4 42.3 40.5 38.6 38.0 44.9 46.0
- Rain Data 47.8 45.3 42.4 40.5 38.6 38.0 44.9 46.0
< 10mph wind speed 46.2 44.5 40.5 40.0 36.7 37.8 42.8 46.0
< 5mph wind speed 46.3 42.6 40.7 38.2 36.9 36.2 41.9 46.0

The comparison of the averaged noise levels for all parameters included in the table above
demonstrate that measured noise levels during the 'On' survey are generally similar to those for the
'Off' survey. In the critical time periods where overall noise levels are reduced (i.e. night-time),
recorded noise levels were no greater during that 'On' survey than for the 'Off' survey.

Figure 4.3 below includes a typical cross-section of data points for the background noise level (i.e.
LA90) for both 'On' and 'Off' surveys during the night-time period only. This dataset has been refined to
exclude all rain and wind affected measurements. The y axis has been altered to give a simple
numerical value to represent the passage of time as opposed to having a date and time. This has
been done for ease of reference and because two datasets are being represented which have different
measurement dates and times.

In the figure below, the background noise levels are quite similar. As in the case of the previous
illustrations on the measured background noise levels, if the live infrastructure was producing any
significant noise, it would be most noticeable as an elevated background noise level during the 'On'
survey at night. The figure below clearly illustrates that there is no elevated background noise levels
at night for the 'On' survey.

53
Evidence Based Studies Noise – Main Report

Figure 4.3: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night at
Location 1 (Betaghstown) on the Dunstown-Maynooth 220kV OHL

The LA5 parameter records the top 5% of noisy activity within a given measurement period. One would
expect that significant increases in corona activity when comparing the 'On' survey with the 'Off' survey
would result in a significant increase in the LA5 parameter for the 'On' survey as compared with the off
survey. Table 4.1 illustrates that for all weather conditions and during all time periods, the LA5
parameter is generally similar during the 'On' survey as compared with the 'Off' survey (in some
instances it is higher for the 'On' survey and in other instances it is higher for the 'Off survey). This
indicates that either random corona discharges were not frequent enough throughout the 'On' survey
to register a noticeable change in the recorded noise level or that if they were relatively frequent, they
were not of sufficient noise energy as to register a change in the overall LA5 measurements.

Either way, the data does not give any evidence that random corona discharge is a significant noise
impact from 220kV overhead lines.

As corona noise is commonly linked with wet weather conditions, an analysis exercise was undertaken
to isolate all of the rain affected LA5 measurement data and compare the 'On' and 'Off' surveys for the
various times of day. However, there wasn't enough rain affected data in each daytime period for
each survey (i.e. 'On' and 'Off') to make a viable comparison.

The dataset for this location provides strong evidence that the 220kV line in this location does not
produce any significant noise emissions at a location immediately adjacent to the OHL. The weather-
adjusted noise levels included in Table 4.4 for the 'On' survey are below the daytime WHO threshold

54
Evidence Based Studies Noise – Main Report

limits for serious annoyance (55dB LAeq) and moderate annoyance (50dB LAeq) for outdoor living areas
and the night-time free-field threshold limit of 42dB (LAeq) for preventing negative effects on sleep.

A review of this data in the context of all 220kV survey data is included in Section 6.2 of this study.

4.2.2 Dunstown-Maynooth 220kV OHL - Location 2 Currabell

Table 4.5 below includes a summary of all averaged data from the analysis of the synchronised noise
and weather data at Location 2 (Currabell) on the Dunstown-Maynooth 220kV OHL.

For each noise parameter, the averaged noise levels for the survey with the infrastructure live (i.e. 'On'
in the table) has been placed beside the averaged noise levels for the survey complete during the
period of outage (i.e. 'Off' in the table). This enables direct comparison to be made between 'On' and
'Off' surveys for each of the time periods and weather conditions.

The 'On' survey produced a dataset of approximately 141 hours for the purposes of analysis, while the
'Off' survey produced a dataset of approximately 188 hours.

Table 4.5: Comparison of Averaged Data from 'On' and 'Off' Surveys at Location 2 (Currabell)
on the Dunstown-Maynooth 220kV OHL

Averaged Noise Levels dB(A)

Dataset LA5 LA50 LA90 LAeq
On Off On Off On Off On Off
All Noise Data 43.0 46.1 39.5 41.3 37.7 39.0 44.6 49.0
- Rain Data 42.1 45.1 38.2 39.8 36.3 37.3 41.0 46.8
< 10mph wind speed 41.7 43.3 37.8 38.0 36.0 35.7 40.5 42.8
< 5mph wind speed 41.5 41.6 37.8 36.3 36.0 33.9 40.5 40.5
Wind speed = 0mph 39.9 39.9 35.7 34.2 33.8 31.9 38.5 37.4

Weekday 46.8 49.6 42.5 44.5 40.6 42.1 47.9 52.1

- Rain Data 45.8 49.6 41.4 43.9 33.4 41.1 43.6 49.9
< 10mph wind speed 45.2 47.3 40.8 41.7 39.1 39.2 42.9 44.3
< 5mph wind speed 45.6 46.8 41.7 41.1 40.1 38.4 43.7 45.8
Wind speed = 0mph 45.8* 49.7 42.3* 42.6 40.7* 39.2 43.1* 42.8*

Night 39.3 43.2 35.8 39.0 34.0 37.0 40.7 46.7

- Rain Data 37.9 40.2 34.2 35.8 32.3 33.7 35.9 41.7
< 10mph wind speed 37.0 38.7 33.3 34.3 32.3 33.7 35.9 35.9
< 5mph wind speed 38.1 37.5 34.5 32.9 32.7 31.7 36.2 35.6
Wind speed = 0mph 37.4 37.2 33.5 31.9 31.7 30.0 36.0 36.8
Rain (Wind <6m/s) 46 47 44 43 41 39 46 45

Quiet Day 43.4 44.9 39.9 39.6 38.2 37.0 42.3 43.7
- Rain Data 42.5 44.6 39.0 39.0 37.1 36.3 40.4 43.1
< 10mph wind speed 42.2 42.0 38.7 38.2 36.9 35.6 40.1 41.9
< 5mph wind speed 41.9 43.0 38.4 37.5 36.6 40.7 40.1 41.2
Wind speed = 0mph 40.0 41.7 36.2 35.8 34.2 33.3 38.3 38.5
Rain (Wind <6m/s) 46 47 43 45 42 44 45 45
* Small dataset, i.e. less than 2.5 hours of data

55
Evidence Based Studies Noise – Main Report

The comparison of the averaged noise levels for all parameters included in the table above
demonstrate that measured noise levels during the 'On' and 'Off' surveys show very little difference.

The data included in Table 4.5 provides strong evidence that 220kV overhead lines do not produce
significant noise emission that contribute to increasing the background (i.e. LA90) or ambient (i.e. LAeq)
noise levels in the vicinity of them. There are marginal differences between all of the measured noise
parameters (i.e. LAeq, LA5, LA50, LA90) for 'On' and 'Off' surveys but these differences are small and fall
within the range of natural variation that would be expected in a low noise environment.

To further illustrate the similarity in the recorded background noise levels for the 'On' and 'Off' surveys,
Figure 4.4 gives a typical snapshot of the background noise data for both surveys at night with rain
and wind affected data removed. This figure follows the same format as explained for Figure 4.3.

Figure 4.4: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night at
Location 2 (Currabell) on the Dunstown-Maynooth 220kV OHL

If the 220kV OHL was producing any significant noise emissions at this location, one would expect an
obvious effect on the graph for the 'On' survey demonstrating a consistently elevated noise level.
Figure 4.4 demonstrates a similar pattern for the background noise level at night for both surveys.

As in the case of the Betaghstown dataset (i.e. Table 4.4), an analysis of the LA5 data in Table 4.5
indicates that the measurements are very similar for the 'On' and 'Off' surveys (actually, marginally
lower for the 'On' survey). Any significant random corona discharge activity would be expected to
result in an increased LA5 noise level during the 'On' survey, therefore there is no evidence to indicate
that random corona discharge presents a significant noise impact on 220kV OHL.

56
Evidence Based Studies Noise – Main Report

As corona noise is commonly linked with wet weather conditions, an analysis exercise was undertaken
to isolate all of the rain affected LA5 measurement data and compare the 'On' and 'Off' surveys for the
various times of day. Comparison of LA5 data during the quietest periods of the day (i.e. night and
quiet day) during rainfall gave a difference of 1dB between the datasets. A similar difference of 1-2dB
was observed between the other parameters (i.e. LAeq, LA50, LA90) with the 'On' survey being greater in
some instances and the 'Off' survey being greater in other instances.

This data does not give any evidence that there is any significant noisy corona activity on 220kV lines
during wet weather conditions.

The weather-adjusted noise levels included in Table 4.5 are below the daytime WHO threshold limits
for serious annoyance (55dB LAeq) and moderate annoyance (50dB LAeq) for outdoor living areas and
the night-time free-field threshold limit of 42dB (LAeq) for preventing negative effects on sleep.

A review of this data in the context of all 220kV survey data is included in Section 6.2 of this study.

4.2.3 Dunstown-Maynooth 220kV OHL - Location 3 Thomastown

Table 4.6 below includes a summary of all averaged data from the analysis of the synchronised noise
and weather data at Location 3 (Thomastown) on the Dunstown-Maynooth 220kV OHL.

For each noise parameter, the averaged noise levels for the survey with the infrastructure live (i.e. 'On'
in the table) has been placed beside the averaged noise levels for the survey complete during the
period of outage (i.e. 'Off' in the table). This enables direct comparison to be made between 'On' and
'Off' surveys for each of the time periods and weather conditions.

The 'On' survey produced a dataset of approximately 44 hours for the purposes of analysis, while the
'Off' survey produced a dataset of approximately 48 hours.

57
Evidence Based Studies Noise – Main Report

Table 4.6: Comparison of Averaged Data from 'On' and 'Off' Surveys at Location 3
(Thomastown) on the Dunstown-Maynooth 220kV OHL

Averaged Noise Levels dB(A)

Dataset LA5 LA50 LA90 LAeq
On Off On Off On Off On Off
All Noise Data 40.2 40.5 34.9 35.7 32.2 33.4 39.1 39.6
- Rain Data 40.2 39.6 34.8 34.8 32.0 33.4 39.1 38.1
< 10mph wind speed 40.2 39.4 34.8 32.3 32.1 32.2 39.1 37.9
< 5mph wind speed 39.5 38.2 34.0 33.4 31.3 31.1 37.9 36.7
Wind speed = 0mph 37.9 37.7 31.9 32.6 29.0 30.1 36.4 34.7

Weekday 43.2 44.6 38.4 39.9 35.9 37.6 41.4 42.3

- Rain Data 42.0 43.9 38.3 39.1 35.9 37.6 41.4 41.1
< 10mph wind speed 42.0 44.3 38.3 39.4 34.8 37.1 41.4 41.4
< 5mph wind speed 42.4 43.0 37.4 38.4 35.1 36.4 40.0 40.7
Wind speed = 0mph 43.3* ND 39.3* ND 37.5* ND 40.3* ND

Night 35.8 35.9 30.1 30.5 27.0 28.0 34.5 33.3

- Rain Data 35.8 35.8 30.1 30.4 27.0 28.0 34.5 33.3
< 10mph wind speed 35.8 35.8 30.1 30.4 27.0 27.9 34.5 33.3
< 5mph wind speed 35.3 35.8 30.0 30.4 27.0 27.9 34.5 33.3
Wind speed = 0mph 36.1 36.5 29.4 31.3 26.4 28.8 34.3 33.9

Quiet Day 40.2 39.0 35.5 35.1 32.8 33.0 37.5 36.6
- Rain Data 40.2 39.0 35.5 35.1 32.8 33.0 37.5 36.6
< 10mph wind speed 40.2 39.0 35.5 35.1 32.8 33.0 37.5 36.6
< 5mph wind speed 40.2 38.3 35.5 34.6 32.8 32.4 37.5 36.2
Wind speed = 0mph 39.5 37.5* 34.7 33.1* 31.9 30.0* 37.1 34.0*
* Small dataset, i.e. less than 2.5 hours of data, ND = No Data

The comparison of the averaged noise levels for all parameters included in the table above
demonstrate that measured noise levels during the 'On' and 'Off' surveys show very little difference.

The data included in Table 4.6 provides strong evidence that 220kV OHL do not produce significant
noise emission that contribute to increasing the background (i.e. LA90) or ambient (i.e. LAeq) noise
levels in the vicinity of them. There are marginal differences between the measured noise parameters
(i.e. LA90, LAeq, LA5, LA50) for 'On' and 'Off' surveys (the measured noise levels for the 'Off' surveys are
marginally greater in this instance) but these differences are small and fall within the range of natural
variation that would be expected in a low noise environment.

Figure 4.5 provides a snapshot of the background noise data during 'On' and 'Off' surveys at night with
rain and wind affected data removed. As with the previous sites as Betaghstown and Currabell, the
pattern is similar for 'On' and 'Off' surveys and illustrates that there is no significant noise being
emitted from the 220kV overhead line.

58
Evidence Based Studies Noise – Main Report

Figure 4.5: Comparison of Background Noise Levels (LA90) for 'On' and 'Off' Surveys at Night at
Location 3 (Thomastown) on the Dunstown-Maynooth 220kV OHL

An analysis of the LA5 data in Table 4.6 indicates that the measurements are very similar for the 'On'
and 'Off' surveys. Any significant random corona discharge activity would be expected to result in an
increased LA5 noise level during the 'On' survey, therefore as with the Betaghstown and Currabell
survey results, there is no evidence to indicate that random corona discharge presents a significant
noise impact on 220kV OHL.

As corona noise is commonly linked with wet weather conditions, an analysis exercise was undertaken
to isolate all of the rain affected LA5 measurement data and compare the 'On' and 'Off' surveys for the
various times of day. However, there wasn't enough rain affected data in each daytime period for
each survey (i.e. 'On' and 'Off') to make a viable comparison.

The weather-adjusted noise levels included in Table 4.6 are below the daytime WHO threshold limits
for serious annoyance (55dB LAeq) and moderate annoyance (50dB LAeq) for outdoor living areas and
the night-time free-field threshold limit of 42dB (LAeq) for preventing negative effects on sleep.

A review of this data in the context of all 220kV survey data is included in Section 6.2 of this study.

59
Evidence Based Studies Noise – Main Report

4.3 SURVEY RESULTS FOR 400KV OVERHEAD LINES

4.3.1 Oldstreet-Woodland 400kV OHL - Ardrums Great

Table 4.7 below includes a summary of all averaged data from the analysis of the synchronised noise
and weather data at Ardrums Great on the Oldstreet-Woodland 400kV line.

For each noise parameter, the averaged noise levels for the survey with the infrastructure live (i.e. 'On'
in the table) has been placed beside the averaged noise levels for the survey complete during the
period of outage (i.e. 'Off' in the table). This enables direct comparison to be made between 'On' and
'Off' surveys for each of the time periods and weather conditions.

The 'On' survey produced a dataset of approximately 240 hours for the purposes of analysis, while the
'Off' survey produced a dataset of approximately 139 hours.

Table 4.7: Comparison of Averaged Data from 'On' and 'Off' Surveys at Ardrums Great on the
Oldstreet-Woodland 400kV OHL

Averaged Noise Levels dB(A)

Dataset LA5 LA50 LA90 LAeq
On Off On Off On Off On Off
All Noise Data 48.0 44.8 44.0 40.7 42.0 38.5 52.2 44.9
- Rain Data 47.8 44.8* 43.9 40.7* 42.1 38.6* 52.0 45.0*
< 10mph wind speed 44.2 44.0* 41.3 40.1* 39.9 38.5* 43.7 44.9*
< 5mph wind speed 43.4 41.9* 40.8 39.4* 39.5 36.9* 42.7 44.9*

Weekday 49.4 49.5 45.0 44.2 43.1 41.4 50.1 47.6

- Rain Data 49.4 49.6* 45.0 44.2* 43.1 41.0* 50.1 47.7*
< 10mph wind speed 47.0 49.5* 43.4 43.6* 41.8 41.3* 45.8 47.8*
< 5mph wind speed 45.8 46.5* 43.0 41.6* 41.7 41.1* 44.6 47.7*

Night 46.3 41.0 42.6 38.2 40.6 36.3 54.3 42.1

- Rain Data 44.8 41.0* 42.3 38.2* 39.4 36.3* 54.0 42.1*
< 10mph wind speed 42.0 41.0* 39.4 37.5* 37.9 35.7* 41.4 42.1*
< 5mph wind speed 42.6 38.6* 39.4 36.7* 38.0 34.9* 40.2 41.8*

Quiet Day 47.9 44.1 44.3 39.8 42.7 37.9 50.6 43.8
- Rain Data 47.0 43.7* 44.2 39.9* 42.0 37.6* 50.6 43.8*
< 10mph wind speed 44.4 44.2* 42.3 39.1* 40.4 37.2* 43.5 43.9*
< 5mph wind speed 44.2 43.6* 41.8 39.2* 40.5 37.3* 44.1 44.1*
* Portions of weather data missing, therefore refinement to dataset only completed where weather data available.

The portion of the weather data was not recorded during the 'Off' survey on account of an error in the
functioning of the weather station; therefore the approach to the analysis of these datasets was a little
different to the strategy outlined for the 110kV and 220kV surveys. As some of the weather data was
missing for the 'Off' survey, it was not possible to refine this dataset to make a direct equivalent
comparison with the 'On' survey. For this reason, the entire dataset for all time periods and including
all rain and wind affected data was used for the purposes of comparison between 'On' and 'Off'
surveys.

60
Evidence Based Studies Noise – Main Report

By presenting graphical representations of both surveys for all time periods, elevated noise levels as a
result of wind/rain affected periods will be clearly visible on the graph. However, of more use to this
study, quiet periods can be directly observed on the graph.

Figure 4.6 illustrates all of the LA90 noise data for all of the time periods without any refinement to
remove rain and wind affected data. There are sections of the data that show clear signs of been
heavily influenced by weather conditions, most notably the significant peaks in the blue graph (i.e. the
'On' Survey. A large portion of the data for both graphs is centred around the 40dB(A) line, indicating
a relative similarity between the core section of the datasets for both surveys. Also, the regular
troughs in the both surveys approaching the 30dB(A) line (which represent the night time LA90 during
this surveys) show a relatively similar drop for both 'On' and 'Off' surveys.

If there was significant steady state noise emissions from the 400kV OHL, one would expect the low
noise troughs in the graph for the 'On' survey to be higher than those for the 'Off' survey as this steady
state noise from the OHL would result in a higher base noise level for the background noise during
quiet periods. This figure provides a good indication that the 400kV OHL is not emitting any significant
detectable steady state noise level which is above the background noise level for the 'Off' survey.

Figure 4.6: Comparison of All Background Noise Levels (LA90) for 'On' and 'Off' Surveys on the
Oldstreet-Woodland 400kV OHL (Rain & Wind Affected Data Included)

The LA5 parameter records the top 5% of noisy activity within a given measurement period. One would
expect that significant increases in corona activity when comparing the 'On' survey with the 'Off' survey
would result in a significant increase in the LA5 parameter for the 'On' survey as compared with the off

61
Evidence Based Studies Noise – Main Report

survey. Figure 4.7 illustrates that for all weather conditions and during all time periods, the LA5
parameter is generally similar during the 'On' survey as compared with the 'Off' survey.

It must be noted that there was no significant period of rainfall during the 'On' survey (i.e. two 15-
minute measurements only) and therefore, it would not have been possible (even if some of the
weather data had not been lost during the 'Off' survey) to make any analysis of possible corona noise
effects during rain affected conditions from this survey. The fact that no measurements were
undertaken on live infrastructure during periods of rain means that no conclusion can be drawn from
this survey in relation to potential corona effects from 400kV overhead lines during rainy conditions.

Figure 4.7: Comparison of All LA5 Noise Levels for 'On' and 'Off' Surveys on the Oldstreet-
Woodland 400kV OHL (Rain & Wind Affected Data Included)

The weather-adjusted noise levels included in Table 4.7 for the 'On' survey are below the daytime
WHO threshold limits for serious annoyance (55dB LAeq) and moderate annoyance (50dB LAeq) for
outdoor living areas and the night-time free-field threshold limit of 42dB (LAeq) for preventing negative
effects on sleep.

A detailed discussion of this data is included in Section 6.3 of this study.

62
Evidence Based Studies Noise – Main Report

4.4 SURVEY RESULTS FOR SUBSTATIONS

4.4.1 Dunfirth 110kV Substation

Noise monitoring was completed at four different locations in the vicinity of the Dunfirth 110kV
Substation, two at alternative distances (i.e. 5m and 10m) from the southern boundary of the
substation and two at alternative distances from (i.e. 5m and 10m) from the western boundary of the
substation site. No measurements were taken at the northern boundary of the substation as the road
would be the dominant noise source at this location. It was not possible to access the eastern
boundary on account of the excessive amounts of vegetation.

Five minute measurements were completed at each location as a steady state noise was being
emitted from the substation. The noise meter was paused to exclude any other noise sources (e.g.
passing traffic on road etc.) during the measurement periods. For the measurements completed at the
southern boundary, subjective observations were recorded that pylon corona noise was audible at low
levels. On the western boundary, the subjective observations were that substation noise, pylon
corona noise and noise from the T210 converter/transformer were audible.

Table 4.8 presents the noise monitoring data from the survey at the Dunfirth 110kV substation on 19th
March 2013.

Table 4.8: Noise Monitoring Data for Dunfirth 110kV Substation

Boundary Distance Date/ LAeq LAmax LAmin LA5 LA50 LA90

from Time [dB(A)] [dB(A)] [dB(A)] [dB(A)] [dB(A)] [dB(A)]
Boundary
(m)
Southern 5 14:59 39 54 32 43 36 34
Southern 10 15:07 39 52 32 43 36 34
Western 5 15:16 37 53 31 40 35 34
Western 10 15:27 38 49 35 41 38 36

As noise propagates from a noise source, the noise energy dissipates at a constant rate as the
distance from the source increases. Therefore, if a particular noise source is dominant, successive
noise measurements at a greater distance from the source should show a progressive reduction in the
measured noise levels as you travel away from the source.

In this instance, most parameters (i.e. LA50, LA5, LA90) including the average noise level (i.e. LAeq) do not
demonstrate any drop in the noise level at measurement locations 5m and 10m from the respective
boundaries. This data indicates that the substation noise stops being the dominant noise source
within a short distance of exiting the substation boundary (i.e. within the first 10m). This does not
mean that the substation noise is not contributing to the ambient noise level in the vicinity of the
boundary, as the subjective observations confirm that corona/substation/ transformer noise is audible.
Nevertheless, it provides valuable information that noise from the 110kV substation is quite low in
close proximity to the substation boundary (i.e. <40dB LAeq).

63
Evidence Based Studies Noise – Main Report

The measured noise levels at the boundary of this substation are below the daytime WHO threshold
limits for serious annoyance (55dB LAeq) and moderate annoyance (50dB LAeq) for outdoor living areas.
They are also below the night-time free-field threshold limit of 42dB (LAeq) for preventing negative
effects on sleep.

Spectral data was recorded throughout each of the measurements completed at Dunfirth 110kV
substation to determine if there were any tonal features in the recorded noise levels. Figure 4.8
illustrates the typical spectral profile recorded during all measurements at the site. The low frequency
range (i.e. 12-315Hz) is dominant in all measurements, while there is very little representation in the
high frequency range. Most importantly in terms of this study, there are no distinct peaks in the data
that would represent a distinct tone. On the basis of this data, it can be concluded that there are no
distinct tonal elements to the noise from a 110kV substation.

Figure 4.8: Typical Spectral Profile of Noise Levels at Dunfirth 110kV Substation

4.4.2 Gorman 220kV Substation

Noise monitoring was completed at eight different locations in the vicinity of the Gorman 220kV
Substation, two at alternative distances (i.e. 5m and 10m) from each boundary of the substation. Five
minute measurements were completed at each location as a steady state noise was being emitted
from the substation. The noise meter was paused to exclude any other noise sources (e.g. passing
traffic on road etc.) during the measurement periods.

64
Evidence Based Studies Noise – Main Report

Subjective observations taken during the measurement survey recorded that substation noise, pylon
corona noise and T210 converter/transformer noise was audible along all four boundaries of the
substation. Table 4.9 includes the noise monitoring data recorded during the survey at the Gorman
220kV substation.

Table 4.9: Noise Monitoring Data for Gorman 220kV Substation

Boundary Distance Date/ LAeq LAmax LAmin LA5 LA50 LA90

from Time [dB(A)] [dB(A)] [dB(A)] [dB(A)] [dB(A)] [dB(A)]
Boundary
(m)
Southern 5 10:13 43 48 40 44 43 42
Southern 10 10:20 41 48 38 42 41 40
Western 5 10:28 39 55 37 40 39 38
Western 10 10:34 38 51 35 40 37 36
Northern 5 10:45 37 51 33 40 36 34
Northern 10 10:54 37 54 32 41 35 33
Eastern 5 11:04 40 51 36 42 39 38
Eastern 10 11:10 41 50 35 45 39 37

As noise propagates from a noise source, the noise energy dissipates at a constant rate as the
distance from the source increases. Therefore, if a particular noise source is dominant, successive
noise measurements at a greater distance from the source should show a progressive reduction in the
measured noise levels as you travel away from the source.

On two of the boundaries (i.e. northern & eastern), most parameters (i.e. LA50, LA5, LA90, LAeq) do not
show any significant drop between boundary measurement locations at 5m and 10m. The data at
these boundaries indicates that the substation noise stops being the dominant noise source within a
short distance of exiting the substation boundary (i.e. within the first 10m).

On the western boundary, there are marginal reductions of 1-2dB(A) between the 5m and 10m
measurements for a number of the parameters (i.e. LA50, LA90, LAeq) and this is marginally more
pronounced on the southern boundary with a different of 2dB(A) between the average noise levels
recorded (i.e. LAeq). This data indicates that substation noise remains more dominant on the western
and southern boundaries (probably on account of the layout of the plant within the site), but on
account of the relatively small decrease in the measured noise levels between 5m and 10m at these
boundaries, substation noise would not remain dominant far beyond the 10m distance from the
substation boundary.

As with the discussion on the 110kV substation, the substation noise will still contribute to the ambient
noise level beyond the point where the substation noise is dominant, however this contribution will
diminish quickly with distance from the substation boundary. Nevertheless, it provides valuable
information that steady state substation noise from the 220kV substation is quite low in close proximity
to the substation boundary (i.e. 43dB LAeq at 5m).

65
Evidence Based Studies Noise – Main Report

The measured data from the Gorman substation survey illustrates that noise levels from a 220kV
substation are well within the daytime WHO threshold limits for serious annoyance (55dB LAeq) and
moderate annoyance (50dB LAeq) for outdoor living areas and the night-time free-field threshold limit of
42dB (LAeq) for preventing negative effects on sleep.

Spectral data was recorded throughout each of the measurements completed at Gorman 220kV
substation to determine if there were any tonal features in the recorded noise levels. Figure 4.9
illustrates the typical spectral profile recorded during all measurements at the site. As in the case with
the 110kV substation measurements, the noise levels from the substation are dominated by noise
levels in the low frequency range. However, unlike the 110kV substation measurements there are a
number of significant peaks in the mid-frequency range (i.e. 300-600Hz) and one at the lower
frequency range (i.e. approximately 80Hz) which fulfil the description of being tonal.

Figure 4.9: Typical Spectral Profile of Noise Levels at Gorman 220kV Substation

Figure 4.9 illustrates that there are tonal qualities to the noise from 220kV substations. This is
discussed further in Sections 6 and 7 of this study.

4.4.3 Woodland 400kV Substation

Noise monitoring was completed at nine different locations in the vicinity of the Woodland 400kV
Substation, two at alternative distances (i.e. 5m and 10m) from the southern, western and eastern
boundaries of the substation and three at alternative distances from the transformer (i.e. 11m, 22 &
47m). Five minute measurements were completed at each location as a steady state noise was being

66
Evidence Based Studies Noise – Main Report

emitted from the substation. The noise meter was paused to exclude any other noise sources (e.g.
passing traffic on road etc.) during the measurement periods.

Subjective observations taken during the measurement survey recorded that substation noise, pylon
corona noise and transformer noise was audible along all four boundaries of the substation. Table
4.10 includes the noise monitoring data recorded during the survey at the Woodland 400kV substation.

Table 4.10: Noise Monitoring Data for Woodland 400kV Substation

Boundary Distance Date/ LAeq LAmax LAmin LA5 LA50 LA90

from Time [dB(A)] [dB(A)] [dB(A)] [dB(A)] [dB(A)] [dB(A)]
Boundary
(m)
Southern 5 14:23 41 53 36 43 40 39
Southern 10 14:38 42 48 37 43 41 40
Western 5 14:52 51 56 49 52 50 50
Western 10 15:00 50 60 48 51 50 49
Transformer 11 15:08 53 55 51 55 53 52
Transformer 22 15:15 52 54 50 53 52 51
Transformer 47 15:24 48 50 46 49 48 47
Eastern 5 15:45 46 57 43 48 46 44
Eastern 10 15:53 49 56 45 51 49 47

As described in Sections 4.4.1 and 4.4.2, if a particular noise source is dominant, successive noise
measurements at a greater distance from the source should show a progressive reduction in the
measured noise levels as one travels away from the source.

The measurement data included in Table 4.10 shows no progressive reduction in the noise levels for
all relevant noise parameters as distance increases from the southern, western and eastern
boundaries. The data for the transformer measurements show a clear progressive reduction in the
noise levels for all parameters with distance from the transformer. The measurement strongly indicates
that the transformer is the dominant noise source throughout the substation and that it remains the
dominant noise source up to distances of at least 50m away from it.

The noise survey at Woodland 400kV substation illustrates how the higher voltage substations are
significantly louder than the 110kV and 220kV substations, producing average noise levels of greater
than 50dB(A) beyond a 20m distance from the boundary of the substation.

Spectral data was recorded throughout each of the measurements completed at Woodlands 400kV
substation to determine if there were any tonal features in the recorded noise levels. Figure 4.10
illustrates the typical spectral profile recorded during all measurements at the site. This figure
demonstrates how the overall noise level from the substation is dominated by noise levels in the low
and mid frequency ranges. The data shows clear tonal peaks at approximately 100Hz, 200Hz and
300Hz, which follow the spectral profile discussed in Section 2.3 of this report.

67
Evidence Based Studies Noise – Main Report

Figure 4.10: Typical Spectral Profile of Noise Levels at Woodland 400kV Substation

Figure 4.10 illustrates that there are significant tonal qualities to the noise from 400kV substations.
This is discussed further in Sections 6 and 7 of this report.

68
Evidence Based Studies Noise – Main Report

5 AECOM SURVEY ON 400KV OVERHEAD LINE

5.1 INTRODUCTION

AECOM were separately commissioned by EirGrid to undertake a long term noise survey in the
vicinity of the Dunstown-Moneypoint 400kV OHL in Cloney, Co Kildare. The report produced by
AECOM is included in its entirety in Appendix B and summarised in the sections below. The
conclusions and recommendations from the AECOM report are included in the sections that follow and
help to supplement the overall conclusions and recommendations from this Evidence Based Study.

5.2 METHODOLOGY

Noise monitoring was undertaken at three locations in the vicinity of the existing 400kV OHL: directly
beneath a tower; at a mid-span location between two towers; and a distance of approximately 200m
from the line as a control location. The control location is intended to capture the general background
noise under the same weather conditions, but excluding noise from the OHL.

All noise measurements were made using Rion NL-52 integrating sound level meters and Rion NC-74
acoustic calibrators. The meters were set to log in 10 minute contiguous periods. Monitoring stations
were set up in each location comprising an all-weather equipped sound level meter mounted between
1.2m and 1.5m off the ground in free-field conditions. At the control site, a weather station was also
erected to record rainfall, temperature, ambient air pressure, wind speed and wind direction. All of
these parameters were logged over one minute and ten minute intervals.

The noise meters were set to automatically log the following parameters: LAeq, LAMax, LAmin, LA05, LA10,
LA50 and LA90. Each parameter was logged at 1-minute intervals, allowing more detailed correlation
with the 1-minute rainfall data.

5.3 MEASURED DATA

Table 5.1 presents that average and absolute measurements for the 10-week survey period. During
the 10-week measurement period, a wide variety of weather conditions were recorded. For the first 4
weeks, conditions were predominantly dry with a total of 10.5mm of rain recorded, while the following
6 weeks had a total of 165mm of rainfall. Wind speeds were variable with individual maximum gust
speeds of up to 15m/s recorded, although 10-minute average wind speeds remained below 5m/s for
over 99.5% of the monitoring period. Temperatures remained between 3.5ºC and 18.5ºC for 90% of
the measurement period. Humidity remained above 70% for 90% of the measurement period.

69
Evidence Based Studies Noise – Main Report

Table 5.1: Summary of Measured Noise Levels Over 10-Week Period

Location Noise Level dB(A)

Average from 1-minute Values Absolute over Entire
Monitoring Period
LAeq LAMax LAmin LA10 LA90 Total Maximum Minimum
LAeq LAMax LAMin
Control 37.8 52.8 27.3 40.6 30.3 42.1 91.7 15.6
Tower 41.4 52.8 35.2 43.2 37.7 45.6 97.8 21.1
Mid-span 36.4 50.2 28.1 38.9 30.7 42.2 99.8 15.8

In analysing the data from the 10-week survey, AECOM concluded that while control and mid-span
noise levels generally decreased at night in dry conditions, this was not the case with the tower
locations. Figure 5.1 is a graphical demonstration of this from the AECOM report (see Appendix B).

Figure 5.1: Typical Dry 24-hour Variation in LAeq, 10min Level for All Locations

Figures 5.2 - 5.4 are also taken from AECOM report and show the distribution of noise levels for the
control, mid-span and tower monitoring locations during dry conditions. These figures illustrate the
relative similar distribution between control and mid span locations but illustrate an increase in the
distributed noise levels for the tower location, which is attributable to corona noise.

The report makes the connection between the increase in humidity levels (especially at night) and the
increase in corona noise levels at the tower and postulates that this may contribute to corona noise
being the dominant noise source at night near the tower structure.

70
Evidence Based Studies Noise – Main Report

Figure 5.2: Distribution of LAeq, 1min Levels at the Control Location (Dry Conditions)

Figure 5.3: Distribution of LAeq, 1min Levels at the Mid-Span Location (Dry Conditions)

71
Evidence Based Studies Noise – Main Report

Figure 5.4: Distribution of LAeq, 1min Levels at the Tower Location (Dry Conditions)

The AECOM Report provides separate analysis of rainfall affected data to determine the potential
corona effects that may be prevalent during periods of rain. Figure 5.5 presents a graphical illustration
of average noise levels plotted against rainfall rate. The greater number of recorded noise levels at
lower rainfall rates allow for the smoother line observed at the lower rainfall rates. The greater
variation at higher rates is a consequence of fewer instances of very high rates of rainfall.

The graph illustrates the increased noise level attributed to corona noise between the control site and
the mid-span and especially tower locations, with reported increase of approximately 6dB observed
when comparing the tower and control locations.

Figure 5.5: Measured Noise Levels Against Rainfall Rate for All Sites

72
Evidence Based Studies Noise – Main Report

Spectral analysis of selected audio recordings identified a tonal component to the corona noise at
100Hz, which tallied with the expected tonal frequency based on an expected tonal component at
twice the electrical frequency of the system (which is 50Hz in Ireland).

Figure 5.6 from the report illustrates a sample 1/24-octave band analysis demonstrating the principal
tones and associated harmonic frequencies.

Figure 5.6: 1/24-Octave Band Analysis of Corona Noise at Mid-Span Locations

73
Evidence Based Studies Noise – Main Report

6 DISCUSSION AND CONCLUSION

6.1 DISCUSSION ON SURVEY RESULTS FROM 110KV OVERHEAD LINES

Section 4.1 of the report includes the survey results and analysis of survey data for the survey
locations adjacent to 110kV OHL. As discussed in Section 4.1.4, no weather data was retrieved for
the 'Off' survey at Golagh Tee - Letterkenny 110kV OHL, and therefore no detailed analysis of the data
for this site was undertaken.

Section 4.1.2 highlighted the large difference between the datasets recorded at C’Fall Golagh Tee -
Letterkenny 110kV OHL and how this may have related to the fact that sheep were placed in the field
subsequent to the meter being placed there. On account of what appears to be significant
contamination to the dataset for the 'On' survey from this noise source, this location has not figured
significantly in terms of drawing conclusions from the surveys.

The discussion of the noise associated with 110kV lines is therefore based predominantly on the
survey results and subsequent analysis of the data from C’Fall Golagh Tee - Letterkenny 110kV OHL
and C’Fall - Srananagh 110kV OHL. For both locations, a full dataset of noise and weather
synchronised data was retrieved for both 'On' and 'Off' surveys and no known source of noise
contamination that could have potentially skewed any of the datasets is suspected.

Tables 4.1 and 4.3 indicate that when rain and wind affected data are removed from each of the noise
parameter datasets (i.e. LA5, LA50, LA90, LAeq), the measured noise levels are similar for 'On' and 'Off'
surveys. Particular attention was paid to the background noise level (i.e. LA90) during the quietest
periods of the day (i.e. night). If the 110kV OHL was producing a constant detectable noise emission,
one would expect that this would be most clearly observed during the night when background noise
levels were at their lowest. The background noise levels were observed to be similar for 'On' and 'Off'
surveys at both locations, with no clear sign of elevated noise levels during the 'On' survey. Figures
4.1 and 4.2 were included in Section 4.1 to illustrate a comparable trend to the measured background
noise levels for 'On' and 'Off' surveys at both locations.

While the comparison of the background noise levels during the quietest periods of the day
demonstrates that there is no clear constant noise source from the infrastructure during the 'On'
survey, this parameter does not account for short random events such as corona discharge which may
occur only occasionally (the LA90 represents the noise level exceeded for 90% of the measured time
during a defined measurement period). The LA5 parameter represents the noise level exceeded for
5% of the measured time during a measurement period and would be expected to capture noisy
corona discharge events subject to them being sufficiently common and noisy.

The LA5 measurements summarised in Tables 4.1 and 4.3 do not provide any evidence that corona
discharge events are sufficiently regular or loud as to result in an increase in the measured LA5 noise
levels.

74
Evidence Based Studies Noise – Main Report

Approximately 717 hours of noise data synchronised with weather data was retrieved from C’Fall
Golagh Tee - Letterkenny 110kV and C’Fall - Srananagh 110kV OHL. The analysis of this data for
each of the noise parameters relevant to this study (i.e. LA5, LA50, LA90, LAeq), demonstrates that there
was no increase in the measured noise levels during the 'On' survey when compared with the 'Off'
survey. The analysis of data from the studies on the 110kV OHL provides strong evidence that there
is no significant noise emissions from 110kV OHL.

The results from the 110kV surveys support the findings of the literature review which indicate that
corona noise only starts to become a significant noise issue at voltages of 350-500kV and above.

6.2 DISCUSSION ON SURVEY RESULTS FROM 220KV OVERHEAD LINES

Section 4.2 of the report includes the survey results and analysis of survey data for the survey
locations adjacent to 220kV OHL. Noise surveys were completed at three different locations along the
Dunstown-Maynooth 220kV line - Betaghstown, Currabell and Thomastown. In all, 666 hours of noise
and weather synchronised data was collected for 'On' and 'Off' surveys at all three locations.

Analysis of the datasets at all three locations (see Tables 4.4, 4.5 & 4.6) demonstrate that when rain
and wind affected data was removed, there is no significant difference between the measured noise
levels for all relevant parameters (i.e. LA5, LA50, LA90, LAeq), for 'On' and 'Off' surveys. The data for
quieter periods of the day (i.e. night) do not show elevated noise levels during the 'On' surveys as
illustrated in Figures 4.3, 4.4 and 4.5.

The LA5 parameter records the top 5% of noisy activities during a measurement period and would be
expected to capture noisy corona discharge events subject to them being sufficiently regular and
noisy. The LA5 measurements summarised in Tables 4.4, 4.5 and 4.6 do not provide any evidence
that corona discharge events are sufficiently regular or loud as to modify the measured noise level or
present any significant noise impact.

Significant periods of rain affected data was recorded during the 'On' and 'Off' surveys at the
Dunstown Maynooth 220kV Currabell site. The data for quiet periods of the day (i.e. night and quiet
day) show no significant difference between 'On' and 'Off' survey for all parameters recorded (i.e. LAeq,
LA90, LA50, LA05). This provides strong evidence that significant rain induced corona discharge is not
evident for 220kV OHL.

The results of the analysis of what is a large dataset for all three locations, provides a strong evidence
base for stating that there are no significant noise emissions associated with 220kV OHL.

The results from the 220kV surveys support the findings of the literature review which indicate that
corona noise only starts to become a significant noise issue at voltages of 350-500kV and above.

75
Evidence Based Studies Noise – Main Report

6.3 DISCUSSION ON SURVEY RESULTS FROM 400KV OVERHEAD LINES

Section 4.3 provides an outline of survey data and analysis for one location (Ardrums Great) along the
Oldstreet-Woodland 400kV OHL. It was only possible to complete surveys around the outage
schedule at one location on the 400kV network on account of the shortage of programmed works on
the 400kV network during the survey period for this project.

As discussed in Section 4.3, a portion of the weather data for the 'Off' survey was not recorded on
account of a fault with the weather station. On account of this, the analysis of the Oldstreet - Woodland
400kV OHL data concentrated on the entire datasets and the use of graphical representation to view
noise level trends in the quieter periods for both 'On' and 'Off' surveys.

Figures 4.6 illustrates that LA90 noise levels in the quietest (night-time) periods are similar for 'On' and
'Off' surveys, indicating that the 400kV OHL was not producing a steady state noise level that was
increasing the background noise level at the monitoring location. Figure 4.7 illustrated that the trends
for the LA5 data were similar for both 'On' and 'Off' surveys. Because there was no significant period of
rainfall during the 'On' survey at this location, it was not possible to make any analysis of potential
corona noise effects from 400kV OHL during rain affected conditions. However, the literature review
indicated that corona noise only starts to become a significant issue at voltages of 350-500kV.

On account of the fact that no rain affected data was recorded during the 'On' survey, no conclusions
can be drawn from the survey at Oldstreet Woodland 400kV OHL (Ardrums Great) in relation to the
potential for corona noise effects from 400kV OHL during rainy conditions. For this reason, more
survey work over a longer period of time (to ensure a significant dataset was recorded during rainy
conditions) was required for 400kV OHL. It is in this context that EirGrid commissioned AECOM to
undertake a 10-week survey adjacent to the existing Dunstown-Moneypoint 400kV OHL at Cloney, Co
Kildare (see Section 5). A discussion of the findings of this survey is included in Section 6.4.

6.4 DISCUSSION ON AECOM 400KV SURVEY

Section 5 provides a description of the methodology and measurements from a 10-week noise survey
at locations adjacent to the 400kV OHL at Cloney, Co Kildare. The significant duration of the survey
ensured that a substantial dataset was recorded, including noise measurements under all weather
conditions. In particular, the survey enabled a significant dataset to be recorded during rain affected
conditions to determine if corona effects were a significant noise source associated with 400kV OHL.

The AECOM report shows the distribution of noise levels for the control, mid-span and tower
monitoring locations during dry conditions, illustrating an increase in the distributed noise levels for the
tower location, which the report has attributed to corona noise. The report makes the connection
between the increase in humidity levels (especially at night) and the increase in corona noise levels at

76
Evidence Based Studies Noise – Main Report

the tower and postulates that this may contribute to corona noise being the dominant noise source at
night near the tower structure.

The AECOM report provides separate analysis of rainfall affected data to determine the potential
corona effects that may be prevalent during periods of rain. The graphical output of this analysis
illustrates the increased noise level attributed to corona noise between the control site and the mid-
span and especially tower locations, with reported increase of approximately 6dB observed when
comparing the tower and control locations. Spectral analysis of selected audio recordings identified a
tonal component to the corona noise at 100Hz, which tallied with the expected tonal frequency based
on an expected tonal component at twice the electrical frequency of the system (50Hz in Ireland).

The more limited dataset gathered as part of the Oldstreet - Woodland 400kV OHL (Ardrums Great)
survey meant that definitive conclusions could not be drawn in relation to the potential for significant
corona effects from 400kV OHL. The more extensive 10-week survey completed by AECOM gave the
substantial dataset that enabled more definitive analysis and conclusions to be drawn in relation to
potential corona effects from 400kV OHL. On the basis of this analysis, more definitive
recommendations can be made in relation to the acceptable proximity of residential receptors to
400kV OHL and this is presented in Section 7 of the report.

6.5 DISCUSSION ON SURVEY RESULTS FROM SUBSTATIONS

Section 4.4 includes details on noise surveys completed at different types of substations (i.e. 110kV,
220kV & 400kV) on the Irish network. Short-term noise measurements were taken at various locations
recording the steady state noise level at different boundaries and at different distances from standard
110kV, 220kV and 400kV substations.

The results of the measurement surveys demonstrate that in the case of 110kV and 220kV
substations, the substation noise (including low level corona and transformer noise) stop being the
dominant noise source at a distance of 10-15m from the boundary of the substation. The noise will
still be subjectively audible and contribute to the ambient noise level beyond this distance; however
the extent of this contribution will reduce significantly with distance.

The spectral analysis of the 110kV and 220kV substations demonstrated that there was no distinct
tonal aspects to the 110kV substation noise; however clear tonal elements could be discerned at up to
three locations on the 220kv frequency spectrum. Any recommendations for 220kV substations
should include consideration of the tonal elements to the noise from these substations.

The noise measurements included in the surveys were conducted on the basis of recording steady
state noise from the substation sites. This does not reflect any potential random corona effects that
may arise under certain weather conditions and which may result in short-term increases in the
recorded noise levels from the substation. The survey data from the 110kV and 220kV OHL and the
findings of the literature review indicate that significant noise impacts from corona effects are not

77
Evidence Based Studies Noise – Main Report

common at these voltages (i.e. 110kV & 220kV). The survey methodology for substations did not
provide an opportunity to confirm this, as outlined in Section 3.2.5.

For both 110kV and 220kV substations, the measured noise levels at 10m are well within the daytime
WHO threshold limits for serious annoyance (55dB LAeq) and moderate annoyance (50dB LAeq) for
outdoor living areas and the night-time free-field threshold limit of 42dB (LAeq) for preventing negative
effects on sleep.

In the case of the noise measurement survey for the 400kV substation, the effect of a higher voltage
transformer can be clearly observed in the noise monitoring results associated with the transformer.
The measurement strategy at various distances demonstrated that the transformer was still the
dominant noise source at distances of greater than 50m. The measured average noise levels (LAeq)
were greater than 50dB(A) at a distance of greater than 20m from the substation boundary, which is
above the daytime WHO threshold limit for moderate annoyance in an outdoor living area.

The measured noise levels from the 400kV transformer are above the WHO night-time free-field
threshold for preventing disturbance to sleep (i.e. 42dB) for measurement distances up to 50m. In
addition to this, the spectral analysis of the 400kV substation data demonstrates that there are clear
tonal elements to the substation noise with clear tonal peaks being identified in a number of low to mid
frequencies. On account of the higher measured noise levels from 440kV substations and the more
discrete tonal elements to the noise from these substations, there is a significantly higher potential for
noise impacts from 400kV substations as compared with 110kV and 220kV substations. This is
reflected in the recommendations included in Section 7 of this report.

As in the case of the 110kV and 220kV discussion, the measurement survey for the 400kV substation
did not reflect any potential random corona effects that may arise under certain weather conditions
and which may result in short-term increases in the noise levels from the substation.

The data from the substation surveys provides valuable information on the steady state noise levels
from substations of various types and gives a good profile of the tonal characteristics associated with
the different types of substations. This data provides a valuable tool to enable recommendations to be
presented in Section 7 of this report.

6.6 CONCLUSION

Based on the survey results and analysis included in Section 4 of this report, it is possible to make
robust statements on the noise emissions from electricity infrastructure, which are strongly supported
by the evidence from the surveys and analysis.

Based on the survey results and analysis completed for the 110kV OHL, there is strong evidence
indicating that there are no significant detectable steady state noise emissions from 110kV OHL. The

78
Evidence Based Studies Noise – Main Report

survey results and analysis completed for the 220kV OHL also indicate that there is strong evidence in
support of there being no significant detectable steady state noise emissions from 220kV OHL.

Corona discharge is a well recognised phenomenon associated with electrical infrastructure under
certain weather conditions. The large datasets collected as part of 110kV and 220kV OHL surveys
provide a strong evidence base that random corona discharge is not sufficiently regular or loud
enough to significantly alter measured noise levels from these types of infrastructure. This finding
corroborates the findings from the literature review which indicated that corona noise is not a
significant noise issue below 350kV.

The survey results and analysis completed for the one 400kV OHL, combined with the additional 10-
week survey completed by AECOM at a separate 400kV OHL location, indicate that significant corona
noise effects are observed from 400kV OHL. Under dry conditions, this is limited to immediately
adjacent to tower locations, while under wet conditions this observed for mid span but most notably
tower locations. A maximum difference of 6dB was observed under wet conditions between tower and
control locations. A tonal component to this noise was also recorded. These observations align with
the literature review which indicated that corona noise starts to become an issue at voltages of 350-
500kV and above.

Noise measurement surveys completed at 110kV, 220kV and 400kV substations recorded the steady
state noise levels in the vicinity of the boundaries of these substations. In the vicinity of the 110kV
substation, average noise levels (i.e. LAeq) of less than 40dB(A) were recorded. The measured LAeq
noise levels at the noisiest boundaries were in the range of 40-45dB(A) for the 220kV substation and
50-55dB(A) for the 400kV substation. Spectral analysis of all substation types demonstrated that there
are distinct tonal elements to the noise from 220kV and 400kV substations

79
Evidence Based Studies Noise – Main Report

7 RECOMMENDATIONS

On the basis of the results of this study, a number of recommendations have been included in this
section for the purpose of protecting the amenity of noise sensitive receptors in the vicinity of any
proposed transmission infrastructure development. The recommendations are presented in the context
of being of use as part of the design and siting process for such future projects.

The noise studies on 110kV and 220kV OHL present a strong database of evidence that indicates that
these lines do not produce steady state noise levels that are likely to result in a significant noise
impacts at receptors in the vicinity of them. There is no evidence that random corona discharge events
are sufficiently regular or loud as to result in significant noise impacts to noise sensitive receptors in
their vicinity. On this basis, the planning of 110kV and 220kV OHL should not be significantly
constrained on the basis of potential noise issues associated with these types of infrastructure.

The noise study on the 400 kV OHL provided a strong database of evidence to indicate that these
lines do produce significant corona noise effects under certain conditions, which are immediately
adjacent to the tower under dry humid conditions and at tower and mid span locations under wet
conditions. The literature review indicates that corona effects start to become significant in noise
impact terms at voltages in the range of 350-500kV, and the evidence presented in this study would
appear to verify this.

While Section 5 does illustrate the potential for significant corona noise under certain conditions
immediately adjacent to 400kV towers and to a lesser extent mid span OHL, the noise levels from
these corona effects are relatively low (i.e. low 40s dB) in the context of other typical environmental
noise sources such as road traffic noise. However, there is potential for noise impacts from such
corona effects under certain conditions such as properties located very close to such infrastructure in
quiet rural locations (especially at night under humid or wet conditions). The potential for noise
impacts is further exacerbated by the tonal characteristics associated from this corona noise.

The standard Irish reference document for undertaking noise assessments is the Environmental
Protection Agency (EPA) Office of Environmental Enforcement (OEE) Guidance Note for Noise:
Licence Applications, Surveys and Assessments in Relation to Scheduled Activities (NG4). NG4
would recommend the use of a tonal correction of +5dB(A) to account for the tonal characteristics of a
noise source such as that from corona noise. Any assessment of noise impacts from 400kV lines
should be undertaken on the basis of the worst-case corona noise levels as included in Section 5 of
this report, but including an additional 5dB(A) penalty to account for the tonal characteristics of the
noise as outlined in the EPA NG4 document.

Setting an absolute recommended minimum distance between a 400kV tower or OHL and a sensitive
property is difficult as it is very much dependent on the background noise levels at that particular
property. As outlined in the previous paragraph, the best approach is to determine an appropriate

80
Evidence Based Studies Noise – Main Report

distance based on worst-case corona noise levels (including tonal correction) assessed against the
background noise level (i.e. LA90) at the particular property in question.

As a practical example, the AECOM report used a control at 200m distance, at which location no noise
impacts from the 400kV OHL were observed. Similarly, the potential noise effects from the tower were
observed to be significantly greater than the mid span OHL. As a rule of thumb for undertaking high
level assessments prior to undertaking more detailed assessment on a property by property basis, it
may be appropriate to use a recommended reference separation distance of 200m between any
property and a 400kV tower and 100m between any property and an OHL.

The survey on the 110kv substation at Dunfirth indicated that measured noise levels (LAeq) were less
than 40dB(A) at 5m from each of the boundaries of the substation. This is below the WHO night-time
free-field threshold limit of 42dB for preventing effects on sleep and well below the WHO daytime
threshold limits for serious and moderate annoyance in outdoor living areas (i.e. 55dB & 50dB
respectively). Spectral analysis of the data recorded at this site demonstrated that there were no
distinct tonal elements to the recorded noise level. To avoid any noise impacts from 110kV
substations at sensitive receptors, it is recommended that a minimum distance of 5m is maintained
between a 110kV substation and the land boundary of any noise sensitive property.

The survey on the 220kv substation at Gorman indicated that measured noise levels (LAeq) were
approximately 43dB(A) at 5m from the most affected boundary of the substation. This is marginally
above the WHO night-time threshold limit for preventing disturbance to sleep (i.e. 42dB). Spectral
analysis of the noise from the Gorman substation demonstrated that there are a number of distinct
tonal elements to noise in the low to mid frequency range. To avoid any noise impacts from 220kV
substations at sensitive receptors, it is recommended that a distance of 20m is maintained between
the nearest site boundary and the nearest sensitive receptor.

The survey on the 400kv substation at Woodland indicated that measured noise levels (LAeq) were
approximately 53dB(A) at 11m from the main transformer in the substation. The measured noise
levels also indicated that the substation/transformer noise remained the dominant noise source at
distances of 50m and greater from the boundary of the substation.

The measured noise data from the 400kV substation is above the WHO daytime threshold limit for
moderate annoyance in outdoor living areas (50dB) at 11m and significantly above the WHO night-
time free-field threshold limit for preventing disturbance to sleep (42dB) at 47m distance from the
substation boundary. Spectral analysis also demonstrated that there was a number of distinct tonal
elements to the measured noise level in the low frequency range.

On account of the significant broadband noise levels and tonal noise elements associated with 400kV
substations, it is recommended that a minimum distance of 150m is maintained between the nearest
substation boundary and the nearest sensitive receptor.

81
Evidence Based Studies Noise – Main Report

8 REFERENCES

Akagi, Y., Sato, K., Iwama, N., Kubota, K., (1998), Low-noise Conductor for Harmonizing 1,000 kV
Transmission Lines with the Environment. U. D. C. 621.315.14.027.89 : 534.11 : [551.556+537.525.3]

Al-Faraj, M. A., Shwehdi, M. H., Farag, A. S., (1997), Environmental Effect on High Voltage AC
Transmission Lines Audible Noise. Energy Conversion Engineering Conference, 1997. IECEC-97.
Proceedings of the 32nd Intersociety, Vol. 3, p. 2082-2087.

Belardo, C. A., Fujimoto, F. T., Jardini, J. A., Bistafa, S. R., Kayano, P., Masiero, B. S., Nascimento, V.
H., Ribeiro, F., Mercato, E., Lima, R. G., Chavez, L., Tamai, E., (2006), Dynamic Absorption of
Transformer Tank Vibrations and Active Canceling of the Resulting Noise. 1st Congress on
Operational Excellence and innovation of AES Businesses in Latin America, 2006.

Bhatti, S. A., Shan, Q., Glover, I. A., Atkinson, R., Portugues, I. E., Moore, P. J., Rutherford, R.,
(2009), Impulsive Noise Modelling and Prediction of Its Impact on the Performance of WLAN Receiver.
17th European Signal Processing Conference (EUSIPCO 2009), Glasgow, Scotland. Aug 24-28, 2009.
pp 1680-1684.

British Standard BS4142: 1997 – Method for rating industrial noise affecting mixed residential and
industrial areas. British Standard Institution.

Chang, Y-H., Hsu, C-H., Lin, H-W., (2009), Design of Low-Noise Power Transformer with the Noise
Effect Survey on the Resident. IMCAS'09 Proceedings of the 8th WSEAS International Conference on
Instrumentation, Measurement, Circuits and Systems, p. 223-227.

Chartier, V. L., Blair, D. E., Easley, M. D., Raczkowski, R. T., (1995), Corona Performance of a
Compact 230-kV Line. USA: IEEE Transactions on Power Delivery, Vol. 10, No. 1, January 1995.

Cigre (1999) High Voltage Overhead Lines - Environmental Concerns, Procedures, Impacts and
Mitigations, Cigre Working Group 22.14.

Department of Energy and Climate Change (2011) National Policy Statement for Electricity Networks
Infrastructure (EN-5), UK Department of Energy and Climate Change.

Environment Agency (2004) Integrated Pollution Prevention and Control (IPPC) Horizontal Guidance
for Noise Part 2 - Noise Assessment and Control, UK Environmental Agency.

Environment Protection Agency (2012) Guidance Note for Noise: Licence Applications, Surveys and
Assessments in Relation to Scheduled Activities (NG4), Ireland: Environmental Protection Agency
Office of Environmental Enforcement (OEE).

Gela, G., LaForest, J. J., Zaffanella, L. E. (1993) HVDC Transmission Line Reference Book. Electric
Power REsearch Institute, 3412 Hillview Avenue, Palo Alto, Califormia, 1993, Section 3, pages 1-188.

82
Evidence Based Studies Noise – Main Report

Leventhall, G., (2003), A Review of Published Research on Low Frequency Noise and its Effects. UK:
Department for Environment, Food and Rural Affairs.

Medeiros, E. B., Borges, E. N. M., (2005) Evaluation of the Global Attenuation Noise Coefficient of an
Active Control System in an Open Space. Forum Acusticum Budapest 2005: 4th European Congress
on Acustics, 2005, PP. 893-897.

Medeiros, E. B., Kroeff, G., (1998), Acoustic Modeling and Simulation of an Urban Substation. Journal
of the Acoustic Society of America, Vol. 103, Issue 5, pp. 2838-2838.

Meyer, R. E., Aldrich, T. E., Easterly, C. E., (1989), Effects of Noise and Electromagnetic Fields on
Reproductive Outcomes. USA: Environmental Health Perspectives, Vol. 81, pp. 193-200.

Masti, R. S., Desmet, W., Heylen, W., (2004), On the Influence of Core Laminations Upon Power
Transformer Noise. Proceedings of the 2004 International Conference on Noise and Vibration
Engineering, Leuven, Belgium. 3851-3862, Sept. 20-22, 2004.

Muhr, M., Pack, S., Schwarz, R., Kornhuber, S., Koerbler, B., (2004), Sound Analysis and PD
Measurement of HV Transmission Lines. ACEDTsinghua University Press, p. 191-194.

Scottish Government (2011) Planning Advice Note PAN 1/2011 - Planning and Noise, UK: Scottish
Government.

Shan, Q., Bhatti, S. A., Glover, I. A., Atkinson, R. C., Portugues, I., Moore, P. J., Rutherford, R.,
(2009a), Characteristics of Impulsive Noise in Electricity Substations. 17th European Signal
Processing Conference (EUSIPCO 2009), Glasgow, Scotland. Aug 24-28, 2009. pp 2136-2140.

Shan, Q., Bhatti, S. A., Glover, I. A., Atkinson, R. C., Portugues, I., Moore, P. J., Rutherford, R.,
(2009b), R., (2009b), Extraction of Impulsive Noise from Measurements in a 400kV Electricity
Substation. International Conference on Energy and Environment, 2009-03-19 - 2009-03-21, India.

Shan, Q., Glover, I. A., Moore, P. J., Portugues, I., Judd, M. D., Rutherford, R., Watson, R. J., (2009c),
Impulsive Noise Environment of High Voltage Electricity Transmission Substations and Its Impact on
the Performance of Zigbee. Paper Presented at XXIX General Assembly of the International Union of
Radio Science, Chicago, Illinois, USA.

Straumann, U., Fan, J., (2009), Audible Noise from AC-UHV Transmission Lines - Theoretical
Comparison of Broadband and Tonal Components.

Straumann, U., Weber, H. J., (2010), Potential Reduction of Audible Noise from New and Aged
Overhead Transmission Line Conductors by Increasing Their Hydrophilicity. Cigre 2010, B2-113.

Task Force of the Corona and Field Effects Subcommittee of the Transmission and Distribution
Committee (1982), A Comparison of Methods for Calculating Audible Noise of High Voltage

83
Evidence Based Studies Noise – Main Report

Transmission Lines, USA: IEEE Transaction on Power Apparatus and Systems, Vol. PAS-101, No. 10
October 1982

Task Force of the Radio Noise and Corona Subcommittee of the Transmission and Distribution
Committee (1975) Audible Noise from Power Lines - Measurement, Legislative Control and Human
Response. USA: IEEE Transaction on Power Apparatus and Systems, Vol. PAS-94, No. 6,
November/December 1975.

Zhang, Z., Zeng, R., Yu, Z., (2009), Measurement of Corona Characteristics and Electromagnetic
Environment of +/-800kV HVDC Transmission Lines under High Altitude Condition. Progress in
Electromagnetics Research Symposium Proceedings, Moscow, Russia, August 18-21 2009, 61-65.

Wordpress (2011), 'What is Corona Effect', Overview [online], available at:

http://electricalnotes.wordpress.com/2011/03/23/what-is-corona-effect/ [accessed 20 Feb 2013].

World Health Organisation (WHO) (1999), Guidelines for Community Noise

World Health Organisation (WHO) (2009), Night Noise Guidelines for Europe

World Health Organisation (WHO) (2011), Burden of Disease from Environmental Noise -
Quantification of Healthy Life Years Lost in Europe.

84
Evidence Based Studies Noise – Main Report

GLOSSARY OF ACOUSTIC TERMINOLOGY

A channel model in which the only impairment to communication is a linear addition

Additive White of wideband (see below) or white noise (see below) with a constant spectral density
Gaussian Noise (expressed as watts per hertz of bandwidth) and a Gaussian distribution of
amplitude (see below).
The totally encompassing sound in a given situation at a given time, usually
Ambient noise composed of sound from many sources, near and far.
The amplitude of a periodic variable is a measure of its change over a single period
Amplitude (such as time or spatial period).
The reduction in level of a sound between the source and a receiver due to any
Attenuation combination of effects including: distance, atmospheric absorption, acoustic
screening, the presence of a building façade, etc.
A quantity that characterises how easily a material or medium can be penetrated by
Attenuation coefficient a beam of sound.
Normal hearing covers the frequency (pitch) range from about 20Hz to 20,000 Hz
but sensitivity is greatest between about 500Hz and 5,000Hz. The “A-weighting” is
A- weighting an electrical circuit built into noise meters to approximate this characteristic of
human hearing.
The steady existing noise level present without contribution from any intermittent
sources. The A-weighted sound pressure level of the residual noise at the
Background noise assessment position that is exceeded for 90 per cent of a given time interval,
T(LAF90,T)

Broadband Sounds that contain energy distributed across a wide range of frequencies.
The statistical measurement of a range of frequencies over which the channel can
Coherence bandwidth be considered flat.
The logarithmic measure of sound level. 0dB is the threshold of normal hearing. 140
Decibel (dB) dB(A) is the level at which instantaneous damage to hearing is caused. A change of
1 dB is detectable only under laboratory conditions.
Decibels measured on a sound level meter incorporating frequency weighting (A
weighting) which differentiates between sounds of different frequency (pitch) in a
similar way to the human ear. Measurements in dB(A) broadly agree with an
dB (A)
individual's assessment of loudness. A change of 3 dB(A) is the minimum
perceptible under normal conditions and a change of 10 dB(A) corresponds roughly
to doubling or halving the loudness of a sound.
The noise level at a Golagh Tee - Letterkenny 110kV m from the façade of a
Façade level building is described by the term façade level , and is subject to a higher noise level
than one in an open area (free-field conditions) due to reflection effects.
These are conditions in which the radiation from sound sources is unaffected by the
presence of any reflecting boundaries or the source itself. In practice, it is a field in
which the effects of the boundaries are negligible over the frequency range of
interest. In environmental noise, true free-field measurement conditions are seldom
Free-field
achieved and generally the microphone will be positioned at a height between 1.2
and 1.5 metres above ground level. To minimise the influence of reflections,
measurements are generally made at least 3.5 metres from any reflecting surface
other than the ground.

Hertz (Hz) The unit of sound frequency in cycles per second.

A noise that is of short duration (typically less than one second), the sound pressure
Impulsive level of which is significantly higher than the background.
A noise level exceeded for 90% of the time during a measurement period, often
LA90 used for the measurement of background noise.
This is the equivalent continuous sound level. It is a type of average and is used to
describe a fluctuating noise in terms of a single noise level over the same period (T).
The closer the LAeq value is to either the LAF10 or LAF90 value indicates the relative
LAeq,T
impact of the intermittent sources and their contribution. The relative spread
between the values determines the impact of intermittent sources, such as traffic, on
the background.

Narrowband Sounds that contain energy distributed across a small range of frequencies.
Sound that evokes a feeling of displeasure in the environment in which it is heard,
Noise and is therefore unwelcomed by the receiver.

Noise emission The noise emitted by a source of sound.

85
Evidence Based Studies Noise – Main Report

Noise immission The noise to which a receiver is exposed.

The ambient noise that remains in the absence of the specific noise whose effects
Residual noise are being assessed.
Physically: a regular and ordered oscillation of air molecules due to a source of
vibrations which creates fluctuating positive and negative acoustic pressure above
Sound and below atmospheric pressure. Subjectively: the sensation of hearing caused by
the ear being excited by the acoustic oscillations described above.
The logarithmic measure of sound power in comparison to a reference sound
2
intensity level of one picowatt (1pW) per m where:

Sound power level

Where W is the rms value of sound power in pascals; and W0 is 1 pW

Sound pressure refers to the fluctuations in air pressure caused by the passage of a
sound wave. It may be expressed in terms of sound pressure level at a point, which
is defined as:

Sound pressure level Lp = 20Log P/P0 dB

Where: P is the sound pressure;

P0 is a reference pressure for propagation of sound in air and has a value of
−5
2x10 Pa.

The lowest amplitude sound capable of evoking the sensation of hearing in the
Threshold hearing average healthy human ear (0.00002 Pa, equivalent to 0dB)
Sounds which cover a range of only a few Hz which contains a clearly audible tone,
Tonal i.e. distinguishable, discrete or continuous noise (whine, hiss, screech, or hum etc.)
are referred to as being ‘tonal’
In communications, a system is wideband when the message bandwidth
Wideband significantly exceeds the coherence bandwidth (See above) of the channel.

White Noise A random signal with a flat (constant) power spectral density.

86
 APPENDIX A

OVERVIEW OF ELECTRICITY TRANSMISSION

INFRASTRUCTURE, INCLUDING TYPICAL CONSTRUCTION
METHODOLOGY
Evidence Based Study Noise - Appendix A

A1 Description of Typical Electricity Transmission Project Designs

The transmission network in Ireland comprises structures and overhead lines, underground cables
and substations. When the need for a new circuit is identified in Ireland, EirGrid will consider all
available solutions for the new circuit. This will include overhead line and underground cable solutions,
considering both High Voltage Alternating Current (HVAC) and High Voltage Direct Current (HVDC)
technology, as appropriate.

Factors which will influence the solution decision include technical, economic and environmental
considerations. It is important to note that each project is different and EirGrid will determine potential
technology solutions on a project-by-project basis. EirGrid will continue to keep technology
developments under review and will consider new technologies as appropriate.

A1.1 Overhead Lines (OHL)

Transmission lines are generally supported on either wooden pole sets or steel lattice towers. Towers
along a straight of the alignment are known as intermediate towers. Angle towers are used where a
line changes direction and conductors are held under tension.

The type and height of structures required will vary according to the voltage of the overhead line, and
the location and type of environment and terrain in which they are placed.

A1.2 Structure Design

For all new electricity transmission projects, efficient, appropriately placed and optimally designed
structures are carefully considered and proposed. The design employed depends on the local
environment, topography and technologies involved, and will vary from 110 kV, 220 kV or 400 kV,
depending on the specific transmission need identified.

The spacing between structures depends on technical limitations and on the topography, particularly
to ensure that conductors maintain a specific minimum clearance above the ground at all times.

Steel Lattice Tower Structures

The weight of conductors and characteristics of 220 kV and 400 kV lines require that they be
supported exclusively on lattice steel structures (this also applies to angle towers along a 110 kV line).
The three phases (conductors) of a circuit are carried in a horizontal plane.

Table A1: Key Design Features: Single Circuit 220 kV and 400 kV overhead line structures

Key Design Features 220 kV Indicative Range 400 kV Indicative Range

Depends on technical details of Depends on technical details of
Height range individual projects but generally individual projects but generally
between 20-40m between 20m -52m
Maximum range of width at
ground level 6m to 12m 7m to 12m
Number of foundations per
structure 4 4
Average span between Approx. 320m (dependent on Approx. 350 (dependent on local
towers local topography) topography)

A1

Evidence Based Study Noise - Appendix A

Example of a 400 kV intermediate tower design along the Dunstown-Moneypoint overhead line, Co Clare

Example of a 220 kV intermediate tower design along the Cashla – Flagford overhead line, Co Roscommon

A2

Evidence Based Study Noise - Appendix A

Single Circuit 110 kV Overhead Lines

1
A 110 kV single circuit overhead line requires that conductors (and earth wires ) are supported on a
combination of steel lattice angle towers and double wood intermediate polesets.

The average span between polesets for a 110 kV single circuit alignment is approximately 180m;
however, the actual span achievable depends on local topography. Again, the three phases of the
circuit are carried in a horizontal plane.

Table A2: Key Design Features of Single Circuit 110 kV overhead line support structures

Key Design Features 110 kV Indicative Range

Height range (double wood polesets) 16m to 23m (incl. buried depth normally 2.3m)
Pole centres 5m
Number of foundations 2
Height range (steel angle towers) 18m to 24m
Maximum width at ground level 4m to 9.8m
Average span 180m

Example of a typical 110kV single-circuit double wood polesets with earthwire (Co Sligo)

On an alignment there may arise a very slight change in direction, and this may necessitate, in the
case of a 110 kV single-circuit line, the use of a braced wood poleset, wherein the space between the
polesets is reinforced with steel members.

1
Lines running above the conductors which protect the conductors from lightning strike.

A3

Evidence Based Study Noise - Appendix A

Braced double wood poleset

Double Circuit Overhead Lines

Overhead alignments can be configured as single circuit or double circuit (two separate circuits
supported on a single structure). This generally only occurs where two single circuit lines are in close
proximity (for example on approach to a substation), or where space is at a premium.

Double circuit alignments, including 110 kV overhead lines, always require to be supported by lattice
steel towers. The average number of structures on a line is 3-4 per km depending on topography. In
addition, the structures are higher, as each circuit must be carried in a vertical plane.

Typical 110 kV double circuit structures

A4

Evidence Based Study Noise - Appendix A

A1.3 Construction of Overhead Lines

Overhead line construction typically follows a standard sequence of events comprising:

• Prepare access;
• Install tower foundations/Excavation;
• Erect towers or wood poles;
• Stringing of conductors;
• Reinstate tower sites and remove temporary accesses.

Prepare Access
It is preferable to have vehicular access to every tower site for foundation excavation, concrete
delivery and a crane to erect towers. With wood pole construction, (on 110 kV single circuits) a crane
is not usually required, as these are normally erected with a digger using a lifting arm.

Access can take various forms and is dependent on ground conditions. In poorer conditions, more
complex access works are required which can vary from the laying of bog mats, or laying temporary
wooden matting, to installing crushed stone roads. Some of this work may entail removal of topsoil.

Access routes may require to be constructed for both the construction and maintenance of the
transmission line, and may be temporary or permanent.

Every effort is made to cause least disturbance to landowners and local residents, and to cause the
least potential environmental impact during construction. As a result, the most direct access route to a
tower installation may not always be the most appropriate.

Example of a newly built access route for a transmission project, Co. Donegal

A5

Evidence Based Study Noise - Appendix A

Install Tower Foundations/Excavation

Tower foundations are typically 2–4m deep with excavation carried out by mechanical excavator.
Excavations are set out specifically for the type of tower and the type of foundation required for each
specific site.

A larger footing may be required in the case of weak soils. Pile foundations may be required in the
case of deep bog. In the case of rock being encountered at shallow depths, reduced footing size
foundations may be required.

Prior to excavation, the foundations for each tower site will be securely fenced off to ensure the safety
of members of the public and livestock. Tower stubs (the lower part of the tower leg) are concreted
into the ground. Once the concrete has been poured and cured, the excavation is back-filled using the
original material in layers. Surplus material is removed from site.

The excavation required for a wooden poleset is typically 1.5m-2m x 3m x 2.3m deep; no concrete
foundations are required for polesets in normal ground conditions. Installation time is approximately
two per day. The average foundation size for a braced poleset is 9.3m x 3.1m x 3.2m deep.

In addition to the excavation required for the poleset itself, where ground conditions dictate, stay lines
may be required. This generally involves excavation of four trenches (approximately 2m x 2m x 1.8–
2m deep) at a distance from the poleset. The installation of stay wires expands the area of
disturbance associated with the erecting a poleset.

Stay lines in place, Donegal 110 kV Project

Concrete foundations are required for all steel towers. Foundation size and type is dependent on
ground conditions and tower type, but is typically 4m x 4m x 3.1m for each foundation pad. The base
installation time is approximately one week.

A6

Evidence Based Study Noise - Appendix A

110kV angle towers at Srananagh Station with exposed substructures

For all transmission lines with earth wires, there is a requirement to install an earth ring or mat at the
base of the structure to ground the structure for safety reasons. The ground around the base of
structures is excavated after conductors and earthwires are in place and the earth ring is installed.

Earth ring on Donegal 110kV Project

A7

Evidence Based Study Noise - Appendix A

Erect Towers or Wood Poles

Materials required for construction are transported around the site by general purpose cross country
vehicles with a lifting device. Excavators are generally of the tracked type to reduce likely damage to
and compaction of the ground. In addition a temporary hard standing may be required for machinery
and this may require the removal of topsoil. Materials are delivered to site storage/assembly areas by
conventional road transport and then transferred to sites.

Tower erection can generally commence two weeks after the foundations have been cast. Tower
steelwork is usually delivered to site and assembled on site.

Installation of tower using a derrick pole at the base

Construction of wooden poleset support structure for Donegal 110 kV Project (Binbane – Letterkenny)

A8

Evidence Based Study Noise - Appendix A

Stringing of conductors
Once angle towers are erected, conductor stringing can commence, installing conductors from angle
tower to angle tower via the line intermediate structures. Conductor drums are set up at one end of
the straight with special conductor stringing machinery, and pulled from one end to the other.

Stringing Machine

Conductor stringing equipment

Reinstate tower sites and remove temporary accesses

The disturbed ground around a tower or poleset location is made good, and all temporary access
materials generally removed.

A9

Evidence Based Study Noise - Appendix A

A1.4 Line Uprating and Refurbishment

In general a transmission line requires little maintenance. It is periodically inspected to identify any
unacceptable deterioration of components so that they can be replaced as necessary. A more detailed
condition assessment on a line is usually carried out when it is approximately 35 years old.

The majority of the existing transmission grid was constructed after 1960; the majority of those lines
constructed prior to 1960 have already been refurbished. There is an on-going programme of line
refurbishment concentrating on older lines.

Refurbishment projects are condition based, and once a line has been identified for refurbishment,
consideration is given to the potential opportunity to upgrade its carrying capacity or thermal rating.
This might involve replacing existing conductors with modern conductors which, while having
effectively the same diameter, can carry significantly greater amounts of electricity.

Often the additional weight of these replacement conductors means associated replacement of
support structures with stronger structures. Where structures require replacement during a line
upgrade or refurbishment, additional excavation may be required particularly where angle towers or
structures require replacement. In general they are replaced within the footprint of the original
structure.

Insulators and conductors are normally replaced after about 40 years, and towers are painted every
15-20 years or as necessary.

A10

Evidence Based Study Noise - Appendix A

A1.5 Underground Cabling (UGC)

High voltage (HV) circuits can only be laid underground using special HV cables designed specifically
for underground use. The conductors in underground HV cables must be heavily insulated to avoid a
short circuit between the conductor and the ground around the cable.

Table A3: Key Design Features: Underground Cabling

Key Design Features HV Cable (typical dimensions)

Cable Trenches c.0.6m wide-1.25m deep for a 110 kV trench,
c. 1.1m wide x 1.25m deep for 220 kV and 400 kV for
a single cable
Joint Bays 6m long, 2.5m wide and 1.8m deep
Excavation trench for Joint Bay 7m long, 3m wide and 2m deep
Average span between joint bays 500m–700m
Directional Drill entry and exit pits 1m x 1m x 2m

The cable is installed directly into the ground in an excavated trench. The majority of high voltage
cable routes are located along public roads and open spaces. It is very unusual for a cable route to
cross private open ground but this may be the case on occasion. The civil contractor will scan the
ground using a cable avoidance tool (CAT), carry out a visual inspection of existing services and
compare the information with the utility service records which they will have obtained from the various
service providers in advance. If any previously unidentified services are discovered the site engineer
will adjust the cable route accordingly.

Typical 110kV Trench Excavation (Ducts in Trefoil Formation)

A11

Evidence Based Study Noise - Appendix A

The overall installation of a cable route over a large distance is broken down into sections of cable
that are connected using a cable joint. Cable joints are installed in joint bays which are typically
concrete structures buried underground, occurring generally every 500–700m along an alignment, and
ranging in size up to 6m long, 2.5m wide and 1.8m deep.

Typical Joint Bay Construction Adjacent to Public Road

If the cable was installed directly in the ground the entire trench from joint bay to joint bay must be
fully excavated. The advantage with installing cable in pre-laid ducts is that only a short section of
cable trench, up to 100m is open at any time. This helps to minimise the impact on the local residents
and minimise traffic impact at any given time.

Typical HV Cable Installation

A12

Evidence Based Study Noise - Appendix A

Once installed, the road surface is reinstated. Where a cable route is in an open area, it is returned to
agricultural/grassland use. Where a cable passes through forested land the route is not replanted with
trees to prevent any damage to the cable by tree root growth.

Re-growth following underground cable construction on agricultural land

A1.6 Substations

Substations connect two or more transmission lines; they take the electricity from the transmission
lines and transform high to low voltage, or vice versa. They contain various electrical equipment,
including voltage switches, transformers, protection equipment, and associated lines and cabling.

The siting of a substation depends on topography; the ground must be suitable to meet technical
standards. With regard to earthing requirements and soil stability, substations are usually constructed
on reasonably level ground, in areas that are not liable to flooding or crossed by significant
watercourses.

A substation site is normally future proofed with the capability to be extended if the need arises.

Substations can take two forms:

An Air Insulated Switchgear (AIS) substation is where the electrical equipment infrastructure is
primarily installed outdoors, with the use of natural air as an insulation between circuits. This option
requires a relatively large compound footprint.

A13

Evidence Based Study Noise - Appendix A

Srananagh 220kV/110kV substation, Co Sligo, example of a typical outdoor AIS substation

A Gas Insulated Switchgear (GIS) substation, is where gas (Sulphur Hexafluoride – SF6) is used as
the insulation between circuits. This requires the electrical equipment to be contained internally, in
buildings of some 11–13m over ground. This allows for a significantly smaller substation footprint.

Both options require the associated provision of access roads off and onto the public road network
and the provision of associated electrical equipment and infrastructure (including underground
cables), as well as ancillary waste water treatment facilities and other site development and
landscaping works. Both are therefore significant civil engineering projects.

Example of a typical indoor GIS substation, Co Limerick

A14

 APPENDIX B

400KV TRANSMISSION LINE MONITORING

NOISE REPORT PREPARED BY AECOM CONSULTANTS

APRIL 2015
400kV
Transmission
Line
Monitoring
Noise Report

April 2015

47071393.NOISE

Prepared for:
EirGrid

UNITED
KINGDOM &
IRELAND
 EirGrid — 400kV Transmission Line Noise Monitoring

REVISION SCHEDULE

Rev Date Details Prepared by Reviewed by Approved by

1 December 2014 Draft Report Conor Tickner Chris Skinner Paul Shields
Graduate Acoustic Associate - Head of Acoustics
Consultant Acoustics

2 April 2015 Draft Report after Conor Tickner Chris Skinner Paul Shields
client comments
Graduate Acoustic Associate - Head of Acoustics
Consultant Acoustics

AECOM
12 Regan Way
Chetwynd Business Park
Chilwell
Nottingham
Nottinghamshire
NG9 6RZ

NOISE REPORT
April 2015
1
 EirGrid — 400kV Transmission Line Noise Monitoring

Limitations

AECOM Infrastructure & Environment UK Limited (“AECOM”) has prepared this Report for the sole use of EirGrid Plc
(“Client”) in accordance with the Agreement under which our services were performed. No other warranty, expressed or
implied, is made as to the professional advice included in this Report or any other services provided by AECOM.
The conclusions and recommendations contained in this Report are based upon information provided by others and
upon the assumption that all relevant information has been provided by those parties from whom it has been requested
and that such information is accurate. Information obtained by AECOM has not been independently verified by AECOM,
unless otherwise stated in the Report.
The methodology adopted and the sources of information used by AECOM in providing its services are outlined in this
nd th
Report. The work described in this Report was undertaken between 2 September and 13 November and is based on
the conditions encountered and the information available during the said period of time. The scope of this Report and the
services are accordingly factually limited by these circumstances.
Where assessments of works or costs identified in this Report are made, such assessments are based upon the
information available at the time and where appropriate are subject to further investigations or information which may
become available.
AECOM disclaim any undertaking or obligation to advise any person of any change in any matter affecting the Report,
which may come or be brought to AECOM’s attention after the date of the Report.
Certain statements made in the Report that are not historical facts may constitute estimates, projections or other forward-
looking statements and even though they are based on reasonable assumptions as of the date of the Report, such
forward-looking statements by their nature involve risks and uncertainties that could cause actual results to differ
materially from the results predicted. AECOM specifically does not guarantee or warrant any estimate or projections
contained in this Report.
Unless otherwise stated in this Report, the assessments made assume that the sites and facilities will continue to be
used for their current purpose without significant changes.
Where field investigations are carried out, these have been restricted to a level of detail required to meet the stated
objectives of the services. The results of any measurements taken may vary spatially or with time and further
confirmatory measurements should be made after any significant delay in issuing this Report.
Copyright
© This Report is the copyright of AECOM Infrastructure & Environment UK Limited.

NOISE REPORT
April 2015
2
 EirGrid — 400kV Transmission Line Noise Monitoring

TABLE OF CONTENTS EXECUTIVE SUMMARY ..................................................................... 4

1 INTRODUCTION ............................................................... 1
2 NOISE SOURCES AND STANDARDS ............................ 1
2.1 Aeolian Noise ................................................................... 1
2.2 Gap Discharge .................................................................. 1
2.3 Corona Noise .................................................................... 1
2.4 Guidance and Standards ................................................ 2
3 MEASUREMENT SITE ...................................................... 2
4 SCOPE AND METHODOLOGY ........................................ 2
4.1 Monitoring Procedures.................................................... 2
4.2 Data Processing Procedures .......................................... 3
5 CORONA NOISE PREDICTION ....................................... 4
5.1 BPA Method ...................................................................... 4
5.2 EPRI Method ..................................................................... 5
5.3 Evolutionarily Formed Methods ..................................... 5
5.4 Rainfall Rate Corrections ................................................ 6
5.5 Predicted Levels .............................................................. 6
5.6 Predicted Rain Noise ....................................................... 6
5.7 Noise Levels with Distance ............................................. 7
6 MEASURED DATA ........................................................... 8
6.1 Overview ........................................................................... 8
6.2 Weather conditions .......................................................... 9
6.3 Dry Noise Levels ............................................................ 10
6.4 Noise Levels during Rainfall ......................................... 12
6.5 Spectral analysis ............................................................ 12
7 COMPARATIVE ANALYSIS ........................................... 13
7.1 Comparison with predicted wet ambient levels ......... 13
7.2 Comparison with predicted corona noise ................... 14
8 DISCUSSION OF FINDINGS .......................................... 16
8.1 Overall levels .................................................................. 16
8.2 Tonality and Impulsivity ................................................ 17
8.3 Comparison with Predicted levels ............................... 17
9 CONCLUSIONS .............................................................. 18
10 RECOMMENDATIONS ................................................... 19
APPENDIX A: NOISE PERCEPTION AND TERMINOLOGY........... 20
APPENDIX B: SITE MAP AND MONITORING LOCATIONS .......... 22
APPENDIX C: TOWER STYLE AND SPECIFICATIONS ................. 23
APPENDIX D: MEASURED NOISE LEVELS BY TIME OF DAY ..... 24
APPENDIX E: DATA LOST OR REMOVED ..................................... 27
NOISE REPORT APPENDIX F: EQUIPMENT CALIBRATION CERTIFICATES ......... 28
April 2015
3
 EirGrid — 400kV Transmission Line Noise Monitoring

EXECUTIVE SUMMARY

In this report, noise emissions from the 400kV overhead power transmission line in Cloney,
Co. Kildare has been quantified based on data from more than 10 weeks of measurements.
Noise measurements were made simultaneously at three locations relative to the line: directly
beneath a tower, at the mid-span position half way between towers, and at a distance of
approximately 200m from the line as a control location.

Measurements showed noise level increases with increasing rainfall at all sites. At the control
location, the increases in noise level closely match standardised predictions of the noise level
increases from rainfall alone (i.e. the sound of rainfall hitting the ground and objects). At the
tower and mid-span locations however, there were significantly greater increases in noise
level than would be expected from the rainfall alone.

It was determined that; at the control location the contribution to the overall noise levels by the
transmission line is low in all weather conditions. At the mid-span location close to the line, the
contribution to the overall noise levels by the transmission line is low in dry conditions but
increases rapidly with rainfall. In high rainfall conditions, the noise from the line at the mid-
span increases significantly and is the dominant noise source.

During dry conditions, noise levels recorded at the tower location were significantly greater
than those recorded at other locations. Noise from the transmission line was particularly
dominant at night. Noise levels are usually expected to decrease at night, and whilst
decreases were observed at night at both control and mid-span locations, levels increased at
night at the tower location. It was noted that at night, humidity levels were often greater and
temperatures were lower, increasing condensation effects which may contribute to noise
generation from the line at the tower location. During a wide range of weather conditions,
noise from the line is the dominant noise source at the tower location. For periods of higher
rainfall rates, the noise from the line at the tower increases significantly.

It has been determined that; the primary sources of noise from the line during dry conditions
are located at the tower. The primary sources of noise from the line in wet conditions are the
conductors along the length of the line, and hence similar levels are experienced at tower and
mid-span locations during high rainfall rates.

Several noise prediction methodologies have been evaluated and comparisons made between
the predicted noise emissions from the line and measured noise levels. Two primary prediction
methodologies, from the Electronic Power Research Institute (EPRI) and the Bonneville Power
Administration (BPA) respectively, were obtained for comparison. An additional set of four
Evolutionarily Formed prediction methodologies outlined in a research paper have also been
included for analysis.

Output parameters of the prediction methodologies are the median noise level during rainfall,
th
the L50, and the 95 percentile noise level during rainfall, the L5. EPRI and BPA methods were
found to be most consistent with each other in their predictions. The Evolutionarily Formed
methods cover a greater range of values and are considered to be less reliable and less
consistent than either of the EPRI or BPA methods.

With the exception of one of the Evolutionarily Formed methods, all predicted noise levels are
greater than those measured. The extent to which the methodologies tend to over-predict the
noise levels depends on the noise indicator and the measurement location.

The most accurate predictions for both parameters occurred for the control location, although
there remains a significant error between predicted and measured levels. Predictions
suggested clearly audible and marginally significant noise impacts from the line at this

NOISE REPORT
April 2015
4
 EirGrid — 400kV Transmission Line Noise Monitoring

location. Conversely, measurements at this location suggested that generally the noise from
the line is not clearly audible and contributes very little to the overall noise levels, suggesting
little noise impact from the line at this location. Predictions for the L50 were generally found to
be more accurate than predictions for the L5.

Noise levels at both the tower and mid-span locations were over-predicted to an even greater
extent. Predictions for the L50 were again generally more accurate than predictions for the L5
at both these locations. The L50 was predicted slightly more accurately at the tower location
whereas the L5 predictions were slightly more accurate at the mid-span. The accuracy of noise
level predictions at the tower and mid-span locations were generally low.

Discrepancies between measured and predicted noise levels from this line could be due to
variations in line load, ageing effects, or other variables. The prediction methodologies may be
best suited for worst-case or new-line noise impact predictions. The BPA and EPRI methods
are considered to be more consistent and therefore more suitable for use in the predictions of
future noise emissions from overhead lines.

NOISE REPORT
April 2015
5
 EirGrid — 400kV Transmission Line Noise Monitoring

1 INTRODUCTION

AECOM has been commissioned by EirGrid to perform long term noise monitoring of a 400kV
overhead power transmission line with the intention of capturing noise from corona discharge
over a range of weather conditions. Of particular interest is the variation of noise at different
positions along the line.

Acoustic measurements were undertaken for a period of approximately 10 weeks under and in
the vicinity of the transmission line. In addition, short measurements were taken at the site to
provide further detail of the sound at various distances from the line. The monitoring took
nd th
place between 2 September 2014 and 13 November 2014.

An explanation of acoustics terminology is provided in Appendix A.

2 NOISE SOURCES AND STANDARDS

When examining overhead transmission lines from a noise perspective, there are a number of
potential noise sources to consider; each depending on the parameters of the line in question.
The primary noise sources are: Aeolian, gap discharge and corona discharge noise. These
are described below.

2.1 Aeolian Noise

Aeolian noise is caused by the movement of air over an obstacle, creating trailing vortices in
the air flow. These vortices have oscillatory periodic components which can cause periodic
fluctuations in pressure, generating steady tones. The frequency of these tones is proportional
1
to the velocity of the air and inversely proportional to the size of the obstacle. A laminar wind
must blow steadily and perpendicular to the lines to set up an Aeolian vibration, which can
produce resonance if the frequency of the vibration matches the natural frequency of the line.

This type of noise is usually infrequent and low-level. No Aeolian noise has been observed by
AECOM staff on site at any time during monitoring, although it is recognised that Aeolian noise
is often difficult to identify and the presence of this noise on cannot be excluded definitively.

2.2 Gap Discharge

Gap discharge is caused by sparking between hardware components. This effect is usually
infrequent and most often removed by the selective design of the overhead line. Gap
discharge is usually a sign of hardware failure or damage and can usually be easily located for
2
repair . Gap discharge is a much more significant phenomenon for radio interference (RI) than
for audible noise and was not observed as a source of audible noise by AECOM staff on site.

2.3 Corona Noise

Corona noise is the most commonly observed noise source from overhead lines. It is most
often observed as a frying, hissing or crackling sound and is a function of various parameters
including the line voltage, relative air density, the electric field on the surface of the conductor,
the line geometry and the condition of the line. It is rarely observed for low voltage overhead
2
lines but is known to be a significant parameter for lines operating above 345 kV .

Corona is induced by the electric field in localised areas where the gradient is greatest (at the
surface of the conductor, usually where an irregularity causes an increased electric field), and

1 Lilien, J-L, ‘Power Line Aeolian Vibrations’, Université de Liege, 2013

2 Electric Power Research Institute, Transmission Line Reference Book 345kV and Above, Second Edition, EPRI, 1982

NOISE REPORT
April 2015
1
 EirGrid — 400kV Transmission Line Noise Monitoring

consists of an electric discharge which makes the surrounding air molecules ionise, or
undergo a change of electric charge. This phenomenon gives rise to pressure fluctuations in
the air surrounding the conductor, generating noise which is generally broadband in character.
As the corona effect increases, such as during high levels of rainfall, the pressure fluctuations
align with the electrical frequency, oscillating at twice the electrical frequency of the system.
This causes tonal components at twice the electrical frequency (and subsequent harmonics) to
be present in the acoustic emissions of the line.

2.4 Guidance and Standards

No specific standard exists for assessing noise from overhead lines. The noise emissions are
categorised under ‘industrial sound’ and the relevant assessment standard is therefore BS
3
4142 . This standard is used to assess and quantify the significance of the impact from
industrial sound sources on sensitive receptors.

BS 4142 makes no explicit reference to overhead line noise and does not give specific
instructions for assessment of noise sources generated during rainfall. A separate guidance
4
document exists; TR(T)94 , which gives an applicable method for assessment of overhead
lines and takes account of the varying background noise levels with varying rainfall rates, but
remains in-line with the fundamental methods and principles of BS 4142. TR(T)94 describes
separate methods for assessing noise levels during both wet and dry weather conditions, and
is better suited to the assessment of sound emissions from overhead lines.

3 MEASUREMENT SITE

The measurement site, in Cloney in rural County Kildare, was selected for the monitoring by
the client, based on the requirements set out in the agreement with AECOM. The site is
enclosed to the east by countryside, with the R417 approximately 500 meters from the
measurement locations. Immediately to the south is a mixture of woodland and fields, with the
R427 approximately 400 meters south of the tower and 200 meters south of the closest
measurement location to the road. Both of the closest roads are relatively minor single
carriageway roads. There is open rural countryside to the north and a mixture of trees and
open countryside to the west. The transmission line runs between the north-east and south-
west through the area, and the area immediately surrounding the line is predominantly flat.

The soundscape at the measurement site is dominated by a low level of distant road traffic
noise during busy periods and by wind and the movement of foliage at other times. Occasional
animal sounds such as from birds, dogs and a horse, were observed at times by AECOM staff
on site.

The plan in Appendix B shows the location of the site and the long term acoustic monitoring
locations.

4 SCOPE AND METHODOLOGY

4.1 Monitoring Procedures

Noise monitoring was undertaken in three locations relative to the power line; directly beneath
a tower, at the mid-span position half way between towers, and at a distance of approximately
200m from the line as a control location. The control location is intended to capture the
general background noise under the same weather conditions, but largely excluding noise
from the line.

3 BS 4142:2014, Methods for rating and assessing industrial and commercial sound
4 National Grid Company PLC, Technical Report TR(T)94, 1993. ‘A Method for Assessing the Community Response to Overhead Line Noise’,

NOISE REPORT
April 2015
2
 EirGrid — 400kV Transmission Line Noise Monitoring

These monitoring locations recorded noise data for a total period of 10 weeks from 02/09/2014
to 13/11/2014, split into five approximately two-week periods listed below:

 Period 1 – 02/09 to 19/09

 Period 2 – 19/09 to 30/09

 Period 3 – 30/09 to 13/10

 Period 4 – 13/10 to 29/10

 Period 5 – 29/10 to 13/11

This breakdown has been used due to the timing of site visits to download data and change
equipment batteries, and does not relate to any characteristics of sound from the transmission
line.

Measurements were undertaken using the following instrumentation:

 Rion NL-52 integrating sound level meters; and
 Rion NC-74 acoustic calibrator.

The meters were set to log in 10 minute contiguous periods, with regular 15-second audio
recordings taken at each period start. In addition, a full time history of 100ms LpA sound
pressure level measurements was logged at each location to allow calculation of noise
indicators over other periods.

The calibration level of each meter was checked using a field calibrator before and after the
measurements, and during each visit for equipment maintenance, with no significant changes
in calibration level detected (with the exception of damaged equipment as noted below).
Calibration certificates are included in Appendix F.

Monitoring stations were set up in each location comprising an all-weather equipped sound
level meter mounted between 1.2m and 1.5m off the ground in free-field conditions. At the
control site, a weather station was also erected to record rainfall, temperature, ambient air
pressure, wind speed, and wind direction. All these parameters were logged over one minute
and ten minute intervals.

At the end of each period, batteries were replaced, data were downloaded and equipment
checks were made. Short-term attended sound measurements in the wider area were taken
during the set-up and at the first maintenance visit.

Due to flooding at the site, one of the meters recorded no data for the last few days of the
monitoring period. A short period of data was also not recorded at the end of Period 1 as
instrument batteries ran flat prior to the planned maintenance visit. Details of lost data are
included in Appendix E

4.2 Data Processing Procedures

The meters were set to automatically log the parameters LAeq, LAmax, LAmin, LA10, and LA90 (see
Appendix A for definitions).

Raw data were processed to return each of those parameters for 1-minute intervals, allowing
more detailed correlation with 1-minute rainfall data.

NOISE REPORT
April 2015
3
 EirGrid — 400kV Transmission Line Noise Monitoring

Data deemed unsuitable for analysis were removed from the dataset. Details of removed data
are included in Appendix E. Reasons for removal were due to contamination by staff during
set-out, collection or maintenance, due to very high wind speeds, or due to the equipment
failure noted above.

The remaining data were separated into contiguous 24-hour periods for initial analysis. This
allowed an overview of the levels with respect to weather conditions and time of day. More
detailed analysis required the data to be separated by time of day and by weather conditions
(wet and dry). This allowed the variation with time of day to be more closely examined, and the
effects of rainfall on the noise levels to be quantified.

5 CORONA NOISE PREDICTION

There are a number of methodologies for predicting corona discharge noise from power lines.
5 6
There are two main methods that have been selected for analysis; the BPA and EPRI
methods. An additional set of prediction methodologies have been obtained for comparison.
These additional methods were Evolutionarily Formed using iterative algorithms to determine
7
the dependence of the noise on a variety of parameters .

A number of parameters are used in all methods to predict noise levels, such as: conductor
8
surface gradient , conductor diameter, the number of subconductors, and the geometric
arrangement of the conductors, i.e. heights, widths and separation distances. The BPA
method also takes altitude into account, resulting in higher predicted noise levels at high
9
altitudes. Since the altitude at the monitoring site is low (approx. 55m above sea level ), this
has no significant effect on the calculations for the site under consideration but could be
significant elsewhere. All parameters were sourced from the client unless alternate source is
given.

Each prediction methodology yields two results; the noise level exceeded for 50% of the time
during rain, or the L50, and the noise level exceeded for 5% of the time during rain, or the L5.
These are defined slightly differently depending on methodology chosen.

It is important to note that the L5 and L50 noise levels defined by these methodologies do not
directly relate to percentile noise levels as normally reported by sound level meters. The
percentile values (LN,T) produced by sound level meters present the noise level which is
exceeded for N% of a defined time interval, T. However the L5 and L50 levels calculated by the
prediction methodologies present the corona noise level which will be exceeded for 5% or
50% of time whilst rain is falling.

5.1 BPA Method

The method established by the Bonneville Power Administration (BPA), calculates the median
audible noise levels during measurable rain. The primary output is the L50 level during rain at a
given measurement point. It was developed from long term measurements on a number of full-
scale operating or test lines.

The calculation is different depending on the number of subconductors in each conductor

bundle; a different calculation is required when there are three or more conductors. The line
examined in this report has only two subconductors for each conductor bundle.

5
Electric Power Research Institute, ‘Chapter 7: Corona and Field Effects’, Transmission Line Reference Book 115-230 kV Compact Line Design, EPRI, 2007.
6
Electric Power Research Institute, ‘Chapter 6: Audible Noise’, Transmission Line Reference Book 345kV and Above, Second Edition, EPRI, 1982, p. 267-318.
7
K-H Yang, et al, ‘New Formulas for Predicting Audible Noise from Overhead HVAC Lines Using Evolutionary Computations’, IEEE Transactions on Power Delivery,
Vol. 15: Issue 4, IEEE, 2000, p. 1243-1251.
8
Method obtained IEE Power Series 17, “High Voltage Engineering and Testing”, Appendix 3.4, Annexure B
9
Irish Grid Reference. 2008 Data. Available at: http://www.gridreference.ie/ [2014].

NOISE REPORT
April 2015
4
 EirGrid — 400kV Transmission Line Noise Monitoring

Where there are multiple phases, the noise level from each phase is calculated separately and
then summed logarithmically. The line examined in this report has three phases with a lateral
separation of 10.5m. A diagram of the tower dimensions is given in Appendix C. The altitude is
approximately 55m above sea level, which makes the altitude correction very small (less than
0.2 dB).

For the L5 level, a fixed correction factor of +3.5 dB is suggested. For fair weather dry
conditions a fixed correction factor of -25 dB is suggested. The L50 is described as
representative of ‘steady rain’ conditions, whereas the L5 is described as representative of
‘heavy rain’ conditions. These conditions are not quantified.

5.2 EPRI Method

The method established by the Electronic Power Research Institute (EPRI) is calculated
empirically from test data similar to the BPA method. It is a function of similar parameters, but
does not account for altitude. The default calculation for this method is the L5 level during rain.

Again, this calculation applies to a single phase. Where there are multiple phases, the noise
level from each phase is calculated separately and then summed logarithmically.

This method does not offer a dry or fair weather prediction, instead quantifying the variation of
empirical data as being ‘up to 9 dB lower’ in dry conditions. Noise levels are considered to be
highly dependent on surface condition parameters such as age, grease, dust, and other
particles. A wet correction factor Pwc is calculated and applied to output the predicted L50 level
during rain.

These calculations are based upon data collected in Pittsfield, USA, which experiences a local
median rainfall of 0.75 mm/h. A local rainfall climate correction factor is provided which can be
manually determined from the median rainfall at any given location.

5.3 Evolutionarily Formed Methods

10
Evolutionarily Formed methods utilise a genetic algorithm (GA) or genetic programming (GP)
process to iteratively modify prediction formulae using empirical data. Calculations are based
on a single-phase (1P) or three-phases (3P) and output the L5 or L50 rain noise level. After
testing a number of the methods against old and new experimental data, the most accurate
prediction formulae were presented. Each method is given a short-code to be identified, for
example the genetic algorithm (GA) single-phase (1P) L50 level (L50) formula is given the short
code GA1PL50. The presented formulae include GA1PL50 as described; along with the
genetic programming single-phase L5 level (GP1PL5), the genetic programming 3-phase L5
level (GP3PL5) and the genetic programming 3-phase L50 levels (GP3PL50).

The outputs of these methods are stated to be representative of stable rain (≤2.7mm/hr) for
the L50 levels and heavy rain (≥7.7mm/hr) for the L5 levels. They do not contain conversion
factors between methods. However, single phase methods can be calculated individually and
logarithmically summed to obtain a three-phase prediction. For comparative purposes, all
evolutionarily methods are presented in three-phase form. No dry or fair weather prediction is
offered in this method.

10
K-H Yang, et al, ‘New Formulas for Predicting Audible Noise from Overhead HVAC Lines Using Evolutionary Computations’, IEEE Transactions on Power Delivery,
Vol. 15: Issue 4, IEEE, 2000, p. 1243-1251.

NOISE REPORT
April 2015
5
 EirGrid — 400kV Transmission Line Noise Monitoring

5.4 Rainfall Rate Corrections

Only the EPRI methodology explicitly states a rainfall rate correction factor. In the BPA
methodology the levels are simply described as ‘steady rain’ and ‘heavy rain’ conditions
without quantification. The evolutionary methods state similar descriptions but quantify them to
mean ≤2.7mm/hr for steady rain, or ≥7.7mm/hr for heavy rain.

The EPRI methodology indicates a local climate correction factor to account for the variation in
th
rainfall rates for the local area. This correction factor is related to the median and the 95
percentile rainfall rates at the location of their empirical study. These values are 0.75 mm/h
and 6.4 mm/hr respectively.

In Cloney, Co. Kildare, the local median rainfall rate was recorded as 0.8 mm/hr, and the local
th
95 percentile rainfall rate was recorded as 5.1 mm/hr. These result in a correction factor of
less than 0.5 dB with respect to either the L5 or L50 calculation. The effect of local climate is
therefore considered to be insignificant to any conclusions which may be drawn.

5.5 Predicted Levels

Using each of the prediction methods and the transmission line specification provided by
EirGrid (see Table C.1 in Appendix C), predictions of the expected noise levels at lateral
distances of 5 and 200 meters from the base of the tower have been made. These locations
are representative of the monitoring locations. Results are shown in Table 5.1.

TABLE 5.1: PREDICTED CORONA DISCHARGE NOISE LEVELS

Distance from Sound Predicted Noise Level, dB

Transmission Level
Line Indicator BPA EPRI GA1PL50 GP1PL5 GP3PL50 GP3PL5

L50 55.2 56.0 57.4 - 54.3 -

5m
L5 58.7 61.2 - 65.3 - 57.3

L50 46.0 44.5 41.6 - 45.8 -

200m
L5 49.5 49.7 - 48.7 - 37.8

The 5m L50 levels are fairly consistent, with about 3 dB difference between all methodologies.
The 5m L5 predictions are less consistent, with 8 dB between all methodologies.

The 200m L50 and L5 levels are less consistent than 5m levels, with more than 4 dB between
L50 levels and almost 12 dB between L5 levels.

It should be noted that the BPA and EPRI methods agree to within 2.5 dB in both locations
and for both parameters, possibly suggesting a greater consistency than the evolutionary
formed methods.

5.6 Predicted Rain Noise

Predicted levels in Table 5.1 relate to the transmission line as the sole source of noise. In
practice, there will be additional noise from the surrounding area. In particular, rainfall is a
potentially significant component of the ambient noise, especially in conditions when corona
noise is likely to be present.

NOISE REPORT
April 2015
6
 EirGrid — 400kV Transmission Line Noise Monitoring

Prediction methodologies for the noise from rainfall itself can be applied based on a dry
baseline noise level and additional predicted rain noise. The rain noise prediction is based on
11
one of series of Miller-curves . An R1 curve has been selected as most suitable for the
measurement location, i.e. a rural area with a grassy ground surface. The R1 Miller-curve for a
40 dB dry ambient level is shown in Figure 5.1 below.

70
Sound Pressure Level, dB(A)

60
50
40
30
20
10
0
0.01 0.10 1.00 10.00 100.00
Rainfall Rate, mm/hr

Figure 5.1: R1 Miller Rain Noise Curve

5.7 Noise Levels with Distance

The identified corona noise prediction methodologies discussed above have been calculated
at a range of distances to produce graphs of predicted sound level against distance. These
levels hence show how the predicted noise levels fall with distance. Figure 5.2 shows the
predicted L5 levels and Figure 5.3 shows the predicted L50 levels, for each methodology with
distance.

70
Sound Pressure Level, dB(A)

60
50
40 BPA heavy rain L5
30 GP1PL5 heavy rain L5
20 EPRI heavy rain L5
10 GP3PL5 heavy rain L5

0
0 100 200 300 400 500
Distance from tower, m

Figure 5.2: Predicted L5 levels with distance from the line

11
National Grid Company PLC, Technical Report TR(T)94, 1993. ‘A Method for Assessing the Community Response to Overhead Line Noise’,

NOISE REPORT
April 2015
7
 EirGrid — 400kV Transmission Line Noise Monitoring

70

Sound Pressure Level, dB(A)

60
50
40 GP3PL50 wet L50
30 BPA wet L50
20 GA1PL50 wet L50
10 EPRI wet L50
0
0 100 200 300 400 500
Distance from tower, m

Figure 5.3: Predicted L50 levels with distance from the line

The L50 predictions are in greater agreement near to the tower, with a greater spread at further
distance, particularly from 100m. Excluding the GP1PL5 method, which appears to result in
notably different levels to the other methodologies, the L5 methods have a generally good
agreement at greater distances, even to 300m and beyond.

6 MEASURED DATA

During the monitoring, large amounts of data were recorded, including weather conditions and
a range of sound level parameters. This section provides an overview and initial analysis of
the measured data.

6.1 Overview

Over the monitoring period, several noise parameters were measured, including the LAeq,
LAmax, LAmin, LA5, LA10, LA50 and LA90. Given the large quantities of data, these parameters can
be presented in a number of ways. As an overview of all 10 weeks of data, the absolute or
average level is preferred (see Appendix A for further explanation of the noise metrics used).
Absolute values more closely match the methods used in presenting data from previous short
term measurements, however as they summarise data over a much longer time period, the
amount of information regarding corona noise which can be drawn from these values is
limited. The average values presented represent typical values measured over the 1-minute
logging periods.

The average and absolute measurements for the whole ten weeks are displayed below in
Table 6.1, with the exception of the absolute L10 and L90 which are unable to be calculated
from periodically logged data. While absolute and average parameters provide a method of
interpreting large periods of data, the variation of values is lost.

NOISE REPORT
April 2015
8
 EirGrid — 400kV Transmission Line Noise Monitoring

TABLE 6.1: 10-WEEK TOTAL NOISE LEVELS SUMMARY

Noise Level, dB

Absolute over entire

Average from 1-minute values monitoring period

Total Maximum Minimum

Location LAeq LAmax LAmin LA10 LA90 LAeq LAmax* LAmin

Control 37.8 52.8 27.3 40.6 30.3 42.1 91.7 15.6

Tower 41.4 52.8 35.2 43.2 37.7 45.6 97.8 21.1

Mid-span 36.4 50.2 28.1 38.9 30.7 42.2 99.8 15.8

* Whilst it is generally not possible to identify the cause of the measured maximum noise
levels, it is considered likely that these are due to other local noise sources, such as animals,
birds, or weather conditions. A review of audio recordings collected at the times of some
peaks in measured noise levels has been undertaken to provide further information on
potential causes of maximum noise levels. This has identified, that, whilst noise from the
power lines is audible at times, the maximum noise levels in all cases reviewed were due to
other sources, such as birdsong, dogs barking, other animals, local human activity, overflying
aircraft, or weather conditions (heavy rainfall/thunder storms).

As there was a wide range of variation in noise levels and weather conditions over the
th th
monitoring period, this report has focussed on presentation of the 5 and 95 percentiles for
1-minute or 10-minute values. This provides the range of values within which the data lie for
90% of the time.

Another more detailed representation of the data considers the hourly variations over a 24-
hour day. Tables of average hour-by-hour levels for each measurement location are shown in
Appendix D.

6.2 Weather conditions

During the monitoring period, a wide variety of weather conditions were recorded. The weather
was predominantly dry for the first 4 weeks where a total of 10.5 mm of rain was recorded.
The following six weeks recorded almost 165 mm of rainfall.

Wind speeds were variable with individual maximum gust speeds of up to 15 m/s recorded.
Despite this, the maximum gust speed was less than 5m/s for over 80% of all 10-minute
periods and the 10-minute average wind speeds remained below 5 m/s for over 99.5% of the
monitoring period and below 2.8 m/s 90% of the time. Where high wind speeds were
considered to have compromised noise measurements these periods were excluded from
subsequent analysis included in this report. See Appendix E for details of excluded data.

Temperatures varied between -2˚C and 23˚C, remaining between 3.5˚C and 18.5˚C for 90% of
the monitoring period. The mean temperature was 11.4˚C.

Relative Humidity was generally high, with a minimum of 49% recorded over the complete
period. Humidity remained above 70% for 90% of the monitoring period and often reached
100%, particularly at night when temperatures dropped.

NOISE REPORT
April 2015
9
 EirGrid — 400kV Transmission Line Noise Monitoring

Ambient air pressure was generally slightly low, with the median air pressure at 1006 mbar,
compared to a standard atmospheric pressure of 1013 mbar. The ambient air pressure
exceeded 1013 mbar only for approximately 25% of the monitoring period.

6.3 Dry Noise Levels

During the dry periods of monitoring, noise levels at the control and mid-span sites varied as
would be expected, with lower levels during the night and higher levels during the day. Level
distribution graphs for the control and mid-span locations are shown in Figures 6.1 and 6.2
below.

Figure 6.1: Distribution of LAeq,1min Levels at the Control location

Figure 6.2: Distribution of LAeq,1min Levels at the Mid-span location

NOISE REPORT
April 2015
10
 EirGrid — 400kV Transmission Line Noise Monitoring

The tower location experienced a quite different set of noise levels. While the control and mid-
span levels generally decreased at night, levels at the tower often did not decrease, and
sometimes increased. This phenomenon (shown in a typical dry 24-hour noise level plot for all
sites in Figure 6.3) has the effect of significantly increasing the distribution of noise levels
upwards, and resulting in a far narrower range of observed levels. The level distribution graph
for the tower is shown in Figure 6.4 below. The tower experiences a greater and narrower
band of noise levels, indicating a relatively constant and dominant source of noise.

It has been noted that when air is cooled, such as at night time, the relative humidity increases
significantly. The rise in noise levels at the tower location corresponds with the increase in
humidity. This is an indication that corona noise is dependent on humidity and could be the
dominant noise source at night near to the tower. Analysis of the recorded audio from this
location confirms the presence of corona noise during these times.

70 100
90
60
80
50 70

Relative Humidity, %
40 60
LAeq, dB

50
30 40
20 30
20
10
10
0 0

Control Tower Mid-span Relative Humidity

Figure 6.3: Typical dry 24-hour variation in LAeq,10min level for all locations

Figure 6.4: Distribution of LAeq,1min Levels at the Tower location

NOISE REPORT
April 2015
11
 EirGrid — 400kV Transmission Line Noise Monitoring

At the tower location, 90% of dry 1-minute LAeq levels are between 38.2 dB and 55.6 dB, a
range of 17.4 dB. At the mid-span location however, 90% of dry 1-minute LAeq levels are
between 32.0 dB and 53.6 dB, a range of 21.6 dB. At the control location, 90% of dry 1-minute
LAeq levels are between 28.7 dB and 52.0 dB, a range of 23.3 dB. This reduced range at the
tower location is consistent with corona noise limiting the lowest noise levels observed at the
tower location.

6.4 Noise Levels during Rainfall

Noise levels during rainfall consist primarily of three components; background noise, rain
noise and corona noise. The impact of corona noise is expected to be minimal at the control
site but significant at the mid-span and tower locations. The average noise level is shown
plotted against rainfall rate in Figure 6.5. In addition, this graph shows the general R1 rain
noise curve presented previously (based on an assumed background level of 40 dB). A
greater number of recorded noise levels at lower rainfall rates allow for a smoother line with
fewer peaks and troughs. The greater variation at higher rates is a consequence of fewer
instances of very high rates of rainfall during the monitoring period, compared with much
longer periods of dry and low rainfall rates. At low rainfall rates, noise levels at the control and
mid-span locations are significantly lower than those at the tower, aligning approximately with
the expected level from background and rain noise alone (represented by the R1 Miller curve).
At all sites there is a general trend towards greater noise levels with greater rainfall rates.

Figure 6.5: Noise level against rainfall rate for all sites.

Control location noise levels can be seen to follow the general Miller curve estimates well,
being on average 0.7 dB lower than the Miller curve and within ±4.5 dB for 90% of the time.
Tower location noise levels were on average 6.1 dB above the Miller curve, varying between 4
and 10 dB higher for 90% of the time. Mid-span noise levels were on average 5.5 dB above
the Miller curve, varying between 3 and 9 dB higher for 90% of the time.

The mid-span location is seen to transition with increasing rainfall. During dry weather, the
measured levels are similar to those at the control location, whilst at higher rainfall rates
(above approximately 1.3 mm/hr), they match those at the tower location.

6.5 Spectral analysis

Spectral analysis of selected audio recordings identified the characteristic presence of a

100 Hz hum. This is due to a well understood principle of corona effects generating strong

NOISE REPORT
April 2015
12
 EirGrid — 400kV Transmission Line Noise Monitoring

12
tonal components at twice the electrical frequency of the system (electrical frequency is 50
Hz in Ireland). A detailed 1/24-octave band analysis is presented in Figure 6.6 and shows the
presence of a tone at 50 Hz, and associated harmonic frequencies in multiples of 50 Hz up to
250 Hz, with 100 Hz being the strongest component. These are seen at both the tower and
mid-span location, with an example from the mid-span location shown in Figure 6.6 below.

An increase in level at 1.9 kHz during rainfall was also noted; whilst this is assumed to be
associated with corona generation, the precise cause of this increase is unknown.

50
45 100 Hz
Sound Pressure Level, dB(A)

40
50 Hz 200 Hz
35
30 150 Hz 250 Hz 1.9 kHz
25
20
15
10
5
0
22.4 Hz
26.7 Hz
31.7 Hz
37.7 Hz
44.8 Hz

107 Hz
127 Hz
151 Hz
179 Hz
213 Hz
254 Hz
302 Hz
359 Hz
427 Hz
510 Hz
600 Hz
720 Hz
850 Hz
53 Hz
63 Hz
75 Hz
90 Hz

1.01 kHz
1.21 kHz
1.43 kHz
1.71 kHz
2.03 kHz
2.41 kHz
2.87 kHz
1/24-octave Frequency Bands

Figure 6.6: 1/24-octave band analysis of corona noise at mid-span location; tones highlighted

7 COMPARATIVE ANALYSIS

7.1 Comparison with predicted wet ambient levels

In order to predict noise levels during rainfall, it is necessary to identify the ambient noise level
at each measurement location in the absence of rain. This has been taken as the average
LAeq,1min noise level at the location during dry periods of monitoring. Such dry ambient sound
levels for each location are given in the Table 7.1 below, with separate values for daytime and
night time.

It should be noted that during these dry conditions, levels at the tower are significantly greater
than those at the mid-span and control locations, with electrical noise prominent throughout
much of the monitoring period. Typical day and night levels as presented in Table 7.1 differ by
less than 1 dB at the tower, compared with 7 and 6 dB differences between daytime and night
time levels at the control and mid-span locations respectively.

12
Electric Power Research Institute, ‘Chapter 6: Audible Noise’, Transmission Line Reference Book 345kV and Above, Second Edition, EPRI, 1982, p. 276

NOISE REPORT
April 2015
13
 EirGrid — 400kV Transmission Line Noise Monitoring

)
TABLE 7.1: DRY AMBIENT SOUND LEVELS (AVERAGE LAEQ,1MIN

Time of Day Control Tower Mid-span

Day
39.9 41.2 37.8
(07:00-23:00)

Night
32.9 40.4 31.8
(23:00-07:00)

An R1 Miller curve is used to predict the background level during rain. Whereas general
assessments can be made on the basis of a general Miller curve, the true curve values
depend on time of day and on each location’s dry ambient sound level. Correcting for these
specific conditions, differences between expected level and measured levels can be
quantified. These are summarised in Table 7.2 below.

TABLE 7.2: DIFFERENCE BETWEEN PREDICTED AND MEASURED LEVEL

Average Predicted Level Above Miller Curve, dB

Weather conditions Control Tower Mid-span

Light rain (≤0.5 mm/hr) 0.6 2.7 3.1

Medium-Heavy rain (≥0.5 mm/hr) 0.2 5.9 6.9

Measured levels at the control location are within 1 dB of the predicted levels in all conditions
considered in this table. The mid-span shows the greatest exceedance of predicted rainfall
levels, but the tower levels are the greatest in absolute terms. It should be noted that there are
higher dry background levels at the tower, especially at night due to the greater electrical
noise present in this location in dry conditions. This has the effect of increasing the predicted
rainfall sound levels and hence reducing the noted exceedances.

By logarithmically subtracting the contribution of the ambient noise levels (including predicted
contribution from rain) from the total noise, the contribution of other sources, primarily corona
noise, can be quantified. This is shown in Table 7.3 below.

TABLE 7.3: CONTRIBUTION OF CORONA NOISE SOURCES

Contribution to total noise level, dB

Weather conditions Control Tower Mid-span

Light rain (≤0.5 mm/hr) 41.6 47.1 44.8

Medium-Heavy rain (≥0.5 mm/hr) 44.7 52.9 50.7

7.2 Comparison with predicted corona noise

The measured L50 and L5 noise levels as defined in the prediction methodologies are
th
calculated by the average measured noise levels at each location corresponding to the 50
th
and 95 percentile rainfall rates of 0.8 mm/hr and 5.1 mm/hr respectively. Only one period of
rainfall exactly matching 5.1 mm/hr was recorded during the monitoring period, so in this

NOISE REPORT
April 2015
14
 EirGrid — 400kV Transmission Line Noise Monitoring

instance, the data for L5 noise level have been averaged for data relating to rainfall between
4.6 mm/hr and 5.6 mm/hr , with an equal number of measurements above and below
5.1 mm/hr. This approach has been taken to provide a sufficiently large sample of data to
allow representative analysis to be carried out. These calculations correspond with the L50 and
L5 noise levels, under the assumption that the average noise levels will increase for any
increase in rainfall.

These measured L50 and L5 levels have been compared with predicted levels at 200m for the
control location and 5m for the tower and mid-span locations. The data for the control location
are shown in Table 7.4 below.

TABLE 7.4: MEASURED AND PREDICTED CORONA DISCHARGE NOISE LEVELS:

CONTROL LOCATION

LAeq, dB

Noise Indicator Measured BPA EPRI GA1PL50 GP1PL5 GP3PL50 GP3PL5

Wet L50 41.5 46.0 44.5 41.6 - 45.8 -

Heavy Rain L5 44.6 49.5 49.7 - 48.7 - 37.8

The calculated levels for almost all methods over-predict the result at the control location. The
L50 predictions are on average 3 dB greater than the measured levels. It should be noted that
the BPA, EPRI and GP3PL50 methods predict 3.9 dB greater than measured levels on
average. The GA1PL50 methodology predicts to a very high accuracy, with only 0.1 dB
difference to the measured level.

The L5 predictions over-predict to an even greater extent, with the exception of the GP3PL5
method, where calculated levels are more than 7 dB lower levels than measured. The
remaining three methodologies agree with each other to within 1 dB but differ from measured
levels by 4.7 dB on average.

Equivalent data for the tower location are shown in Table 7.5 below.

TABLE 7.5: MEASURED AND PREDICTED CORONA DISCHARGE NOISE LEVELS:

TOWER LOCATION

LAeq, dB

Noise Indicator Measured BPA EPRI GA1PL50 GP1PL5 GP3PL50 GP3PL5

Wet L50 48.4 55.2 56.0 57.4 - 54.3 -

Heavy Rain L5 51.6 58.7 61.2 - 65.3 - 57.3

The calculated levels for almost all methods over-predict the result at the tower location. The
L50 predictions are on average 7.3 dB greater than the measured levels.

The L5 predictions are on average 9 dB greater than measured levels. It is noted that the BPA,
EPRI and GP3PL5 methods predict an average of 7.5 dB greater. The GP1PL5 predicts a
level 13.7 dB greater that measured.

NOISE REPORT
April 2015
15
 EirGrid — 400kV Transmission Line Noise Monitoring

Equivalent data for the mid-span location measured against predicted levels are shown in
Table 7.6 below.

TABLE 7.6: MEASURED AND PREDICTED CORONA DISCHARGE NOISE LEVELS:

MID-SPAN LOCATION

LAeq, dB

Noise Indicator Measured BPA EPRI GA1PL50 GP1PL5 GP3PL50 GP3PL5

Wet L50 47.0 55.2 56.0 57.4 - 54.3 -

Heavy Rain L5 52.8 58.7 61.2 - 65.3 - 57.3

Predicted levels for the mid-span are the same as those for the tower as they are based on
the same distance from the line. Predicted levels may increase if line sag is taken into
consideration (this will be greatest at the mid-span), however this would further increase the
degree to which noise levels are over-predicted. However, the measured sound levels show a
lower L50 at the mid-span location than that at the tower, but a greater measured L5 level
(although this is still an average of 7.8 dB lower than the predicted values).

8 DISCUSSION OF FINDINGS

8.1 Overall levels

Near to the tower, noise levels are greater in magnitude and consistency than compared to the
other sites. Levels typically do not reduce significantly at night and often increase, despite the
significant drop in background noise levels (as seen at the control location). Tower levels are
typically greater than at the control and mid-span locations during the day and significantly
greater at night. Noise levels away from the tower are lower during dry conditions, particularly
at night.

It was noted that there was an increase in noise levels during periods of higher humidity. This
suggests that humidity is a significant factor in the production of noise in dry conditions. It is
likely that this increase is due to higher levels of condensation on the tower insulators and the
conductors.

It has been suggested by EirGrid that the use of composite insulators (as opposed to the glass
insulators found on the tower under investigation) may influence the noise produced close to
the tower, but this has not been quantified as part of this study.

During rain, the control location experienced some increase in levels and rain noise became
the dominant noise source above a rainfall rate of approximately 1 mm/hr. Corona noise was
determined to contribute to a less than 1 dB increase in average noise levels, but can be
audible in the worst-case situations.

Noise levels at the tower rise rapidly with increasing rainfall due to the onset of corona
discharge. The prominence of noise from the tower is most significant during low and medium
rainfall rates (up to 5 mm/hr), where levels are 6-10 dB greater than background levels at the
control location. Higher rainfall rates resulted in increased background noise levels and hence
lessened the impact of corona noise. At these high rainfall rates, corona noise was 3-5 dB
greater than background levels.

Noise levels at the mid-span monitoring location behaved similarly to the control location in dry
conditions and in light rain but experienced a more rapid increase in level with rainfall. Above a

NOISE REPORT
April 2015
16
 EirGrid — 400kV Transmission Line Noise Monitoring

rainfall of 1.3 mm/hr, the mid-span noise levels closely match those at the tower. This
indicates that in high rainfall, the conductors are the primary source of noise rather than the
insulators at the tower.

The noise source levels obtained from this work can be used to predict levels at a variety of
distances from power lines for the assessment of new or proposed lines. It is recommended
that such calculations be carried out using detailed noise modelling, to account for local
geography, ground cover, screening, etc.

8.2 Tonality and Impulsivity

Based on the measured noise levels during rainfall at this location, there is potential for the
noise to be considered a significant adverse impact (as described in BS 4142:2014) on
potential receptors close to the line. This would depend on the precise distance of receptors
from the line and the receptors’ sensitivity to noise. The presence of the 100 Hz tonal
component increases the audibility of the sound.

The impact of this tonal component has been considered following the methodology set out in
13
BS 4142:2014 . According to this method, a noise impact penalty of up to 6 dB can be
applied depending on the prominence of the tone under consideration. An additional penalty
can also be applied when a sound is found to be impulsive, resulting in a further penalty of up
14
to 9 dB .

Detailed sound recordings from a sample audio file have been processed according to the
methodologies set out in BS4142. This identified that a tonality penalty of 6 dB and an
impulsivity penalty of 4 dB would be applicable under this standard, although these penalties
could vary depending on the situation under consideration, such as the distance from the line
at which the assessment is being carried out. These can be calculated for other sites and
other distances from appropriate measurements or by using suitable acoustic calculation
formulae or calculation software.

8.3 Comparison with Predicted levels

Two sets of level predictions were made using a range of prediction methodologies; the
prediction of expected levels without the line, i.e. the background noise and rainfall noise, and
the prediction of noise levels from the line, i.e. corona noise.

Predictions of noise from rainfall itself apply for all conditions at the control location and can be
used as a comparative measure for wet conditions at the tower and mid-span. It was found
that in dry conditions and low rainfall rates, the control and mid-span locations closely match
the expected levels. The tower experienced much greater levels than expected, typically 1-3
dB greater during the daytime and 7-9 dB during the night time.

During periods of rainfall, measured noise levels at the control location continued to match
predicted levels from rainfall alone. The mid-span location experienced expected noise levels
for low rainfall rates but levels quickly increased with higher rainfall, similar to those at the
tower. Both the mid-span and tower locations experienced noise levels approximately 6 dB
greater than expected levels during significant rainfall. This difference is due to the additional
corona noise which affects these two measurement locations.

Predictions of corona noise were found to be higher than measured levels. However it is
important to consider that the site under investigation is an aged line and it has been

13
BS 4142:2014 Annex D: Objective method for assessing the audibility of tones in sound: Reference method
14
BS 4142:2014 Annex E: Objective method for measuring the prominence of impulsive sounds and for adjustment of LAeq

NOISE REPORT
April 2015
17
 EirGrid — 400kV Transmission Line Noise Monitoring

suggested that corona noise levels from a new line can decrease by over 10 dB over the first 3
15
years of a line’s lifetime . This is due to a smoothing effect of weather conditions and pollution
on the conductors over time. Excess pollution, dirt, insects and other irregularities can become
built up on the line during extended dry weather, increasing noise levels. These are often
washed away during heavy rain, restoring previous lower noise levels. Irregularities due to
damage to the line, inadequate maintenance or poor design can also increase noise levels.
Although prediction methodologies are specified to apply to ‘aged conductors’, ‘aged’ is not
quantified. This may be a significant factor in the large error of predictions.
th
Two parameters were predicted, the L50 and the L5, corresponding to the median and 95
percentile noise levels during rainfall respectively. The results from the various prediction
methodologies were generally consistent, with strong agreement (within 3 dB) between
methodologies at distances under 100m for the L50 and between 100-300m for the L5. Despite
some strong agreement between methods, measurements were significantly lower than
predicted levels, by over 10 dB in some cases. There was some variation between the
individual prediction methodologies, but the BPA and EPRI methods were found to be
consistent with each other. It is considered that these methods are more reliable and suitable
for future use than evolutionarily formed methods, although the differences between the output
from the prediction models and measured levels are of some concern.

This over-prediction of levels may influence any noise assessment of a transmission line, and
result in impacts being considered to be more significant than would occur in practice. Despite
this, it is also important to consider that initial noise levels from a new line are expected to be
significantly greater in the first few years following installation than after some time, making
these over-predictions a possible worst-case prediction method.

9 CONCLUSIONS

In this report, noise from the 400kV transmission line in Cloney, Co. Kildare has been
quantified based on data from over 10 weeks of measurements. Noise measurements were
made simultaneously at three locations relative to the power line; directly beneath a tower, at
the mid-span position half way between towers, and at a distance of approximately 200m from
the line as a control location during a wide range of weather conditions. A number of
methodologies for predicting transmission line noise levels have been evaluated against
measured data, and the findings reported.

It has been shown that 200m from the line, at the control location, the contribution to the
overall noise levels by the transmission line is low, even in the worst-case weather conditions,
when noise emissions from the line are at their greatest difference from ambient levels. Close
to the line, near the mid-span point between towers, the contribution to the overall noise levels
by the transmission line is low in dry conditions but increases rapidly with rainfall. In high
rainfall conditions, the noise from the line at the mid-span increased significantly and was the
dominant noise source. Noise levels at the tower were significantly greater than at other
locations during dry conditions, particularly at night when humidity levels were often higher.
This may be due to condensation effects. During a range of weather conditions, noise from the
line at the tower was the dominant noise source, and for periods of higher rainfall rates the
noise from the line at the tower increased significantly.

It has been determined that; the primary sources of noise from the line during dry conditions
are located at the tower. The primary sources of noise from the line in wet conditions are the
conductors along the length of the line.

15
Electric Power Research Institute, ‘Chapter 6: Audible Noise’, Transmission Line Reference Book 345kV and Above, Second Edition, EPRI, 1982, p. 299.

NOISE REPORT
April 2015
18
 EirGrid — 400kV Transmission Line Noise Monitoring

Prediction methodologies including the BPA method, EPRI method and a number of
Evolutionarily Formed methods (GP3PL50, GA1PL50, GP3PL5 and GP1PL5) have been
evaluated against measured data. With the exception of the GP3PL5 method which predicted
significantly lower levels than measured, all predicted noise levels were greater than those
measured.

The significant variation between measured and predicted values could be due to variations in
line load, or variations in line ageing effects. The prediction methodologies may be best suited
for worst-case initial noise impact predictions. The BPA and EPRI methods are considered to
be more consistent and therefore more suitable for use in the predictions of future
transmission line noise.

10 RECOMMENDATIONS

It is recommended that more research takes place to further quantify noise from high-voltage
transmission lines. Evaluation of the influence of different tower specifications may also allow
better identification of the source of dry weather noise at the towers. Long term measurements
over several years at a new installation would also allow quantification of variation in sound
levels with the aging of conductors.

Assessment of an alternative tower with composite insulators would provide comparative data
to determine and quantify any effect of using different insulators, although it has not been
confirmed whether insulator selection significantly effects corona noise generation.

Further investigation into the error and accuracy of the BPA and EPRI prediction
methodologies would provide useful information for the evaluation of future transmission line
developments and quantify or remove a large portion of uncertainty in predicting future noise
levels.

The currently adopted method for assessing the impact of transmission line noise, BS 4142,
has recently been updated (2014), adopting significantly greater noise impact penalties for
tonality and impulsivity than was evident in the previous version of the standard. An
investigation into the impact of corona noise on noise sensitive receptors in proximity to
transmission line noise would allow further context to be brought to any noise impact
assessment for new or modified transmission lines. This may either take the form of the
provision of further guidance relating to the implementation of BS4142:2014 or an update to
the 21-year old transmission line noise impact guidance given in TR(T)94.

Finally, the development of a new prediction methodology based upon acoustical foundations
and noise impact (as opposed to the electrical engineering ‘audible noise phenomenon’ based
approach of current prediction methodologies) could be an exceedingly useful tool for future
assessments.

NOISE REPORT
April 2015
19
 EirGrid — 400kV Transmission Line Noise Monitoring

APPENDIX A: NOISE PERCEPTION AND TERMINOLOGY

Between the quietest audible sound and the loudest tolerable sound there is a million to one ratio in sound
pressure (measured in Pascals, Pa). Because of this wide range, a noise level scale based on logarithms is
used in noise measurement called the decibel (dB) scale. Audibility of sound covers a range of
approximately 0 to 140 dB. The human ear system does not respond uniformly to sound across the
detectable frequency range and consequently instrumentation used to measure noise is weighted to
represent the performance of the ear. This is known as the ‘A weighting’ and annotated as dB(A). Table A.1
lists the sound pressure level in dB(A) for common situations.
TABLE A.1: NOISE LEVELS FOR COMMON SITUATIONS
Typical noise level, dB(A) Example
0 Threshold of hearing
30 Rural area at night, still air
40 Public library, refrigerator humming at 2m
50 Quiet office, no machinery
60 Normal conversation
70 Telephone ringing at 2m
80 General factory noise level
90 Heavy goods vehicle from pavement
100 Pneumatic Drill at 5m
120 Discotheque – 1m in front of loud speaker
140 Threshold of pain

The noise level at a measurement point is rarely steady, even in rural areas, and varies over a range
dependent upon the effects of local noise sources. Close to a busy motorway, the noise level may vary over
a range of 5 dB(A), whereas in a suburban area this variation may be up to 40 dB(A) and more due to the
multitude of noise sources in such areas (cars, dogs, aircraft etc.) and their variable operation. Furthermore,
the range of night-time noise levels will often be smaller and the levels significantly reduced compared to
daytime levels. When considering environmental noise, it is necessary to consider how to quantify the
existing noise (the ambient noise) to account for these second to second variations.

Statistical Noise Levels

Statistical noise levels can be obtained from measurements, such as L5, L10, L50, L90 etc. Each LN level
represents the level exceeded for N% of the time during the logging period. Typically A-weighted, these can
be written in the form LN dB(A) or LAN dB. The LA50 represents the median A-weighted level. The LA0 and
LA100, usually written LAmax and LAmin respectively, are the maximum and minimum recorded A-weighted levels
for each logging period.

Background Noise Levels

A parameter that is widely accepted as reflecting human perception of the ambient noise is the background
noise level, L90, which is usually A- weighted and can be written as L90 dB(A) or LA90 dB. This is the noise
level exceeded for 90 % of the measurement period and generally reflects the noise level in the lulls between
individual noise events. Over a one hour period, the LA90 will be the noise level exceeded for 54 minutes.

NOISE REPORT
April 2015
20
 EirGrid — 400kV Transmission Line Noise Monitoring

Ambient or Activity Noise Levels

The equivalent continuous A-weighted sound pressure level, LAeq (or Leq dB(A)) is the single number that
represents the total sound energy measured over that period. LAeq is the sound level of a notionally steady
sound having the same energy as a fluctuating sound over a specified measurement period. It is commonly
used to express the energy level from individual sources that vary in level over their operational cycle.

Noise Changes

Human subjects are generally only capable of noticing changes in noise levels of no less than 3 dB(A). It is
generally accepted that a change of 10 dB(A) in an overall, steady noise level is perceived to the human ear
as a doubling (or halving) of the noise level. (These findings do not necessarily apply to transient or non-
steady noise sources such as changes in noise due to changes in road traffic flow, or intermittent noise
sources).

Frequency Spectrum

Frequency is the rate at which the air particles vibrate. The more rapid the vibrations, the higher the
frequency and perceived pitch. Frequency is measured in Hertz (Hz).

A young person with average hearing can generally detect sounds in the range 20 Hz to 20,000 Hz (20 kHz).
Figure A.1 below illustrates the range of frequencies, for example, the lowest note on a full scale piano, ‘A’,
has a fundamental at 28 Hz, and the highest, ‘G’, a fundamental at 4186 Hz (there will be higher order
harmonics). Human speech is predominantly in the range 250 Hz - 3000 Hz.
The musical term ‘octave’ is the interval between the first and eighth note in a scale and represents a
doubling of frequency. A series of octave and one-third octave bands have been derived, as shown on
Figure A.1 and these are commonly used in noise measurements where it is necessary to describe not only
the level of the source noise but also the frequency content. The frequency content of a noise source can be
useful for identifying acoustic features such as a whine, hiss or screech. One-third octave bands can be
further subdivided into smaller intervals, such as one-sixth octave, one-twelfth octave or one-twenty-fourth
octave bands, etc. One-twenty-fourth octave bands are often utilised for spectral analysis to identify tonal
components in a signal.

Figure A.1: 1 octave and 1/3 octave frequency bands

NOISE REPORT
April 2015
21
 EirGrid — 400kV Transmission Line Noise Monitoring

APPENDIX B: SITE MAP AND MONITORING LOCATIONS

Tower monitoring location ~ 500m

Mid-span monitoring location

~ 200m

Control monitoring location ~ 400m

NOISE REPORT
April 2015
22
 EirGrid — 400kV Transmission Line Noise Monitoring

APPENDIX C: TOWER STYLE AND SPECIFICATIONS

This appendix presents the specification for the power line monitored, as provided by EirGrid.

The transmission line uses Type 401 towers; dimensions of the tower within the measurement site are given
below in Figure C.1. The maximum permissible sag on the line at the mid-point between towers is 9 meters.

The line nominally operates at 400kV but is rated for up to 420kV, with the load on the line varying with
demand. The specification for the conductor is given in Table C.1.

Figure C.1: Tower style and dimensions

6.5 m

6.0 m

4.1 m

10.5 m

40.1 m

34.1 m Maximum Allowable Line Sag: 9 m

TABLE C.1: CONDUCTOR SPECIFICATION

2
Size (approx. mm ) 600
2
Actual Area (mm ) 592
Code Name Curlew
Conductor diameter, d (cm) 3.168
Mass per unit length, w (kg) 1.98
2 6
Young’s Modulus, E (kg/m/s ) 0.65 x 10
-1 -6
Coefficient of Linear Expansion, α (˚C ) 19.5 x 10
Nominal Breaking Load (kg) 16600

NOISE REPORT
April 2015
23
 EirGrid — 400kV Transmission Line Noise Monitoring

APPENDIX D: MEASURED NOISE LEVELS BY TIME OF DAY

TABLE D.1: 10-WEEK CONTROL NOISE LEVELS BY HOUR

Average 10-minute Noise Level, dB Absolute Noise Level, dB

Hour LAeq LAmax LAmin LA10 LA90 LAeq LAmax LAmin

00:00 – 01:00 33.5 46.6 21.5 34.2 23.3 41.1 82.1 16.2
01:00 – 02:00 32.4 45.6 21.0 32.9 22.7 38.5 68.2 15.6
02:00 – 03:00 31.1 44.2 21.0 31.5 22.6 36.5 66 15.6
03:00 – 04:00 30.2 43.1 20.9 30.6 22.4 35.8 64.3 16.1
04:00 – 05:00 31.0 44.0 21.0 31.4 22.5 35.1 62.9 16.8
05:00 – 06:00 34.1 46.6 22.0 34.8 23.9 37.5 71.2 16.8
06:00 – 07:00 38.0 50.4 25.0 38.6 27.9 40.9 77.1 17.3
07:00 – 08:00 41.1 52.5 28.1 41.7 31.6 44.1 82.8 18.5
08:00 – 09:00 42.0 53.6 30.0 42.2 33.3 44.6 89.2 20.0
09:00 – 10:00 40.4 51.5 29.1 40.4 32.0 42.7 78.5 21.2
10:00 – 11:00 39.5 50.0 27.5 38.8 30.2 42.1 88 20.9
11:00 – 12:00 39.6 50.1 27.6 39.2 30.4 42.2 88.9 19.6
12:00 – 13:00 39.9 51.7 27.8 40.0 30.8 42.8 91.7 20.3
13:00 – 14:00 39.8 52.4 28.2 40.6 31.3 42.8 77.9 17.5
14:00 – 15:00 40.3 52.1 28.7 40.9 31.8 43.2 74.9 17.5
15:00 – 16:00 40.7 52.2 28.3 41.0 31.5 43.5 89.4 19.7
16:00 – 17:00 40.8 51.6 28.2 40.7 31.4 44.0 81.3 20.0
17:00 – 18:00 41.1 52.1 28.9 41.5 32.4 44.1 88.9 20.0
18:00 – 19:00 41.4 52.6 28.7 41.9 32.7 44.1 76 19.8
19:00 – 20:00 41.1 52.4 27.9 41.7 31.8 43.5 88.2 19.4
20:00 – 21:00 39.7 50.6 26.2 40.5 30.0 42.1 82.5 18.4
21:00 – 22:00 38.3 50.2 24.1 39.1 27.6 42.0 81.1 17.2
22:00 – 23:00 36.9 49.0 22.7 37.9 25.5 41.6 69.8 16.6
23:00 – 00:00 35.0 47.8 21.8 35.9 24.0 39.9 75.8 16.0

NOISE REPORT
April 2015
24
 EirGrid — 400kV Transmission Line Noise Monitoring

TABLE D.2: 10-WEEK TOWER NOISE LEVELS BY HOUR

Average Noise Level, dB Absolute Noise Level, dB

Hour LAeq LAmax LAmin LA10 LA90 LAeq LAmax LAmin

00:00 – 01:00 40.2 48.0 20.9 41.2 37.2 47.0 76.8 16.1
01:00 – 02:00 40.5 47.5 21.0 40.9 37.3 44.8 77.6 16.8
02:00 – 03:00 40.7 47.2 22.0 41.0 37.5 43.8 75.5 16.8
03:00 – 04:00 40.4 46.6 25.0 40.6 37.5 45.9 82.6 17.3
04:00 – 05:00 40.5 47.2 28.1 40.8 37.5 43.0 72.6 18.5
05:00 – 06:00 41.2 48.6 30.0 41.6 37.7 43.4 66.3 20.0
06:00 – 07:00 42.9 51.8 29.1 43.4 38.9 44.7 74.6 21.2
07:00 – 08:00 44.4 54.2 27.5 44.9 39.7 45.8 84.3 20.9
08:00 – 09:00 44.6 55.4 27.6 44.8 39.8 46.3 85.8 19.6
09:00 – 10:00 43.1 53.9 27.8 43.3 38.4 45.7 89.7 20.3
10:00 – 11:00 41.9 53.4 28.2 41.9 36.5 45.2 97.8 17.5
11:00 – 12:00 41.4 52.0 28.7 40.8 35.0 45.9 89.4 17.5
12:00 – 13:00 41.1 52.9 28.3 41.5 35.0 45.1 90.3 19.7
13:00 – 14:00 40.9 54.1 28.2 42.1 35.1 48.0 97.6 20.0
14:00 – 15:00 41.4 54.7 28.9 42.9 35.7 46.2 86.8 20.0
15:00 – 16:00 41.5 54.8 28.7 43.1 35.5 47.0 89.1 19.8
16:00 – 17:00 41.6 55.8 27.9 42.8 35.4 47.3 88 19.4
17:00 – 18:00 41.9 55.1 26.2 43.1 36.2 45.7 93.5 18.4
18:00 – 19:00 41.7 54.0 24.1 43.2 36.3 45.1 83 17.2
19:00 – 20:00 41.4 52.6 22.7 43.0 36.4 44.3 77.4 16.6
20:00 – 21:00 40.6 50.6 21.8 42.4 36.2 43.1 79.3 16.0
21:00 – 22:00 40.3 50.2 33.7 41.7 36.2 44.5 79.9 22.7
22:00 – 23:00 40.3 49.4 34.2 41.9 36.6 46.1 76.8 23.4
23:00 – 00:00 40.1 48.4 34.5 41.3 36.9 45.3 76 21.3

NOISE REPORT
April 2015
25
 EirGrid — 400kV Transmission Line Noise Monitoring

TABLE D.3: 10-WEEK MID-SPAN NOISE LEVELS BY HOUR

Average Noise Level, dB Absolute Noise Level, dB

Hour LAeq LAmax LAmin LA10 LA90 LAeq LAmax LAmin

00:00 – 01:00 32.1 43.0 22.9 32.9 24.7 42.9 75.6 16.4
01:00 – 02:00 31.4 43.0 22.5 32.1 24.2 40.5 76.2 16.3
02:00 – 03:00 31.2 42.7 22.6 31.6 24.3 39.3 68.9 15.8
03:00 – 04:00 30.6 41.5 22.7 30.9 24.4 42.4 99.8 16.8
04:00 – 05:00 31.3 42.3 22.7 31.7 24.5 37.7 68.5 17.0
05:00 – 06:00 33.5 44.5 23.5 34.2 25.5 37.1 78.6 17.0
06:00 – 07:00 37.3 47.9 26.0 36.9 28.5 40.2 79.9 17.7
07:00 – 08:00 39.7 50.6 28.3 39.0 31.4 42.8 89.3 19.7
08:00 – 09:00 40.5 51.5 29.7 39.5 32.6 45.6 90.9 22.0
09:00 – 10:00 39.0 49.5 29.2 38.4 31.7 42.4 85.2 21.6
10:00 – 11:00 38.3 49.2 28.2 37.7 30.7 44.2 95.5 20.7
11:00 – 12:00 38.7 48.2 27.7 37.2 30.1 43.4 85.7 19.6
12:00 – 13:00 38.5 48.6 28.3 38.1 30.8 42.3 85.6 19.2
13:00 – 14:00 38.4 47.8 27.6 37.4 30.2 42.8 85.7 17.5
14:00 – 15:00 38.9 48.8 28.6 38.5 31.3 42.9 79.4 16.9
15:00 – 16:00 39.2 49.1 28.4 38.7 31.1 44.7 93.8 19.3
16:00 – 17:00 38.8 48.9 28.3 38.4 31.1 42.3 80.8 20.3
17:00 – 18:00 39.0 49.1 29.0 38.9 31.9 41.9 77.4 20.0
18:00 – 19:00 39.0 49.5 28.6 39.0 31.7 42.1 83.6 19.1
19:00 – 20:00 38.8 49.6 27.9 38.8 31.1 43.3 99.5 19.1
20:00 – 21:00 37.1 46.7 26.5 37.5 29.7 40.5 82.4 18.7
21:00 – 22:00 35.8 46.6 24.7 36.4 27.6 40.3 81.4 16.8
22:00 – 23:00 34.4 44.5 23.5 35.4 26.1 41.4 75.9 16.8
23:00 – 00:00 33.1 44.0 22.9 34.1 25.1 40.6 82.7 16.1

NOISE REPORT
April 2015
26
 EirGrid — 400kV Transmission Line Noise Monitoring

APPENDIX E: DATA LOST OR REMOVED

 Data lost due to battery failure in Period 1:

o Equipment batteries at the control location failed after 13 days of a 17-day monitoring period.
Equipment batteries at the tower location failed approximately 36 hours before collection.
o Shorter periods between maintenance schedules were implemented. No further data losses
due to battery issues were encountered.
o Incomplete logging periods immediately before failure were also excluded.

 Data lost due to flooding in Period 5:

o 07/11/14 01:40 onwards was removed from the mid-span location data, it was determined
that flooding contaminated the recordings shortly after this time. A total of 52 hours were
excluded prior to the equipment failure. A further 106 hours of data were not captured
between failure and collection of the equipment. Approximately 8.5 days of data were
captured during this period, with a further 6.5 days excluded or not captured.

 Data removed due to high wind speeds:

o Data corresponding to wind speeds above 5 m/s were excluded from analysis. A total of 5
hours of data were excluded.

 Data removed due to contamination by staff:

o Initial set-out; the first 30-minutes of data were removed for all locations.
o Subsequently, the first and last periods of data were removed (usually 10-minutes but were
sometimes extended if staff were on site for a greater period of time).
o Approximately 2.5 hours of data were removed in total.

 Total data removed & remaining:

o Data lost at control location: Approximately 96 hours
o Data excluded at control location: Approximately 7.5 hours
o Good data remaining at control location: Approximately 1728 hours (94%)
o Data lost at tower location: Approximately 36 hours
o Data excluded at tower location: Approximately 7.5 hours
o Good data remaining at tower location: Approximately 1788 hours (98%)
o Data lost at mid-span location: Approximately 106 hours
o Data excluded at mid-span location: Approximately 58 hours
o Good data remaining at mid-span location: Approximately 1667 hours (91%)

NOISE REPORT
April 2015
27
 EirGrid — 400kV Transmission Line Noise Monitoring

APPENDIX F: EQUIPMENT CALIBRATION CERTIFICATES

Equipment calibration certificates are given for the following equipment used:

 Rion NL-52 Sound Level Meter (Tower)

 Rion NL-52 Sound Level Meter (Mid-span)

 Rion NL-52 Sound Level Meter (Control)

 Rion NL-52 Sound Level Meter (Spot measurements)

 Rion NC-74 Acoustic Calibrator

 Weather Station conformance certificates

NOISE REPORT
April 2015
28