AQUO

AQUO
Achieve QUieter Oceans by shipping noise footprint reduction
FP7 - Collaborative Project n° 314227
WP 3: Measurements
Task T3.1
European URN Standard Measurement Method
Deliverable Reference Issue Dissemination status
yes no D 3.1 Rev 2 PU
File name D3.1 European URN Standard Measurement Method_Rev2.Doc
Name Organisation Function Date
Issued by Alfonso Moreno TSI Task leader 21/04/2014
Checked by Publio Beltrán TSI WP leader 21/04/2014
Approved by Christian Audoly DCNS Coordinator 23/062014

WP 3 - Measurements
D 3.1 European URN Standard Measurement Rev 2
Method
REVISION HISTORY
ISSUE DATE NAME ORGANISATION MODIFICATIONS
Alfonso Moreno
Draft document for
Rev 0 15-4-2014 Raúl Salinas TSI
comments
Publio Beltrán
Rev 1 21-4-2014 Publio Beltrán TSI Verification
Minor modifications in
Rev 2 23-4-2014 Christian Audoly DCNS
Annex A
Dissemination level: PUBLIC 2
© AQUO Project Consortium 2012 - all rights reserved

WP 3 - Measurements
Method
LIST OF CONTRIBUTORS
CHAPTER SECTION NAME ORGANISATION
Chapter 1 1.1 Christian Audoly DCNS

Raúl Salinas TSI
Publio Beltrán TSI
Raúl Salinas
1.2 TSI
Alfonso Moreno
Raúl Salinas
Publio Beltrán
1.3 TSI
Ángela López
Alfonso Moreno
Raúl Salinas
Chapter 2 Publio Beltrán TSI
Alfonso Moreno
Raúl Salinas
3.1 Alfonso Moreno TSI
Álvaro Pérez
3.2 Christian Audoly DCNS
Raúl Salinas
3.3 Publio Beltrán TSI
Alfonso Moreno
Raúl Salinas
Publio Beltrán
Chapter 3 3.4.1 David Cordero TSI
Óscar Zaba
Alfonso Moreno
Christian Audoly
3.4.2 DCNS
Celine Rousset
3.4.3 Jan Hallender SSPA

Raúl Salinas
Álvaro Perez
3.4.4 TSI
David Cordero
Alfonso Moreno
Raúl Salinas
TSI
Publio Beltrán
TSI
Álvaro Pérez
TSI
David Cordero
TSI
Chapter 4 All Ángela López
TSI
Alfonso Moreno
DCNS
Christian Audoly
SSPA
Jan Hallender
BV
Eric Baudin

WP 3 - Measurements
Method
CHAPTER SECTION NAME ORGANISATION
Raúl Salinas
Álvaro Pérez
Chapter 5 All TSI
Publio Beltrán
Alfonso Moreno
Christian Audoly
Annex A All DCNS
Celine Rousset

WP 3 - Measurements
Method
CONTENTS
SUMMARY ................................................................................................................. 9
1. INTRODUCTION ............................................................................................... 11
1.1. AIM AND SCOPE ................................................................................................................................. 11
1.2. METHODOLOGY ................................................................................................................................ 11
1.3. NEEDS .................................................................................................................................................... 12
2. DEFINITIONS .................................................................................................... 14
3. REVIEW OF EXISTING STANDARDS AND PROCEDURES........................... 18
3.1. KEY PARAMETERS ............................................................................................................................ 18
3.2. LIST OF EXISTING STANDARDS AND PROCEDURES CONSIDERED................................... 19
3.2.1. STANAG 1136 ................................................................................................................... 19
3.2.2. ANSI-ASA S12.64-2009/PART 1 AND ISO DOCUMENTS IN PROGRESS .................... 20
3.2.3. DNV SILENT CLASS NOTATION. .................................................................................... 22
3.3. SYNTHESIS OF THE EXISTING REFERENCE DOCUMENTS .................................................. 23
3.3.1. INITIAL CONSIDERATIONS. ............................................................................................ 23
3.3.2. HOW THE “KEY PARAMETERS” ARE TACKLED IN THE STANDARDS....................... 24
3.3.3. THE EXISTING UNCERTAINTIES IN THE STANDARDS. .............................................. 27
3.3.4. CONCLUSIONS AND RECOMMENDATIONS ................................................................. 30
3.4. PARAMETER EVALUATION AND IDENTIFICATION OF DEFICIENCIES IN THE

CURRENT MEASUREMENT PROCEDURES.............................................................................................. 31
3.4.1. DEVICE ............................................................................................................................. 31

WP 3 - Measurements
Method
3.4.2. ENVIRONMENT ................................................................................................................ 55
3.4.3. VESSEL PARTICULARS .................................................................................................. 70
3.4.4. POST-PROCESSING........................................................................................................ 73
4. PROPOSAL FOR A NEW URN MEASUREMENT PROCEDURE .................... 79
4.1. INTRODUCTION.................................................................................................................................. 79
4.2. SUMMARY OF THE REQUIREMENTS........................................................................................... 80
4.3. INSTRUMENTATION ......................................................................................................................... 82
4.3.1. HYDROPHONES, SIGNAL CONDITIONING AND RECORDER ..................................... 82
4.3.2. DISTANCE MEASUREMENT SYSTEM ........................................................................... 82
4.4. TEST SITE ............................................................................................................................................. 83
4.4.1. TEST SITE REQUIREMENTS .......................................................................................... 83
4.4.2. SEA SURFACE CONDITIONS.......................................................................................... 83
4.5. HYDROPHONE DEPLOYMENT ....................................................................................................... 83
4.5.1. GRADE A1......................................................................................................................... 83
4.5.2. GRADE A2......................................................................................................................... 84
4.5.3. GRADE B1......................................................................................................................... 84
4.5.4. GRADE B2......................................................................................................................... 84
4.6. TRANSMISSION LOSS ....................................................................................................................... 85
4.6.1. MODELLING ACTIVITIES................................................................................................. 85
4.6.2. SIMPLE LAW ..................................................................................................................... 87
4.7. URN MEASUREMENT ........................................................................................................................ 88
4.7.1. BACKGROUND NOISE..................................................................................................... 88

WP 3 - Measurements
Method
4.7.2. VALIDATION OF THE MEASUREMENT LOCATION ...................................................... 88
4.7.3. CONFIGURATION OF THE DIFFERENT RUNS.............................................................. 88
4.7.4. REPEATABILITY CHECKING........................................................................................... 90
4.7.5. UNCERTAINTY CHECKING ............................................................................................. 90
4.7.6. VESSEL CONDITIONS ..................................................................................................... 91
4.7.7. TEST SEQUENCE ............................................................................................................ 92
4.8. POST-PROCESSING............................................................................................................................ 93
4.8.1. ACOUSTIC CENTRE ........................................................................................................ 93
4.8.2. DATA WINDOW ................................................................................................................ 93
4.8.3. BACKGROUND NOISE..................................................................................................... 93
4.8.4. SPECIFIC POST-PROCESSING ...................................................................................... 95
4.8.5. NARROW BAND ANALYSIS............................................................................................. 97
4.8.6. DIRECTIVITY .................................................................................................................... 99
4.9. UNCERTAINTY AND REPEATABILITY STUDY........................................................................ 101
4.9.1. THEORETICAL STUDY .................................................................................................. 101
4.9.2. EXPERIMENTAL STUDY................................................................................................ 102
5. CONCLUSIONS .............................................................................................. 105
REFERENCES ....................................................................................................... 107
A. ANNEX A - EUROPEAN WATERS ENVIRONMENTAL DATA...................... 109
A.1 GENERAL DATA ............................................................................................................................... 109
A.2 ACOUSTICAL DATA......................................................................................................................... 114
A.3 PROPAGATION LOSSES.................................................................................................................. 117

WP 3 - Measurements
Method

WP 3 - Measurements
Method
SUMMARY
This study has been realized in the scope of AQUO, a collaborative research project
supported by the 7th Framework Programme through Grant Agreement N°3142 27, whose
final goal of AQUO project is to provide to policy makers practical guidelines to mitigate
underwater noise footprint due to shipping, in order to prevent adverse consequences to
marine life. The present document is the deliverable D3.1 project “European URN Standard
Measurement Method”.
The URN (Underwater Radiated Noise) is the physical quantity that allows quantifying the
underwater noise emission of a ship or a maritime system. Thus the availability of reliable
measurement procedures is needed in order to determine properly the underwater noise field
created by a vessel or more generally ship traffic in a maritime area of interest.
These are already some standards for the ship URN measurement, but they suffer some
limitations:
- the requirement on measurement parameters and of the quantification of
uncertainties and repeatability is not sufficient, as only few justification documents are
available,
- they address mainly measurements in deep waters, whereas many European
maritime areas are in shallow waters.
The purpose of the present study is to fill this gap. The report is organized in different
sections:
• The introduction, section 1, reminds the objectives of the study and expresses the
needs.
• General and specific terminology useful for this study is defined in section 2.
• In section 3, a review of the existing standards and procedures for ship URN
measurements is done and their limitations are identified. Then, the main part of this
section consists in identifying the key parameters driving the uncertainty and
repeatability of the measurement. An in-depth analysis is performed, giving examples
and also by quantifying, as far as possible, the impact of each parameter. The
parameters are of different types: frequency range of interest and disturbances such
as cable vibrations, signal processing issues, determination of accurate distance
between the ship to be measured and the sensors, influence of underwater acoustic
propagation and reflection of sound waves on sea surface and sea bottom, etc.

WP 3 - Measurements
Method
• In section 4, a proposal for a new procedure for ship URN measurements is given.
Two grades are defined: Grade A for engineering purposes, with high accuracy and
repeatability, and Grade B for comparison to noise limits, with medium accuracy and
repeatability. Furthermore, these two grades are split into grades A1/B1 and A2/B2,
for use in shallow waters and deep waters, respectively. The main evolutions by
comparison to existing procedures are: the recommendation to deploy the sensors
from sea bottom in case of shallow water measurement, the use of specific methods
for determining the actual distance between ship and sensors and Grade A. One key
issue is the determination of sound waves propagation loss between the noise source
(i.e. the ship) and the sensors, which is done by using adequate predictive models for
Grade A, instead of simple laws (such as spherical spreading).
• Section 5 consists in the validation of the proposed procedure. This has been done
first by using numerical simulation, allowing verifying the impact of different
parameters on the accuracy. Then a real experiment has been performed on a ship.
Results show a very good repeatability, confirming the robustness of the proposed
procedure.
In conclusion, this study has allowed to study in detail the effect of different key parameters
on the uncertainty and repeatability of URN measurement of ships, both in deep and shallow
waters. One of the most important parameters is the accurate determination of sensors
locations and estimate of sound propagation loss. A new procedure is defined, split into two
grades and two variants for deep and shallow waters. The results of the study prove that the
needs expressed for accuracy and repeatability can be achieved. It is thought that the work
carried out here is a significant contribution to the improvement of ship URN measurement
techniques, which can be used to build a new standard or to contribute to the work done in
the scope of the international standardization organizations.

WP 3 - Measurements
Method
1. INTRODUCTION
1.1. Aim and scope
This study has been realized in the scope of AQUO, a collaborative research project
supported by the 7th Framework Programme through Grand Agreement N°314227, whose
final goal of AQUO project is to provide to policy makers practical guidelines to mitigate
underwater noise footprint due to shipping, in order to prevent adverse consequences to
marine life. The present document is the deliverable D3.1 project “European URN Standard
Measurement Method”.
Underwater radiated noise (URN) is the physical quantity that represents the underwater
noise emitted by a ship or a naval system, due to acoustic radiation of the hull excited by
internal machinery, to the propeller, or to other noise sources such as flow noise.
For a long time, URN has been a matter of concern mainly for military vessels, in order to
limit the risk of detection by adverse passive sonar systems. In the civilian domain, acoustic
requirements on research vessels have been imposed for a few decades. Nowadays, the
need for URN mitigation is foreseen also to any type of vessel in the context of
environmental concerns regarding ship traffic, for the preservation of fish resources and
biodiversity of marine fauna.
Then, appropriate standards have to be defined for the measurement of ship URN levels in
accordance with the needs. Different measurement procedures for Underwater Radiated
Noise are already available but suffer from some limitations. The aim of the present study is
the development of European standard measurement methods for ship URN.
1.2. Methodology
The approach used to perform the study is as follows:
• Identification of the needs with regards to URN control, measurement and

assessment of the environmental impact in terms of underwater noise emitted by
ships.
• Analysis of the current standards and measurement procedures to identify the

advantages and possible deficiencies.
• Exhaustive review of all the parameters affecting the measurement with special
emphasis on the assessment uncertainty associated to them.
• Initial proposal for the procedure, according to the needs;
• Validation of proposed procedures through experiments at sea (not detailed in this

deliverable);
• Development of a final proposal.
• Provide experimental and numerical support for the uncertainty and repeatability of
the measurement procedure.

WP 3 - Measurements
Method
Finally, with the aim to describe properly the approach to cover the scope of the present
document, the Figure 1.1 shows the flow chart of the different activities considered, by
decomposing the activity into different subtasks.
Figure 1.1.- Flow chart of the activities considered.-
1.3. Needs
The aim is the development of a European standard measurement method for ship URN
source levels that can be used in both, shallow and deep water. The case of shallow waters
is an important point, as in a large proportion of European maritime areas, no deep waters
are available. This standard proposal, which is expected to fulfil the needs of industry and
end-users, can be summarized as follow:
• Accurate source identification for engineering purposes.

WP 3 - Measurements
Method
• Assessment of the underwater environmental impact of the vessel in terms of

underwater noise.
• Besides, the possibility of estimating the underwater noise signature of the vessel
without disturbing its route has been considered.
Having in mind these needs the following grades have been defined:
• A-Applications. Engineering purposes: For accurate source identification and

assessment of the URN signature of the vessel for further comparison with limits:
o Grade A1 (Measurement Uncertainty 4dB, Repeatability 1.5dB, Frequency

range: 10Hz-50KHz): For sources identification and accurate quantification.
The measurement procedure and configuration of the device are meant to
allow narrow band analysis over the range 10Hz-50KHz. This grade is
conceived for achieving high accuracy and repeatability in the identification of
the sources below 20 Hz, aiming at building silent vessels. It also focuses on
measurements in shallow water. In spite of the fact that important noise
sources are sometimes below 10Hz, it is unfeasible to measure them due to
technical reasons (as we will detail later).
o Grade A2 (Measurement Uncertainty 4dB, Repeatability 1.5dB, Frequency

range: 20Hz-50KHz): It covers similar needs than the previous grade but it is
conceived for measurements in deep water. Due to technical limitations in the
measurement device used in this kind of measurements, there is an increase
of the lower limit of the frequency range. This is why, this grade is
recommendable only for vessels whose main harmonics are foreseen to be
above 20Hz.
• B-Applications. Comparison with limits: Procurement of the URN signature of the

vessel for traffic management:
o Grade B1 (Measurement Uncertainty 7dB, Repeatability 2dB, Frequency

range 20Hz-50KHz): This grade is conceived to obtain an indicator (signature)
of the underwater environmental impact of the vessel. This grade is conceived
for measurements in shallow water.
o Grade B2 (Measurement Uncertainty 6dB, Repeatability 2dB, Frequency

range 20Hz-50KHz): It tries to cover similar needs as the previous grade but it
is conceived for measurements in deep water.
C-Applications – Ocean monitoring: It is conceived to obtain rough estimates of

underwater noise pollution of the vessel without disturbing its route. Those estimates can be
later used also for traffic management, deriving URN patterns, etc. This topic is outside the
scope of the research carried out in this study.

WP 3 - Measurements
Method
2. DEFINITIONS
• Acoustic centre: Position on the ship where it is assumed that all the noise sources
are co-located as a single point source
• Arithmetical average: Average expressed in dB of different and comparable signals

referred in dB under the same reference value.
• Background noise: Noise from all sources (biotic and abiotic) other than ship being
measured.
• Beam aspect: Direction to either side of the ship under test.
• Calibrated underwater source: Device transforming an electrical signal into an

underwater sound whose frequency and level are known if the frequency and level of
the electrical signal are known.
• Closest point of approach: The point with the minimum horizontal distance (during
a test run) between the acoustic centre of ship under test and the hydrophones.
• Data window(s): Different sets of data for which an averaged spectrum will be
computed.
• Data window length: Distance covered by the acoustic centre of the vessel during a
data window.
• Dipole source: Type of model of the vessel as noise source that assumes that the
source is made up of two points sources in opposite phase symmetrically positioned
with respect to the sea surface.
• Directivity enveloping spectrum: The spectrum resulting from the envelope of the
spectra corrected to 1m of a given side taken at different angles between the acoustic
centre of the vessel and the sea surface above the hydrophones.
• Dynamic range: It is the ratio between the maximum and the minimum signal the
system can distinguish.
• End data location (FINEX): Position of the acoustic centre of vessel under test
where data recording is finished.
• Experimental field calibration: Procedure applied in situ to evaluate the actual

transmission loss of the test site.
• Far field: Field where the ship under test can be considered to behave as a point
source collocated for all frequencies with a low error.
• Floating configuration: Deployment configuration where the hydrophones are

suspended from a floating buoy and not anchored on the bottom.

WP 3 - Measurements
Method
• Frequency range: It is determined by the minimum and maximum frequency where

the measurement system shall record the noise level with an uncertainty lower than
required.
• Frequency resolution: Minimum required change of frequency between two spectral

components so they can be distinguished in the narrow band spectrum of the time
signal recorded by the measurement device.
• Frequency response: Sensitivity of the measurement chain along the frequency.
• Hydrophone bandwidth: It is given by the frequency range where the frequency

response of the hydrophone does not vary more than a certain quantity.
• Hydrophone directivity: It will be determined by the hydrophone sensitivity for a

given frequency and angle.
• Hydrophone sensitivity: It is the magnitude of the noise signal required to produce

a specified output signal. It is expressed in dB (ref 1V/µPa).
• Insert voltage calibration: Known, calibrated and traceable stimulus in the form of
an electrical input injected at the input (or other stage) of a measurement system in
order to ascertain that the system is, in fact, responding properly to known stimulus.
• Measurement Uncertainty (MU): Expected difference between the measured

resulting signature source level and the true signature source level stated in decibels
for a given measurement system and procedure, in each one-third octave band.
• Measurement Repeatability (MR): Expected difference between the resulting

signature source levels, stated in decibels, in one-third octave bands, taken from a
same vessel at different times, under the same vessel and environmental conditions
in terms of what can be quantified, following the same measurement procedure and
using the same measurement equipment.
• Monopole source: Type of model of the vessel as noise source that assumes that
the source is made up of one point source collocated for all frequencies.
• Narrow band component: It is a clear frequency line in the spectrum of the signal
measured by the noise recording system, commonly related to the machinery or the
propeller of the vessel.
• Nominal distance to CPA (dncpa): It is the distance between the hydrophones and
the CPA of the closest run to the hydrophones.
• Omni-directional hydrophone: Underwater sound pressure transducer that

responds equally to sound from all directions.
• Power average: Average expressed in dB of different and comparable signals

referred in a linear scale.
• Range average transmission loss: It is the transmission loss along the horizontal
distance for a given frequency and receiver depth resulting from the logarithmical

WP 3 - Measurements
Method
interpolation of the numerical values of the transmission loss obtained from the
corresponding numerical model at the receiver depth and different horizontal
distances.
• Range-source depth average transmission loss: It is the arithmetic average of the

range average transmission loss of different source depths for a given frequency and
receiver position.
• Receiver depth-average sound pressure level: It is the power average of the

underwater sound pressure level of the hydrophones placed at different depths.
• Resulting signature source level: Final underwater sound pressure level after
removing the effects of sea bottom and sea surface and referred to a distance of 1m
adjusted using the proper transmission loss. It is also referred to as a “source level” or
“signature”. It is stated in decibels (dB) whose reference is 1µPa/√Hz.
• Self-generated noise: It refers to the artificial noise recorded by the hydrophones

due to wave induced motion, electronic noise and any spurious signal caused by the
intrinsically physical properties of the measurement device.
• Ship-draught after: Distance between the vessel’s waterline and the lowest point of
the vessel measured in the perpendicular of the stern.
• Ship-draught forward: Distance between the vessel’s waterline and the lowest point
of the vessel measured in the perpendicular of the bow.
• Ship-length: It is the maximum length of the vessel's hull measured parallel to the
waterline.
• Signal plus noise to noise ratio (SNR): It is the difference stated in dB between the
signal measured during a run (signal plus noise) and the background noise measured
during the trial.
• Sound celerity profile: Sound speed along the water depth for a certain place and a
given time.
• Source depth average transmission loss: It is the arithmetic average of the

transmission loss of different source depth for a given frequency and horizontal
distance.
• Source-receiver depth average sound pressure level: It is the arithmetic average

of the receiver depth-average sound pressure level measured for different source
depths.
• Source-receiver depth averaged transmission loss: The difference, stated in dB

for a given frequency and horizontal distance, between the source level generated by
an emitter at different depths and the source-receiver depth average sound pressure
level measured by the column of the hydrophones.
• Start data location (COMEX): Position of the acoustic centre of vessel under test
where data recording is started.

WP 3 - Measurements
Method
• Test site: Location where the underwater noise measurements are performed.
• Tilt angle: Angle between the vertical axis and the line created by the cable
supporting the hydrophones.
• Transmission loss: Difference, expressed in dB, between the source level at 1m

produced by an emitter placed at a given depth and the underwater sound pressure
level measured by a receiver at a given horizontal distance from the emitter and
vertical distance from the sea surface.
• Transmitting voltage response (TVR): The output sound intensity level generated
at 1m range by the transducer per 1V of input Voltage.
• Underwater noise true signature source level: Underwater sound level stated in
decibels represented in one-third octave bands of a point noise monopole source
whose free far field noise is equivalent to the free far field noise of the vessel, i.e. no
interaction with the sea surface or sea bottom are considered.

WP 3 - Measurements
Method
3. REVIEW OF EXISTING STANDARDS AND PROCEDURES
3.1. Key parameters
Once the requested applications of the standard have been defined, it is necessary to
identify the variables that can affect the measurement procedures and devices.
In order to be able to define proper methodologies and equipment for the applications, the
parameters to be taken into account have been classified as follow:
Post-Processing
Device (D) Environment (E) Vessel constraints (V)
(PP)
Frequency Range of
1 Shallow / Deep Water Vessel Speed Bandwidth
the device
Frequency
AC (Acoustic center)
2 Resolution of the Mirror effect Vessel Length
Definition/ location
device
Relative distance
Propeller & Machinery sources
3 measurement Sea State Data adquisition Window
type and working conditions
equipment
Background Noise
4 Tilt. Propagation Model Vessel load conditions
Correction
Celerity Profile
Hydrophone
5 (Temperature & Family vessel Directivity
movement
Salinity)
6 Distance to the CPA Currents Availability Number of runs
Number of
7 Seabed type
hydrophones
8 Cost Background noise
9 Handling
Independency
(Auxiliary vessel,
10
Previous underwater
hull works, Diver….)
11 Directivity
Accuracy of the
12
instrumentation.
13 Dynamic range
14 Electrical noise
Table 3.1.- Parameters that can influence the measurement procedures and devices.-
The following step will consist on defining requirements for each parameter in order to
determine procedures and devices for A-Applications and B-Application purposes.

WP 3 - Measurements
Method
3.2. List of existing standards and procedures considered

• NATO Standardization Agreement N°1136 (STANAG), “S tandards for use when
measuring and reporting radiated noise characteristics of surface ships, submarines,
helicopters, etc. in relation to sonar detection and torpedo risk, May 29, 1995
(reference [1]).
• AMERICAN NATIONAL STANDARD - Quantities and Procedures for Description and

Measurement of Underwater Sound from Ships – Part 1: General Requirements.
ANSI-ASA S12.64-2009/Part1 (reference [2]).
• ISO/DPAS 17208-1 "Acoustics – Quantities and procedures for description and

measurement of underwater sound from ships – Part 1: General requirements".
• ISO/DIS 16554.2 “Ships and marine technology – Marine environment protection –

Measurement and reporting of underwater sound radiated from merchant ships”.
• DNV. Part 6, Chapter 24 - Silent Class Notation. Rules for Classification of Ships. s.l.
: DNV, January 2010 (reference [9]).
3.2.1. STANAG 1136
The scope is to standardize measurement and report of the radiated noise characteristics of
surface ships, submarines, helicopters, etc, in relation to sonar detection and torpedo
acquisition risk.
The initial goal was to fulfil needs for NATO Navies (North Atlantic Treaty Organization), to
characterize the risk for military vessels of being detected by adverse systems using
underwater acoustics means. However, the document is “NATO Unclassified”, so that it is
made available and can be used for other purposes.
The contents of this document are mainly:
• Terminology units, such as frequencies and noise levels, reference distance, etc;
• The definition of the format for reporting the measurement (this part is detailed in an
appendix);
• Guidelines to conduct the trial and to process the data.
The standard deals both with narrowband and one-third octave band analysis. For the
measurement, one or several hydrophones are considered, and it is recommended to
perform the trial in deep waters. For some applications, measurement can be done in
relatively shallow waters, and a keel aspect hydrophone is considered.
Note also that this STANAG applies both to surface ships and submarines (or underwater
vehicles). For the analysis in section 1.3, we shall consider only the case of surface ships.

WP 3 - Measurements
Method
3.2.2. ANSI-ASA S12.64-2009/Part 1 and ISO documents in progress
The scope is the description of general measurement systems, procedures, and

methodologies used for the measurement of underwater sound pressure levels from ships at
a prescribed operating condition. It contains methodology for the reporting of one-third
octave band sound pressure levels. The context for the production of this standard is the
need for reduction of underwater noise impact on marine life due to man activity, by an
appropriate characterization of ship underwater radiated noise at sea. This it is in close
relationship with the aim of AQUO project.
The contents of this document are mainly:
• Terms and definitions;
• Instrumentation;
• Measurement requirements and procedure;
• Post-processing;
• Measurement uncertainty;
• Reporting example (table of contents for an URN report).
An important issue in this standard is the definition of three grades (A, B, and C)
corresponding to different levels of accuracy and/or completeness of information. The main
purpose is measurement of one-third octave noise levels. However, narrowband analysis is
considered in grades A and B for deeper analysis.
Grade C method, which is intended for survey, uses only one hydrophone (Figure 3.1).
Grades A (precision method) and B (engineering method) use three hydrophones, as shown
on Figure 3.2. The minimum measurement distance is 100 m, or the ship length if it is
greater than 100 m.
Figure 3.1.- Hydrophone location for Grade C method

WP 3 - Measurements
Method
Figure 3.2.- Hydrophone location for Grade A and B methods
The hydrophones are fixed on a vertical line deployed under a surface buoy, or using a
bottom anchor (Figure 3.3).
Figure 3.3.- Examples of hydrophone deployment arrangements
Using ANSI-ASA S12.64-2009 as a start, two ISO Technical Committees have prepared draft
standards for measurement of ship URN in deep waters:
• ISO Committee 8-2 (Marine technology and protection of environment): ISO 16554
• ISO Committee 43-3 (Underwater Acoustics): ISO 17208

WP 3 - Measurements
Method
The situation is not very clear, as two committees have been working on the same topic. The
latest orientation followed is that two ISO documents will be derived from ANSI-ASA S12.64:
ISO 16554 close to Grade C, and ISO 17208 covering the merging of Grades A and B.
These two documents are starting the approval process in 2014.
In all cases, these documents apply to deep waters measurements only (bottom reflection
and non standard underwater propagation are not considered), the water depth being
typically greater than ship length and 100 metres. Work is in progress or planned in ISO
Technical Committees to address the issue of ship URN measurements shallow waters.
3.2.3. DNV Silent Class Notation.
The main goal of this document, issued by a classification society, is to define URN limits for
some classes of ships. Associated to these requirements, a measurement procedure is
defined.
Section 1 includes an introduction, general definitions and the contents of the documents to
be produced for a ship URN measurement.
For the Silent Class Notation, the following classes of ships are considered (section 2 of the
document):
• seismic survey vessels;
• fishery ships;
• research vessels;
• environmental (any vessel demonstrating a controlled environmental noise emission).
The limit noise curves are expressed in one-third octave band levels. Narrowband
measurements are optional.
Section 3 provides general requirements and guidance for the measurement. Appendix A
describes the procedure.
As shown on Figure 3.4, the method uses a single hydrophone placed on sea bed, with a
typical CPA distance comprised between 150 m and 250 m. Then, the measurement is
carried out in relatively shallow waters. However, the document requires a minimum 30 m
water depth under the keel.
The measurement hydrophone must be at a maximum distance of 0.2 above bottom. A -5 dB

correction factor is applied to account for acoustic wave reflection on seabed.

WP 3 - Measurements
Method
Figure 3.4.- DNV document measuring situation
3.3. Synthesis of the existing reference documents
3.3.1. Initial considerations.
The Standards and regulations under study, listed in paragraph 3.2, are the most commonly
used in Europe by the scientific community to carry out underwater noise measurements.
The analysis to identify and evaluate the uncertainties revealed that the strict revision,
standard by standard, of the detected deficiencies would be too tedious and useless because
most of the listed standards are based on similar assumptions, thus with several common
points in their contents. The main problem is the reliability of the assumptions, which can
lead to underestimated or even neglected uncertainties. In general, no document is available
regarding these justifications and the estimation of uncertainties.
The analysis of the standards has detected some similarities among them, which has
permitted to concentrate the analysis on three standards, STANAG [1] , ANSI/ASA [2] and
the Silent Class notation of the DNV [9], making more practical and friendly to the reader the
presentation of the results.
The most relevant difference among them is that the STANAG and ANSI/ASA standard have
been developed to perform deep water measurements while DNV class notation is defined
for shallow water.
This fact is important to weight properly the relevance of the reflections and mirror effect as
well as the configuration of the recommended devices, mainly floating configurations for
deep water and bottom anchored for shallow water.
Finally, it is important to pay attention to the global uncertainty and repeatability of each
standard. These two parameters show the quality of the different standards under study.
Moreover, they will be used to help us to understand the applicability and usefulness of each
methodology.

WP 3 - Measurements
Method
3.3.2. How the “Key Parameters” are tackled in the standards.
Once the most significant “Key parameters” have been listed in the previous paragraphs, it is
now important to spend time presenting the image of how the standards under study tackle
all the parameters affecting the underwater noise measurement.
The following tables summarize, by representative data, the definition of each parameter in
the standards under study. For those parameters without clear definition in the standards, a
“question mark” has been included to identify the lack. This exercise permits, at first glance,
the identification from a qualitative point of view of the number of issues undefined that can
also generate uncertainties. Moreover, the fact that some explicit information exists in the
standard related to some parameters does not guarantee that their corresponding
uncertainty was completely annulled.
Device group
Focusing on the definition of an easy procedure to review all this information, a subdivision
by standards is proposed. In general terms all the matters related to cost, handling and
independency to external support means in the device group have not been considered. It is
not the matter of this paragraph to assess the applicability and cost-effective of the different
devices and procedures to the industry.
a. STANAG standard.
Regarding the device, this standard does not deal with all the key parameters identified
within this topic. Beyond the general comments included as introduction for the device
paragraph, some lack remains present in terms of hydrophone movement and tilt suffered by
the floating configuration. Besides, the aspects regarding the instrumentation and their
capabilities are not specified which implies lack of standardisation of the uncertainty because
of this matter.
b. ANSI/ASA standard.
As happens in the STANAG standard some issues remain undefined in terms of accuracy
required for the measurement of the relative distance between the targeted vessel and the
hydrophones as well as the uncertainty induced by movements in the hanging line. The
complete definition of the instrumentation is also partially defined, anyway a general
indication about uncertainty levels are mentioned to ensure the compliance with the general
uncertainty and repeatability of the standard definition.
c. DNV Class Notation.
Due to the fact that this standard has been defined to measure in shallow water, the
proposed configuration of the device is bottom anchored which does not suffer from the
issues related to tilt and hydrophone movements. On the other hand the instrumentation
remains no completely defined as the accuracy of the instrumentation as well as the self-
noise have not been mentioned, leaving these aspects open.

WP 3 - Measurements
Method
Device
Nº Parameters STANAG ANSI/ASA DNV
G-A: 10 to 50 kHz
1 Frequency Range 10-100kHz G-B: 20 to 25 kHz By vessel Families
G-C: 50 to 10 kHz
G-A: Less than 1 Hz
No Narrow band
2 Frequency Resolution 1Hz & 3% G-B: Not defined
Analysis
G-C:No Narrow band Analysis
2% distance to CPA
Relative measurement
3 +/- 1dB 2% distance to CPA +/- 5m
distances
5% distance to CPA
4 Tilt (for floating devices) ? Max 5º Bottom
5 Hydrophone movement Track + Beam ? Bottom
6 Distance to the CPA >100 m Greater 100m or 1xLoa +/- 10% 150m to 250m
G-A: 3
7 Number of hydrophones 2 (100 m/ 30 m keel) G-B: 3 1
G-C: 1
8 Cost ? ? ?
9 Handling ? ? ?
10 Independency ? ? ?
11 Directivity 1 to 3 dB ? ?
Accuracy of the
12 ? Aprox 1.3 dB ?
instrumentation
13 Dynamic range ? ? more than 90 dB
14 Electrical noise Mentioned ? ?
Table 3.2.-Device Group.-
Environment group.
At this stage of the document it is important to note that one of the most important and
uncontrolled parameter is the propagation model. As seen previously, sea surface and sea
bottom will have a considerable impact on the propagation of the sound, especially for
shallow waters and low-medium frequencies. Moreover the medium is dissipative, especially
for high frequencies. All of this makes the transmission loss vary with frequency, receiver and
source depth and sea depth. However, the three standards simplify this problem defining a
simple equation only function of distance and constant for the whole frequency range to
make these corrections. The dependency on how each frequency propagates is directly
linked to the depth water in the measurement location, which is the reason why the
propagation law assumed by DNV is more similar to the cylindrical propagation. Besides,
other parameters affecting the transmission loss as sound celerity profile have not been
considered. Finally, the minimum distance between vessel and hydrophones from which the
vessel can be considered as a point source, remains not well empirically supported.
a. STANAG standard.
The following assertion included in the standard definition, “The depth of water must be
sufficient to ensure that the level of bottom reflection is insignificant”, does not remove the
uncertainty related to this aspect because it does not specify from what depth this assertion
holds. There is also a lack of how to proceed in terms of sea state, currents and sea bed.

WP 3 - Measurements
Method
b. ANSI/ASA standard.
The most important uncontrolled aspect in this case is the currents that can affect the noise
measurements in the hydrophones. The effect of this parameter can be limited by the use of
a specific control procedure.
General assertions, included in the standard have to be also considered: “The measurement
methods mitigate the variability caused by Lloyd’s Mirror surface image coherence effects,
but do not exclude a possible influence of bottom reflections”.
c. DNV Class Notation.
Due to the proposed bottom anchored device in this case, the currents are not considered as
problematic issue oppositely to the correction of 5 dB considered to account for bottom
reflection in all the cases.
Environment
G-A: Greater 300 m or 3xLoa
Shallow / Deep More than 30m or 0.64V^2 for
1 0.6*√(gh) G-B: Greater 150 m or 1,5xLoa
Water high speed vessels
G-C: Greater 75 m or 1xLoa
2 Mirror effect Mentioned Mentioned but not separately studied ?
100 m and above: Wind Speed less than 20
3 Sea State ? Max Beaufort 4/Sea State 3
Knots. Small vessels not specified
4 Propagation Model 20Log 20Log 18 Log
5 Celerity Profile ? ? ?
6 Currents ? ? ?
Not perfectly flat (correction -5
7 Seabed type ? Nothing-Deep water
db)
8 Background noise Defined Defined Defined
Table 3.3.-Environment Group.-
Vessel group.
In general terms for the three standards, this group can be considered as the worst defined.
The matter of fact is that the achievement of the fix image of the vessel under study is
absolutely necessary to be able to ensure a minimum level of test repeatability. Moreover,
some standards require to perform several runs then to average them, so it is really
important to define what vessel conditions must be under control. In particular, the proper
definition of all the possible underwater sources of the vessel, including the machinery on
board, as well as the accurate load conditions should be requested to define how the vessel
is working as speaker.
This is one of the major lacks regarding the requirement definition in the standards. Despite
of this fact, no supplementary uncertainties are included because of this.
The adjustment of the requirements depending on the final purpose of the measurement, the
vessel typology, vessel availability and other parameters is not considered in the standards
excepts for ANSI/ASA, for which is partially achieved through grades.

WP 3 - Measurements
Method
Vessel constraints
Through
1 Vessel Speed ? (through water or ground) Through water & Ground
Water
2 Vessel Length Mentioned Without restriction ?
Propeller &
3 Mentioned Mentioned Mentioned
Machinery
Vessel load To be
4 ? Normal Load Range
conditions described
5 Family vessel Military Underway surface vessels Acoustic/Seismic/Fishery/Research/Environment
Three grades of accuracy (A,B &
6 Availability ? ?
C)
Table 3.4.-Vessel group.-
Post Processing group.
Once all the data have been collected, different post processing strategies can deliver
different results for the same data. In this sense, it is important not only to define univocally
the way to proceed but also it is crucial to review if all the particularities of the vessel have
been really taken into account. For example, depending on the directivity of the vessel the
highest radiation may occur at different angle from around 90º beam angle, so window
symmetric around the closest point of approach does not necessarily represent the worst
condition. Such effect has not been properly addressed by the current standards.
Vessel constraints
G-A: 1/3 o.b & Narrow
Narrow &
1 Bandwidth G-B: 1/3 o.b & Narrow 1/3 o.b (Narrow Optional)
1/3 o.b
G-C: 1/3 o.b & Opt.Narrow
Determined During Test
AC (Acoustic centre)
2 Keel/ Hull Half way E.R and Propeller 0,7xR propeller
Definition/ location
Half way E.R and Propeller
Half or vessel length depending on
3 Data acquisition window ? +/- 30º
speed
Background Noise
4 ? Defined Defined
Correction
5 Directivity Defined ? ?
G-A: 3+3
6 Number of runs 2/speed G-B: 3 2/speed in opposite directions
G-C: 1
Table 3.5.-Post Processing Group.-
3.3.3. The existing uncertainties in the standards.
The first issue that has to be clarified is the concept of global uncertainty as well as
repeatability. The only standard that tackles these essential concepts is the ANSI/ASA.
Unfortunately, the poor treatment of these global parameters in the standards produces a
high unawareness about the quality of the measurement itself as well as their stability when
a new set of measurements are performed. This general approach, always complicated
because many uncertainties exist, is unavoidable and much more important than particular
uncertainties corresponding with specific parameters.

WP 3 - Measurements
Method
As it has been mentioned at the beginning of this paragraph, the aim is not to review and
assess the uncertainties of the current standards through a quantitative point of view but
following a qualitative approach to highlight the current scenario in terms of quality of the
results of the underwater noise measurements. Moreover, this analysis will permit to
continue with the ongoing work aimed at defining a new European Standard for the
Underwater Measurement Procedure.
With the aim to deal with the analysis in a proper way, some “Key Groups” have been
defined regrouping “Key Parameters” according to their common effects in the same area of
interest. Some of the parameters included in the four groups introduce uncertainties to the
procedure itself while others are only useful to define and assess the measurement
conditions, validity of the measurement, measurements methodology definition and
cost/applicability to the current shipbuilding industry. Anyway the uncompleted definition of
some criteria can add a supplementary uncertainty not foreseen.
The following paragraphs try to present in an easy way the final uncertainties affectation
present in the different parameters of the current standards under study.
Uncertainties due to undefined criteria.
Those parameters with red crosses are not sufficiently well defined and produce insufficient
uncertainty assessment. The most relevant parameters which are not well determined are
those that can invalidate the measurement.
Uncertainties of the defined criteria.
The following table tries to summarise the affectation produced by those parameters that add
uncertainty by themselves due to an inexistent or unsuitable definition in the standards.
As it can be seen in Table 3.7, many issues are adding uncertainties to the full-scale
measurement tests. Therefore it is not so realistic to deliver accurate results without dealing
with complementary analysis of the uncertainty and repeatability obtained during the
measurement campaign. Following the expertise of the authors involved in the creation of
this document, the most important challenge that has to be accomplished is the better
knowledge about the “transmission loss”. As an example of that the Figure 3.5 shows the
source level spectra obtained for the same vessel and same conditions corrected from three
different distances. As can be seen the difference between 50Hz and 200Hz are up to 10dB,
which is a clear proof of the level of uncertainty introduced by the current measurement
procedures. This is why, as it is normally done in the noise and vibration engineering field,
we have to attack firstly the major source of uncertainty.

WP 3 - Measurements
Method
Key
Description Key Group STANAG ASA DNV
parameters
Frequency range of
Frequency range
the device
Basic requirements
Frequency
Frequency resolution
resolution of the
of the device
device
Distance to the CPA Distance to the CPA
Measurement
conditions Sea depth Sea depth
Sea state Sea state
Background noise
Parameters that can
invalidate the Signal to noise ratio Hydrophone
measurement movement
Electronic noise
Parameters that Number of runs

Number of
improve the
measurements Number of
repeatability
hydrophones
Table 3.6.-Uncertainties due to undefined criteria.-
URN source level taken at different ranges

190
Presure dB (ref 1 µ Pa) @ 1 m - 1/3 Octave Band (1 Hz)
180
170
160
150
140 Hydrophone A 30m Hydrophone B 30m
130 Hydrophone A 72m Hydrophone B 72m
120
Hydrophone A 82m Hydrophone B 82m
110
10 100 1 000 10 000 100 000
Frequency HZ
Figure 3.5.- URN source level obtained at different ranges.-.-
In-depth studies regarding the uncertainties definitions for each parameter have been
developed in previous paragraphs. Specific numerical and experimental tests have been
defined to get a final approach to the whole uncertainty and repeatability of the new
developed URN procedure. In general terms, several laboratory and full-scale measurement
have been scheduled to solve, group by group, the deficiencies detected in the current
measurement standards.

WP 3 - Measurements
Method
Description Key Group Key parameters STANAG ASA DNV
Distance Hydrophone line tilt

measurement
accuracy Relative measurement
distances
Hydrophone Directivity
Accuracy of the
instrumentation
Dynamic range
Noise recording
accuracy
Electronic noise
Hydrophone movement
Currents
Mirror effect
Sound propagation
Parameters
introducing Transmission loss Sea bed reflections
uncertainties
Celerity profile
Vessel length
Vessel speed
Propeller and machinery

Vessel conditions
Load conditions
Currents
Acoustic centre
Data Window definition

Post-processing
Directivity of the vessel
Bandwidth
Table 3.7.-Parameters inducing Uncertainties.-
3.3.4. Conclusions and recommendations
The uncertainty and repeatability of the measurement procedures as a whole have been
identified as crucial to know the quality of the data obtained. Otherwise, we will not be able to
know the accuracy of the results, which is fundamental information in all experimental
activities as well as the “stability” of the data obtained along the time.

WP 3 - Measurements
Method
The only current standard that deals with this request is the ANSI/ASA [2] standard. Anyway,
according to the qualitative results shown in the previous paragraphs of this document, it
could be interesting to review in detail with ANSI/ASA team the proposed uncertainties and
repeatability values that seems to be a bit optimistic in some cases. In this way, two
assertions included in the standard ANSI/ASA [2] - §7 have to be recovered:
“The estimates (the overall measurement uncertainty & repeatability for the
measurements) given above are provided as representative values for guidance and
should not be considered to be exact.”
But at the time:
“With careful implementation of the methodology described in this standard, the

overall uncertainties described above are achievable.”
ANSI/ASA [2] is in our opinion the most relevant standard in use to carry out properly
underwater measurements and it have to be considered as the best starting point for future
improved standards.
One first conclusion is that some specific on-site tests have to be included in the procedure
to finalize the measurements with enough information to be able to guarantee the uncertainty
and repeatability levels certifying the quality of the final signature delivered.
3.4. Parameter evaluation and identification of deficiencies in

the current measurement procedures
3.4.1. Device
1 Frequency range and disturbances:
The frequency range required will depend on the purpose of the measurement and which
frequency components need to be captured. So, the device and procedure must allow to
measure within this frequency range with uncertainty and repeatability according to what is
specified. This requirement will strongly affect the configuration of the device and therefore,
the problems and phenomena causing these limitations must be understood, the root causes
identified, and it must be properly quantified.
Globally, the frequency range will be limited by the signal plus noise to noise ratio achieved
and the frequency response of the measurement device. The first one will depend on the
arrangement of the system, the sea state and the source pattern of the vessel, whereas the
second one will depend exclusively on the technical characteristics of the measurement
device.
In the low frequency one of the main problems arising is the parasitic signals due to:
• The movement of the hydrophones.
• The cable vibrations.
• The mechanical chafing between components.

WP 3 - Measurements
Method
• The change of depth (consequently changes on hydrostatic pressure) caused by the

hydrophone movement and the waves.
These phenomena produce in the hydrophones what is known as self noise. They do not
generate uncertainties on its own but mainly generates background noise in the
measurement that may mask the actual sound received from the vessel and thereby limiting
the frequency range for a given distance to her.
The Figure 3.6 shows the existence of these phenomena in a real measurement performed
with a floating configuration. In this figure the slight difference between background noise
and the measurement of the two hydrophones for a certain run highlights the problems in the
low frequency. In fact, for this particular measurement the results should be discarded below
50Hz.
Hydrophone 20m (white)
Hydrophone 40m (red)
Figure 3.6.- Comparison between background noise and real measurement.-
Figure 3.7.- Comparison of background noise level of different configurations under sea state 2.-
Moreover, in the Figure 3.7 shows the importance of the arrangement of the measurement
device. This graph shows the background noise for a given sea state of a bottom supported
configuration and of a floating configuration.
Below there is a description of the main causes generating excessive background noise in
the low frequency.

WP 3 - Measurements
Method
Movement of the hydrophones: The movement of the hydrophone will be studied in the
item 5, but as a summary, the major source for this phenomenon is the wave motion of the
sea surface and the currents existing into the sea. Decoupling the hydrophones from the sea
surface movement is a way of reducing the effect of sea state in floating configurations,
whereas performing the measurement in the absence of strong currents can prevent from
excessive self-generating noise in the hydrophone affected.
Cable vibrations: The cable vibrations are another important source of self noise in the low
frequency range for hanging measurement systems (floating configuration). These vibrations
are induced by vortices appearing behind the cables by the effect of the water flow. There
are two main types of motions to be considered: gross oscillatory motion at frequencies lower
than 5Hz controlled by the effect of the gravitational forces and hydrodynamic forces around
the buoy, and higher frequency vibratory motion of the cable controlled by the cable tension
and the hydrodynamic forces affecting the cable.
In the reference [12] a comparison between the noise records of floating hydrophones and a
bottom supported hydrophone without any cable (see Figure 3.8 left) are presented (Figure
3.8 right). The noise spectrum levels of the three hydrophones are almost the same in the
frequency range 64Hz to 1000Hz. However in the 10Hz to 63Hz frequency band the noise
spectrum level measured by the bottom supported hydrophone were 3dB to 25dB lower than
the levels measured by the floating ones.
Figure 3.8.-Averaged underwater noise spectrum level of the three hydrophones.-
Moreover, this difference is remarkable in the range between 10Hz and 25Hz. Figure 3.9
shows the recorded amplitude for the 20Hz and 100Hz components at different current
speeds to compute the current speed dependency at the mentioned noise frequency range.
This results in a neglected correlation for the non-cabled hydrophone put onto the seabed,
and a large correlation for the cabled hydrophones. Besides a larger correlation for 20Hz
noise component appears showing that the effect is more prominent for lower frequencies.
The fact that there is no dependency between noise spectrum level and current speed for the
bottom supported hydrophone may be a proof that part of the increase in the noise level for
low frequencies is due either to vortex-induced vibration on the cable or to flow induced
noise.

WP 3 - Measurements
Method
Figure 3.9.- Underwater noise level at selected frequency versus current velocity.-
Solutions to this problem already exist. Some of them are based on the disruption of the
characteristic vortex flow field. These solutions are attached along the entire length of the
cable and tend to break up the coherence of the vortices by causing a variable separation
point. The vibrations are also reduced if the boundary layer is induced to turbulent since the
pressure gradient is then reduced.
Other solutions consist on standing-off the hydrophones from the supporting ropes by flexible
materials to reduce the influence of strumming.
Mechanical chafing between components: It is important to take care of the coupling of

the different elements (hydrophones, cable,…) because mechanical contact induces
vibration which generates parasitic signals and sound. In order to avoid undesirable effects,
flexible plastic supports are used for the hydrophones to minimise mechanical contact as well
as no metal parts in the deployment to avoid metal-on-metal contact. With sufficient care this
phenomenon will not be an important source of self noise so it does not need to be
estimated.
Water surface motion: The change of depth of the hydrophones due to water surface
motion caused by waves generates hydrostatic pressure fluctuation that is picked up by the
hydrophones. The Figure 3.10 (reference [13]) shows the distance between two
hydrophones with different nominal depths to the sea surface. As can be seen, mean
variation of up to 1 metre was recorded for this case, which means a difference of 0.1 bar on
the hydrostatic pressure, generating a sound level of about 200dB in the hydrophones.
On the other side, the frequency components of this phenomenon are really low. As can be
seen in the Figure 3.11 (reference [14]) the main components are below 1Hz. Anyway this
phenomenon could limit the frequency range and force the use of high-pass filters to
eliminate this signal that may overload the hydrophone invalidating the measurement.

WP 3 - Measurements
Method
Figure 3.10.- Distance of the hydrophones to the sea surface along time.-
Figure 3.11.- Energy spectrum of the sea state.-
Cut-off frequencies: One effect in shallow water channels is that they do not allow the
propagation of low frequency signals. Because of the depth, the low frequency noise
components are trapped between the sea surface and sea bottom causing that sound waves
do not propagate below a cut-off frequency so they cannot be present in the noise field
beyond ranges equivalent to a few water depths. Figure 3.12 (reference [16]) shows the cut-
off frequency as a function of depth for a shallow water channel whose sea bottom has
certain features (one infinitively rigid, the other soft enough). One obvious effect of this
phenomenon is that it limits the minimum sea depth required and the maximum distance of
the measurement device to the target vessel depending on the lowest frequency to be
caught. In particular, we can see that it is impossible to measure signals below 10Hz in
shallow waters.
Considering a homogeneous water column of depth and sound speed overlying an

homogeneous bottom of sound speed the cut-off frequency is given by:
Factors limiting the highest frequency: As said before the highest frequency will be
determined by:

WP 3 - Measurements
Method
• The frequency sample of the measurement system.
• The hydrophone characteristics.
Figure 3.12.- The lower cut-off frequency as a function of depth for a certain shallow water
channel.-
On the one side the frequency sample must be at least more than twice the highest limit of
the frequency range required (Nyquist limit), and on the other side the hydrophones should
be suitable and must present a flat frequency response, according to the uncertainty
requirements, within this frequency range.
After what have been seen in this point, the following actions shall be taken in the definition
of the measurement procedure:
• In order to be able to measure in the low frequency range the use of bottom
supported configurations for the measurement device should be required, as long as
it is technically feasible. This is true for measurement in shallow waters. In deep
waters, deploying a measurement device with this configuration will be complicated
and extremely difficult, so the procedure should contemplate the use of floating
configurations with a system to decouple the sensors from the sea movement.
• Besides, after what has been seen with regard to the cut-off frequency phenomenon,
at least 60m water depth should be required.
• Finally, the technical characteristics of the measurement equipment as well as the
frequency sample acquisition should be defined in accordance with the frequency
range required.
2 Frequency resolution.
Depending on the aim of the measurement certain frequency accuracy will be required. For
instance, if the main aim of the measurement is the source identification, high frequency
resolution will be required for narrow band spectra analysis whereas if the aim is to obtain
the sound levels for traffic management an analysis in one-third octave bands will be
enough.
This characteristic does not introduce any uncertainty but it is a specific requirement of the
measurement. In order to achieve certain frequency accuracy the duration of the sample and
the following post-processing (window selection) must be suitable to ensure that components

WP 3 - Measurements
Method
of the spectrum whose difference in frequency is larger than the frequency resolution
required can be distinguished.
In order to have a frequency accuracy of fa the duration of the sample must be larger than:
Two effects should be taken into account:
• Doppler effect: It affects the frequency accuracy, especially for high frequency
components and/or high speed vessels. Sometimes this effect can prevent from
obtaining some spectral lines.
• Wrong distance correction due to the fact that the distance over the time sample
used to obtain the spectra is not constant, thus, uncertainties in the source energy
level appears (they are presumed to be low, however in the current standards the
distance used along the whole data window is the same so the error due to this
phenomenon is larger)
Doppler effect: The Doppler Effect is the change in frequency of a wave for an observer
with relative movement to its source. Considering that the relative speed of the source in the
medium is constant and equal to the frequency of the signal that the hydrophones will
record is going to be related to the actual frequency of the signal by:
Where is the sound speed in the sea water (approximately 1500m/s), the actual
frequency of the sound and the frequency measured by the hydrophones.
In the Table 3.8 the frequency range observed by a measurement device is calculated
according to a given vessel speed and data window angle, for different frequencies.
Data Window Angle = +/-30º

Vessel speed = 11knots Vessel speed= 22knots
Observed frequency Observed frequency
Actual frequency Actual frequency
range range
25 Hz 24.95Hz-25.05Hz 25Hz 24.9Hz-25.1Hz
50 Hz 49.9Hz-50.1Hz 50 Hz 49.8Hz-50.2Hz
100 Hz 99.8Hz-100.2Hz 100 Hz 99.6Hz-100.4Hz
500 Hz 499Hz-501Hz 500 Hz 498.1Hz-501.9Hz
Table 3.8.- Maximum source speed required to achieve a given frequency accuracy.-
As can be seen in this table the effect of vessel speed begins to be important for frequencies
higher than 100 Hz. In order to understand better this phenomenon the actual situation of the
vessel with respect to the hydrophones in the current standards are shown in the Figure
3.13 which causes the variation in distance (approximately) shown in the Figure 3.14. As
can be seen in these figures, the source speed to be taken into account is lesser than the
vessel speed and its maximum value occurs in the extremes of the Data Window Length
(DWL).

WP 3 - Measurements
Method
Figure 3.13.- Actual position of the vessel and the hydrophone line.-
Figure 3.14.- Distance between the vessel and the hydrophone line for a given run.-
Figure 3.15.- Illustration of the Doppler effect before CPA.-
Therefore the relative velocity of the vessel regarding the hydrophone is not constant, and it
changes from negative to positive when the acoustic centre of the vessel is in the closest

WP 3 - Measurements
Method
point of approach (CPA). To illustrate this effect an example is shown in the Figures 3.15
and 3.16.
The Figure 3.15 shows the source spectrum around a discrete component for a given
vessel. The sample used to obtain it corresponds with the two seconds period around the
CPA. The Figure 3.16 shows the source spectrum obtained using the sample corresponding
to the two seconds period starting five seconds before the vessel reaches the CPA. A
frequency shift for this discrete component of the spectrum is observed.
Figure 3.16.- Illustration of the Doppler effect after CPA.-
The Figure 3.17 (reference [19]) shows a very clear spectral line modulated due to the
Doppler effect.
Figure 3.17.- Spectral line modulated due to Doppler effect.-
Because of the actual speed profile of the vessel, what really happens is that when the
averaging is done along the DWL, and considering only the effect in a single tune, an energy
spread of this tune is produced. This distortion will be more important as frequency, vessel
speed and data window angle increases as Table 3.8 shows. This may be important

WP 3 - Measurements
Method
because the spectral line could be hidden in the continuous spectrum if the power of this line
is spread over a broad frequency integration band.
Therefore the Doppler Effect shall be taken into account. The simplest way of avoiding its
impact on the measurement is to define a variable frequency resolution with the frequency:
the higher the frequency, the wider the frequency resolution.
Wrong distance correction: Other effect that must be studied is the variation in range
throughout the passage of the vessel through the data window used to compute the spectra
which is related to the frequency accuracy. The influence of that in the accuracy of the
measurement is presumed to be low, but in any case, it is tried to be quantified below.
The Table 3.9 shows the maximum error made with different range variations (in terms of
percentage) assuming a spherical transmission loss law ( ).
Maximum Range variation Maximum error in the source level

1% 0.08dB
5% 0.42dB
10% 0.82dB
15% 1.21dB
Table 3.9.- Maximum error expected in the source level due to range variation during the
measurement window.-
On the other hand, the range variation during each single sample will depend on the distance
to the closest point of approach (CPA), the duration of the time window used as well as the
speed of the vessel during the run (Figure 3.18). The Table 3.9 shows some values of this
mathematical relationship in order to see the importance of this effect.
d2 2s data window
Taking this distances is d1
proposed. dcpa ANSI/ASA considers this same
distance along the DW
Hydrophones
Figure 3.18.- Relationship between the distance to the closest point of approach and the distance
at the end of the measurement window.-
As can be seen in the Table 3.10 small range variations are expected, therefore there should
be small errors in the source level due to this effect. However special care should be taken
when the distance to the closest point of approach is short or when the vessel speed is high.
Anyway the uncertainties introduced by this phenomenon are expected to be really low so no
countermeasures shall be taken.

WP 3 - Measurements
Method
Finally, current standards only use the distance to the CPA along the whole data window
which is composed by a larger data window than the length of the samples used to perform a
narrow band analysis as shown in Figure 3.18 (even though different distance could be used
among the different samples used to compute the spectra). This phenomenon is similar to
the one previously explained but since the time window could be much larger the error is
higher and could be relevant (see Data Acquisition Window definition).
Time window = 2s (0.5Hz)

Speed = 11knots Speed = 22knots
Distance to CPA Range variation Distance to CPA Range variation
300m 0.01% 300m 0.06%
100m 0.1% 100m 0.6%
50m 0.6% 50m 2.4%
Time window =8s (0.125Hz)
Speed = 11knots Speed = 22knots
Distance to CPA Range variation Distance to CPA Range variation
300m 0.2% 300m 1.11%
100m 2.5% 100m 9.7%
50m 9.7% 50m 34.8%
Table 3.10.- Range variation for different cases.-
On the other hand, no special countermeasures shall be taken to avoid the fact that the
same distance correction must be used for the smallest data window required to compute the
spectrum for a given frequency resolution as the error because of this is presumed to be
negligible.
3 Relative distance measurement equipment.
A correct measurement of distance between the vessel and the hydrophones should be
achieved in order to properly estimate the propagation loss transfer function. Two important
sources of uncertainties associated to this parameter can be distinguished when a GPS or
similar system is used:
• The uncertainty due to the measurement of the distance between the location of the
GPS receiver and the actual location of the hydrophone/s in the sea.
• The uncertainty due to the fact that despite of being able to idealize the vessel as a
point source and to estimate its location along the length of the vessel, it will be
probably highly difficult to estimate its location over the breadth of the vessel. In the
current measurement procedure standards the side of the hull is the reference taken
from which the distance to the hydrophones is measured, which is sometimes
ambiguous since the hull of the vessel is not straight.
The uncertainties associated to the first item, when a GPS system is used, are made up of
two factors:

WP 3 - Measurements
Method
• The uncertainty in the distance between the buoy and the vessel supply where the
GPS antenna receiver is located (Figure 3.19). This will depend on the accuracy of
the GPS used. Commonly, a Differential GPS achieves an accuracy higher than 3m
for most of European waters, whereas simple GPS systems achieves an accuracy
about 10m.
• The lack of awareness of the actual position of the hydrophones relative to the buoy.
Currently, it is generally assumed that hydrophones are placed vertically in relation to
the buoy. Some measurement standards claims that the tilt angle of the hanging line
does not need to be taken into account if it does not exceed 5º, otherwise it should be
considered but the standard does not specify how. Anyway, this phenomenon will be
studied in the next item.
Figure 3.19.- Diagram of the support vessel and the surface buoy.-
For the second item, the further the vessel passes from the hydrophone, the less important
this error will be. In any case, this source of error is considered to be low compared to other
sources of error.
The tilt angle and the accuracy of the GPS system should be defined according to the
required accuracy for the measurement procedure and the value of other sources of
uncertainties. In order to do this, numerical studies to quantify the uncertainty of the source
level obtained for a given tilt angle have been performed in this task. Moreover, the use of
other methods to measure the distance which do not suffer from the drawbacks of the GPS
system, specially the tilt angle, should be analysed.
4 Hydrophone line tilt.
Due to the sea currents the hydrophone line is not totally straight but a drift angle appears
which mainly affects the measurement of the distance between the target vessel and the
hydrophones (it also may have undesirable effects on other parameters).

WP 3 - Measurements
Method
Figure 3.20.- Measurement layout.-
The uncertainties caused by this phenomenon have been already studied in reference [13]
for the layout shown in Figure 3.20. In order to estimate the hydrophone line tilt during the
measurement, the distance between the hydrophones, the sea surface and the sea bottom
were measured. The results are shown in Figure 3.21. It can be seen that there is a medium
deviation from the nominal depth of the hydrophones (see Table 3.11) caused by the
hydrophone line tilt. The small variations in the graphs of the Figure 3.22 are due to waves in
the sea surface and the drifting of the measurement device.
According to these values the graph of the Figure 3.22 shows the maximum error (up to
3dB) what may be made depending on the distance to the closest point of approach. As it
can be seen the error decreases as the distance increases.
Figure 3.21.-Distance of the hydrophones to the sea surface and sea bottom.-

WP 3 - Measurements
Method
Nominal (m) Derived from data (m)
Vertical position Average Standard deviation
H1 40 34.4 0.56
H2 60 52.36 0.55
Horizontal position Average Standard deviation
H1 0 20.38 1
H2 0 29.27 1
Table 3.11.- Comparison between actual and theoretical position (reference [13]).-
However this error may be higher if the actual transmission loss is taken into account instead
of the one obtained from assuming a spherical transmission. As an example, the Figure 3.25
shows a computation of the transmission loss for a given environment. It can be seen that a
difference of a few meters can determine if the hydrophone is in a shadow zone or not which
could imply differences up to 5 dB just due to this phenomenon.
There are three ways to reduce this uncertainty:
• Increasing the distance to the CPA as much as possible. Other undesirable effects
may appear, such as low signal-to-noise ratio.
• Measuring the angle tilt during the measurement. For this case, the uncertainties will
be reduced to the ones caused by the variation of this angle (given by the standard
deviation in table 3.11) which will be much smaller than the previous one. However it
could be not sufficient if more than one hydrophone is deployed because the shape of
the line into the sea is not known.
• Using another distance measurement system obtaining directly the distance between
the source and the receiver, such as methods based on the Doppler Effect.
Due to the importance of this factor its influence in the noise source level measured has
been internally studied in this task.
Figure 3.22.- Error in dB versus distance to CPA (reference [13]).-

WP 3 - Measurements
Method
Figure 3.23.- Transmission loss at 500Hz numerically computed.-
5 Hydrophone movement.
One of the main problems in some of the current devices in use is the movement of the
hydrophone in the sea. In fact the American National Standard (reference [2]) mentions that
special measures should be taken to isolate the hydrophone from the movement of the
waves in order to avoid self generating noise in the receivers.
As it can be seen in the Figure 3.11 the spectrum content of the wave motion are
concentrated in the very low frequency, however mechanisms are present in the system for
the generation of high frequency accelerations given only forcing a much lower frequency.
Three types of mechanisms that generate this type of noise can be distinguished:
• Noise generated by snap loadings.
• Hydrophone acceleration induced noise.
• Flow induced noise.
Snap loads occur when one or more segments of the cable go slack and then retension due
to wave induced motions. This phenomenon generates an impulsive load with broadband
frequency content.
One of the possible solutions is to increase the sinker weight to add extra tensional level to
the cable to avoid that goes slack. The Figure 3.25 shows the acceleration time signal of the
subsurface buoy of the Figure 3.24. The Figure 3.26 shows the band pass filtered
subsurface float acceleration where the phenomenon described above can be seen. Finally
the Figure 3.27 shows the time signal recorded by one of the hydrophone of the system of
the Figure 3.24, which has very large spikes corresponding to the occurrence of snap loads.

WP 3 - Measurements
Method
Figure 3.24.-Measurement configuration with a subsurface float.-
However according to reference [18] this source of self generated sound is much lower than
the other two mentioned above.
Most of hydrophones are sensitive to accelerations, so they record relevant signals when
they have acceleration. Figure 3.28 (reference [18]) shows an acceleration equivalent noise
level diagram computed using the acceleration recorded by an accelerometer and the known
acceleration response of the hydrophones.
Figure 3.25.- Acceleration time signal of the subsurface float.-
Figure 3.26.- Bandpass filtered subsurface float acceleration.-

WP 3 - Measurements
Method
Figure 3.27.-Time signal recorded by one hydrophone.-
On the other side the noise level measured by the same hydrophone is shown in the Figure
3.29. The total noise of the hydrophone is much higher than the component due to
hydrophone acceleration induced noise, which means that the flow induced noise is the
major source of self generating noise caused by the movement of the hydrophones.
However, the levels for low frequencies and high vertical speed of the acceleration induced
noise are high enough to mask the noise that we want to measure so it must also be
removed.
Finally, flow noise results for the pressure fluctuation generated by the flow around the
hydrophone which excites vibrations in it. Taking into account that the noise level expected
for these frequencies during the measurement is 85dB, the fact that for a vertical velocity of
0m/s a 90dB noise level was recorded suggests that flow around the hydrophones due to
sea currents must be considered as well.
All these phenomena generate background noise that can mask the actual noise coming
from the source (the vessel) making the measurement impossible for low frequencies.
Figure 3.28.- Acceleration equivalent noise level.-
As seen in the item Frequency range, the movement of the hydrophones shall be minimized
so the bottom supported configuration is recommended. Moreover, the sea state strongly
affects this item so the measurement procedure will define a minimum acceptable level in
terms of sea state.

WP 3 - Measurements
Method
Figure 3.29.- Diagram of the noise level recorded by one hydrophone.-
6 Distance of the hydrophones to the Closest Point of Approach (CPA)
The distance of the hydrophones to the CPA is a really important key parameter, what,
despite of not introducing uncertainties by itself, will strongly influence factors introducing
uncertainties, such as the frequency range where the signal is valid, the transmission loss
factor, the importance of the background noise compared to the vessel noise, the duration of
the data window, etc. This parameter will be influenced by:
• The signal-to-noise ratio expected, in particular, the level of noise emitted by the
vessel.
• Difficulty of the computation of the transmission loss. If a transmission loss of a point

source is considered, this distance shall be large enough so the assumption of the
vessel behaving as a point source can be realistic.
On the one hand, the maximum distance will be limited by the expected signal-to-noise ratio
as the further the CPA is the lower the signal-to-noise ratio is expected. Since the
background noise must be 10dB lower than the vessel noise to be able to ignore it, the
distance to the CPA cannot be larger than a value what depending on the kind of vessel, the
transmission loss inherent to the medium and the background noise present in the area,
which is mainly determined by the sea state.
On the other hand, the acceptable minimum distance will be limited by the hypothesis about
the vessel as a source, which means that if the vessel is considered as a point source the
measurements must be taken far enough from the vessel. This consideration is assumed in
all the current standards because it eliminates from the physic model the complications
appearing in the near field where the model of the source as a monopole or dipole point
source is no longer valid.
There is no clear agreement about the distance where the assumption of far field
measurements applies. It is considered that it is necessary to measure at ranges larger than
one ship length to reach this condition however, it may be influenced by the depth of the
location. If this condition is not complied, there will be bigger uncertainties due to the
disagreement with either the physic model (in case a theoretical propagation law is

WP 3 - Measurements
Method
considered) or the experimental model (in case a calibrated point source is used to estimate
the propagation loss).
Even though the distance will be mainly defined according to these two factors, the larger the
distance is the better to reduce the undesirable effects on the measurements caused by the
hydrophone line tilt and the Doppler Effect. Besides, larger distances allow longer data
windows for a given Data Window Angle. However, very large distance also complicates the
computation of an accurate transmission loss.
If the background noise is low enough along the frequency range, then the distance can be
higher. That is why not only the sea state must be appropriate during the measurement
(environment) but also the self generated noise of the measurement chain should be
reduced as much as possible. The sources of background and self generated noise are
studied in other items within this paragraph.
Current standards already defined this requirement. The reference [2] and [3] require a
distance larger than 100m or once the vessel length, whichever is greater for the maximum
grade of accuracy and the reference [9] requires a distance from 150m to 250m.
On the other hand, this requirement may imply that, especially for silent vessels, the distance
to the vessel is too large since the noise coming from it may be the same order as the
background noise, which makes the measurement not valid. For example, the Figure 3.30
shows the spectrum measured for a fishing research vessel where the distance to the closest
point of approach was about 42m. It can be seen that the levels are 10 dB higher than the
background noise for frequencies between 40 and 75 Hz and less than 10dB higher between
30 and 45Hz. If the distance requirement had been in accordance to what is specified in
current standards the level measured would have been 7.5dB lower in the best case
(considering a transmission loss factor equal to which would have made the
measurement not valid due to the excessive background noise. That is why the vessel type
should be taken into account somehow in the definition of this requirement.
Other effects may appear if the distance to the CPA is too short, such as large spurious
signal recorded by the hydrophones caused by the wave pattern produced by the testing
vessel or interaction between the vessel and her surroundings.
Figure 3.30.- Noise spectrum measured by two hydrophones at a given distance.-

WP 3 - Measurements
Method
Therefore, a minimum distance between the hydrophones and the CPA should be defined
taking into account the constraints regarding the assumption of the vessel as a point source.
Due to the importance of this parameter, the suitability of this requirement has also been
experimentally verified.
7 Number of hydrophones
The references [2] and [3], for grades of accuracy A and B, deploy more than one
hydrophone at different depths and then take the power average of their outputs to obtain the
final measurement of one run. The purpose of this is to average as much as possible the
effects caused by the Lloyd Mirror effect as well as averaging the error in the transmission
loss.
Besides the hydrophones shall be positioned vertically in the water column at depths which
results from nominal 15º, 30º and 45º from the sea surface at a distance equal to the nominal
distance at CPA. However, it is very difficult to obtain an actual distance to the CPA close to
the nominal one. Therefore, the hydrophones would be positioned at different angles from
the ones specified in the standard, so a more realistic target should be provided, in terms of
tolerance for this angle, in case more than one hydrophone are used, or different approaches
to take the Lloyd Mirror effect into account should be proposed.
Other possible advantage of deploying more than one hydrophone is that the transmission
loss can be estimated from the measurements taken by different hydrophones if they are
positioned at different ranges although the deployment of this configuration is quite more
difficult. This study can also be done fixing different nominal distance to the CPA for each run
performed. Moreover the repeatability of the measurement increases as the number of
hydrophones increases.
Solution: The grades with the most demanding requirements with regard to accuracy should
require the use of several hydrophones, whereas B-application whose accuracy
requirements are less demanding, will only require the use of one hydrophone, as deploying
more than one hydrophone is costly.
8 Cost, Handling & Independence
The cost associated to the measurement will strongly influence the scope of it. On the one
hand it must be in accordance with the overall cost of the ship building for new vessels, on
the other hand, the performance of the test should not imply an excessive time where the
vessel is out of service. That means for example that neither the cost of the test can be
excessive for small vessels (like fishing vessels) nor it can imply that a large merchant vessel
must spend too many days of navigation to go to a proper location to perform the test
according to the standard.
The overall cost can be split into the next items:
• Cost associated to the device: This item only includes the cost associated to the
instrumentation (hydrophones, amplifiers, recorders, the distance measurement
system, etc). It will strongly be affected by the requirements specified for it. The
specifications given in the standards [2] and [3] about the instrumentation are not
detailed and depend on the grade of accuracy. The main difference is the number of
hydrophones, the frequency range, the bandwidth required for them and the distance
measurement system accuracy as can be seen in Table 3.12.

WP 3 - Measurements
Method
ANSI/ASA Grade A Grade B Grade C
Number of hydrophones 3 3 1
Sensitivity As needed As needed As needed
Dynamic range As needed As needed As needed
Frequency range 10Hz to 50kHz 20Hz to 25kHz 50Hz to 10kHz
Bandwidth Narrow band Narrow band One-third octave
Distance ranging uncertainty 2% 2% 5%

Table 3.12.-Requirements for the instrumentation in the standards [2] and [3].-
The requirements specified in the standard of DNV [9] are given in Table 3.13.
DNV
Number of hydrophones 1
Sensitivity As needed
Dynamic range 90dB
Frequency range Depend on the operational group
Bandwidth One-third octave bands
Distance ranging error +/-5m

Table 3.13.-Requirements for the instrumentation in the standard [9].-
It can be seen that the cost of the instrumentation required in the standard of the
reference [9] will be similar to the one associated to the grade C of accuracy of the
references [2] and [3], whereas the cost of grade A and B will be higher, mainly
because of the number of hydrophones and the bandwidth required.
Apart from the factors mentioned in the previous tables there are other factors
affecting the cost of the device, which are:
• Necessary instrumentation to measure environmental variables, such as

temperature, hydrostatic pressure, sound speed profile, or other kind of
parameters such as orientation of the hydrophones. Within this category is
also included all the measurements to be performed to evaluate the
transmission loss (none of these variables are required to be measured in
the current standards).
• Requirements for the autonomy of the measurement chain which will

depend on the time required to perform the test (number of runs, data
window period).
• Cost associated to the deployment of the measurement chain: There are

different arrangements to deploy the hydrophones. The Figure 3.31 (references [2]
and [9]) shows the main configurations currently in use.

WP 3 - Measurements
Method
Figure 3.31.- Different arrangement to deploy the hydrophone.-
The diagram on the left shows a hanging line which is the easiest way to deploy the
hydrophones although it suffers from a severe movement of the hydrophones. The
middle diagram shows a bottom anchored deployment which insulates the
hydrophones from the sea surface movement although it does not eliminate all the
problems in the low frequency such as the effects of the currents or cable strumming.
It is also more expensive and requires the recorder to be into the sea. Finally the
figure on the right shows a bottom mounted deployment which requires ways to
recover the hydrophones and suffer from other disadvantages as the strong influence
of the reflection coming from the seabed, which must be taken into account. Besides
it will probably require divers to put the device onto the sea bottom and ways to
measure and visualize where it is exactly.
Due to all the reasons mentioned above the bottom mounted option implies a higher
cost of deployment than if the hydrophones are deployed in a hanging line.
In general the deployment cost will be according to the requirements in terms of

accuracy of the position of the hydrophones, the necessity of external help (like the
use of divers, cranes, supply vessel, etc.) which will be accordingly to the handling
and independence of the device and the complexity of the measurement chain
(number of hydrophones, etc.)
• Cost associated to auxiliary measurements in the vessel: In order to assure a

proper comparison among different measurements it is necessary to know the vessel
conditions, speed, etc. In this item the cost associated to these auxiliary
measurements is included and it will be equal to achieve a good accuracy no matter
what procedure and device.
• Personnel cost: This item includes the cost associated to the personnel directly
involved in the completion of the test. No big differences are expected.
• Cost of diversion of the ship: This item includes the cost associated to the
navigation of the vessel to the proper location, the vessel crew, the cost necessary to
navigate and perform the test and, when it is applicable, the cost associated to the
fact that the vessel is out of service during the test.

WP 3 - Measurements
Method
Figure 3.32.-Depth map in the Mediterranean sea.-
This cost is going to be determined by the kind of vessel, the time spent to perform
the test, and by how far the closest proper location is from the shore. The last point is
the most important and it will depend on the depth required according to the
procedure. The Figure 3.32 and 3.33 show depth maps of the sea around Europe.
Figure 3.33.- Depth map in the North sea.-
The references [2] and [3] require for grade A of accuracy a minimum water depth of 300m
or three times the overall length of the vessel (for some vessel this requirement implies more
than 600m depth). As can be seen in the maps the German, the Adriatic and Baltic country
shores do not comply with this requirement, which enforce to navigate long distances.
11 Hydrophone directivity.
The directivity of the hydrophone influences the uncertainties associated to the

measurements. If the directional characteristics are not accounted in the final data post-
processing it should show no more than a certain variation in the whole frequency range of
the measurement to assure a certain level of uncertainties. The Figure 3.34 shows a

WP 3 - Measurements
Method
directivity diagram for different frequencies of a hydrophone model which in principle is omni-
directional:
Figure 3.34.-Directivity characteristics of a hydrophone.-
If directional hydrophones are going to be used, specific data post processing to account the
directivity should be defined as well as the necessary measurements to know the actual
position and orientation of the hydrophones.
12 Accuracy of the instrumentation
The frequency response is the output-to-input ratio of the hydrophone. As it was said in the
item dealing with frequency range, one of the factors determining the frequency range of the
measurement chain is the range where the frequency response is flat, which means the
range where the hydrophone outputs the same amount of voltage per amount of acoustic
excitation regardless the frequency. The Figure 3.35 shows an example of frequency
response of one hydrophone.
The frequency range of the hydrophone will be determined according to the acceptable
deviation from the flat response which is directly related to the acceptable uncertainty for the
isolated hydrophone.
Finally this parameter varies with temperature and pressure even though the uncertainties
associated to this effect is much lower than others described in this paragraph.
Figure 3.35.-Frequency response of the hydrophone type 8103 of B&K.-

WP 3 - Measurements
Method
Figure 3.36.-Hydrophone sensitivity for different hydrostatic pressure.-
13 Dynamic range
It is the ratio between the largest and smallest components that the measurement chain can
measure. In underwater acoustics, the smallest signals that the measurement device is
able to measure are not only limited by the dynamic range of the measurement chain but
also by the background noise.
The references [2] and [3] do not specify the dynamic range for the measurement chain,
unlike the reference [9]. Even though the background noise will be the most limiting factor
determining the ratio between the biggest and smallest signal that can be measured, this
may not be true for all kind of vessels (specially the noisy ones) and, consequently a
suitable dynamic range may have to be specified in future standards.
14 Electronic noise
Electronic noise is a random fluctuation in an electrical signal, characteristic of all circuits. It

may affect the signal-to-noise ratio if the electronic noise of the electronic components of the
measurement chain is high, however it is presumed low compared to the background noise
of the measurement location, except for very high frequencies.
Therefore the procedure shall specify a minimum level of performance of the instrumentation
with regard to the points of frequency response, dynamic range and electronic noise.
3.4.2. Environment
The effect of environment on underwater detection has been addressed for a long time,
mainly for the assessment of the detection range of sonar systems in the military domain. A
very abundant literature exists and it is of course not possible to present here in detail all the
phenomena and implications. The reader can find more information in some references such
as [15], [22] and [23]. However we should note that here we are more interested in short
propagation ranges, because URN measurements are carried out at maximum distances of a
few hundreds of meters, as military sonar detection ranges can achieve several tens of
kilometres.

WP 3 - Measurements
Method
The main issues regarding the effect of environment on sound propagation are:
• Celerity profile and reflection of acoustic waves on sea surface and sea bottom,
which affect the propagation loss between the acoustic source and the detection
devices.
• Ambient noise level, which affects the quality of the measurement on the signal to
noise ratio point of view.
• Other environmental effects such as currents, wind, etc, which will disturb the
deployment of equipment and produce self noise in the hydrophones, so introducing
inaccuracy on the actual distance between the acoustic source and the detection
devices.
1 Shallow/deep water.
Water depth influences URN in different ways:
• Hydrodynamics: shallow waters can alter hydrodynamic behaviour of the propeller,

and possibly the hull, thus affecting underwater radiated noise;
• Reflection on sea bottom: refer to item 6
• Sound propagation: refer to item 3.
Relation with ship speed: Referring to STANAG 1136 [1], the hydrodynamic effect is
related to an interaction between the waves produced by the moving vessel and the water
layer of height h. A warning speed vw, not to be exceeded during the trial, is defined as:
, where vc is the critical speed defined as:
For the ISO standard part I [3], shallow effects are not considered, because the
measurement is assumed to be done in deep waters. The minimum water depth, for Grade
C, which is the less accurate, is 75 m or greater than ship length. Ship speeds up to 50 kts
are considered, although it may raise some practical or safety issues.
In the DNV document, reference [9], it is required the minimum water depth to be greater
than 30 m and 0.64v2.
For comparison, the criterion from the STANAG gives approximately 0.3v2. In both cases, a
squared speed law is used, but the criterion from the DNV document is more stringent.
Solution: The depth required by the procedure should also account for this phenomenon.
Knowledge of sea bottom profile: If the measurement is not performed in deep waters, a
good knowledge of sea bottom profile is necessary. This information can be taken from
cartography, if detailed information is available, or using an adequate echo sounder survey. It
is recommended to include a map of the area in the measurement report. In addition, care
must be taken to sea level variations which can vary significantly along time due to tides.

WP 3 - Measurements
Method
2 Mirror effect.
Because of the high acoustic impedance contrast between air and water, sea surface
behaves like a pressure-release boundary condition, on the acoustical point of view. For the
radiation of a sound point source, the problem can be modelled by introducing a second
fictive source, opposite in phase, symmetrically above sea surface (Figure 3.37).
Assuming sea surface to be a perfect planar surface, the acoustic pressure measured at the
receiver is the algebraic sum of the waves coming from the sound source and from its image:
Where is source strength, the acoustic wavenumber and the speed of sound.
For a pressure-release surface, , then the system behaves as an acoustic dipole
with separation, leading to acoustic interferences between direct and reflected paths,
producing alternatively constructive and destructive interferences at the measuring point.
This is known as the “Lloyds mirror effect” (reference [15], page 128).
The Lloyd mirror effect is expected to depend on:

• frequency
• immersion depth of the sound source
• distance and immersion depth of the receiver(s)
• size of the sound source
• sea state (possible variation of sea surface shape due to waves, affecting refection
coefficient)
• avenging effects
o narrow band / wide band
o time average (corresponding to measurement angle with respects to CPA)
Figure 3.37.- Lloyd Mirror effect.-

WP 3 - Measurements
Method
-30.00
-35.00
Pressure level (dB) -40.00
Direct
-45.00 Total (narrow band)
Total (1/3 octave)
-50.00
-55.00
-60.00
10 100 1000 10000
Frequency (Hz)
Figure 3.38.- Mirror effect on the radiation of a point source.-
In the limiting case of perfect boundary, the interference effect is illustrated in Figure 3.38,
for a noise source immersed at 5 m and a measurement at 100 m distance. When noise level
is averaged over 1/3 octave frequency bands, this interference pattern is less visible, except
at low frequencies, and a deviation of about 2 dB at high frequencies.
Two methods can be considered to separate the reflected path to the direct path:
• time separation, because the reflected signal comes later than the direct path,
• use of a directive sensor.
Unfortunately, these methods cannot be used easily in practice. Time separation is not
possible because all methods used in practice require to average signal over time (typically
several seconds). Exploitation of directivity would require a large vertical array with
numerous hydrophones, which increases cost, requires complex post-processing and in any
case the main lobe aperture is not small enough to separate sources at low frequencies.
Vertical arrays are sometimes used for military applications for the measurement of very
silent vessel such as submarines running at low speeds, in order to increase measurement
signal to noise ratio.
Coming back to the Lloyd mirror effect, the actual effect in realistic conditions is not as acute
except from the low frequency. Different mitigation effects are:
• Sound radiated from a ship does not come from a single noise source, but from several
noise sources distributed on the hull, or from the propeller. As an example, the Figure
3.39 shows the difference between considering a point source at 2m depth and a
distributed source whose mean depth is 2m and standard deviation of 1m (reference
[24]). It can be seen that the effect of the sea surface for the two hypotheses is the same
up to the first peak, then the distributed source shows an increasing attenuation of the
higher frequencies peaks and nulls.
• sea surface is not perfectly smooth, due to the waves (depending on weather conditions)
so that the level of the coherent wave scattered by sea surface is significantly lower than
the one produced by a perfect mirror;

WP 3 - Measurements
Method
• frequency average over 1/3 octave bands;
• time averaging during measurement.
In order to decrease the measurement deviation due to Lloyds mirror effect, some
procedures or standards require an average on several measurement points. This is the
case in particular for the Grades A and B of ISO standard, where the measurement is
averaged on three points on a vertical line (Figure 3.40).
As the wave interference pattern will not be the same for the three points, the average will be
less sensitive. To assess the efficiency of this averaging process, a simulation model has
been built. Figure 3.41 shows the comparison of the real URN 1/3 octave-averaged spectra
with the simulated measured level with only one hydrophone (Grade C) and three
hydrophones.
Figure 3.39.- Lloyds Mirror Pattern Interference of a sound point source and a distributed
source.-
As expected, the grades A and B reduce the Lloyd Mirror effect compared to grade C.
To check the influence of the sea surface, the previous calculation is redone with a different
value of reflection coefficient R. In this example it is assumed that R remains close to one at
low frequencies, up to 100 Hz, then diminishes linearly down to 0.2 at 1500 Hz and constant
at higher frequencies. This approach is justified by the fact that for a given wave height, there
will be some destructive interference of the acoustic waves reflecting on sea surface,
reducing the effective reflection coefficient. In that case, we see that the deviation introduced
by the comparison to free-field radiation occurs only for low frequencies (lower than 300 Hz
in that example).
Figure 3.40.- Measurement points in ANSI-ASA standard, Grades A and B.-

WP 3 - Measurements
Method
Figure 3.41.- Lloyd Mirror effect with 1/3 octave band average.-
Figure 3.42.- Mirror effect simulation with an effective surface reflection coefficient.-
It is also important to take into account that the current standards provide results as the
Dipole Source Level (termed “affected” source level by ANSI/ASA 12.64 reference [2])
whereas if the Lloyd Mirror effect is accounted in the transmission loss the results obtained
correspond to a monopole source. The conversion between a dipole and monopole source
level (considering a perfect reflexion surface and a point source) is given by Ainslie
(reference [25]) as:
Where is the wave number, the depth of the source and θ is the depression angle
relative to the surface. The correction to obtain the ANSI affected source level dipole from
the monopole source level, including the averaging from the three depression angle is plotted
against frequency in Figure 2.45 (reference [17]).

WP 3 - Measurements
Method
Figure 3.43.- The correction applied to the monopole Source Level at 4m depth to convert to the
ANSI/ASA S12.64 affected source level averaged over depression angles of 15º, 30º and 45º.-
3 Sound propagation into water.
General propagation rules: Basic propagation rules in a fluid medium are plane waves,
spherical waves and cylindrical waves. Plane waves can be used locally or to describe wave
propagation in pipes, but is not relevant for outdoor propagation. Figure 3.44 presents the
geometry corresponding to spherical and cylindrical propagation.
As a start, we consider that the fluid medium is not dissipative. In that case, for spherical
propagation, acoustic pressure decays as the inverse or the radial distance r, so the level N
at observation point M as :
Figure 3.44.- Coordinates for spherical and cylindrical propagation.-
For cylindrical propagation, if the height H of the water layer is small by comparison to the
distance R, the pressure field decays as R-1/2, then , instead of R-1:
Influence of sound absorption: In addition the previous loss factors, which are only due to
geometrical spreading of waves, some additional loss can be due to dissipative effects
related to the presence of particles or chemical components in sea water. Many others have
dealt this issue, one reference model for deep waters being Thorp’s model:

WP 3 - Measurements
Method
Different models are compared on Figure 3.45. The orders of magnitude of these losses at
200 meters from acoustic sources are:
• 0.01 dB/km at 100 Hz
• 0.1 dB/km at 1 kHz
• 1 dB/km at 10 kHz
• 30 dB/km at 100 kHz.
Figure 3.45.-Sound attenuation in sea water.-
As a consequence, if we consider a ship URN measurement distance set at 200 meters,

propagation loss will be negligible up to 10 kHz (lower than 0.2 dB), but on the other hand it
can exceed 5 dB at 100 kHz. It should be noted that this effect is not clearly addressed in the
existing standards, leading to an underestimate of URN level.
Propagation loss in real environment: In real environment, propagation loss will depend
also on celerity profile (discussed in item 4), and reflection on boundaries (bottom and sea
surface). As a consequence, propagation loss generally does not match exactly either
spherical or cylindrical law.
In deep waters, propagation field for short distances is approximately spherical. This can be
seen on the example presented on Figure A-9. However, it can be seen that for this
example, propagation loss at 1 km distance is approximately 57 dB, instead of 60 dB for
spherical decay.
In shallow waters, propagation tends to be cylindrical at sufficiently large distance R. This

pattern is clearly visible on Figure A-11.In this example, propagation loss at 1 km distance is
approximately 54 dB, to be compared to 60 dB for spherical propagation and 30 dB for pure
cylindrical propagation. As a consequence, for shallow waters, a possible approach is to
consider spherical propagation up to a short distance r0, then cylindrical propagation for r > r0
[7].

WP 3 - Measurements
Method
Then the expression of acoustic pressure level at distance r is:
For the STANAG and ISO standards, the distance correction factor is assumed to be
spherical, i.e. . For the DNV standard, the correction factor is , justified by
the fact that measurements in that case occur in shallow waters, at a typical distance of 200
meters.
Figure 3.46.- Shallow waters short range propagation experiment (d=60m).-
Relatively high errors or deviations can occur if the distance correction factor law is not
appropriate. For example, if the actual correction factor is instead of , the
URN will be underestimated by ,
Errors can be even more important in very shallow waters, because propagation is complex
and subject to important environmental variations. The study reference [11] gives such an
example. Figure 3.48 shows that the correction factor with respects to spherical propagation
at 60 meters can exceed 5 dB. In addition, it is important to note that it depends on
frequency, by about +/-2dB.
There are several models, based on different physical principles, to compute the
transmission loss with distance. Some standard models are listed below:
• BELLHOP: ray tracing based model.
• Kraken: normal model propagation code.
• KrakenC: same as Kraken but allowing complex value data.
• RAMGeo: parabolic equation model that uses a split-step Padé algorithm.
• RAMsGeo: same RAMGeo but incorporating shear.
• Scooter: wave number integration model.

WP 3 - Measurements
Method
• OASES: wavenumber integration model.
Figure 3.47.- Propagation loss uncertainty due to the unawareness of the parameters of the
model in shallow waters (yellow) and deep waters (red).-
All these models need the introduction of a number of parameters to model the environment.
Most of the time the value of these parameters are not known and they are estimated or
taking from bibliography sources. Within the internal studies, a Monte Carlo Analysis was
performed to evaluate the uncertainties associated to the unawareness of the parameters of
the model (celerity profile, hydrophone depth, source depth,…), showing difference up to +/-
5dB (Figure 3.47).
4 Celerity profile.
The value of sea water celerity (i.e. speed of sound) depends significantly upon temperature,
depth, and salinity. An approximate formula is:
Seawater celerity profile with respects to depth, which can be routinely measures using
appropriate equipment, influences greatly sound propagation. The effect is particularly visible
at large distance, the reason for what it is a key issue for determination of sonar system
performance. Because of the variation of celerity, acoustic rays do not propagate along
straight lines, but present some curvature, generally dependent on the gradient of celerity.
Furthermore, celerity profile varies generally, with the period of the year, and can even vary
during daytime in shallow waters environment. Figure A-6 in Annex A gives examples of
celerity profiles. From the examples in Annex A, we can note that:
• in deep waters in summer, the acoustic rays tend to plunge to sea bottom,
• in other cases, interferences occur due to reflections on sea bottom and sea surface.
A special case occurs when celerity reaches a maximum for a given depth, generally a few
tens below sea surface, due to the warmer water layer close to the surface, and a “surface
propagation duct” appears. This effect, explained in reference [7] and represented
schematically on Figure 3.48, leads to a situation where some acoustic energy is “trapped”
near sea surface.

WP 3 - Measurements
Method
An example of propagation loss computation in such a situation is given on Figure A-12 of

Annex A. It can be seen that some minimum distance is needed for the phenomenon to
appear. Then, it is not expected to influence greatly URN measurements.
Figure 3.48.- Surface propagation duct.-
Figure 3.49.- Uncertainty due to changes in the celerity profile.-
In order to see if the celerity profile affects significantly the transmission loss at ranges typical
of this kind of measurements, Within internal studies a numerical simulation was performed
for different water depths (shallow and deep) to compute the standard deviation of the results
obtained considering different celerity profiles. The Figure 3.49 shows the results of this
study. This figure reveals that the influence of celerity profile on transmission loss is much
more important in deep waters than in shallow waters.
5 Currents.
The main effect of currents is to displace the equipment used for the measurement, mainly
buoys and underwater cables and instruments, by comparison to the expected location. The
main consequence is a bias or an inaccuracy of the determination of the actual distance
between the hydrophone(s) and the ship to be measured.
ANSI/ASA and ISO standards ([2] and [3])addresses this issue by a requirement of accuracy
for the determination of that distance, which is 2% for Grades A and B, and 5% for Grade C.
Regarding the effect of currents, guidance is given by requiring that the tilt angle of the cable
supporting the hydrophone should not exceed 5°. A s uggested method to estimate the
deviation between actual position and vertical is to measure the static pressure at

WP 3 - Measurements
Method
measurement location, giving the actual depth to be compared with the length of the cable.
This principle is shown on Figure 3.50. We can note that this method is not very reliable: the
current may vary with depth, acting on the cable in such a way that the tilt angle is not
constant. For example with a parabolic shape the approximate relationship between the
lateral deviation ∆L and the depth variation ∆h is (3.h.∆h)1/2 instead of (2.h.∆h)1/2 for a straight
cable tilt. In addition, the error on measurement distance will be greater if the lateral deviation
direction is perpendicular, rather that parallel to ship course.
The STANAG requires some accuracy for determination of measurement distance but does
not address particularly the effect of currents.
The DNV document does not address this topic because the measurement hydrophone is
placed on sea bottom. The hydrophone is placed in a cage, sufficiently large not to be
displaced by currents.
The effect of the current on the vessel working conditions should be also accounted since
this could have a great impact on the underwater radiated noise of the vessel. This effect is
studied in the item regarding the vessel.
Finally, currents increase the self-generated noise of the hydrophones in the low frequency
range as seen in paragraph 3.
Figure 3.50.- Effect of current on measurement equipment.-
6 Seabed effect.
Reflection of acoustic waves on sea bottom will modify the acoustic pressure and introduce
some deviation or uncertainty on ship URN measurement.
If water depth is sufficient (typically greater than 100 m and greater than measurement
distance), the travel distance of the acoustic wave reflected on sea bottom is about two times
the direct wave travel distance R (Figure 3.53). If it is assumed that seabed reflects totally
the incoming wave, which is a rare limit case, the maximum error due to bottom reflection will
be, in narrowband:

WP 3 - Measurements
Method
In practice, bottom losses for acoustic reflection on seabed, if grazing angle is not too small,
is at least 6 dB, then the error will be reduced to 0.75 dB.
A shallow waters configuration, close to the one defined in DNV standard, is also shown on
Figure 3.51. Acoustic path N°3 contribution is expected to be small because of multiple
reflection, and in any case it is taken into account in the propagation loss factor discussed in
its respective item. In the case where the acoustic sensor is on sea bottom, acoustic
interference occurs between acoustic paths N°1 and N°2. Using modelling from reference
[22], Figure 3.52 gives the reflection coefficient of a perfect plane seabed composed of mud,
sand and clay, then the total pressure, along frequency, sensed by a hydrophone put at
20cm from bottom on a sandy seabed. For this measurement configuration, the grazing
angle is expected to be about 30° or smaller. Conse quently, the expected reflection
coefficient is close to one (0 dB), except in the case of muddy seabed. In the case of a more
reflective seabed, such as sand, interference can occur, and total pressure oscillates
between a maximum of about +5 dB with respects to direct path, and lower values when
destructive interference occurs. This is probably why the DNV document introduces a -5 dB
correction factor to account for seabed effect. However, we can sea that this correction factor
is not reliable in all cases, because it depends on seabed characteristics and frequency.
Figure 3.51.- Effect of seabed on acoustic propagation paths.-
In reality, sea bottom is not perfectly smooth. In case of strong rugosity, the wave impinging
on sea floor is scattered in all directions, as represented on Figure 3.52. In that case, the
specular reflection of the impinging wave is lower, and the interference effect shown
previously is mitigated. This is probably the reason why the DNV document recommends the
sea floor not to be too flat.
In the case of the STANAG standard, there is a possibility to put a sensor on sea bottom, for
ship URN measurement at keel aspect. However, in this case, the grazing angle is close to
90°, then the reflection coefficient and consequent ly the interference effects are lower than in
the previous case. .

WP 3 - Measurements
Method
Figure 3.52.- Sea bottom roughness.-
Other important effect of the seabed in shallow waters is the bad transmission of lower
frequencies. This effect was already explained in the item frequency range.
7 Background noise.
The presence of background noise affects the signal to noise ratio, and so the accuracy of
the measurement. For that reason, all existing standards require to make a measurement
without the ship to be measured, then with the ship running in the acoustic range.
If we discard electronic noise and self hydrophone noise, of the measuring and processing
equipment which should be appropriately low, ambient noise at sea is due to different
components:
• turbulences (very low frequencies)
• sea noise related to sea state and wind
• remote shipping noise
• thermal noise (very high frequencies).
A reference model for ambient noise level at sea in deep waters is Wenz model represented
on Figure 3.53 (see also reference [15] page 210). This model introduces as an input the
sea state level (which can be related to wind speed) and a shipping traffic index “Tracom”
(low, medium, heavy, very heavy).
Before selecting a measurement site, it is also possible to get some on-site ambient noise
measurement data, such as data shown on Figure A-7 in Annex A.
To validate a ship URN measurement, the following signal to noise ratio (in dB) is first
calculated for all frequencies of interest:
If S/N is greater than 10 dB, the measurement is valid and the influence of background noise
is considered to be negligible.

WP 3 - Measurements
Method
If S/N is between 3 dB and 10 dB, the URN can be corrected using the following formula:
110
100
90
Noise level (dB ref. µPa^2/Hz)
80 Turbulence
Tracom 1
70 Tracom 4
Tracom 7
60
Sea state 1
50 Sea state 3
Sea state 6
40 Thermal
30
20
10
1 10 100 1000 10000 100000 1000000
Frequency (Hz)
Figure 3.53.- Wenz model for ambient noise at sea.-
If S/N is lower than 3 dB, the measurement is not valid and the corresponding data must be
discarded. If possible, the disturbing noise source should be investigated or removed, if
possible, or the experiment postponed in order to benefit from better environmental
conditions.
Conclusion
As this topic must be dealt globally, the guidelines for the definition of the measurement
procedure drawn from this study are analysed at the end of the point “Environment”.
The effect of the environment on the measurement is confined to two items:
• Transmission loss.
• Background noise.
Regarding the transmission loss, this paragraph has shown that:
• The effect of the Mirror effect is important for low frequencies. Moreover, this
phenomenon is one of the causes of the poor signal to noise ratio of the vast majority
of the vessels in the very low frequency range.
• The water depth affects significantly its value.
• It is not constant along the frequency span.
• The sea bottom characteristics shall be accounted in the case of shallow waters or
receivers located close to the sea bottom in measurements of deep waters.
• The celerity profile has a non-negligible effect on the transmission loss for the case of
measurements in deep waters.
Due to the amount of parameters affecting the transmission loss, the procedure should
consider a different and more accurate approach to compute it. Initially, a procedure for

WP 3 - Measurements
Method
obtaining experimentally the transmission loss of the test site was defined and tested in real
measurements. However, the tests did not give good results. Therefore, a numerical
approach must be followed to obtain a more accurate transmission loss for A-Applications. It
has been seen in internal studies performed that some input parameters shall be measured
or estimated. As we have seen here, the most important ones are:
• Sea bottom characteristics for shallow waters.

• Celerity profile for deep waters.
Moreover, due to the high uncertainty with regard to the characteristics of the sea bottom, it
is recommended not to deploy a receiver close to it.
Regarding the background noise, it is clear that the measurement should be performed with
a proper sea state.
3.4.3. Vessel particulars
The URN strongly depends on the vessel particulars like the ship type, speed and loading
condition. The questions are how these parameters will influence the measurement,
especially the measurement uncertainty, and if the current measurement standards account
for these parameters.
1 Vessel speed:
The influence of vessel speed on other parameters is not commented in the existing
standards. DNV Silent-E class notation prescribes 11 kn as quiet cruise condition if the
overall vessel length is larger than 50 m (8 kn if LOA≤50 m). The same applies to the
SILENV Green label proposal.
The noise radiated by a ship with a fixed pitch propeller (FPP) does generally increase with
vessel speed. The noise from a ship propeller will depend whether it is cavitating or not. The
lowest speed at which cavitation occurs is called the Cavitation Inception Speed (CIS). Naval
vessels and quiet research vessels are designed to be free from cavitation below a certain
operating speed. Naval vessels are often designed to have as high CIS as possible, typically
around 15-16 kn. However, all propellers will cavitate above a certain speed (if the installed
power is enough), no matter how well designed the hull and propeller are. Below the CIS,
other noise sources grow in importance. Thus, for naval vessels and quiet research vessels it
becomes important to reduce e.g. the non-cavitating propeller noise. The CIS can be
improved by increasing the blade area and by reducing the loading at the blade tips and hub.
Unfortunately, both these measures cause reduced propeller efficiency. Merchant ships are
designed primarily for optimum fuel efficiency and do thus have some amount of cavitation
when operating at design speed. The CIS for a merchant ship is likely to be 10 kn or lower
[Leaper & Renilson, 2012].
In the literature [Ross, 1976; Arveson & Vendittis, 2000], one can found models that suggest
a linear relationship between spectrum levels in dB and the log of the speed for a cavitating
propeller.
A ship with a controllable pitch propeller (CPP) could become noisier with decreasing speed
[Wittekind 2009]. This is particularly the case if they have a shaft generator and operates at
constant rpm. In this case, the speed is adjusted by the propeller pitch. This gives poor

WP 3 - Measurements
Method
hydrodynamic conditions that lead to increasing face cavitation and tip vortex cavitation with
increasing divergence from design pitch.
In summary, the dominating source and the frequency characteristics of the sources changes
with speed.
2 Vessel length:
In the current standards the vessel length influences the sea depth required. These
requirements will be analysed in terms of accuracy and measurement cost implications.
Moreover the vessel length will influence other phenomena such as the Mirror effect
observed.
In the existing standards, direct relationship with ship length is not mentioned. It is expected
that ship length has indirect influence:
• a bigger ship is expected to have a larger draft, then reducing water depth under keel;
• a longer ship will influence the time window for data acquisition and can introduce a bias
on the acoustic pressure levels coming from widely separated noise sources on the hull.
The vessel length does also influence the acoustic centre location (distance between
propeller and machinery location).
The propulsive power needed increases with the ship length. [Ross, 1976] suggest a linear
relationship between spectrum levels in dB and the log of the ship length (proportional to
20 log L).
3 Propeller and machinery source type and working conditions
The primary working condition should be the ones that propeller and machinery combination
is designed for. The DNV and SILENV procedures propose 85% MCR which usually
corresponds to design speed for a merchant ship. Secondary working conditions are other
typical operating conditions, e.g. passage in an area with restricted speed, which are off-
design working conditions.
The cavitation characteristics are determined by the propeller loading which is characterised
by a certain combination of advance ratio (J) and cavitation number (σ):
Where p0 is the local pressure, pv is the vapour pressure of water, ρ is the density of water,
VA is the inflow velocity to the propeller, n is the shaft speed and D is the propeller diameter.
The parameters influencing the advance ratio and cavitation number are mainly: draft, speed,
power and shaft speed. Thus, to be able to reproduce a certain condition, one must have
control over these parameters. Environmental parameters such as wind, waves and current
may violate the possibility to repeat an exact combination of speed, power and shaft speed.

WP 3 - Measurements
Method
In field, one has to accept some deviation from at least one of these. If a deviation in these
parameters will significantly change the cavitation characteristics and hence the URN will
vary very much from case to case. For a given propeller, the inception limits for different of
cavitation can be plotted as a function of advance ratio and cavitation number, Figure 3.54.
The design point for a naval propeller is usually in the non cavitating area while a commercial
propeller typically has tip vortex cavitation and some amount of sheet cavitation on the
suction side. Going from design draft to ballast draft decreases the cavitation number, i.e.
one moves downwards in the graph which normally increases the amount of cavitation. If the
point of operation is close to one of these inception limits, the cavitation characteristics might
be sensitive to a deviation in parameters.
Figure 3.54.- Inception limits for different types of cavitation.-
The number of propeller blades and the typical rpm of the machinery type affect the
frequency range of the radiated noise. For example, a water-jet unit has typically a higher
number of rotor blades and a higher rpm than a conventional propeller designed for a ship of
equal size. Thus, the fundamental blade rate frequency will be considerably higher for the
water-jet vessel.
Different types of engines (two-stroke, four-stroke, low speed, medium speed, steam turbine,
gas turbine, etc.) have different frequency characteristics.
4 Silent vessel:
It is a specific characteristic of certain kind of vessel that strongly influences the

measurement. The sound level of the vessel will be an important factor determining the
maximum distance where the device can be achieving a good signal-to-noise ratio. Moreover
it will influence the magnitude of other parameters such as the Mirror effect of the reflections
coming from the seabed.
5 Availability
The current procedures require certain conditions for the measurement location. For certain
vessels this requirement implies going far from the coast which means several days of
unavailability of the vessel.
For a commercial vessel it is almost out of the question to taking it out of operation in order to
go to a designated test area. The DNV procedure seems to be well suited to available
European conditions while the ISO procedure is very hard to achieve close to European
coasts. The ISO procedure is maybe too influenced by US Navy procedures.

WP 3 - Measurements
Method
3.4.4. Post-processing
1 Bandwidth
Once the data are collected the way of reporting the results will depend on the scope of the
measurement. The standards [2], [3] and [9] specify that the results must be reported using
one-third octave bands and the use of narrow bands is optional whereas the reference [1]
does not require a specific bandwidth.
The election of the way to report the results for the standard to be developed in this
document must depend on the required application defined in paragraph 2.1. For A-
applications narrow band representation is necessary because the aim of this application is
the identification of the sources, although it is true that the narrow band analysis is more
relevant for low than high frequencies since tonal components are more likely to appear in
this frequency range. On the other hand, for B-application one-third octave band will be
enough because its aim is to evaluate the source level of the vessel.
For some purposes is convenient to use the power spectral density per unit bandwidth and
the most widely accepted method to convert the spectrum level expressed in acoustic
pressure to a spectrum level in terms of power is to divide the in-band noise pressure by the
square root of the bandwidth. The assumption behind this method is that the sound energy is
uniformly distributed across the measuring band.
2 Acoustic centre definition/location.
The current standards assume that the vessel is a point source and therefore, they idealise
that the sound comes from an acoustic centre with the same location for all frequencies. This
point is important since the distance between the vessel and the hydrophones are measured
from it.
In the reference [9] the acoustic centre is assumed to be at 0.7 propeller radius when the
blade is pointing upwards. In this case the predominant source is assumed to be the
propeller. However, depending on the machinery characteristics this assumption may not be
true. On the other side the references [2] and [3] propose two ways of estimating the
acoustic centre depending on the grade of accuracy. For grade B and C the location of this
point is assumed to be halfway between the centre of the engine and the propeller and for
grade A it is estimated experimentally using the time signal recorded by the hydrophones.
However there is no specification on how to do it but it is a possible solution to the problem. It
proposes that this point can be defined by the maximum broadband hydrophone output
during each run. Nevertheless such thing is sometimes difficult to determine for some kind of
vessels because the signal may have parasitic signal or the vessel may not be noisy enough,
which make this task difficult and subjected to uncertainties. In fact, as seen in the Table
3.14, the results of this approach are not consistent. Indeed, this table shows the location of
the acoustic centre for the same vessel using the records of different runs. As can be seen
there is any coherence in the results.

WP 3 - Measurements
Method
Run Distance to CPA Distance from AC to GPS

Portside 1 96m -37.8m
Portside 2 92m -10m
Portside 3 105m -48.6m
Portside 4 185m 48.6m
Portside 5 207m 0m
Table 3.14.- Results of the experimental location of the acoustic centre.-
Besides, the idealisation of the existence of an acoustic centre implies to allocate in the
same place all the different sources of the vessel (propeller, machinery, etc.).
Anyway for sufficiently large distance (see the item distance of the hydrophones to the CPA)
the error made assuming this hypothesis is presumed to be low compared with other sources
of uncertainties. However the idealization of the existence of an acoustic centre is not
appropriate if the measurements are taken in the near field since the interaction among the
different sources are important in the noise transmission at short ranges.
3 Data acquisition window definition.
The data acquisition window determines the amount of data used to compute the source
spectrum of the vessel. The longer it is the larger the computed average will be, however the
fact that the average is larger does not necessary imply a better repeatability as other factors
such as the directivity of the source may strongly affect if they are not properly considered.
Figure 3.55.- Definition of the data acquisition window in the references [2] and [3].-
The references [2] and [3] define the data window fixing the data window angle to ±30º
(Figure 3.55) but the distance to the CPA is not fixed and it will depend on the length of the
vessel and, even though this distance were fixed, it would be really difficult to achieve it in a
real test. That definition leads to different data window period which, as can be seen in
Figure 3.55, implies different sound levels obtained in the post-processing depending on the
distance to the CPA. Such factor introduces uncertainties if no proper distance corrections
are taken due to the variation in the distance, as it can be seen in the Table 3.15 considering
a variation equal to 15%. However in order to properly take distance correction it is important
to have good estimation of transmission loss, otherwise the distance correction may
introduce more uncertainties than if it is not used.

WP 3 - Measurements
Method
Figure 3.56a.- Difference in the data window period depending on the distance to the CPA.-
30º
Hydrophones Hydrophones DW1 DW2
Figure 3.56b.- Difference in the data window period depending on the distance to the CPA.-
The speed of the vessel during the test is also important because it is going to determine the
data window period, therefore the number of terms in the average. The reference [9] defines
the averaging time considering the vessel speed and length.
According to that, the higher the vessel speed the shorter the time window period and the
fewer terms in the average could be used for the same frequency accuracy. Besides, this
definition usually results in shorter averaging time than the definition provided by [2] and [3].
However If everything is properly controlled more terms in the average implies better
repeatability. Therefore, in order to assure the same grade of repeatability the procedure
must be defined to achieve similar data window periods for different tests. Nevertheless
different vessels will have different operational speeds so the distance to the CPA is the only
parameter that can change (under the limitations explained in the above item) to achieve the
previous condition.
Finally, the directivity of the source may need being taken into account for long data window
since it may affect the final average. A way of accounting this is to define data windows at
different angles.

WP 3 - Measurements
Method
4 Background noise correction.
This item was explained from a physical point of view in the paragraph 3.4.2. The
measurement should be taken far from other sources so the background noise can be much
lower than the vessel noise measured by the hydrophones. Besides the sea state must be
suitable to assure, among other things, low background noise. The references [2] and [3]
specify that if the background noise is 10dB lower than the signal measured by the
hydrophones during the run no correction shall be made and if the difference is between 3dB
and 10dB it should be applied the following correction.
Where,
∆Ls+n/n is the signal-plus-noise-to-noise-ratio.
Ls+n (hi) is the sound pressure level, in dB, recorded by the hydrophone hi when the
vessel is performing one run.
Ln(hi) is the sound pressure level recorded by the hydrophone hi during the record of
the background noise.
L(hi,rj,f) is the sound pressure level by the hydrophone hi after applying the background
noise correction.
However if the difference between the vessel and background noise is less than 3dB the
results must be discarded.
The reference [9] proposes a similar specification but it recommends that the measurements
must be discarded when the background noise is not 5dB lower than the vessel plus
background noise.
This way to proceed has several drawbacks.
• Background noise should be sufficiently stationary so the results of the correction

applied can have enough accuracy. As a matter of fact, the background noise used
for the correction is the one measured previous or after the runs, and it may be
different from the actual one during the run. In this way, in order to have a given
accuracy such difference cannot be higher than a certain value, which can be higher
as the higher the signal plus background noise to background noise ratio is. Figure
3.57 shows how stationary the background noise should be to achieve 1dB of
accuracy.

WP 3 - Measurements
Method
Applicability of background noise correction
6
5,5
Variation of background noise (dB)

5
4,5
4
3,5
3
2,5
2
1,5
1
0,5
0
0 2 4 6 8 10 12 14 16 18 20 22
Signal plus noise to noise ratio (dB)
Figure 3.57.-Variation of background noise maximum to achieve 1dB of accuracy in the background
noise correction.-
5 Directivity.
When the vessel is modelled as a point source it is observed that this source has directivity
which depends on the value of the frequency. Generally the vessel emits less at bow aspect
than at beam aspect due to the shadow effect caused by the hull. Moreover differences are
appreciated when the measures are taken at keel aspect. The Figure 3.58 shows an
example of a directivity diagram of a vessel for a given frequency.
The directivity of the source may affect the extraction of the averaged source spectrum at 1m
depending on the duration of the data window. Furthermore, in order to fully characterize the
vessel as a source this directivity should be obtained and the new standards should take it
into account and provide methods to compute it, however the current standards do not
explicitly consider this phenomenon.
merchant5: Source levels for 125 Hz
0
330 30
300 60
190 200
170 180
160
270 90
240 120
210 150
180
Figure 3.58.-Directivity diagram of a vessel for a frequency of 125Hz.-
In the current state-of-the-art there are several methods to compute these kinds of diagrams.
However, all of them require the following points:
• Computing accurately the angle and range between the vessel and the hydrophones.
• Computing accurately the transmission loss for each third-octave band covered in the
analysis. No spherical or cylindrical transmission loss can be applied because induce

WP 3 - Measurements
Method
large errors becoming more important in the process of obtaining the directivity
diagrams.
• Having the suitable distance to the CPA in order to achieve a good signal-to-noise
ratio to be able of computing source spectrum averages for angles far from 90 or
270º.
• Depending on the speed of the vessel it may be necessary to perform a Doppler

Effect correction; otherwise the spectrum lines may be hidden. It requires an
identification of the tonal components.
• If the directivity along the vertical angle is required it is necessary to take

measurements at different depths or distances to the CPA.
6 Number of runs
Taking several underwater radiated noise measurements under the same conditions and
then taking the average as the final result is a way of reducing the uncertainties associated to
the measurement and improving its repeatability.
That is why the number of runs should be fixed for a given accuracy grade as the references
[2] and [3] do for each side of the ship and as the reference [9] does with the two
measurement taken, one per each side of the ship. The average of the different runs is
usually arithmetically averaged in dB.
Moreover, performing runs at different distances is a way of reducing the uncertainty with
regard to the transmission loss. This novel approach is not followed by current standards,
and the studies performed in this task show its high potential.
Besides, taking several measurements for different runs may give us a way to evaluate the
repeatability and uncertainty of our measurement for the actual conditions and to establish a
reject criterion for the runs according to their deviation from the average.

WP 3 - Measurements
Method
4. PROPOSAL FOR A NEW URN MEASUREMENT PROCEDURE
4.1. Introduction
This paragraph describes the measurement procedure developed in the Task 3.1 of the
AQUO project. This procedure is the result of the studies performed in the previous
paragraph as well as the experimental validation of the modifications of this procedure with
regard to the existing ones. These experimental verifications were performed within internal
studies of the project. In particular, the following items were experimentally verified:
• Measurement uncertainty and repeatability. In fact this study is also shown in the
paragraph 4.9 of this document.
• The number of runs.
• The goodness of performing runs at different distances.
• Distance to consider that the vessel behaves as a point source.
• Analysis of the influence of the directivity of the vessel to define properly the data
window.
• Analysis of the repeatability achieved by different types of post-processing.
• Transmission loss. Suitability of the procedure to obtain experimentally the
transmission loss of the test site with portable equipment.
• Transmission loss. Suitability of the numerical models, its sensitivity to the uncertainty
of the input parameters required and experimental assessment of the methodology by
computing the repeatability achieved.
• Numerical verification of the mathematical formula used to compute the transmission
loss for B-Applications.
• Sea state required. Background noise for different measurement device
configurations subject to different sea states. Assessment of the variation of the
background noise during a trial.
• Number and position of the hydrophones by means of an assessment of the
repeatability achieved.
• Experimental assessment of the accuracy of the distance measurement when a GPS
system is used.
• Experimental verification of the technical feasibility of an alternative way of measuring
the distance using the Doppler Effect.
• Definition of the acoustic centre.
• Definition of the limits of the repeatability checking defined in the paragraph 4.7.4.
• Definition of the limits of the uncertainty checking defined in the paragraph 4.7.5.
• Study of the different sources of uncertainties:
o Hydrophone line tilt.
o Influence of celerity profile.
o Influence of the source depth.
o Influence of the sea bottom in deep and shallow waters.
o Sensitivity analysis of the numerical models.
This standard procedure is applicable to surface vessels whose operating at speeds lower
than 50 knots. Besides, it lies on the assumption that the vessel behaves as a point source,
which requires the measurements to be conducted in the far field and then, normalized to 1m

WP 3 - Measurements
Method
distance from the centre plane of the vessel for further comparison. This standard provides
ways of obtaining the vessel signature source level at beam aspect of the equivalent
monopole source where the effects of the free surface and sea bottom are taken into
account.
4.2. Summary of the requirements

The following table summarizes the requirements for each application and grade within the
scope of this document:
Application A B
Grade A1 A2 B1 B2
Global
Measurement
4dB 4dB 7dB 6dB
Uncertainty
Measurement
1.5dB 1.5dB 2dB 2dB
Repeatability
Minimum Frequency 0.5Hz if f<50Hz

One-third octave band
resolution 1%f if f>50Hz
Frequency range 10Hz-50KHz (*) 10Hz-50KHz (*)(**) 20Hz-50KHz (*) 20Hz-50KHz (*)(**)
Test site
50 m 50 m
Minimum sea depth 0.3v2 under the keel. 0.3v2 under the keel. 100m or 1 x ship
200m or 2 x ship
(***) 3 x ship draught aft 3 x ship draught aft length.
under the keel. under the keel.
Sea state Sea state 3 Sea state 2 Sea state 3 Sea state 2
General
Flat seabed. homogeneous, not seabed features
characteristics
Noise recording
Configuration Bottom mounted Floating Bottom mounted Floating
Minimum number of
3 3 1 1
hydrophones
Hydrophones depth See Figure 7.1a See Figure 7.1b See Figure 7.2a See Figure 7.2b
Hydrophone
<3dB <3dB <4dB <4dB
sensitivity uncertainty
Hydrophone directivity <2dB
Dynamic range 80dB

Hydrophone
Every two (2) years Every three (3) years.
calibration
System verification Prior to the measurements by means of insert voltage calibration, calibrated sources,etc...

WP 3 - Measurements
Method
Application A B
Grade A1 A2 B1 B2
Distance measurement
Distance uncertainty
between the vessel +/-2m +/-10m
and surface buoy
Maximum tilt angle 10º
Field calibration
Modelling / Simple
Modelling Modelling Simple law Simple law
law
Runs and vessel characteristics
Nominal distance to
200m or 1 x ship length whichever is larger.
CPA
Number of runs 5 3
Vessel Speed (***) Log speed
Power provided by the main engines

Rotational speed, pitch and load of any propeller or thrusters
Conditions to keep
Load conditions
constant (****)
Fore and aft draught
Machinery working during the test
Post-processing
Acoustic centre Halfway between the main engines and the Halfway between the main engines and the
location propeller or propeller.
Data window length 100m or 1 x ship length whichever is larger.
Data window angle -45º to 45º each 5º for each side.
Narrow band analysis Required Optional
Directivity analysis Optional
(*) Upper frequency recommended 100KHz.

(*) Due to the measurement device configuration used in deep waters, practical frequency range for
this grade will go from 20/30Hz to 50KHz.
(***) Deviation to this requirement should be explicitly indicated.
(****) According to contract specifications.

WP 3 - Measurements
Method
4.3. Instrumentation
4.3.1. Hydrophones, signal conditioning and recorder
The hydrophones shall have a flat frequency response over the frequency range required for
the application and shall be omnidirectional. In particular the hydrophones shall be linear with
a maximum variation of:
• +/-3dB for grade A.
• +/-4dB for grade B.
And their directivity diagram shall not present a variation higher than 2dB. Moreover, the
hydrophones should have proper sensitivity, bandwidth and dynamic range to measure the
ship under test. The required performance shall be obtained with the full cable length that will
be used during the test.
The recorder and signal conditioning shall be capable of recording data from the
hydrophones without loss of accuracy.
The full measurement system shall be laboratory calibrated every 24 months according to
ANSI S1.20 or IEC 60565 for all required one-third octave bands for A-applications and every
36 months for B-applications. They will also be verified in the field prior to, and daily
throughout, the measurements series using either insert voltage methods for all one-third
octave bands, capacitance metre, Megger or a single frequency device, such as pistophone.
4.3.2. Distance measurement system
A distance measurement system is required to determine, continuously throughout the

acquisition of the data, the distance between the hydrophones and the acoustic centre of the
vessel.
The distance can be measured in two ways. The first one consists on measuring the
horizontal distance between the vessel and the buoy located above the hydrophones. In this
case, the position of the vessel and the buoy surface shall be recorded every second with an
error of +/-2m for A-applications and +/-10 for B-applications. The first one implies the use of
differential GPS whereas the latter one requires the use of standard GPS. For bottom
supported configurations the buoy swing shall be lower than 10m. For bottom supported and
suspended configuration the tilt angle shall be lower than 10º. If the hydrophone cable tilt
angle exceeds this quantity, it can be reduced attaching weight to end of the cable.
The first method is subject to many uncertainties so for grade A it is recommended to employ
the second way of measuring the distance. This method consists in measuring directly the
distance between the hydrophone and the vessel to avoid the uncertainty related to the tilt
angle. In this case, the GPS cannot be used as the hydrophones are located into the sea,
therefore other systems based on the Doppler effect, radar, etc. must be used. If this is the
case, the maximum error shall be lower than 5% of the distance to the CPA.

WP 3 - Measurements
Method
4.4. Test site
4.4.1. Test site requirements
The place where the measurements will be performed shall be far from ship lanes or areas
with dense maritime traffic so the measurement is not affected by external ships.
The sea bottom shall be as flat as possible and neither sloping seabed nor seabed features
shall be present in the test site as much as possible. The sea bottom shall have the same
characteristics over the location where the runs will be carried out.
It is recommended to measure in the absence of currents as far as possible since they may
increment the level of background noise recorded by the hydrophone.
The sea depth of the location shall be higher than 50m for grades A1 and B1. For grade A2,
it shall require 200m or twice the ship length, whichever is larger. Grade B2 requires a see
depth of 100m or once ship length whichever is greater. Besides and in any case, the depth
of the location shall be enough so as to have under the keel 0.3v2(being v the vessel speed
in m/s) or three times the ship draught after, whichever is greater, to assure a representative
hydrodynamic performance of the vessel.
4.4.2. Sea surface conditions
The sea surface conditions may strongly affect the measurements so they must comply with
some requirements. Rough seas may increment the background noise in the area limiting the
frequency range free of background noise corrections. Moreover it may cause added self
generating noise in the hydrophones since the hydrophone line will be excited. Finally, the
repeatability of the surface vessel’s source level will be affected since rough seas cause
added instability of the ship under test and its propulsion system.
For all these reasons the sea state shall be 3 or lower if bottom supported hydrophones are
used (grade A1 and B1) and the length of the ship under test is larger than 50m, otherwise
the sea state should be 2 or lower.
4.5. Hydrophone deployment
4.5.1. Grade A1
The hydrophones shall be arranged vertically in the water column and located to measure
the beam aspect of the vessel under test. At least three omni-directional hydrophones shall
be deployed in a bottom supported line as shown in Figure 4.1a.
The grade A1 requires the deepest hydrophone to be at least 3m above the sea bottom to
avoid undesirable effects due to the pulls to the base. The hydrophones shall be separated
15/20m from each other and always the closest hydrophone to the sea surface shall be more
than 15m away from it. Measures shall be taken to mitigate the effect of cable strum.

WP 3 - Measurements
Method
Figure 4.1.-Hydrophone deployment for grade A1 and A2.-
4.5.2. Grade A2
The hydrophones shall be arranged vertically in the water column and located to measure
the beam aspect of the vessel under test. At least three omni-directional hydrophones shall
be deployed, suspended from a buoy which shall be uncoupled somehow from the sea
surface as shown in Figure 4.1b. The closest hydrophone to the sea surface shall be more
than 40m away from the sea surface, and the hydrophones shall be separated more than
30m from each other. Finally, the deepest hydrophone shall be located not deeper than half
the sea depth to minimize the effects of the sea bottom.
4.5.3. Grade B1
At least one hydrophone shall be deployed not rigidly mounted on the base and shall be
approximately half depth from the sea bottom (see Figure 4.2a).
4.5.4. Grade B2
At least one hydrophone shall be located at 50m depth from a floating buoy, which shall be
uncoupled somehow from the sea surface as shown in Figure 4.2b.

WP 3 - Measurements
Method
Figure 4.2.-Hydrophone deployment for grade B1 and B2.-
4.6. Transmission loss
In order to express the noise level measured at a reference distance for further comparison a
correction shall be applied taking into account how to the sound propagates in the test
location (transmission loss).
This procedure envisages two possibilities to compute the transmission loss:
• Modelling activities: By means of the use of mathematical models that require input
data such as sound speed profile, bathymetry, sea bottom characteristics, etc .
• Simple law: It is a mathematical formula
4.6.1. Modelling activities
Grade A1 and A2 will compute the transmission loss by means of the appropriate numerical
models. These models require the input of the following data:
• Sound speed profile, water density and salinity during the sea trials: This is more
important for deep waters than for shallow waters, which is why measurements of
celerity profiles is required for grade A2. Anyway, due to its not negligible importance
in shallow waters, measurement of the sound speed profile during the sea trials is
recommended for grade A1.
• Bathymetry of the test site, sea bottom density and acoustic characteristics:
The bathymetry and the acoustical properties of the sea bottom should be obtained
from the Total Sediment Thickness of the World's Oceans & Marginal Seas2
database or a similar one. Unless information that is more specific is available, the
sea bottom will be considered flat with only one semi-infinite layer with acoustical
properties equivalent to the actual seabed structure of the test site. This information is
much more relevant for grade A1 than for grade A2.

WP 3 - Measurements
Method
• Source depth: The numerical models shall be able to handle individual point
sources. Due to its importance in the results of the models, several source depths
must be considered according to the following:
o For vessels whose ship draught is below 8m two different source depths must
be considered. The depth of the first one shall be 1/3 of the actual aft draught,
and the depth of the second one shall be 2/3 of the actual aft draught of the
vessel.
o For vessels whose ship draught is above 8m three different source depths
must be considered at 1/4, 1/2, and 3/4 of the actual draught of the vessel.
These data shall be measured with enough accuracy. The extension of the area that shall be
modelled to compute the transmission loss shall cover enough distance for the computation
of the URN signature and directivity diagram of the vessel.
The transmission loss shall be computed at each hydrophone depth every one (1) meter
from 50m to the maximum distance required for the computation of the URN signature or the
directivity diagram.
For each point, source depth, and centre frequency of all the one-third octave bands within
the frequency range, the transmission loss at the position of the receivers (hydrophones) will
be computed.
Then, for each point and centre frequency of all the one-third octave bands the source
average transmission loss shall be computed arithmetically averaging the transmission loss
obtained from all the source depths considered for each point and frequency.
Where,
is the source depth average transmission loss in dB, at horizontal

distance ,frequency and for the hydrophone .
is the depth of each source depth considered.
is the number of source depths considered.
Once it has been done for all the points of the grid, the values obtained for each frequency
and hydrophone shall be used to perform a logarithmical regression analysis to obtain a
range-source depth average transmission loss. This accounts the fact that, on the one hand,
the vessel is a moving point source and on the other hand, the transmission loss computed
previously is for a given frequency and it shall be used for the whole one-third octave band.
Where,

WP 3 - Measurements
Method
is the range-source depth average transmission loss for a given frequency

and hydrophone .
is the horizontal distance between the source and the receiver.
and are the parameters obtained from the logarithmical regression analysis.
Two models will be used, one for low frequencies (<1KHz) and other for high frequencies
(>1KHz). The following models are recommended:
• Low frequencies: Scooter + Fields.
• High frequencies: Bounce + Bellhop.
4.6.2. Simple law
.The following mathematical relationship should be applied to obtain the transmission loss:
• Water depth lower than 100m.
o For frequencies below 100Hz : .
o For frequencies above 1000Hz: .
o For frequencies between 100Hz and 1000Hz: where

will vary linearly with the frequency being 19 for 100Hz and 20 for 1000Hz.
Water depth above 100m:
o For all frequencies:
Where
is the distance from the acoustic centre of the vessel to the hydrophone.
is the transmission loss without accounting the mirror effect.
Due to the fact that the aim of this procedure is to obtain the source level considering the
vessel as a monopole source a low frequency correction will be made to account for the
Mirror effect. Therefore, the final transmission loss to apply during the post-processing is
obtained according to the following equation:
Where,
is the frequency.

WP 3 - Measurements
Method
is the sound speed in water (1515m/s).
is the source depth and equal to 2/3 of the aft draught of the ship during the
measurement.
is 20º.
4.7. URN measurement
4.7.1. Background noise
The background noise shall be recorded at least before and after the runs. If possible, it is
recommended to record the background noise after each run. Each record shall be at least
two minutes long and the vessel shall be far enough or with its machinery completely
stopped during the record.
4.7.2. Validation of the measurement location
Some actions shall be taken to validate the measurement location before performing the
required measurements according to this procedure. They are the following:
Sea state: Sea state shall be Beaufort 3 or lower when the hydrophone bottom supported
configuration is used and the ship length is larger than 50m. Otherwise, the sea state shall
be Beaufort 2 or lower.
Background noise: If the background noise measured by the hydrophone(s) is not 10dB
lower for a reasonable quantity of one-third octave bands than the expected signal at the
location of the hydrophones the test location in that moment is not valid.
4.7.3. Configuration of the different runs
The vessel under test shall transit a straight line course to achieve the required distance at
CPA for this run. Recording of data shall be performed from approximately the minimum of
800m or 4 minutes before the front of the vessel reaches the CPA to the minimum of 800m
or 4 minutes after the whole vessel has passed the CPA. Before the vessel reaches the
starting point of the record, she shall have achieved the required run conditions, and unless
required by the test plan, vessel speed, machinery conditions, etc. shall be constant.
Depending of the application, the configuration of the runs is different:
A-applications:
Six runs per each side and vessel conditions shall be performed at distances to CPA
according to the Figure 4.3 with a margin of +/-10m. The nominal distance to CPA (dncpa)
shall be 200m or 1 x ship length, whichever is greater. For some very silent vessels, this
requirement may cause problems in terms of the SNR. In these cases, a shorter dncpa can be
considered taking into account that the closer the distance, the less realistic the assumption
that the vessel is as a point source is.

WP 3 - Measurements
Method
The direction of the course of each run shall not vary more than 5º among them to minimize
the differences in the vessel hydrodynamic performance due to the effects of the currents.
Figure 4.3.-Test course configuration for A-applications.-
B-Applications:
Three runs per each side and vessel conditions shall be performed at distance to CPA
according to the Figure 4.4 with a margin of +/-10m. The nominal distance to CPA (dncpa)
shall be 1 x ship length or 200m, whichever is greater. For some very silent vessels, this
requirement may cause problems in terms of the SNR. In these cases, a shorter dncpa can be
considered taking into account that the closer the distance, the less realistic the assumption
that the vessel is a point source is.
The direction of the course of each run shall not vary more than 5º among them to minimize
the differences in the vessel hydrodynamic performance due to the effects of the currents.

WP 3 - Measurements
Method
Figure 4.4.-Test course configuration for B-applications.-
4.7.4. Repeatability checking
This paragraph is aimed at describing the procedure to check that the repeatability of the
actual measurement is according to what is required for the grade considered. This test is
only applicable for grade A.
As seen in Figure 4.3, three runs at 1 x dncpa should be performed. The aim of this is to
check that there are no significant variations in the measured source levels obtained from
these runs under the same conditions. In particular, the average of the differences in each
one-third octave band among the resulting source levels obtained from the runs performed at
the closest distance shall not be higher than 1.5dB.
4.7.5. Uncertainty checking
As can be seen in the equation below the uncertainty of a given measurement procedure will
be mainly determined by the uncertainty of the distance measurement system, the noise
measurement system, the transmission loss and the variability of the vessel as a source.
This paragraph will detail a way of quantifying the actual uncertainty of the measurement
using the data obtained from the very same measurement. This test consists on comparing
the source level obtained from the runs performed at different distances. In particular, the
average of the differences in each one-third octave band among the resulting source levels
obtained from these runs shall not be higher than 3dB for grade A and 6dB for grade B.

WP 3 - Measurements
Method
4.7.6. Vessel conditions
Vessel conditions shall be fixed before the test. The following parameters shall be defined
and kept constant during the test:
• Vessel speed through water
• Power provided by the main engines
• Rotational speed, pitch and load of any propeller or thrusters.
• Load conditions.
• Fore and aft draught.
• Machinery that are working during the test.
The vessel conditions for the test shall be defined according to contract specifications.
Commonly it will be as close as possible to the expected operating profile of the vessel. The
vessel shall carry a load within her normal load range.
For vessels with large variations in the load conditions during her normal operation,
measurements at two loading conditions (full load and ballast) are recommended since the
underwater noise radiated by the vessel will strongly depend on her operational conditions
(different depth of the propeller, power requirements, etc). Therefore, these two
measurements will represent the characteristic range for the underwater radiated noise level
of the vessel.

WP 3 - Measurements
Method
4.7.7. Test sequence
The following table tries to summarize the sequence of actions to perform during this
measurement procedure.
Action Comments
Check the weather conditions
Deployment of the hydrophones The same location as the previous day.
Background noise measurement Record of the background noise

Measurement of sound celerity
profile and bottom When applicable.
characteristics.
Checking that the level of background noise is 10dB lower than the
Verification of the level of
expected noise of the vessel at the hydrophone position for a
background noise
reasonable amount of one-third octave bands.
Configure the ship according to run
Check the distance measurement system
Notify that the measurement system is operational.

Notify that the ship configuration is consistent with the test procedure,
start the run.
Runs Start the recording at “start data record point” position.
Inform the vessel crew that the recording has started.

Stop the recording at “end data record point” position and notify the end
of the run.
Inform the vessel crew that the recording has stopped.
Inform the measured ship if the run is OK or need to be repeated.
Background noise measurement Record of the background noise

Measurement of sound celerity
When applicable.
profile
Check the weather conditions Note variations in the weather conditions during the measurement
Table 4.1.-Test sequence.-

WP 3 - Measurements
Method
4.8. Post-processing
4.8.1. Acoustic centre
This procedure will consider that the acoustic centre is collocated for all frequencies and it
will be halfway between the propeller and the main engines on the centre plane of the vessel
and 2/3 ship draught depth.
4.8.2. Data window
Multiple data windows centred at different angles from -45º to 45º (see Figure 4.5) will be
defined to obtain for each one of them an averaged spectrum to take into account somehow
the directivity of the vessel. The data window length (DWL) for each data window will be
equal to the ship length (SL) or 100m whichever is greater, and its middle instant occurs
when the acoustic centre goes by the centre of the respective data window.
Figure 4.5.-Definition of the data window.-
For each data window,a linear average one-third octave band spectrum of the sound
pressure level measured by each hydrophone will be obtained in accordance with the IEC
1260:1995 and ANSI S1.11-2004.
4.8.3. Background noise
The fact that background noise will be always present during the measurement of the URN
signature of the vessel introduces error whose magnitude will depend on the signal-plus-
noise-to-noise ratio (SNR). When this value is smaller than a certain value, corrections can
be applied, however the magnitude of the error of this correction will depend on how much
the background noise varies.
This procedure fixes the limit of the SNR from which the background noise can be neglected
and how much the background noise can vary so the corresponding correction can be
applicable to obtain a maximum uncertainty because of this phenomenon of 2dB for A-
applications and B-applications.
For instance, if the signal plus background noise measured is 100dB and the background
noise in the area is 90dB, the actual noise coming from the vessel has a level of 99.5dB, so

WP 3 - Measurements
Method
an error of about 0.5dB would be made if the background noise was neglected. In general,
the error caused by not considering the background noise is:
So, the signal-plus-noise-to-noise-ratio of each hydrophone shall be at least 10dB for each
one-third octave band in order not to take it into account. See equation below to obtain the
SNR for each hydrophone.
Where,
is the signal-plus-noise-to-noise-ratio.
is the sound pressure level, in dB, recorded by the hydrophone when the
vessel is performing one run.
is the sound pressure level recorded by the hydrophone during the record of
the background noise.
If it is between 3dB and the 10dB and sufficiently stationary, corrections according to the
following equation shall be made in the affected one-third octave bands and properly
indicated in the report.
Where,
is the sound pressure level by the hydrophone after applying the

background noise correction.
As it was said before, the background noise correction can be performed when background
noise is sufficiently stationary. That relies on the fact that when the correction is performed
an error is made because the background noise recorded before the trials will not exactly be
equal to the background noise existing during the run. That error will depend on how much
the background noise has varied according to the following equation:
Where,
is the estimated variation of the background noise. It can be computed as the

difference between the background noise at the beginning and at the end of the trial.

WP 3 - Measurements
Method
Considering the equation above and taking into account that the maximum error permitted
because of the background noise correction is fixed to 2dB, the graph of the Figure 4.6
shows the maximum variation of the background noise per each one-third octave band
allowed so the correction can be performed. Otherwise, it would be a deviation from this
standard.
For example, if we have a SNR equal to 4 dB for a particular one-third octave band; then, we
cannot have a variation between the level of background noise measured at the beginning
and at the end of the trial for that particular band higher than 2dB, so we can perform the
background noise correction.
Figure 4.6.-Applicability of background correction.-
4.8.4. Specific post-processing
For each hydrophone and data window a distance correction to refer the measurements at
1m shall be made using the numerical range-source depth average transmission loss
according to this equation.
Where
is the one-third octave band spectrum of the sound level received by

each hydrophone
is the one-third octave band spectrum of the source level computed

from the hydrophone
is the horizontal distance from the hydrophones to the midpoint of the corresponding
data window for grade A or the total distance for grade B.
is the centre frequency of the corresponding one-third octave band.
is the horizontal angle of the data window.

WP 3 - Measurements
Method
is the transmission loss at distance .for the one-third octave band whose
centre frequency is obtained for the hydrophone
Then, the power average of the source level obtained from each hydrophone is obtained
according to the following equation..
Where
is the power average one-third octave band spectra of the source level
obtained by each hydrophone .
is the number of hydrophones deployed.
Then, the envelope of the resulting power average of the source levels obtained for each
data window shall be computed according to the following equation:
Where
is the envelope, corresponding to the run , of the family of curves made up

by the resulting power average the source levels obtained for each data window whose
angle is .
Finally, all the runs of data are then arithmetically averaged to determine the final sound
source value for each side.
Where
is the resulting signature source level for the side , which can be either
portside or starboard.
is the number of runs per each side.

WP 3 - Measurements
Method
Figure 4.7.-How to obtain the directivity enveloping spectrum per each side of the vessel.-
Finally the power spectral density of both directivity enveloping spectra will be computed to
obtain the resulting signature source level per each side:
Where,
is the bandwidth of the one-third octave band considered.
is the power spectral density of the resulting signature source level per
each side .
The Figure 4.7 summarizes this way to proceed.
4.8.5. Narrow band analysis
For A applications narrow band analysis shall be performed for noise signals recorded and
the level of the main spectral narrow band components shall be showed in the corresponding
report. The inherent bandwidth of lines in radiated noise spectra will be function mainly of the
stability of the source and the Doppler Effect.
Each data window will be divided into 2s samples and for each one a narrow band spectrum
shall be obtained windowing the signal by means of a Hanning window.
Then before averaging the spectra, the frequency resolution of each one shall be adapted as
follow to avoid that the frequency lines are hidden because of the Doppler Effect:
• 0.5Hz for frequencies below 50Hz.
• 1% for frequencies above 50Hz.

WP 3 - Measurements
Method
If the user wants to perform narrow band analysis for B-applications wider frequency
resolution is valid.
Spectra obtained for a given data window will be RMS averaged with linear weighting. The
spectral narrow band component measured by each hydrophone deployed shall be corrected
to 1m using the transmission loss applicable according to the following equation:
Where
is the narrow band component obtained from the signal measured by the
hydrophone at distance and data window whose angle is .
is the transmission loss at distance .for the one-third octave band the
frequency of the component belongs to obtained for the hydrophone
is the narrow band component obtained from the signal measured by the
hydrophone and data window whose angle is referenced at 1m from the source.
If more than one hydrophone is used then the spectra of the different hydrophones shall be
power average as follow:
Where
is the power averaged narrow band spectrum taken and data window
whose angle is referenced at 1m from the ship hull.
Then, the maximum of the amplitudes obtained for this component among all the data
windows shall be computed:
Where
is the maximum amplitude obtained among all the data windows for the
narrow band component of interest.
Finally the final amplitude for the narrow band component of interest will be computed
averaging arithmetically the amplitudes obtained for each run according to the previous
equation:

WP 3 - Measurements
Method
Where
is the final amplitude for the narrow band component of interest at either
portside or starboard (which is defined by .
4.8.6. Directivity
Due to the minimum required distance between the hydrophones and the CPA of the run and
the sea-depth allowed by this procedure, just the slice of the hemisphere representing the
directions the vessel emits at can be computed (see Figure 4.8). Fortunately, at larger
distances of the vessel (commonly tenths of sea depths) the beam aspect of the vessel is
much more important than the keel aspect. For this reason the horizontal directivity will be
obtained averaging the values along the vertical.
The directivity diagrams shall be obtained for all the one-third octave bands of the frequency
range applicable unless the user requires something different.
In order to obtain the directivity diagrams of the vessel the following post-processing shall be
applied.
1. Each run will be divided into two second samples whose horizontal angle is αi and
horizontal distance to the hydrophones is di (see Figure 4.9).
2. For each sample and hydrophone deployed, the one-third octave band spectrum of
the sound pressure received by the corresponding hydrophone shall be obtained in
accordance with the IEC 1260:1995 and ANSI S1.11-2004.
3. The spectra obtained from the previous step shall be referenced to 1m using the
transmission loss applicable and the distance di specified in the step 1. Then the
power average of the results obtained from each hydrophone shall be computed.
4. Then, the amplitude of the one-third octave bands will be extracted obtaining the
value .
5. Finally, the amplitude of each frequency of interest of all the samples with certain ,
and whose signal to noise ratio is above the limits specified will lead to obtain the
directivity diagram as shown in the Figure 4.10.

WP 3 - Measurements
Method
Figure 4.8.-Directivity hemisphere covered by the procedure.-
Since runs at four different horizontal distances will be performed, four directivity diagrams
will be obtained, each one corresponding to each horizontal distance. These four diagrams
can be compared with each other to check the validity of the assumption of the vessel acting
as a point source.
Figure 4.9.-Post-processing for directivity diagrams.-
Figure 4.10.-Directivity diagram.-

WP 3 - Measurements
Method
If values of the vessel directivity close to bow and stern aspects are required, it is necessary
that the CPA of the run will be closer than the nominal distance to CPA given by the
procedure. The worse the signal to noise ratio is the closer the hydrophones need to be.
Anyway, the distance di shall always be higher than the nominal distance to CPA so the
spectrum obtained for the corresponding sample could be valid.
4.9. UNCERTAINTY AND REPEATABILITY STUDY
4.9.1. Theoretical study
This paragraph will summarize the study of the uncertainties associated to the procedure
proposed in this paragraph:
• Firstly an evaluation of the individual sources of error will be performed. To this end,
different studies, both theoretical and experimental have been done. In summary, the
following items have been considered as a source of error:
o Uncertainty regarding the distance measurement: It was divided into two

items, one associated to the accuracy of the GPS system used, and other
associated to the uncertainty caused by the hydrophone line tilt.
o Uncertainty associated to the instrumentation; both, the one associated to the

frequency response of the hydrophones and the one associated to their
directivity. In spite of indicating these values of uncertainty (according to what
is specified in the procedure) they are not considering in the computation of
the final uncertainty as they commonly affects only the very high frequency.
o Uncertainty associated to the transmission loss. With the aim to estimate this
value a sensitivity analysis of the model was performed considering the
uncertainty of its input parameters.
o Uncertainty associated to the post-processing. Only, the uncertainty

associated to the background noise correction (when it has been applied) has
been considered as the other factors (acoustic centre location, effect of vessel
direcitivity,…) have shown to have negligible importance with regard to other
factors.
• Finally, the final uncertainty associated to the measurement is obtained combining the
individual uncertainties obtained previously.
Before detailing the development of this work, it is important to take into account that in this
study the uncertainty is computed so 95% of a hypothetical large list of URN measurements
of the same vessel are within the limits made up of , where is the measured
URN source level of the vessel and is the uncertainty associated to this measurement.
The Table 4.2 shows the final values obtained from this theoretical study for the different
source of uncertainties as well as the final value for the measurement procedure. The Table
4.3 shows similar information for the analysis of the repeatability of the measurement.

WP 3 - Measurements
Method
Grades A1 A2 B1 B2
Distance Measurement Accuracy 0.5dB 1dB 1.5dB 1.5dB
Noise recording accuracy 2.5dB (1) 2.5dB (1) 4.3dB (1) 4.3dB (1)
Transmission loss 3dB 3dB 7dB 6dB
Vessel 1dB 1dB 1.2dB 1.2dB
Post processing 2dB 2dB 2dB 2dB
Total 4dB 4dB 7dB 6dB

Table 4.2.- Final uncertainty for each item in each one of the grades defined.-
(1) Due to the fact that this uncertainty is only important for high frequencies, it is not
accounted in the value of the final uncertainty of the measurement.
(2) Grades A1 A2 B1 B2
Distance Measurement Accuracy 0.5dB 1dB 1.5dB 1.5dB
Noise recording accuracy 0dB 0dB 0dB 0dB
Transmission loss 0dB 0dB 0dB 0dB
Vessel 1dB 1dB 1.2dB 1.2dB
Post processing 2dB (1) 2dB (1) 2dB (1) 2dB (1)
Total 1.2dB/2.2dB (1) 1.5dB/2.3dB(1) 2dB/3dB(1) 2dB/3dB(1)

Table 4.3.- Final repeatability of the grades defined in the URN measurement procedure.-
(1) Only when the background noise correction is applied.
4.9.2. Experimental study
This paragraph will detail the tests performed in the FS4 aimed at the experimental
verification of the real uncertainty of the measurement of the measurement procedure
described here. The aim of this trial is, on one hand, to support experimentally the suitability
of the procedure, and on the other side, to obtain an experimental value of the actual
repeatability and uncertainty of the measurement which supports the theoretical values
obtained in the paragraph 4.9.1. Firstly, the experiment will be described and finally, the
results will be shown and analysed.
During the second full scale measurements the underwater radiated noise of the vessel FS4
was measured twice according to the measurement procedure in the first day to assess the
repeatability and once more the following day under as different conditions as possible (sea
state, heading, variable bathymetry, etc.), to assess the uncertainty of the measurement. The
Figure 4.11 shows the different locations and the schematic directions of the runs performed
in these three measurements. As can be seen in this figure, the runs performed during the
first day were perpendicular to the ones performed during the second day. The first one
corresponds to constant bathymetry and the second one to variable bathymetry.

WP 3 - Measurements
Method
Figure 4.11.- General overview of the trial to quantify uncertainty and repeatability.-
The Figure 4.12 shows the two final URN signatures obtained, after applying the whole post-
processing, during the first day of the trial. Due to the rough sea state during the first day the
frequency range below 100Hz had to be discarded as it was background noise affected. This
is why a red shaded zone is present in this frequency range. The Figure 4.12 clearly shows
the high repeatability of the measurement procedure.
Figure 4.12.- Signatures taken during the same day to evaluate repeatability.-
The Figure 4.13 shows the signature taken during the second day and one of the signatures
taken during the first day so we can assess the uncertainty of the measurement. This figure
shows worse accordance between these two signatures than what Figure 4.12 shows. This
fact was expected and confirms that there were changes in the condition of the measurement
(celerity profile, effect of bathymetry, sea state, etc.). Moreover, the average difference
between these two signatures is about 2dB, which shows that the uncertainty of the
measurement procedure looks reasonable and does not contradict the uncertainty obtained
theoretically in the paragraph 4.9.1.
Therefore these results seem to be in accordance with the theoretical values obtained for the
repeatability (1.5dB) and uncertainty (4dB) and support, by means of experimental data, the
final definition for the underwater radiated noise measurement procedure.

WP 3 - Measurements
Method
Figure 4.13.- Signatures taken during different days to evaluate the uncertainty.-

WP 3 - Measurements
Method
5. CONCLUSIONS
A preliminary study to identify the needs concludes that the measurement procedure shall
address three topics: Sources identification, assessment of the URN of the vessel to
compare with future limits and measurements of rough estimates of the URN of the vessel
for traffic management and development of URN patterns. The measurement procedure
developed here covers the first two topics.
Other important need identified comes from the fact that large areas of the European
continent are surrounded by shallow waters, so the procedure shall address measurements
in shallow waters. On the other hand, in other areas, performing measurements in shallow
waters is rather difficult so the measurement procedure shall also address measurements in
deep water.
After carrying out the study of the key factors/parameters affecting the measurement of the
underwater noise radiated by the vessels, reviewing how the current standards deal with
them and developing of a new measurement procedure suitable for both, shallow and deep
water, the following conclusions can be drawn:
• One of the most important deficiencies/setbacks of the current standards is the lack
of a paragraph dealing with the uncertainty and repeatability associated to the
procedure or the lack of experimental data or facts supporting the values claimed by
some of them, which in some cases are rather optimistic.
• Consistent theoretical, numerical and experimental studies have been performed to
support the uncertainty and repeatability claimed by this procedure. Moreover,
individual estimations for each one of the main topics of the measurements has been
obtained. This enables us to see which topics require more effort and which do not. In
particular, the computation of the transmission loss is the most important source of
uncertainty as well as the measurement of the distance.
• Finally, the procedure provides means of checking experimentally that the value of
uncertainty and repeatability of the measurement is under control, and therefore the
uncertainty specified in the procedure is valid for the final measurement.
• Most of the modifications introduced by the procedure developed in this deliverable
as well as certain key requirements coming from existing standards have been
experimentally or numerically verified. Moreover, the levels of uncertainty and
repeatability of the measurement procedure have been specifically checked during a
real measurement being in accordance with the values obtained by the analysis
performed to study these topics.
• After the experimental results, we can conclude that the final measurement procedure
developed in this deliverable complies with the initial aims. It is also true, that some
topics can be improved (see below) but this procedure definitively is a real progress
beyond the current state of the art. In particular, the exhaustive study of its
uncertainty and repeatability and the experimental verification of some important
topics like the minimum distance to the CPA for one run, the improvement regarding
with the requirement of performing runs at different distances, the study of the
transmission loss and the low frequency masking.

WP 3 - Measurements
Method
On the other hand, there are also topics that can be addressed in the future to support and
improve the procedure for measuring the underwater radiated noise of the ships, in
particular:
• This deliverable has not dealt with the need of measuring rough estimates of the URN
signature of the vessel for traffic management without disturbing its route. It was a
need identified initially in this deliverable that, in spite of not being specifically studied
in this task, should be covered with a standard procedure to perform this kind of
activities.
• On the other hand, more experimental data of uncertainties and repeatability, as the
one described in the paragraph 4.9.2, should be collected to support more strongly
the uncertainty and repeatability claimed by the procedure developed here. Indeed,
due to the budget and time constraints, the uncertainty and repeatability of the
measurement procedure was experimentally verified with the measurement of four
vessels in two places. In order to have a more statistically significant support, more
data of this type are required.
• Transmission loss has been identified as the biggest source of uncertainty. The
measurement procedure developed here introduces important novelties with regard
to the computation of this topic and similarly, what is probably more important,
estimates the uncertainty associated to it. Anyway, deeper studies of how to model
the sea bottom, how to deal with a moving source and possible different models of
the source may be beneficial to develop a way of computing the transmission loss
more accurately.
• One of the studies carried out in this deliverable suggests that the current
measurement layouts (bottom supported and floating configuration) are not able to
measure below 10Hz, which makes impossible to measure the blade pass frequency
for certain kind of vessels. Moreover, rough sea states (3 or higher) makes difficult to
measure in the low frequency range (below 40Hz) in the case of silent vessels in
compliance with the ICES Nº 209. The design of new measurement devices
insensitive to the sea state as well as developing a procedure to measure more
closely the vessel (near field) should be studied to overcome this problem. The study
of the last topic will imply to develop source models accounting for the characteristics
of the vessel as a source in the near field.
• Other possible solution to deal with the problem previously described may be the
introduction of the concept of distance of acoustic invisibility, which is the maximum
distance at which the vessel can be “listened” for a certain frequency range. That is,
the maximum distance where the sound pressure level measured in this frequency
range when the vessel goes by at that distance is sufficiently above the background
noise measured by the same recording system. This concept, among other things,
would permit to assess the goodness of the signature of the vessel in the low
frequency range, which is really difficult to measure at the large distances required by
the current and new procedure as seen in previous paragraphs of this deliverable.

WP 3 - Measurements
Method
References
[1] NATO Standardization Agreement N°1136 (STANAG), “Standards for use when
measuring and reporting radiated noise characteristics of surface ships,
submarines, helicopters, etc. in relation to sonar detection and torpedo risk, May 29,
1995.
[2] AMERICAN NATIONAL STANDARD - Quantities and Procedures for Description

and Measurement of Underwater Sound from Ships – Part 1: General
Requirements. ANSI-ASA S12.64-2009/Part1
[3] ISO/DPAS 17208-1 "Acoustics – Quantities and procedures for description and
measurement of underwater sound from ships – Part 1: General requirements"
[4] Ainslie, M.A., TNO-DV 2011 C235 - Standard for measurement and monitoring of
underwater noise, Part I: physical quantities and their units
[5] Madsen 2005 (JASA)
[6] ANSI 1994 (Acoustical Terminology)
[7] Hazelwood, R.A. and Connelly, J. “Estimation of underwater noise – a simplified

method.” Int. J. Soc. Underwater Tech., Vol. 26, No. 3, 2005, pages 51 – 57.
[8] Michael Bathiarian, M., A standard for the measurement of underwater noise.
NOAA Vessel Quieting Symposium.
[9] DNV. Part 6, Chapter 24 - Silent Class Notation. Rules for Classification of Ships.
s.l. : DNV, 2010
[10] Gloza I., 2008. Transmission of low frequency sounds from ships into water
environment. Acoustics 08.
[11] Alzir, C., Marin-Curtoud P., Correction factor for radiated noise measurement in
shallow waters. Undersea Defense Technology Conference, Hamburg, 2006.
[12] Bong-Chae Kim et al. Measurement of low-frequency underwater noise by a self-

recording hydrophone.
[13] T. Gaggero, M. Van Der Schaar, E. Rizzuto and M. Andre. Ship underwater noise
emissions: Uncertainties in the measurements and in the effects on the marine
environment.
[14] J.R. Vidal Bosch. Teoría de olas y comportamiento del buque del mar.
[15] R.J. Urick. Principles of Underwater Sound. Peninsula Publishing, New York, 1983.
[16] F. Jensen, W. Kuperman, M.Porter. Computational Ocean Acoustics. American

Institute of Physics, New York 2000.

WP 3 - Measurements
Method
[17] Measurement of underwater noise arising from marine aggregate dredging

operations. Final Report.
[18] Jason I. Gobat Reducing Mechanical and flow-induced noise in the surface
suspended acoustic receiver.
[19] Deliverable 2.1.2 - On site Measurements (SILENV Project).
[20] Xavier Lurton. An introduction to Underwater Acoustics. Principles and applications.
[21] International Association of Oil and Gas producers. Fundamentals of underwater

sound. Report nº 406. May 2008.
[22] L. Brekhovkikh, Y. Lisanov, Fundamentals of Ocean Acoustics. Springer-Verlag,

1982.
[23] W.S. Burdic, Underwater acoustic system analysis. Prentice-Hall Signal Processing
Series, 1984.
[24] Mark V. Trevorrow, Boris Vasiliev & Svein Vagle. Directionality and maneuvering
effects on a surface ship underwater acoustic signature.
[25] Ainslie M. A., Springer-Praxis, Chichester. Principles of Sonar Performance

Modelling (2010).
[26] ICES Cooperative Research Report Nº 209. Underwater noise of research vessels,
review and recommendations. ISSN 1017-6195, May 1995.
[27] Deliverable 5.2. - Noise & Vibration label proposal (SILENV Project).
[28] MARINE STRATEGY FRAMEWORK DIRECTIVE, Task Group 11 Report,

Underwater noise and other forms of energy, APRIL 2010, M.L. Tasker, M.
Amundin, M. André, A. Hawkins, W. Lang, T. Merck, A. Scholik-Schlomer,
J.Teilmann, F. Thomsen, S. Werner & M. Zakharia. Prepared under the
Administrative Arrangement between JRC and DG ENV (no 31210 –2009/2010),
the Memorandum of Understanding between the European Commission and ICES
managed by DG MARE, and JRC’s own institutional funding Walree.

WP 3 - Measurements
Method
A. ANNEX A - EUROPEAN WATERS ENVIRONMENTAL DATA

The purpose of this annex is to present some general environmental data in relationship with underwater acoustics, and a few examples in
order to support the analysis of the present study. Data is split into general data (sea bottom profile and characteristics, meteorological data,
etc…), and data directly related to acoustic propagation and measurement issues (profile of speed of sound along depth, ambient noise). All
information shown here must be considered as statistically averaged values, as data can vary significantly depending on local weather
conditions and sea state.
A.1 GENERAL DATA
Figure A-1 shows a contour plot of sea bottom level. It allows distinguishing shallow waters and deep waters, and the geographical limit of
continental shelf. Isovalues for the contour plots are: 50 m, 100 m, 500 m, 1000 m, 2000 m, 2500 m, 3000 m, 3500 m, 4000 m and 4500 m. If
we take 100 m as the limit for shallow waters, we can see that they are mainly located in the English Channel, the North Sea and the Baltic
Sea. Figure A-1 shows also two sample locations (red dots) where acoustical data is available, one for deep waters, one for shallow waters.
Figure A-2 shows the characteristics of seabed. Rock is in red colour, Gravel or small stones generally in green or brown colour. Sand is shown
generally in orange or yellow colour. Other cases refer to different kinds of mud. In most cases, deep waters seabed is muddy, and shallow
waters seabed composed of sand or gravel.
Figures A-3 shows an example of map of water temperature at sea surface. This data is of importance regarding heterogeneity of water masses
and sound propagation in the ocean.
Figure A-4 gives an average density of ships. This data can be related to sea ambient noise regarding traffic.
Other data, not shown here, could be displayed, such as average wind speed, sea currents (permanent or tidal), etc.

WP 3 - Measurements
Method
Figure A-1.-Sea bottom level.-

WP 3 - Measurements
Method
Figure A-2.-Geological seabed map, colour corresponding to different types of seabed (source : http://jncc.defra.gov.uk/euseamap).

WP 3 - Measurements
Method
Figure A-3.-Sea surface water temperature acquired on 24 May 2006 (source : http://www.esa.int/spaceinimages).-

WP 3 - Measurements
Method
Figure A-4.- Density map of ship traffic based on AIS data from mid-December 2012 to mid-January 2013. The colour scale is logarithmic and represents the
2
average number of ships present in a 1km cell at any time. Source Quiet-Oceans based on Marine Traffic website.-

WP 3 - Measurements
Method
A.2 ACOUSTICAL DATA
Data will be shown here for the two sample locations identified on Figure A-1, one for deep waters, and one for shallow waters.
Acoustic propagation at sea depends greatly on sound speed profile along depth. Speed of sound depends mainly on salinity, depth, and
temperature. As temperature varies along time, the sound speed profile varies also. Then, it is important to get this information for different
periods of the year. Figure A-5 gives the information (average, minimum, maximum) for the two sample locations and three months in the year
(March, August and November). Note the variability for the shallow waters.
Ambient noise level along frequency is key information when performing acoustic measurements at sea, the quality of the result depending on
signal to noise ratio. Figure A-6 gives the isotropic equivalent ambient noise spectral level (dB ref. µPa2/Hz) as a function of frequency, for the
two sample locations. Isotropic equivalent noise means the noise sensed at a given location by an omnidirectional hydrophone. Noise level for
shallow waters is higher in the low frequency range because it is affected by ship traffic noise.

WP 3 - Measurements
Method
Figure A-5.-Sound speed profiles.-

WP 3 - Measurements
Method
Figure A-6.-Ambient noise level as a function of frequency.-

WP 3 - Measurements
Method
A.3 PROPAGATION LOSSES
A computer model has been used to predict propagation losses (in dB) as a function of depth and range at a given frequency (1000 Hz), for
different environments presented previously. Note that the scale for depth and range are not the same. The model uses ray theory and takes
into account acoustic reflection on sea surface and seabed. In each case, two figures are shown, one for large propagation distances
(commonly used for sonar system performance studies) and one for smaller distances (more relevant for URN measurement purposes). The
depth of the sound source is 5 meters.
For the deep waters case in summer (Figure A-7), the acoustic rays tend to plunge towards sea bottom, and a convergence zone appears. For
low distances, sound propagates regularly with a spherical spreading behaviour. In winter (Figure A-8), the propagation loss pattern is more
complicated due to the shape of the sound speed profile. On the other hand, at short distances, propagation remains similar.
For the shallow waters case (Figures A-9 and A-10), multiple reflections on sea surface and bottom change strongly the propagation losses by
comparison to the deep waters case. Propagation loss is smaller close to the surface. We recall that here the frequency is 1000 Hz. At lower
frequencies, another model should be used, for example based on normal modes. At shorter distances, the acoustic pressure tends to be
constant on a vertical line, which is characteristic of cylindrical propagation instead of spherical propagation. This effect is more visible on
Figure A-10).
In some cases, a layer of water close to the surface has a speed of sound slightly smaller than at greater depths. Then, a surface propagation
duct appears, with a concentration of acoustic waves and higher sound levels. This effect is shown on Figure A-11 with the limit of the duct at
30 m depth.

WP 3 - Measurements
Method
Figure A-7.-Propagation losses for location N°1 in summer at 10 00 Hz.-

WP 3 - Measurements
Method
Figure A-8.-Propagation losses for location N°1 in winter at 10 00 Hz.-

WP 3 - Measurements
Method
Figure A-9.-Propagation losses for location N°2 in summer at 10 00 Hz.-

WP 3 - Measurements
Method
Figure A-10.-Propagation losses for location N°2 in winter at 10 00 Hz.-

WP 3 - Measurements
Method
Figure A-11.-Propagation losses in shallow waters at 1000 Hz with a surface propagation duct

AQUO - D3.1 European URN Standard Measurement Method - Rev2

Uploaded by

Copyright:

Available Formats

AQUO - D3.1 European URN Standard Measurement Method - Rev2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AQUO - D3.1 European URN Standard Measurement Method - Rev2

Uploaded by

Copyright:

Available Formats

Achieve QUieter Oceans by shipping noise footprint reduction

FP7 - Collaborative Project n° 314227

European URN Standard Measurement Method

Deliverable Reference Issue Dissemination status

yes no D 3.1 Rev 2 PU

File name D3.1 European URN Standard Measurement Method_Rev2.Doc

Name Organisation Function Date

Issued by Alfonso Moreno TSI Task leader 21/04/2014

Checked by Publio Beltrán TSI WP leader 21/04/2014

Approved by Christian Audoly DCNS Coordinator 23/062014

ISSUE DATE NAME ORGANISATION MODIFICATIONS

Rev 1 21-4-2014 Publio Beltrán TSI Verification

Dissemination level: PUBLIC 2

© AQUO Project Consortium 2012 - all rights reserved

Chapter 1 1.1 Christian Audoly DCNS

3.4.3 Jan Hallender SSPA

Dissemination level: PUBLIC 3

© AQUO Project Consortium 2012 - all rights reserved

CHAPTER SECTION NAME ORGANISATION

Dissemination level: PUBLIC 4

© AQUO Project Consortium 2012 - all rights reserved

1.1. AIM AND SCOPE ................................................................................................................................. 11

1.2. METHODOLOGY ................................................................................................................................ 11

1.3. NEEDS .................................................................................................................................................... 12

3. REVIEW OF EXISTING STANDARDS AND PROCEDURES........................... 18

3.1. KEY PARAMETERS ............................................................................................................................ 18

3.2. LIST OF EXISTING STANDARDS AND PROCEDURES CONSIDERED................................... 19

3.2.1. STANAG 1136 ................................................................................................................... 19

3.2.2. ANSI-ASA S12.64-2009/PART 1 AND ISO DOCUMENTS IN PROGRESS .................... 20

3.2.3. DNV SILENT CLASS NOTATION. .................................................................................... 22

3.3. SYNTHESIS OF THE EXISTING REFERENCE DOCUMENTS .................................................. 23

3.3.1. INITIAL CONSIDERATIONS. ............................................................................................ 23

3.3.2. HOW THE “KEY PARAMETERS” ARE TACKLED IN THE STANDARDS....................... 24

3.3.3. THE EXISTING UNCERTAINTIES IN THE STANDARDS. .............................................. 27

3.3.4. CONCLUSIONS AND RECOMMENDATIONS ................................................................. 30

3.4. PARAMETER EVALUATION AND IDENTIFICATION OF DEFICIENCIES IN THE

3.4.1. DEVICE ............................................................................................................................. 31

Dissemination level: PUBLIC 5

© AQUO Project Consortium 2012 - all rights reserved

3.4.2. ENVIRONMENT ................................................................................................................ 55

3.4.3. VESSEL PARTICULARS .................................................................................................. 70

4. PROPOSAL FOR A NEW URN MEASUREMENT PROCEDURE .................... 79

4.2. SUMMARY OF THE REQUIREMENTS........................................................................................... 80

4.3. INSTRUMENTATION ......................................................................................................................... 82

4.3.1. HYDROPHONES, SIGNAL CONDITIONING AND RECORDER ..................................... 82

4.3.2. DISTANCE MEASUREMENT SYSTEM ........................................................................... 82

4.4. TEST SITE ............................................................................................................................................. 83

4.4.1. TEST SITE REQUIREMENTS .......................................................................................... 83

4.4.2. SEA SURFACE CONDITIONS.......................................................................................... 83

4.5. HYDROPHONE DEPLOYMENT ....................................................................................................... 83

4.5.1. GRADE A1......................................................................................................................... 83

4.5.2. GRADE A2......................................................................................................................... 84

4.5.3. GRADE B1......................................................................................................................... 84

4.5.4. GRADE B2......................................................................................................................... 84

4.6. TRANSMISSION LOSS ....................................................................................................................... 85

4.6.1. MODELLING ACTIVITIES................................................................................................. 85

4.6.2. SIMPLE LAW ..................................................................................................................... 87

4.7. URN MEASUREMENT ........................................................................................................................ 88