770 Aa

REPORT
BSEE Offshore Wind Recommendations

GUIDELINES FOR STRUCTURAL HEALTH MONITORING FOR OFFSHORE WIND TURBINE
TOWERS & FOUNDATIONS
DOC.NO. 20160190 16-1036

REV.NO. 3 / 2017-06-15
Neither the confidentiality nor the integrity of this document
can be guaranteed following electronic transmission. The
addressee should consider this risk and take full responsibility
for use of this document.
This document shall not be used in parts, or for other purposes

than the document was prepared for. The document shall not
be copied, in parts or in whole, or be given to a third party
without the owner’s consent. No changes to the document
shall be made without consent from NGI.
Ved elektronisk overføring kan ikke konfidensialiteten eller

autentisiteten av dette dokumentet garanteres. Adressaten
bør vurdere denne risikoen og ta fullt ansvar for bruk av dette
dokumentet.
Dokumentet skal ikke benyttes i utdrag eller til andre formål

enn det dokumentet omhandler. Dokumentet må ikke
reproduseres eller leveres til tredjemann uten eiers samtykke.
Dokumentet må ikke endres uten samtykke fra NGI.
Project
Project title: 20160190 BSEE Offshore Wind Recommendations
Document title: Guidelines for Structural Health Monitoring for Offshore Wind
Turbine Towers & Foundations
Document no.: 16-1036
Date: 2017-06-09
Revision no. /rev. date: 3 / 2017-06-15
Client
Client: BSEE Environmental Enforcement Division
Client contact person: Altan Aydin
Contract reference: Contract No. E16PC00006, Requisition 0040246879
for NGI
Project manager: Per Magnus Sparrevik
Prepared by: Per Magnus Sparrevik
Reviewed by: Amir Rahim
NGI Inc. T +1 281 752 4667 NGI Inc. NGI Oslo

WWW.NGI.NO/HOUSTON [email protected] 2000 S. Dairy Ashford, Ste 322 PO Box 3930 Ullevaal Stadion
Houston, TX 77077, U.S.A. NO-0806 Oslo, Norway
c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

Date: 2017-06-15
Rev.no.: 3
Page: 4
Abstract
The value in a Structural Health Monitoring (SHM) program is determined by the use
and the benefits that can be obtained by the data. The monitoring system is only a tool
and an investment to gather the data. The instrumentation plan should be based on the SHM
data that is required for verification and investigation of uncertainties in design, as well as
provide input for possible future design optimization. Other monitoring objectives may be
early warning of progressing degradation that allows for condition based maintenance and
status evaluation for possible life time extension or changed operational conditions (for
example turbine modifications).
Reduced total costs and added value to the data can be achieved by applying an integrated
SHM system approach instead of installing independent and specialized monitoring systems
(geotechnical, structural, corrosion, met ocean, etc.). It is important that the set-up of
instruments is configured such that all linked/correlated parameters are recorded.
The stakeholders must define WHY instrumentation is needed and HOW the collected data
can be used. These aspects are discussed in the report which describes the purpose of
integrated monitoring and specific monitoring aspects for different types of Offshore Wind
Turbine (OWT) foundations. The following foundation alternatives are included in the
study: mono piles, mono buckets, gravity base, jacket/tripod with piles or buckets and
moored floaters.
Offshore SHM monitoring can be challenging both with respect to rough offshore
conditions and costs. Experience is important when selecting the appropriate monitoring
solutions and hardware. Recommendations and examples of monitoring set up for the
defined parameters of interest to monitor are thoroughly described in the report. Finally,
practical advices for execution of an offshore SHM project are given in terms of planning
and design as well as practical installation work and how to make the instrumentation
surviving as long as possible in a rough offshore environment.

Date: 2017-06-15
Rev.no.: 3
Page: 5
Rough environmental conditions can be expected at many wind farm locations

Planning and configuration of the instrumentation set-up should be done at an early stage
and incorporated into the structural design for simplified and cost effective installation
during fabrication of the structure. Offshore retrofit is difficult and costly. If mobile
instrumentation is relevant, preparations should be done at the yard in order to simplify
offshore work.
Accessibility for installation of instrument is usually better at the yard than in the field

Date: 2017-06-15
Rev.no.: 3
Page: 6
Contents
1 Introduction 8
1.1 Definitions 9
2 Overview of areas relevant to monitor 9
2.1 Types of OWT foundations 10
2.2 Corrosion 11
2.3 Structural integrity and response 11
2.4 Geotechnical and Soil-Structure Interaction (SSI) 12
2.5 Correlated parameters 12
2.6 Installation Baseline 12
3 Integrated Monitoring 13
3.1 General considerations for an integrated SHM system 13
3.2 The value of integrated monitoring 16
3.3 Limitations and failures in a monitoring program 18
3.4 Reasonable extent of SHM monitoring 19
3.5 An extreme event on tape 19
4 Specific SHM objectives for OWT foundations 22
4.1 Objectives for Foundation Response Monitoring in the field 22
4.2 Specific monitoring aspects for Mono piles 26
4.3 Specific monitoring aspects for Mono bucket foundations 42
4.4 Specific monitoring aspects for Gravity base foundations 43
4.5 Specific monitoring aspects for Piled jackets or tripods 43
4.6 Specific monitoring aspects for tripods/jackets with bucket foundations 47
4.7 Specific monitoring aspects for floating OWT foundations 48
4.8 Summary of recommended monitoring parameters 53
5 SHM instrumentation solutions for OWT structures 55
5.1 Wave height 56
5.2 Current 58
5.3 Scour monitoring 59
5.4 Dynamic motion monitoring 63
5.5 Tilt - Verticality of the tower 67
5.6 Structural strain, stress and fatigue 69
5.7 Pore pressure monitoring 82
5.8 Earth pressure (effective stress) 86
5.9 Static displacement and settlement 87
5.10 Mooring loads 91
5.11 Monitoring TLP anchors 95
5.12 Corrosion 98
6 Data acquisition system and interface electronics 112
6.1 DAQ system 113
6.2 Signal transmission 117
7 Practical advices for SHM instrumentation of OWT foundations 118

Date: 2017-06-15
Rev.no.: 3
Page: 7
7.1 Introduction 118

7.2 “Ten Commandments” for offshore and subsea instrumentation 120
7.3 Executing a larger SHM project 132
7.4 Guidelines for selection of sensors 132
7.5 Guidelines for purchase and calibration of sensor systems 135
7.6 Instrument corrosion and preservation 138
7.7 Pressure vessels and pressure seals 141
7.8 Connectors and cables 144
7.9 Electric system design 147
7.10 Quality control and documentation 150
7.11 Offshore and Subsea installation considerations 152
8 Standards and guidelines 157
9 References 157
Review and reference page

Date: 2017-06-15
Rev.no.: 3
Page: 8
1 Introduction
Long term Structural Health Monitoring (SHM) is an important tool for reducing risks and
costs in development and management of offshore wind farms. An Offshore Wind Turbine
(OWT) may not be a gigantic structure. However, systematic flaws or problems can be very
expensive as a wind farm normally contains a large amount of OWT's, see Figure 1-1.
Consequently, potential savings can be big if the design can be improved and optimized.
Figure 1-1 Presently the London Array wind farm located in the North Sea outside Kent is the largest
in Europe with 175 OWT's producing 630 MW of power
Three main objectives are relevant for SHM:

Design improvement (verification of novel design or optimization of existing
designs)
Early warning of degradation and condition based maintenance
Possible extension of operational life span or changed design conditions
A well implemented an integrated monitoring approach can continuously track the "state of
health" of a foundation and identify in a very early stage the onset of potential structural
problems. This allows for early, less costly repairs and optimal preparation of the offshore
operations, as well as follow-up of the repair measures themselves (rectification effect). As
such, a significant risk reduction can be achieved over 20 to 25 operational years, the
foreseen life span of a wind farm.

Date: 2017-06-15
Rev.no.: 3
Page: 9
1.1 Definitions
An Offshore Wind Turbine (OWT) includes the complete structure with Rotor and Nacelle
Assembly (RNA), Tower, Substructure and seabed foundation, see Figure 1-2. The
"Foundation" includes the sub-structure below the base of the tower (Transition Piece,
Jacket/Tripod or Gravity base) and the seabed foundation (piles, buckets or raft foundation).
Figure 1-2 Parts of an Offshore Wind Turbine (OWT)
2 Overview of areas relevant to monitor

A proper remote monitoring scheme begins with identifying those components and/or
phenomena that needs to be monitored for a certain structure at certain conditions. Other
criterions for decision making about investment in an SHM system are the risks related to
potential issues with structural components, the amount of information that can be acquired
by instrumentation and the benefits from understanding the monitored structural response,
for example in conjunction with design verification.

Date: 2017-06-15
Rev.no.: 3
Page: 10
2.1 Types of OWT foundations

Presently the main alternatives for offshore wind turbine foundations can be divided into
four categories:
1. Mono piles (or monopods if suction bucket is used) which normally used down to
water depths of 30-40m
2. Gravity base structures with skirts, normally used from 25-60m
3. Jackets or tripods, piled, pre-piled or with suction buckets, normally used from 25-
60m
4. Floating foundations, normally relevant from 50m and deeper
These four foundation categories are illustrated in Figure 2-1.
Figure 2-1 Main alternatives for Offshore Wind Turbine (OWT) foundations
Specific soil or environmental conditions, certain design issues and critical structural
aspects may apply for the different types of foundations. Follow up of specific parameters
by long term monitoring in the field can be beneficial (or even required). Thus, the optimal
monitoring configuration will vary from site to site. The tower is more or less similar for all
types of foundation. However, the response of the combined tower and foundation system
may be different. The following areas and issues relevant to monitor must be considered as
general recommendations with the specific relevance evaluated from case to case.

Date: 2017-06-15
Rev.no.: 3
Page: 11
2.2 Corrosion
Corrosion is a major issue with respect to the operation life of an OWT foundation. The
following phenomena are of interest, both for immediate follow-up and with regard to their
evolution in time. These apply for more or less all steel structures, but especially mono piles
(see also section 4.2.5):
Degradation of organic coatings at the exterior part of the foundation, such as splash
zone and higher parts, as this represents the onset of corrosion damage and the
associated material loss.
The degree of protection offered by the Cathodic Protection (CP) system in place at
the foundation, as overprotection may result in hydrogen embrittlement, thus
structural issues, while insufficient protection may lead to unforeseen corrosion
activity.
The occurrence of specific, mainly localized forms of corrosion such as pitting,
water-level corrosion, microbiological corrosion (MIC), etc., as they give rise to
stress concentrations or other structural problems. Better solutions to monitor
mudline MIC on piles are in demand as inspection is difficult.
Composition of isolated bodies from water as this has a significant influence on the
type of corrosion damage that can occur and also on how to deal with it.
2.3 Structural integrity and response

Structural integrity is important for most foundations, the following parameters may be of
key interest to monitor:
Natural resonant frequencies of the structure, mainly regarding their relation to the
frequencies of ambient excitations such as wind, waves, etc.
Damping characteristics in the different vibrational modes of the structure, to verify
with design assumptions. Especially important for mono piles
Mode shapes and frequencies of the tower + foundation system
The strain levels at those locations experiencing the highest stress concentrations
(onset of permanent deformation). Design dependent.
Strain (static as well as dynamic) in bolts in new foundation types or flanges
subjected to potential damage during installation.
RMS amplitude of vibrations, also connected to the vibrational modes
Fatigue cycles on vulnerable locations, mainly welds.
Stress field in the foundation and/ or transition pieces, including the resulting
deformations and the evolution thereof.
All parameters above are linked with directions of the loads (wind and waves) which
should be monitored as well and location of secondary steel structures.
For floating structures monitoring of the mooring loads and dynamic motion of the
floater are important in order to understand the response to waves and wind.

Date: 2017-06-15
Rev.no.: 3
Page: 12
Further structural issues deserving special attention:

The integrity of the grout in grouted connections, transition piece connection for
mono pile foundations and pile connection for pre-piled jacket foundations.
Relative displacement in these connections
Inclination of the tower.
2.4 Geotechnical and Soil-Structure Interaction (SSI)
Scour around the foundation and the evolution thereof.
Seabed stiffness, especially for large mono piles in deeper waters
For bucket or gravity based foundations the pore pressure response to loading is
important as well as the effective stresses
2.5 Correlated parameters

For a correct, unambiguous interpretation of the obtained data it is also essential to record
metocean data such as wave height, tide, current, wind speed and direction as well as turbine
data such as, RPM, pitch and yaw angle. For corrosion monitoring the environmental
parameters inside and outside the foundation are also important to monitor (oxygen levels:
gas phase and dissolved in entrapped water, pH, electric potentials, temperature, water
levels. etc.). For monitoring of dynamic response and tracking of Eigen frequencies, scour
monitoring may be relevant (or vice versa) as scour will affect the stiffness of mono pile
foundations. All linked parameters must be thoroughly considered when setting up a
monitoring program.
2.6 Installation Baseline

Finally, but not the least, monitored data during installation of the foundation in the field
provides a very important baseline for interpretation of subsequently recorded SHM data.
Installation data may be essential for calibration/offsetting of permanent SHM sensors and
better understanding of the long term behavior of the structure recorded by the SHM system.
The SHM recordings in the field should start as soon as possible. If possible, the SHM
system or parts of it should be in operation also during installation of the structure. The
initial structural response recordings under multiple conditions are very important for datum
readings and comparison with future measurements in order to ascertain the possibility or
nature (progress) of damage and/or potential life extension.

Date: 2017-06-15
Rev.no.: 3
Page: 13
3 Integrated Monitoring
For each of the phenomena described, a dedicated monitoring setup containing a specific
set of sensors can be designed and installed. However, an intelligent combination of sensors
can result in an integrated follow-up of almost all these phenomena at once. Moreover,
isolated monitoring setups next to each other cause loss of value in many other domains.
Examples include:
The reduced ability for an efficient remote follow-up of the monitoring units
themselves. Timing synchronization issues making it difficult to correlate the data
recorded by different systems with each other.
Overlap in recorded data, resulting in an enhanced consumption of bandwidth on the
fiber optic link and storage space.
Multiple contracts and points of contacts, resulting in an administrative overload.
Difficulties with scaling, e.g. when expanding an existing setup using additional
sensors.
Inefficient management of backup power and backup storage due to the overlap of
remote ‘logic units’.
Turbine and meteorological data is normally included as standard output included in the
Supervisory Control and Data Acquisition (SCADA) system as part of the turbine supply.
The other parameters should be monitored by an integrated SHM system. Thus, monitoring
the foundation and the turbine are normally performed with two different systems, however
with the ability to share data. Data integration can be done offshore using a common data
link to shore or transmitted separately if sufficient transmission capacity is available (for
example with fiber optic communication cables)
3.1 General considerations for an integrated SHM system

In order to obtain real and advanced information on the state of the offshore structures, the
acquired data needs to be of sufficiently high quality (reliable, representative, precision,
resolution, sampling rate, noise levels etc.). To obtain high quality data to the end user,
multiple aspects (described in sections 3.1.1 to 3.1.8) must be taken into account in the
specification and supply of an integrated SHM system.
3.1.1 Instrumentation plan

The instrumentation plan includes selecting the required sensors and where to locate them.
This needs to be done in close collaboration with the designer and/or developer as decisions
taken in the design process influence the decision on where to monitor and how frequently.

Date: 2017-06-15
Rev.no.: 3
Page: 14
Next to the sensor selection, the instrumentation plan is governing for the architecture of
the monitoring units themselves: what kind of data transfer to use, local backup power and
backup storage, backup communication, remote management, timing of acquisition, etc.
3.1.2 Sensors
Next step is to select the exact sensors to use. Next to their functionality, the choice is guided
by sampling interval, accuracy required, environmental conditions the sensors need to
operate in, susceptibility for EM noise, expected operational life, ease of replacement, ease
of maintenance, stability, etc.
3.1.3 Sensor interface

The choice of the logic unit is guided by many of the choices made during the Design and
Sensor phase. The type of sensors (digital, analog 0-10V or 4-20mA), the number of sensors,
their location, the accuracy required, all guide the blueprint of the logic unit. Moreover, this
needs to be in line with the requirements put forward regarding data management, backup
power and storage as well as remote management.
3.1.4 Data logging

This covers logging of all data acquired offshore. Elements of relevance are backup storage
in case communication lines are disturbed, as well as initial treatments such as filtering,
down sampling and matching with predefined alarm levels.
3.1.5 Data transfer

All data is prepared for transfer to the onshore data storage (the next 2 elements in the chain).
The main data transfer is depending on the communication lines available. Usually fiber
optic cable links are available with less restrictions on the available bandwidth.
3.1.6 Data storage

The data storage is the place where data obtained at multiple foundations in the wind farm
will be stored for the long term. As a back-up, limited local data storage is recommended to
be available on board the OWT foundation. Long term storage is essential as some
phenomena will only occur after long (multiple years) periods of time. With respect to
storage, emphasis lies on preventing data loss, as well as on availability. This is coupled
with the next phase

Date: 2017-06-15
Rev.no.: 3
Page: 15
3.1.7 Reporting
In reporting, the acquired data is used to generate reports for use by the operator and
appointed parties. Reports can be monthly overviews of the situation, as well as more
profound analysis of cross-correlations. They cover both the evolution of one specific
structure and comparison between different structures. Long-term trends are reported, but
also relations with regards to predefined threshold values. If issues are detected, advanced
analysis based on long datasets can be required. This will involve sourcing long datasets
from the data storage. Another type of report is relevant if lifetime extension becomes
relevant, based on experience from the O&G industry long term SHM records has proven
to be important documentation for continued operation beyond the life span initially
anticipated in design.
An important part of the reporting is the evaluation of the data quality, this assessment
should preferable be done by the instrumentation specialist and before more advanced
structural analyses are done by other disciplines.
3.1.8 Alarms and Follow up

The alarms are the last link in the chain, and are to be generated in two physical locations:
one is on the structure itself, when threshold values are exceeded. The other location is the
onshore data storage, arising from a more profound analysis of acquired data. These
structure-based issues are however not the only use of alarms. Issues regarding data
acquisition, transfer and storage need to be brought to the attention as early as possible. This
should however not be the concern of the operator, so these alarms are directed towards the
provider/maintainer of the monitoring system such that the normal situation can be restored
as fast as possible.
This follow-up of the monitoring process itself thus becomes an inherent part of the
Observation and Measurement (O&M) process of the farm. In case of advanced monitoring,
the resulting knowledge will be used to guide the O&M activities (planning and contents),
and servicing the monitoring setups will be part of the O&M process itself. The amount of
additional transfers needs to be limited however by implementing advanced remote
management features and using almost maintenance-free sensors and monitoring hardware.
The approach proposed here is universal: foundation types, turbines and installation
methods may become more and more uniform, but these are not the only factors contributing
to the state-of-health of a foundation structure. Wind regimes (thus loads), temperatures,
water movement, soil type and soil movements are more difficult to control, as they are
inherently governed by nature. Also, variations in material properties, quality of
construction and components play a relevant role. All factors result in an inherent variability

Date: 2017-06-15
Rev.no.: 3
Page: 16
that only can be determined by implementing an integrated, intelligent continuous

monitoring system.
3.2 The value of integrated monitoring

Proper installation and commissioning of the sensors and monitoring setups are essential
parts for obtaining the desired information from an advanced monitoring package.
Therefore, dedicated training of the operator staff and a follow-up by continuous remote
(onshore) assistance is vital.
Depending on foundation type, location and experience of the operator, the layout of the
monitoring setup may differ. However, this should not result in additional costs as the same
concept applies and our setups all are based on interchangeable building blocks. A final
result of the approach offered is in maintaining state-of-the-art setups. The remote
management schemes implemented to the setups have made it possible to continuously
upgrade the setup from the shore. As such every setup located offshore can, at minimal cost,
be made more intelligent during the 20 or 25 years of operational life, based on the
knowledge gained during the long-term follow-up of the farm, as well as from advanced
analysis performed on the data as well as experience in other farms.
An example of a customized
monitoring set-up is given in Figure
3-1. In this case the primary
objectives for the SHM system was
to evaluate the in-place behavior of
the novel bucket foundations and to
monitor the strain and dynamic
response of the tripod jacket structure
in order to allow for further structural
design optimization.
Figure 3-1
Example of integrated SHM system
with 141 sensors implemented on Dong
Energy Wind Power's prototype Suction
Bucket Jacket BKR01 installed in 2014.
(SHM system provider: NGI)

Date: 2017-06-15
Rev.no.: 3
Page: 17
The integrated monitoring approach can allow for cost reductions in multiple ways as
follows:
As the sensors are monitored remotely, the amount of offshore inspection trips can
be reduced. As inspections provide only data in discrete time points, the information
content obtained is also limited, resulting in the need for additional trips for further
interpretation. Continuous monitoring results in continuous data sets. Interpretation
is much more straightforward and more trips offshore can be avoided
In reality, a limited number of inspections will still be needed to complement the
data obtained. Using the monitored data and knowledge of the setup, inspection trips
will however be more targeted and better prepared and can be guided from the
monitoring center to obtain the highest value possible.
Due to integrated monitoring and trend analysis, deviations of the behavior from the
predicted behavior will be detected in a very early stage. As multiple parameters are
monitored an accurate image of origin and result can be assembled. This allows
early, minor, thus relatively cheap interventions to be set up and stabilize the issue.
The same monitoring setup will be used in the follow-up stage to assess the long-
term performance and efficiency of the repair.
An advanced monitoring setup will allow making the transition from planned
maintenance to condition based maintenance. This will reduce the total cost of
maintenance operations, although condition based maintenance is more relevant for
the turbine compared to the foundation structure.
As deviations from the normal or expected behavior are captured in a very early
stage, chances are high that the situation can be restored to normal within a very
short time frame. This results in a significant reduction of the risk associated to the
foundation structures.
The same risk reduction is also a benefit in the financing phase. More and more
external, risk-averse sources of financing are sought in the offshore wind sector. As
proper monitoring significantly lowers the uncertainty regarding the long-term
structural integrity, this will be reflected in the premiums asked by external parties
to invest in the farms.
At a certain stage in the life of an offshore wind farm, the permitted period to operate
the park will come to an end. At that point, the operator may choose to apply for an
extension of the permit or for a new one, coupled with a repowering project. This
process will be based on an extensive report on the state of the foundation. Imagine
how much more straightforward this process can be based on 20 years of continuous
data and the associated knowledge, compared to having to set up a dedicated
inspection campaign offshore.
Last but not least: savings related to future constructions. As the advanced
monitoring leads to a profound knowledge of the in-operando behavior of the
foundations, the origin of unexpected phenomena can be fully analyzed and
subsequently remedied in future designs.

Date: 2017-06-15
Rev.no.: 3
Page: 18
By mean of continuous monitoring, the risk related to the management of offshore

foundation structures can be kept at a minimum for the foreseen life of 20 years and beyond.
In the design stage, all measures are taken to ensure this lifetime is reached, but in practice
the real state of health of the structures still comes with a degree of uncertainty. New
concepts, assumptions regarding soil conditions and loads as well as unforeseen events all
have an influence on how the structure or some of its key components degrade. This in turn
influences the real state of health of the foundations. The only way of having a continuous
records on the foundation's state of health together with its evolution in time is by equipping
it with an appropriate multi-sensor monitoring solution that continuously collects essential
data, but also translates these data into indicators quantifying the various aspects of the state
of health of the structure.
The structural integrity monitoring serves as a tool to guide O&M operations. Additionally,
the long-term monitored structural and operational data can aid in providing a lifetime
extension or repowering (a part) of the farm.
3.3 Limitations and failures in a monitoring program

A lot of arguments for integrated monitoring have been listed, but there are also limitations
and challenges to be considered. There are numerous examples of less successful SHM
projects where, in spite of the ambitions, the observational method could not be used. Some
of the reasons may be:
Faults (quality) and poor workmanship
Ambiguous data is recorded. Many times the reason for dubious data cannot be
linked to direct faults (sensors, cables, calibration etc.). It is unfortunately quite
common that the sensor/recording system works perfect but does not measure the
parameter of interest or is influenced by external conditions affecting the recorded
signal.
Unrealistic expectations. In-place monitoring cannot be compared with controlled
load/model testing. The loads are determined by Mother Nature and would rarely
get close to ULS (Ultimate Limit State) conditions and certainly not failure. The
recorded structural response will be significantly less compared to the design load
response and scaling up this information to ULS conditions may not be correct.
Some parameters such as soil stiffness are not linear and signal to noise ratio may
be poor for small strain signals. Noise and other errors will also be up scaled. For
small signals, it may be uncertain what phenomena are actuality measured. For
example is it displacement of the foundation or only local flexing of the steel?
Missing links. Recordings of all linked parameters are not included in the
monitoring scheme, for example monitoring strain without any information about
the loads does not make sense. Integrated monitoring thoughts and concept should
be implemented.

Date: 2017-06-15
Rev.no.: 3
Page: 19
How to avoid some of these shortcomings and failures are discussed in this document.
Figure 3-2 shows examples of rare ULS conditions which most likely never will occur,
Figure 3-2 Do not expect ULS events to be captured in the SHM data
3.4 Reasonable extent of SHM monitoring

The reasonable extent obviously depends on specific aspects such as challenging
environmental or geotechnical conditions, introduction of new technology or foundation
concepts, optimized design etc.
It is costly and not realistic to instrument every foundation in a wind farm with integrated
multi sensor SHM systems. Only spot checks and a few percent of the foundations can be
fully instrumented. For limited types of instrumentation mobile sensor systems allows for
periodic monitoring of several foundations by rotating the instrumentation between the
OWT's in the wind farm. If new concepts and design are introduced, a demonstrator OWT
foundation may be installed ahead of full implementation in a wind farm. Implementation
of an integrated SHM system on the demonstrator is important in order to maximize the
information which can be derived from the demonstrator project.
3.5 An extreme event on tape

The following example is a recording obtained from an integrated SHM system installed on
an offshore Oil & Gas platform. Statoil's "Draupner" platform was the first jacket in the
world with bucket foundations when installed in 1994 at 70 m water depth in the North Sea.
As Draupner was a prototype, an extensive monitoring campaign was executed in order to
verify the novel foundation design. At New Year's Day, half a year after installation, an
extreme event was recorded by the instruments on the unmanned platform.

Date: 2017-06-15
Rev.no.: 3
Page: 20
The weather was rough but not severe by North Sea standards when suddenly the down
looking wave radar registered one single wave with a peak of almost 20m above mean sea
level. The reading would probably have been dismissed as a noise spike in the readings if it
was not for the fact that the event was traced by almost all SHM sensors on the platform,
see Figure 3-3.
Figure 3-3 Some of the sensors readings during impact of the Monster wave to Statoil's Draupner
platform (data and illustrations: NGI)
All readings of significance had a logical correlation to the impact of the monster wave with
respect to structural response, the wave hit the top of the platform like a sledgehammer. The
overturning moment caused by the wave generated an over pressure response in the toe
foundation and suction pressure in the heel foundation. By comparing with the forces
distributed through the legs (recorded by strain gauges) it was concluded that most of the
"punch" by the wave was taken by the hydraulic reaction forces in the buckets.
Topside accelerometers and those on the buckets (seabed level) indicated severe vibrations
in the deck while the bucket foundations hardly moved. No permanent deformations (tilt)
was recorded by the inclinometers on deck after the incident. Those few seconds of recorded
data immediately proved the new concept, and also demonstrated that the capacity for

Date: 2017-06-15
Rev.no.: 3
Page: 21
transient loads of the bucket foundation was much higher than the design codes would allow
at that stage (the seabed consisted of dense sand).
The story is an excellent example demonstrating the value of an integrated SHM system.
However, recording such events can only be expected "once in a life time", in fact the loads
were higher than ULS conditions and would statistically be defined as a once in 10 000 year
event. These readings are also the only case were a rogue wave has been directly measured
and consequently proved the theory behind (quantum physics), rogue waves could no longer
be dismissed as fairy tales, see Figure 3-4.
Figure 3-4 The Draupner wave recorded by the on-board SHM system compared with quantum
physical modelling demonstrating how and that rouge waves suddenly can occur

Date: 2017-06-15
Rev.no.: 3
Page: 22
4 Specific SHM objectives for OWT foundations

The described objectives for structural health monitoring and instrumentation systems apply
for mono piles, jacket and gravity base foundations. Moored floating systems are also
discussed. Both piles and suction buckets are considered as relevant seabed foundation
alternatives. Presently piled foundations are the dominating alternative for OWT’s in
Europe. The suction bucket concept is not yet fully implemented in the Offshore Wind
Industry but frequently used by the offshore Oil and Gas industry. Monitoring aspects and
solutions for buckets (shallow foundations) also to a large extent apply for Gravity Base
Structures with skirts.
4.1 Objectives for Foundation Response Monitoring in the field

Although the scope in this report does not include structural or geotechnical analyses,
general design aspects which are expected to also apply for OWT foundations in US waters
are discussed as relevant for monitoring in the field. The main objectives for monitoring
foundation response is design verification and possible future optimization in addition to
early warning of possible development of weaknesses.
For mono piles, horizontal loads and moment are directly transferred to horizontal soil
reactions, as shown in the left-hand side of the Figure 4-1. As the pile is not fixed at the top,
it is free to rotate and translate. This horizontal load transfer usually dictates the pile length,
the pile must be driven long enough to mobilize enough soil over its length to transfer all
loads and prevent "toe-kick" displacement of the tip of the pile. It is however the lateral
stiffness of the pile and not bearing capacity which is governing for the design.
For multi-pile foundations (jackets and tripods), the overturning moment is mainly
transferred as axial loads (compression and tension) to opposing foundation piles as shown
in Figure 4-1. For this type of foundation, either the vertical bearing or pull-out capacity of
the piles is governing for the length of the piles. Similar distribution of loads the seabed
foundations applies for mono or multi-bucket OWT structures.

Date: 2017-06-15
Rev.no.: 3
Page: 23
Figure 4-1 Load transfer to the soil for a mono pile (left) and tripod/jacket (right) foundations
Similar overturning loads are also
transferred to the soil for gravity base
foundations, however the soil reactions
are different. Usually tension loads are
not present beneath the gravity base
foundation, the dominating loads are
compression and horizontal shear.
Figure 4-2 Load transfer to the seabed for

a gravity base foundation
Bucket foundations are also shallow foundations, however compared to a gravity base
structure, the bucket foundation have significant tensile capacity obtained by the suction
response inside the close bucket, especially for dynamic (short term) loads.
A floating wind turbine supports a large payload (wind turbine and tower) with large
aerodynamic loads high above the water surface and challenges basic naval architecture
principles due to the raised center of gravity and large overturning moment. The static and
dynamic stability criteria can be challenging to achieve as the hull weight must be limited
for a cost competitive foundation concept. The motion of the floating wind turbine is a
function of the restoring forces from the mooring and the environmental forces, see Figure
4-3. The dampening effect by the mooring lines has significant impact on the dynamic
response of the wind turbine, the stiffness of the anchors is negligible in this context.

Date: 2017-06-15
Rev.no.: 3
Page: 24
Figure 4-3 Catenary mooring loads and dynamic response of a floating OWT foundation
Independent of foundation type, an offshore wind turbine is a tall, slender with a heavy top
structure. Possible accumulated cyclic deformations or bending of the foundation will result
in a tilt of the tower which may become critical for operation of the wind turbine. Measuring
tilt (inclination) at the base of the tower is therefore recommended as a standard parameter
to be included in a monitoring scheme. In addition, the turbine will shut down if the dynamic
motions exceed the operational limits, the motions in the nacelle are monitored as part of
the SCADA system. For floaters, the static (mean) tilt is relevant for possible trimming by
ballasting/de-ballasting and/or mooring tensioning. The towers with nacelles are based on
similar structural arrangement as for land based wind turbines, although they are usually
much taller and bigger in offshore windfarms, see Figure 4-4.
Figure 4-4 Dimensions of Dong Energy's 7MW Offshore Wind Turbines considered for installation
at the Hornsea wind farm (UK)

Date: 2017-06-15
Rev.no.: 3
Page: 25
A main difference for offshore foundations compared to onshore wind turbines is the
dynamic loading and response of the foundation. Both tower and foundation must be
included in the modal analysis, see Figure 4-5. It is difficult to properly model the response
of the seabed in such analysis, field monitoring of the dynamic response is therefore
relevant.
Figure 4-5 Dynamic modelling of mode shapes for a jacket foundation with tower
Monitoring the environmental load conditions is relevant for all foundation types:
Wind and temperature (usually integrated as part of the turbine control system)
Wave height
Note that the extreme loads may not occur during extreme weather and for winds exceeding
the cut-out speed (typical 25m/sec), but may for example occur if big waves hit the
foundation just before cut-out wind speeds occur and/or in conjunction with an emergency
stop of the turbine. Figure 4-6 shows simulated wave action against a typical mono pile.

Date: 2017-06-15
Rev.no.: 3
Page: 26
Figure 4-6 Simulated wave action against an offshore wind turbine mono pile foundation
(animation by Kostack Studio)
4.2 Specific monitoring aspects for Mono piles

The failure modes for a mono pile foundation are illustrated in Figure 4-7. The ULS and
SLS failure conditions are usually not critical for design and therefore of less importance to
monitor. The exemption is Tilt of the tower which always is of critical concern. The
dynamic response and stiffness are the main design drivers and of paramount interest to
monitor. In addition, local and site related effects such as scour, corrosion and connection
of the Transition Piece can be relevant to monitor.
Figure 4-7 Failure modes for Mono piles (from L Aranya et al)

Date: 2017-06-15
Rev.no.: 3
Page: 27
4.2.1 Dynamic response

The dominating and critical loading condition for a mono pile foundation is the alternating
lateral loads at mudline, consequently the cyclic stiffness and damping of the foundation are
very important design parameters governing for the natural frequency of the OWT, see
Figure 4-8.
A natural frequency close to the rotor frequency may result in excessive vibrations that will
cause shut down of the turbine. The natural frequency of a mono pile is decreasing with
larger water depth and size of turbine, approaching the rotor revolution frequency. Thus, it
is of paramount interest to monitor the in-situ modal response of the complete foundation
and tower when challenging the allowable frequency limits. Even if the critical parameter
is the magnitude of vibrations shutting down the turbine, it is important to understand the
dynamic response of the complete structural system in order to optimize future design. This
includes monitoring the dynamic behavior (stress and motion) of the tower with turbine,
transition piece connection to the pile and the pile-soil stiffness (P-Y response).
Figure 4-8 Load frequencies for a typical mono pile OWT and simulation of dynamic radial
dampening in the soil
For mono pile wind turbines, the aerodynamic damping Daero is the main damping
contributor in the Fore-Aft (FA) direction due to the rotor blades. Daero plays only a very
little role in sideways (SS) motion were the soil damping plays an important role, see
Figures 4-9 and 4-10.

Date: 2017-06-15
Rev.no.: 3
Page: 28
Figure 4-9 Principal tower vibration modes for an OWT with mono pile foundation
Figure 4-10 Simulation of dynamic radial dampening in the soil (SS mode)

Date: 2017-06-15
Rev.no.: 3
Page: 29
The soil stiffness for a mono pile is traditionally derived based on API's P-Y approach, see
Figure 4-11. There are many indications (from various monitoring programs and lateral load
tests) that this approach is conservative, i.e. the foundation is stiffer than predicted according
to API guidelines. This is especially important for the large diameter mono piles. A stiffer
foundation will increase the fatigue life of the pile and (more important) keep the resonant
frequency above the rotor frequency (1P) when the water depth and turbine size are
increased.
Figure 4-11 P-Y stiffness approach from API (left) and a comparison of API predicted response with
measured (and stiffer) response measured during lateral pile load tests in stiff clay (source NGI)
The moment distribution used to determine the soil stiffness is monitored by strain gauges
along the embedded part of the pile, see Figure 4-12. The moment is derived from diametric
opposite strain gauges (compression and tension), normally along X-Y sections of the pile
such that the distribution of maximum bending moments can be computed for loading in
any direction.

Date: 2017-06-15
Rev.no.: 3
Page: 30
Figure 4-12 Example of mono pile instrumentation with strain gauges along the X-Y sections in the
embedded part, the tensile and compressive strain generated by the moment is measured along
the diametric opposite sections. Top right photo shows resistance wire gauges and lower right
photo is from a mono pile instrumented with FBG strain sensors. In both cases the embedded sensor
sections must be protected by channels with cover plates and driving shoes
The moment distribution derived from the strain gauge measurements (mean values)
recorded in the field can be plotted against a parameter representative for the applied load,
for example average wind speed. Based on the measured moment at each instrumented level
of the pile, the moment distribution with depth is plotted for different applied loads
(moments) and compared with predicted values. The moments in the pile diminish faster
with depth if the soil response is stiffer than predicted, see Figure 4-13.

Date: 2017-06-15
Rev.no.: 3
Page: 31
Figure 4-13 Moment measured at different depths of a mono pile and plotted against wind speed.
Note that the range of the Y (Moment) scales are decreased with depth. The inserted plot shows
measured (blue dots) and predicted moment distribution with depth for 10.5-11-5 m/s wind speed
where maximum moments are recorded. Data from Horns Rev, Dong Energy.

Date: 2017-06-15
Rev.no.: 3
Page: 32
As large dynamic motion of the

turbine is the main problem during
operation it is equally important to
monitor and understand the
dynamic response and modal
shapes of the tower itself using
accelerometers at different levels
in in order to track the FA and SS
modes, see Figure 4-14.
Figure 4-14
Configuration of accelerometers
along the mono pile tower (proposed
by OWI-lab)
Resonance frequencies can be tracked during operation in the field, see Figure 4-15 and
damping can be derived from transient loading for example over speed stop of turbine, see
Figure 4-16.
Figure 4-15 Tracking resonant frequencies of a large GBS foundation (Statoil's Troll C platform
through a storm. Data was compared to storm readings 7 year later in order to verify that the
stiffness of the foundation remained unchanged

Date: 2017-06-15
Rev.no.: 3
Page: 33
Figure 4-16 Damping of dynamic motions of Suction Bucket jacket after transient excitation
(example from NGI/Dong Energy)
Aerodynamic, hydrodynamic, structural damping

and, if used, tuned mass damping can be calculated
based on known material properties and dimensions.
The soil stiffness and dampening properties are as
described uncertain and more field data is needed in
order to improve empirical methods and correlations
to different soil conditions.
As monitoring on the pile itself (under water and

below the mudline) is difficult and costly, some
researchers claims that these measurement can be
omitted by means of a combination of modelling and
measurements on the tower (i.e. extrapolating data to
the mono pile), see Figure 4-17. The "virtual sensing"
concept is based on a range of assumptions and
empirical relations which in many cases probably are
representative. The uncertainty can however be
discussed for example if degradation occurs in the
grouted TP connection. It is also impossible to
evaluate difference stiffness of the soil in different
layers without strain measurements along the pile.
Figure 4-17 Virtual sensing concept as presented by OWI

Lab, extrapolating data from the tower to hotspots below
the mudline

Date: 2017-06-15
Rev.no.: 3
Page: 34
Evaluation of the lateral loading stiffness and dampening of the pile, after installation and
with time, is a paramount objective for monitoring. The instrumentation may include:
Dynamic motion (accelerometers) along the structure (above seabed)
Stress (strain measurements) in the pile at the TP connection and from mud-line and
downwards
If relevant in certain soil conditions, cyclic pore pressure build up (affecting the soil
stiffness) with piezometers along the pile
4.2.2 Scour
Scour is another feature which will affect the stiffness (and resonant frequency) of the mono
pile foundation, see Figure 4-18. Usually, lowering of the natural frequency due to scour is
more critical for operation of the wind turbine than the reduced bearing capacity. Thus,
unless scour protection is applied (or before possible scour protection is installed) in-situ
monitoring of possible scour development is relevant if the current and seabed conditions
are susceptible for scour. For shallow foundations (buckets) scour may also introduce a
bearing capacity or undermining problem, see Figure 4-19. Scour monitoring should be
complemented with current recordings in order understand the scouring process and add
value to the observations.
Figure 4-18 Measured scour development around a mono pile during large scale model testing
(from www.fzk.uni-hannover.de)

Date: 2017-06-15
Rev.no.: 3
Page: 35
Figure 4-19 Scour pattern after basin testing of a mono bucket foundation (HR Wallingford)
Scour may occur immediately after installation of the foundation in the seabed. It is
therefore important that the monitoring system is operational as soon as possible after
completed installation. Note that scouring is many times a cyclic process dependent on the
current direction. If the foundation is subjected to reversing tidal currents, scour may be
followed by re-sedimentation. Therefore continuous monitoring is recommended.

Date: 2017-06-15
Rev.no.: 3
Page: 36
4.2.3 Bearing capacity

The ultimate bearing capacity of a mono pile is
normally not a critical issue for design, it is the lateral
stiffness which usually is the design driver.
However, cyclic degradation and pore pressure build

may be an issue for impermeable soil conditions. The
pore pressure build-up during pile driving (or vibro
piling) and the subsequent dissipation (pile setup) time
may also be an issue for design and operation.
Figure 4-20 Excess pore pressures generated during driving and dissipation 6 months after
installation of a steel pile in seabed consisting of clay in conjunction with the Femern Large Scale
field tests (left). Deep rotation failure for a mono pile in clay modelled by NGI (right)
The radial effective stresses (earth pressures) may for some soil conditions be an important
parameter to monitor both with respect to set-up effects and lateral stiffness. It is however
difficult to obtain reliable lateral earth pressure readings for a driven steel pile. Total
pressure cells are used for these measurements and the data must be compensated
(subtracted) by the pore pressure measured at the same position such that effective stresses
can be determined. In order to obtain representative readings, it is absolutely essential that
the membrane of the total pressure cell is integrated with the pile wall with a curvature
corresponding to the pile diameter, which implies a customized design.
4.2.4 Grouted connections of the transition piece

The grouted connection is used to fix the Transition Piece (TP) to the mono pile as shown
in Figure 4-21. The transition piece is placed on top of the monopole and initially rests on
temporary supports. During installation, the temporary supports on the TP are jacked up to
correct the verticality of the transition piece before the grouting is carried out.

Date: 2017-06-15
Rev.no.: 3
Page: 37
After curing of the grout, the jacks are removed, leaving a gap between the temporary
supports and the top of the pile.
Figure 4-21 Old transition piece connection with temporary supports and straight grouted section
(illustration DNV-GL)
The large bending forces occurring at the grouted connection may eventually degrade the
grout (cracking and crushing), see Figure 4-22. If the TP settles (slip) down to the temporary
supports, a different force distribution than that intended at the design stage occurs between
the structural elements. Eventual force transfer through the temporary supports (not
designed for permanent loading) can be critical with respect to fatigue cracking in the
transition piece structure. This fundamental shortcoming in design was discovered in April
2010 when it was observed that the transition piece for some mono pile foundations had
slipped down by up to 25mm.
An extensive monitoring program was initiated in order to alert wind farm operators if the
strain in temporary support brackets exceeded acceptable limits. A Joint Industry Project
was executed by DNV in order to analyze the slippage problem and to improve on the design
practice. Reference is made to "The summary report from the JIP on the Capacity of grouted
connections in Offshore wind turbine structures DNV Report No: 2010-1053, rev. 05" which
can be downloaded from the internet.

Date: 2017-06-15
Rev.no.: 3
Page: 38
Figure 4-22 Structural behavior of the grout due to horizontal load transfer from the transition piece
to the pile (temporary supports not shown)
Accounting for the large dynamic bending moments on mono piles, a more robust design
has been developed using conical shaped connections. According to this concept, the pile
top and transition piece are fabricated with a small cone angle in the grouted section with
the narrower side pointing upwards. The coned connection will expose the grout to
compression loads, rather than shear forces as on traditional designs. Also, relative
deformations between the pile and transition piece will be reduced in case of grout failure
and prevent load transfer directly to the temporary supports.
The report: DNVGL-ST-0126 Support Structures for Wind Turbines includes the new
recommendation for the fixture of the transition piece using conical shaped connections, see
Figure 4-23. There are also other new solution for the TP-Pile connection in the market. As
this is an important structural part with a problematic history, new solutions are relevant to
monitor (until fully field proven).
Deformations between the pile and TP in the grouted connection can be monitored using
crack/joint meters (miniature extensometers) or long base strain gauges. The sensors can be
fixed between the top of the mono pile and the inner wall of the transition piece or across
the vertical gap between the temporary supports and the pile top, see Figure 4-24. The stress
in the grout may also be monitored by embedded load cells. Strain gauges in the transition
section of the pile and TP can also be used to monitor the load transfer.

Date: 2017-06-15
Rev.no.: 3
Page: 39
Figure 4-23 Recommended transition piece connection with conical grouted section. Based on
DNVGL-ST-0126
Figure 4-24 Example of extensometer configuration for monitoring relative deformations in 3D

between TP temporary supports and the pile (illustration NGI)

Date: 2017-06-15
Rev.no.: 3
Page: 40
4.2.5 Corrosion
Figure 4-25 External corrosion on a transition piece at Scroby Sands wind farm (UK)
Corrosion is a major factor limiting the operation life of offshore wind structures.
Refurbishment and repair can be very expensive if the corrosion is progressing without
control. For a mono pile the internal corrosion at sea level (transition piece) is the most
important aspect. Low uniform corrosion rates in a closed compartment are normally
anticipated in design. However, sea water ingress and supply of oxygen to the closed
compartment below the airtight deck have been detected for several old mono pile
foundations and caused accelerated rates of corrosion, see Figure 4-26.

Date: 2017-06-15
Rev.no.: 3
Page: 41
Figure 4-26 Closed compartment at Mono pile (MP) and Transition Piece (TP) joint (left) Observed
internal corrosion on Transition Piece and top of Mono Pile (right). Source: Force Technology
Corrosion mechanisms and problems relevant for mono piles are summarized as follows:
Degradation of organic coatings at the exterior part of the foundation, such as splash
zone and higher parts, as this represents the onset of corrosion damage and the
associated material loss.
The degree of protection offered by the Cathodic Protection (CP) system in place at
the foundation, as overprotection may result in hydrogen embrittlement, thus
structural issues, while insufficient protection may lead to unforeseen corrosion
activity.
The occurrence of specific, mainly localized forms of corrosion such as pitting,
water-level corrosion, microbiological corrosion (MIC), etc., as they give rise to
stress concentrations or other structural problems. Better solutions to monitor
mudline MIC on piles are in demand as inspection is difficult.
Composition of isolated bodies from water as this has a significant influence on the
type of corrosion damage that can occur and also on how to deal with it.
Inspections are necessary to evaluate the current corrosion state, prevailing mechanisms,
cause of changed conditions, and whether areas with risk of stress concentration, e.g. at
welds, are susceptible to corrosion fatigue. Monitoring campaigns increase the
understanding of the conditions under which wind turbine foundations must function in the
years to come and document effect of a given change. The gained knowledge should be
integrated in future designs, thus simple corrosion monitoring arrangement should be
installed at the time of construction and in operation directly after the foundation is installed
at the field.

Date: 2017-06-15
Rev.no.: 3
Page: 42
4.3 Specific monitoring aspects for Mono bucket foundations

The stiffness is also important for mono buckets foundations (monopods). However, for this
type of shallow monopod foundation it is the rotational stiffness (not lateral) in the soil
which is critical for the motions of the turbine, see Figure 4-27.
The general objectives for monitoring are similar as for a mono pile with the following
exceptions:
Strain in the bucket walls is not important for soil stiffness evaluation as lateral
bending of a shallow foundation is not an issue. From a structural point of view the
fixture between the bucket and the tower itself is subjected to large stresses that may
be relevant to monitor.
In combination with cyclic moment loads, a monopod bucket may be subjected to
cyclic settlements in conjunction with a progressing rotational failure. Thus, in
addition to monitoring the dynamic rotational motion of the foundation,
accumulated settlements are relevant to monitor.
Load distribution to the base (top lid) of the bucket foundation may be important for
some soil conditions in order to obtain the required overturning capacity and
rotational stiffness. Therefore under base grouting is normally used. If other
solutions are implemented to increase load transfer along the base to the seabed,
monitoring total stress to the seabed becomes relevant.
Pore pressures along the skirt walls are relevant to monitor in some soil conditions,
especially if the response to loading (drained, partially drained or undrained) is
uncertain. The pore pressure response close to skirt tip level is of primary interest.
Figure 4-27 Mono bucket OWT foundation and critical rotational failure mechanism in clay

Date: 2017-06-15
Rev.no.: 3
Page: 43
4.4 Specific monitoring aspects for Gravity base foundations

A gravity base foundation mainly relay on the weight to obtain sufficient over turning
stability. Usually the base of the foundation have shallow skirts.
Figure 4-28 Gravity base foundation from Seatower consisting of a steel transition piece with a
concrete gravity base
The monitoring aspects for a gravity base foundation (Figure 4-28) are in many perspectives
similar as for a mono bucket foundations. Soil contact stresses and pore pressures under the
base may even be more important to monitor as the bearing capacity of this type of
foundation to a large extent depends on the conditions at the base-seabed interface. Due to
the weight, consolidation settlements may also be relevant to monitor for certain soil
conditions (clay). In addition, the long term condition of the concrete structure (strain,
cracks etc.) is relevant to monitor, especially the joint between a steel transition piece and
the concrete base may be a "hot spot".
4.5 Specific monitoring aspects for Piled jackets or tripods

As described earlier the overturning moment for a multi legged structure is transferred as
axial loads to opposing foundation piles. The axial capacity (compression and tension) of
driven piles is well documented in several design methods and the axial soil/pile stiffness is
usually not a critical design issue. As these structures are normally placed in deeper water
(from 30m) the structural stiffness may be more important to monitor.

Date: 2017-06-15
Rev.no.: 3
Page: 44
Pore pressure build up in certain layers along the piles during cyclic loading may be an issue
that can be relevant for monitoring (dependent on soil conditions). Thus, in addition to
corrosion and tilt of the tower, the specific monitoring parameters may include pore pressure
along the piles and structural health/fatigue life monitoring such as dynamic motion and
strain in critical members/connections of the jacket/tripod.
For pre-piled foundations, the grouted stab connections between the piles and the structure
may be a weak link and depends on the quality of the grouting operations as well as long
term degradation (see also Section 2.2.1 Grouted connection of the transition piece).
Grouted stab connections have been used widely in the offshore industry. The legs of the
structure are either outfitted with a male stabs inserted into the piles or with sleeves around
the piles sticking up from the seabed. The piles and stabs/sleeves are equipped with shear
keys (weld beads) for improved bonding of the grout, see Figure 4-29.
Figure 4-29 Example of grouted male stab connections to pre-driven piles (OWEC tower)

Date: 2017-06-15
Rev.no.: 3
Page: 45
The grouted connections can be

monitored using subsea extensometers
which must be retrofitted by diver or
ROV, see Figure 4-30. It may also be
possible to use inductive contacts
sensors which would detect any
spacing developing between the pile
top and the flange of the stab. Provided
that the structure is designed with
adequate coating and cathodic
protection system there is no special
concern and requirements for
corrosion monitoring.
Figure 4-30 NGI Pile stab extensometers

for detection of relative vertical
movement between pile and jacket stab
if degradation of the grouted connection
is suspected
Post piled jackets/tripod have mud

mats with pile sleeves which also are
fixed to the piles by grouting. Such
arrangement has been used within the
Oil & Gas industry for more than 50
years and the long term performance is
well exploited. Normally these
connections have longer sleeves and
grout bonding sections with larger
margins against degradation failure.
The same applies for piles driven
through the jacket/tripod legs.
Nevertheless, all grouted connections
may be subjected to long term
degradation with a potential need for
monitoring. If serious degradation has
occurred and relative pile/sleeve
motions may occur, retrofitted
extensometers with high resolution
can be used for monitoring, see Figure
4-31.
Figure 4-31 Example of a retrofitted extensometer for measuring relative vertical deformations in
a traditional sleeve-pile grouted connection for a jacket foundation

Date: 2017-06-15
Rev.no.: 3
Page: 46
For the load distribution and fatigue life of a four-legged foundation it is important that the
pile top elevations are accurately measured after installation and the stab flanges are
shimmed to compensate for possible deviations. The required accuracy in these
measurements are normally better than +/-10mm.
Scour is less important to monitor as increased pile stick-up from the seabed will have
limited effect on the Eigen frequencies of the structure and little impact on the axial bearing
capacity of the piles (which normally are designed with ample margins against scour in the
top soil).
In order to optimize steel design and decrease the cost for jacket structures, stress
measurements (strain gauges) at critical locations may be relevant.
For special jacket design such as Louisiana based Keystone's twisted jacket, see Figure 4-
32, special monitoring aspects may apply for design verification. In this case, monitoring
strain in critical members is probably relevant and in addition to lateral response (dynamic
motion), torsional stiffness of the foundation may also be of interest to monitor in the field.
Figure 4-32 The twisted jacket from Keystone Engineering Inc.

Date: 2017-06-15
Rev.no.: 3
Page: 47
4.6 Specific monitoring aspects for tripods/jackets with bucket

foundations
Similar monitoring objectives as for the mono bucket tower apply for the seabed
foundations. However, the load transfer to the buckets is mainly in axial compression and
tension (rotation is prevented by the stiff connection to the legs), see Figure 4-33. In terms
of structural health, the leg connections to the buckets are subjected to large stresses and
may be relevant for strain monitoring. For sandy soils the drainage conditions during cyclic
loading is important, especially for the ULS pull-out capacity. Thus, monitoring pore
pressure at different depths along the skirt walls may be relevant.
Figure 4-33 Tripod bucket foundation and critical failure mechanism in clay (model: NGI)
In order to obtain a statically determinant system, three legged (tripod) foundations are
preferred. For a four-legged foundation, it is important that all four buckets are penetrated
with the same rate or that the corners are kept in a straight plane. For large buckets the
driving forces generated during suction penetration can be significant. Potential distortion
or bending of the jacket during penetration would induce large stresses in the structural
members. Thus, the elevation of each bucket must be monitored during installation as the
tower may remain vertical even if the base is distorted, see Figure 4-34.

Date: 2017-06-15
Rev.no.: 3
Page: 48
Figure 4-34 Vertical four-legged bucket foundation with levelled or distorted base
With respect to the jacket/tripod structure itself, the monitoring aspects are similar to a piled
jacket/tripod, however without grouted connections.
4.7 Specific monitoring aspects for floating OWT foundations

Floating OWT foundations can be an alternative solution for wind farms located in deeper
waters (>50m water depth). The different concepts include Spar type or semi-submersible
floaters with catenary moorings or Tension Leg Platforms with taut moorings, see illustrated
examples in Figure 4-35. The catenary moored floaters are stabilized by ballast and
buoyancy, while the TLP is stabilized by the tension leg moorings.

Date: 2017-06-15
Rev.no.: 3
Page: 49
Figure 4-35 Different types of OWT floaters (Illustration: DNV-GL)

The mooring response and the turbine motions are the primary design issues for a floating
OWT and relevant for monitoring. Dependent on the type of floater the critical motions to
monitor may be different and more complex compared to fixed foundations. Free-floating
bodies have six (6) degrees of freedom. The three (3) translational movements like heave,
surge, sway and the three (3) rotational motions in pitch, roll and yaw, see Figure 4-37. To
reduce extreme loads acting on the turbine especially heave, roll and pitch motions should
be avoided or reduced. The TLP structures achieve the best performance and have low
motion response, especially with respect to heave, pitch and roll. These motions are more
or less eliminated through the taut tension leg mooring system. For deep draft floaters with
catenary moorings like the spar or the semi-submersibles, heave pitch and roll motions are
minimized but not eliminated. Surge, sway and yaw motions applies for both types of
moorings.

Date: 2017-06-15
Rev.no.: 3
Page: 50
Other stabilizing systems for the floaters

may be hull trim systems allowing ballast
water to be pumped to different
compartments to compensate for vertical
misalignment of the OWT caused by the
wind force (thrust). The three columns on
the WindFloat foundation are outfitted with
water entrapment plates acting like brakes
with significant dampening of the dynamic
heave motions. The correlation between the
vertical motions and the hydrodynamic
loads taken by the water entrapment plates
is of interest to monitor during field
operation.
Figure 4-36 Water-entrapment plate on WindFloat (Source: Principle Power Inc.)
The seabed anchors for catenary moorings

(normally driven piles, drag or suction
anchors) are proven technology from the
O&G industry and may be less relevant for
monitoring as they seldom experience
loads with detectable response of the
anchor (also challenging with cabling to
the surface). TLP anchors are however
always subjected to tension load and even
if most of the average tension usually is
balanced by the dead weight of the
anchors, they are more actively loaded
compared to catenary mooring anchors and
therefore vital for the stability of the TLP.
Thus, in place monitoring is relevant for
TLP anchors and can provide useful data
for later design optimization, see section
5.11.
Figure 4-37 Definition of possible floater

motions. Note for a tension leg foundation (as
illustrated) the heave, pitch and roll motions
will be restricted

Date: 2017-06-15
Rev.no.: 3
Page: 51
Figure 4-38 Suction anchors for catenary mooring to a large floater installed 1997 in the North Sea
Gridline connections extend from the floater to the seabed and their operational behavior
can be compared to flexible risers in the O&G industry, see Figure 4-39. Motion tracking
of the riser cable provides important input for fatigue design of the cable assembly with
bend restrictors and seabed anchoring points. Vortex induced vibration due to sea current
drag forces acting on free cable spans is also a critical phenomenon that can affect the
service life of exposed grid sections.
Figure 4-39 Hywind mooring system and gridline connection (illustration Statoil)

Date: 2017-06-15
Rev.no.: 3
Page: 52
Combined and novel mooring systems as outlined in Figure 4-40 involve complicated
modelling design verification by in-field monitoring would most likely be required.
Figure 4-40 Concept with combined mooring system for multiple OWT's (Illustration DNV-GL)
The design of Statoil's Hywind Demo spar floater (Figure 4-41) was verified by means of a
comprehensive monitoring program including more than 200 sensors, the following data
was recorded in the field:
Waves wind and current (magnitude and direction)
Motion and position of floater
Mooring line tension
Strain gauges at tower and hull (4 levels – bending moments and axial force)
Rotor speed, blade pitch and generator power
Flap- and edgeways rotor bending moments
Motion (tower pitch) /blade pitch controllers

Date: 2017-06-15
Rev.no.: 3
Page: 53
Figure 4-41 Statoil's Hywind Demo spar OWT floater
4.8 Summary of recommended monitoring parameters

The following parameters may be relevant for long term monitoring of the different types
of foundations:
All types of foundations (Metocean and Turbine data)

Wind and Wave height
Current
Rotor speed, blade pitch and generator power
Turbine vibrations/motions
Driven Mono pile foundations:
Tilt of tower
Lateral dynamic motion (FA and SS response) of the structure at different levels
Corrosion TP and pile top
Scour and currents (if the seabed is prone for sediment transport)
Axial strain along the pile (P-Y behavior/soil stiffness) and grouted connection (load
transfer)
Deformations in grouted connection transition piece (if critical)
Cyclic pore pressure along the pile (dependent on soil conditions)
Lateral earth pressure along the pile (difficult)
Mono bucket foundation:
Tilt of tower
Lateral dynamic motion (FA and SS response) of the tower at different levels and
dynamic rotation of the foundation bucket
Strain in TP connection to bucket lid
Corrosion in exposed members and CP protection

Date: 2017-06-15
Rev.no.: 3
Page: 54
Cyclic pore pressure along the skirts and inside bucket

Load distribution, vertical earth pressure along the base of the bucket (if not
grouted)
Settlement (shake down)
Gravity base foundation:
Tilt of tower
Dynamic motion (rotation of the gravity base)
Scour and currents (if the seabed is prone for sediment transport and without scour
protection)
Cyclic pore pressure beneath the base
Contact stresses beneath the base
Settlement (differential)
Stress in reinforcement and cracks in concrete, if relevant connections to steel TP.
Piled jackets/tripods:
Tilt of tower
Strain in critical members and joints (fatigue)
Cyclic pore pressures along the piles (for particular soil conditions and pile design)
Integrity/deformation in grouted connections (pre-piling)
Jackets/tripods with bucket foundations:
Tilt of tower
Dynamic motion of structure (lateral on TP and vertical on buckets)
Strain in connection legs to buckets and critical members/joints
Cyclic pore pressure along the skirts of the buckets
Load distribution, vertical earth pressure along the base of the bucket (if not grouted)
Settlement (shake down)
Floating OWT foundations:
Tilt of tower
Dynamic motion of floater and turbine
Position (offset) of floater
Mooring loads
TLP anchor response
Strain in critical structural members

Date: 2017-06-15
Rev.no.: 3
Page: 55
5 SHM instrumentation solutions for OWT structures

An offshore wind farm comprises numerous foundation structures, for monitoring
campaigns normally only a few selected OWT’s can be instrumented with integrated SHM
systems, see also section 3.4. In order to gain useful experience for the complete wind farm
development, instrumented structures should preferable be installed in advance. Metmasts
are installed some years ahead of full development and these structures are sometimes also
outfitted with SHM systems in addition to the meteorological instruments. It should
however be noted that the loads on a Metmast are significantly different compared to full
scale wind turbine.
In order to add value to the instrumentation, the design should if possible allow for retrofit
and re-use (of parts) on other structures in order to maximize the utilization and expand the
data base obtained from the monitoring system. Some subsea instruments may be possible
to install with a hoisting (guide wire) system allowing for deployment and recovery without
under water intervention. Topside, battery operated wireless sensors may be used in some
applications. Such sensors are not dependent on infrastructure such as cabling etc. and
therefore mobile. It should be noted that the EM noise and strong magnetic fields around
the turbine may affect radio transmission and must be checked out, wireless data
transmission between outside and inside of the steel structure may also be limited.
For dynamic monitoring synchronization between different wireless units (for example
accelerometers) may require special broadcast synchronization protocols. If possible,
mobile or replaceable sensors should be considered. For example, in some cases permanent
strain gauges can be replaced by long-gauge sensors (like an extensometer with very high
precision). These are installed between brackets or studs on the structure, with mountings
that allow for replacement, see Figure 5-1.
Figure 5-1 Example of long-gauge deformation sensors, fiber optic strain sensor by Smartech (left)
and fiber optic extensometer by Opsens (right)
In many cases, the required integration of the sensor in the structure and restricted
accessibility in the field, limit the mobility. Therefore the base case is normally to pre-install
the instrumentation at the foundation construction yard. For "permanent" instrumentation it
is important to consider the operational life, especially for sensors exposed to the
environment and subsea. Materials and packaging are usually the limiting factors, following
the general rule of thumb: higher cost-better quality-longer life. A realistic expectation may
be 5 years of operation in such conditions without special maintenance or replacement.

Date: 2017-06-15
Rev.no.: 3
Page: 56
Examples of instrumentation and sensor solutions for relevant SHM parameters are
described in the following sections. The list of options is not complete and recommendations
are mainly based on NGI's experience from more than 35 years with offshore structural and
geotechnical instrumentation projects.
5.1 Wave height

A cost-efficient instrument to monitor wave height from a fixed position is to use a down-
looking microwave radar measuring the air gap between the instrument and sea. It is also
possible to range the water surface from below (instrument mounted below sea level) using
acoustic Doppler's. The most commonly used down-looking wave radars are the Maritime
Microwave Altimeter from MIROS or the Rex Wave Radar from Rosemount. The wave
radars must be mounted with free sight to the sea surface and at sufficient distance out from
wave breaking structures to monitor representative wave height, see Figures 5-3 and 5-4.
Figure 5-3 (Above) Rosemount wave radar on a mono

pile OWT. The radar can be mounted directly on the TP
guardrail with a hinged frame for easy maintenance
from deck
Figure 5-2 Left) Positions of Down looking microwave radar (1) or Up looking acoustic Doppler (2)

Date: 2017-06-15
Rev.no.: 3
Page: 57
Figure 5-4 Down looking micro wave radars from Miros (left) and Rosemount (right)
The up-looking Acoustic Doppler
Current Profilers (ADCP’s) usually
provide multi-directional wave height
measurements in addition to current
profiling data. The most commonly used
acoustic Doppler instruments are
manufactured by TeledyneRDI or Nortek
(AWAC)
Figure 5-5 ADCP's from Teledyne-RDI (left) and typical wave and current data (right)

Date: 2017-06-15
Rev.no.: 3
Page: 58
The disadvantage with a submerged

instrument is limited accessibility and
possible biofouling, the advantage is
that wave and current measurements
can be combined for one instrument.
Figure 5-6 AWAC from Nortek mounted

on a mono pile foundation
5.2 Current
Dedicated current measuring instruments are divided into current meters and current
profilers (ADCP). If ADCP's are used, the current profiling and wave height measurements
can be combined in one instrument, see previous section.
Single point current meters can be based on mechanical (impeller) or electromagnetic

sensing principles, see Figure 5-7.
Figure 5-7 Mechanical (impeller) and electromagnetic current meters from Valeport
For current monitoring in conjunction with scour assessment only the sea-bottom currents
are of interest thus a 3D current meter can be used (placed close or at the seabed). A cost-
efficient instrument is the Aquadopp current meter from Nortek which is delivered in Delrin
housing for shallow water applications (down to 300m depth), see Figure 5-8.
Figure 5-8 Norteks Aquadopp 3D current meter

Date: 2017-06-15
Rev.no.: 3
Page: 59
5.3 Scour monitoring

Scour development around an OWT foundation can be
monitored by echo sounders or sonars scanning the
seabed. The instruments (3) are fixed to the structure well
below the water line and continuously measure the
distance to the seabed around the foundation. Echo
sounders or sonars are ranging the distance to the seabed
reflector by measuring the travel time of submitted
acoustic signals reflected from the seabed. The acoustic
beams can vary in width and consequently in spatial
resolution, the first arrival (shortest travel time) will
usually be registered as the measured distance. The
acoustic beam from the sonars must not be obstructed by
structural elements as they will create an acoustic shadow
or cause false reflections.
The acoustic frequency band suitable for this purpose is

typical between 500-1000kHz, with a usable range up to
50-150m. With higher frequency a higher resolution is
obtained but the range becomes shorter. The acoustic
beam width is usual between 3-6°. Note that the acoustic
sensors do not work in air and should therefore be mounted
well below the splash zone. Entrapped air bubbles in front
of the transducer head may interrupt the measurements.
Standard echo sounders or altimeters are single point

instruments. Nortek has combined 4 acoustic beams for a
cost-efficient solution in their scour monitoring sonar. The
output from these instruments will be numeric ranging
data, see Figure 5-10.
Figure 5-9 (Left) Positions of scour monitoring sonars on a mono pile foundation

Date: 2017-06-15
Rev.no.: 3
Page: 60
Figure 5-10 Single point echo sounder from Tritech and Norteks multi-head sonars for scour
monitoring
For even more information about the seabed topography scanning (profiling) sonars can be
used providing a continuous image of the seabed profile. The MS1000 scanning sonar
system from Kongsberg-Mesotech represents the industry standard for this type of sonars
and is frequently used for scour surveys, see Figures 5-11 and 5-12.
Note that seabed profiling with sonars is performed by scanning in vertical mode.

Date: 2017-06-15
Rev.no.: 3
Page: 61
Figure 5-11 Two scanning sonars used for Mono pile scour imaging along two baselines
Figure 5-12 Sonar image of scour/seabed profile across a bridge pier using Kongsberg-Mesotechs
MS1000 scanning sonars located at either side of the pier scanning in the vertical plane

Date: 2017-06-15
Rev.no.: 3
Page: 62
For a full 3-D image of the scour development the scanning sonars can be mounted on
rotators (obtaining dual axis rotation of the single beam), the scanned profiles can be
merged in a 3D image during post processing using dedicated software, see Figure 5-13.
Figure 5-13 Kongsberg dual axis scanning sonar and post processed 3D image from rotated sonar
profiles

Date: 2017-06-15
Rev.no.: 3
Page: 63
5.4 Dynamic motion monitoring

Understanding the dynamic response of the
structure and configuration of the
accelerometers are perhaps the most important
aspects when monitoring dynamic motion. The
accelerometer data is used for modal analysis of
the structure and tracking of Eigen frequencies.
The data is used to verify stiffness and
dampening of the foundation, changes in the
dynamic response can also indicate structural
degradation.
For an OWT foundation dynamic motion is

usually monitored at 3-4 different levels (see
Figure 5-14), namely Foundation, Tower
base/midpoint and Turbine elevations. For tall
towers, an accelerometer can also be located at
the tower midpoint in order to monitor the 2nd
mode of vibrations. For fixed foundations,
mainly horizontal motions are of interest (X-Y),
for jackets/tripods vertical accelerometers shall
be used at the seabed level as the response to
overturning moments mainly is in the vertical
direction at the seabed. For mono piles, the main
focus is on the horizontal dynamic motions.
Figure 5-14 Key dynamic motion monitoring
elevations for an OWT jacket structure
The sensitivity of the accelerometers must be tuned to the expected amplitudes of motion.
Therefor very sensitive sensors must be used at the seabed foundation where the motions (if
any) are expected to be small. Although the turbine usually is equipped with accelerometers
integrated in the control system, a dedicated system should be used for monitoring the
dynamic behavior of the complete structure (SHM). For later analysis, it is important that
identical sensors are used at all locations and that the data from all sensors is synchronized.
In some cases, it is also of interest to monitor the yaw or twist of the foundation or tower.
Angular rate gyros to monitor rotation are usually less sensitive and more expensive
compared to linear accelerometers. Therefore, two linear accelerometers located at either
side of a baseline are normally used to detect rotation. For dynamic SHM monitoring of
typical OWT structures the frequency band of interest is typical from DC up to a few Hz. A
range of suitable accelerometers is available in the market.

Date: 2017-06-15
Rev.no.: 3
Page: 64
For SHM applications, high sensitivity and low noise accelerometers able to record slow
accelerations down to DC must be used. Suitable types of accelerometers are force balanced
(usually highest sensitivity), MEMs (Micro Electro Mechanical systems), Variable
Capacitance (VC) or Piezo Resistive (PR) sensors. Note that the common piezoelectric type
accelerometers are not DC.
In order to select a suitable accelerometer, the required sensitivity or resolution in motions

must be specified (how small displacement must be traced) within the frequency band of
interest. For practical reasons, reliable tracking of motions less than 0.25-1mm may be
difficult even if the sensitivity of the accelerometer itself is good. In addition to the
sensitivity, the sensor noise level must be considered as this will be the lower limit in the
measuring range, see Figure 5-15.
Figure 5-15 Comparison of noise density spectra's made by NGI for different accelerometers
In addition to the sensor noise, EM noise picked up by the cables must be considered. If
longer cables are used, accelerometers with digital output are recommended. Instrument
units with integrated accelerometers in the required directions allows for simplified
installation and cabling. The initial accelerometers readings should be at high rate, band
pass filtered and re-sampled at minimum 10 Hz rate (20 Hz is recommended).
Normally a measuring range of ± 1 to 2 g and frequency range of DC-200 Hz are sufficient

for monitoring of the dynamic response of fixed foundations, examples of suitable sensors
are given in Figure 5-16.

Date: 2017-06-15
Rev.no.: 3
Page: 65
Figure 5-16 Examples of accelerometers suitable for SHM, from left to right: Force balanced
accelerometers from Geosig with analogue output, Force balanced accelerometer from Jewel
instruments with digital output and combined 3D mems accelerometer and dual axis mems
inclinometer unit with Ethernet output from SENSR.
Mobile accelerometer units with simple hook-up (like the one from SENSR) can be moved
between OWT foundations in order to check Eigen frequencies. The accelerometer units
can be fixed with magnetic footings, although some suppliers claims that such fixtures not
are sufficiently rigid. Welded/glued double plates or directly bolted fixtures are better.
Figure 5-17 Left, accelerometers with magnetic footings (photo OWI-Lab) and mounting
arrangement from NGI (combined accelerometer and inclinometer holder which can be turned)
The DC and gravity referred accelerometers can easily be checked by the tap and flip tests,
checking the response when gently taping the enclosure and flipping the sensor (sensitive
axis) vertical the oupt should be +1g pointing downwards and – 1g pointing upwards. Note
that for vertical measurements the sensor range must be larger than +/-1 g as Earth's gravity
will be included in the measurements,

Date: 2017-06-15
Rev.no.: 3
Page: 66
As described in section 4.7 floaters have more degrees of freedom compared to fixed
foundations and accelerometer units (Motion Reference Units – MRU's) with 5-6
components may then be used, see Figure 5-18. It must however be checked if the MRU
has sufficient sensitivity to monitor the motions of interest (the sensitivity usually not
sufficient for fixed foundations).
Figure 5-18 Motion reference Units from Kongsberg- Seatex, normally used on larger DP vessels
In order to monitor motions of
subsea power umbilical's for
connection to the grid, small self-
contained dynamic loggers may be
used. These are equipped with
accelerometers, data logger,
memory and internal battery. The
motions loggers are
strapped/clamped directly to the
umbilical by divers (or ROV), see
Figure 5-19. The small standalone
loggers are made for temporary
monitoring have limited capacity
with respect to battery and memory,
but can be re-deployed after
charging.
Figure 5-19 Subsea dynamic motion

logger "IntegriPod" from Pulse
deployed on unsupported sections of a
subsea grid cable at the Gunfleet Sands
wind farm (DONG Energy) in the UK
sector of the Southern North Sea. The
investigation was to determine the
impact of tidal activity (Vortex induced
vibrations) on the fatigue life of the
cable

Date: 2017-06-15
Rev.no.: 3
Page: 67
The following operational capacity applies for the IntegriPod dynamic motion logger from
Pulse outfitted with three accelerometers and shown in Figure 5-19:
• Continuous logging @ 10Hz 12 days (limited by memory)
• Continuous logging @ 4 Hz 16 days (limited by memory)
• Continuous logging @ 2kHz 32 days (limited by memory)
• 20 minute logging every 2 hours @10Hz 76 days (limited by memory)
• 20 minute logging every 2 hours @ 4 kHz 96 days (limited by memory)
5.5 Tilt - Verticality of the tower

Tilt is measured by a biaxial inclinometer sensing
the inclination along X and Y baselines. For
small deviations from the horizontal, the
maximum tilt can be derived from the vector sum
of X and Y inclination. The heading of maximum
tilt is derived based on the headings of the X and
Y baselines and their magnitude of inclination.
Figure 5-20 Above: Definitions Dip angle = Maximum

tilt and Strike = Heading of maximum tilt
Figure 5-21 Left: The biaxial inclinometer is

preferable located at the base of the tower (5)

Date: 2017-06-15
Rev.no.: 3
Page: 68
Many high precision inclinometers are available in the market for this purpose. As for
accelerometers the force balanced (servo) and gravity referred inclinometers have the best
precision. However, today most of the precision MEMs based inclinometers have sufficient
accuracy. Normally, an overall accuracy better than 0.01° and a range within ±3-5° is
sufficient for fixed foundations. As tower inclination measurements relate to small
deviations from the vertical axis, full scale accuracy and linearity is of less importance. Zero
point drift and temperature sensitivity are the critical parameters for specification of suitable
inclinometers for monitoring tower tilt. For in-situ calibration (zero point check) it is an
advantage if the instrument can be turned 180⁰ (as shown in Figure 5-17), the same reading
with opposite sign should be obtained when the sensor is turned (unless the zero point has
drifted).
The common error when measuring inclination is that the mounting offset (alignment to the
structure) is not properly determined. Also the mounting location may not be representative
for the tilt of the structure, local deformation due to temperature gradients (direct sunlight)
or structural flexing may generate false readings. Solid mounting and screening from
sunlight is important. These aspects are in most cases more important for the overall
measuring accuracy than the sensor performance. For large structures, it can be challenging
to determine the offset in mounting alignment compared to the baseline of the overall
structure, thus a reference baseline must be used (for example the mounting flange to the
tower).
Figure 5-22 NGI biaxial inclinometer assembly with capacitive inclinometers (blue), the unit can be
mounted horizontally or vertically by changing the internal position of the two inclinometers. These
enclosures are intended for subsea use and without turning option (zero point check)

Date: 2017-06-15
Rev.no.: 3
Page: 69
5.6 Structural strain, stress and fatigue

Structural strain is measured in order monitor derived stresses and to assess distribution of
forces but also to assess structural health at critical locations (joints etc.), this includes stress
magnitude with respect to yield and stress history with respect to fatigue.
The recorded strain is converted to stress using Hooke's law:
σ=E∙ε
where σ represents stress, ε represents strain and E represents Young's modulus of elasticity.
The dynamic stress response is important for SHM on offshore foundations, the strain gauge
monitoring solution must allow for dynamic sampling rates. See Figure 5-23.

Date: 2017-06-15
Rev.no.: 3
Page: 70
Figure 5-23 Example of structural stress monitoring. In the left figure the strain gauged sections (6)
are used to monitor foundation loads distributed from the legs of a jacket (axial stress). For the
mono pile (right) the strain gauges (7) are used to monitor the moment distribution with depth
along the pile
5.6.1 Resistance wire strain gauges

The most common type of strain gauges are the resistance wire gauges (glued foils or spot
welded). These strain gauges are applied for point measurements and must be pre-installed
on the structure, see Figure 5-24.

Date: 2017-06-15
Rev.no.: 3
Page: 71
Figure 5-24 Glued Foil strain gauge (left) and micro spot welded strain gauge (right). Some models
of spot welded strain gauges have the resistance wire protected inside a water and pressure
resistant tube allowing them to be immersed in seawater
In order to measure strain correctly extremely small changes in resistance of the strain wire
is sensed with high accuracy using bridge measurement circuits. A strain gauge bridge
circuit indicates measured strain by the degree of imbalance, and uses a precision voltmeter
in the center of the bridge to provide an accurate measurement of that imbalance. The
simplest bridge configuration is the quarter bridge, see Figure 5-25.
Figure 5-25 Quarter-bridge strain gauge circuit

In this configuration, the hook-up wire resistance (Rwire) has a significant impact on the
operation of the circuit. This effect can be reduced by adding a third wire directly between
the strain gauge and the voltmeter. The strain gauge is very sensitive for temperature
changes, which not are compensated with this configuration. However, by replacing resistor
R2 with a dummy (unstressed) strain gauge, the effect of temperature changes can be
cancelled, see Figure 5-26.

Date: 2017-06-15
Rev.no.: 3
Page: 72
Figure 5-26 Quarter-bridge strain gauge circuit with temperature compensation

For measuring applications were the upper strain gauge is exposed to the opposite force
compared to the lower gauge (i.e. when the upper gauge is compressed, the lower gauge
will be stretched, and vice versa) the circuit illustrated in Figure 5-26 can be used as a half-
bridge (both gauges are stressed and temperature effects cancelled). However, such
configurations are normally not practical applicable for strain measurements on structures.
A full bridge configuration is when all resistors in the bridge circuit are replaced by strain
gauges, see Figure 5-27. NGI use a full bridge configuration with unstressed (dummy)
gauges for temperature compensation and 6 wire configuration (extra sense lines controlling
the excitation voltage at the bridge for compensation of variations in cable resistance).
Excitation and sense lines are twisted pairs for cancelling of EM noise in the cables. An
example of a water proof full bridge and temperature compensated strain gauge assembly is
shown in Figure 5-28
Figure 5-27 Temperature compensated full bridge configuration (4 wire system)

Date: 2017-06-15
Rev.no.: 3
Page: 73
Figure 5-28 HPI waterproof strain gauge for micro spot welding. The temperature compensating
elements are inside the cable transition piece to the left in the picture
For excitation of the bridge circuit and amplification/read-out of the signal, strain gauge
amplifiers are used. The strain gauge amplifiers are rather costly and each unit should serve
several measuring channels, normally the strain gauge amplifiers must be located remotely
from the strain gauges themselves, therefore a robust bridge configuration is important at
the measuring point.
5.6.2 Fiber optic strain gauges

Fiber optic strain sensors are in many aspects an advantageous alternative to the traditional
resistance wire gauges. Several optical sensing principles exists, however for dynamic
measurements the Fiber Bragg Grating (FBG) sensor is the most common type. By means
of operating at different wave lengths several FBG sensors can be multiplexed and
implemented along the same optical fiber. The Fiber Bragg Gratings are wave specific
reflectors (mirrors) in the optical fiber and strain is recorded when the distance between the
"mirrors" are changed, see Figure 5-29. By means of using gratings sensitive to different
bands of wavelengths several FBG sensors can operate along the same fiber.
Other common types of fiber optic strain gauges for point measurements are based on the
Fabry-Perot (FP) cavity sensing principle. The sensing part is essentially composed of two
parallel mirrors separated by a cavity, see Figure 5-29. The spacing between the mirrors (or
length of the cavity) is strain (and temperature) dependent and determined by an
interferometer. These strain gauges are single ended (must be located at the end of the fiber)
and use white light or low coherence interferometry which is less sensitive to dampening in
connectors and splices.

Date: 2017-06-15
Rev.no.: 3
Page: 74
5-29 Operating principles for FBG (left) and FP (right) fiber optic strain gauge
True distributed fiber optic sensor system based on Brillouin or Rayleigh scattering derives
the strain data based on many repeated interrogations and are mainly used for static
measurements, the long response time is not suitable for SHM on OWT foundations.
The FBG and FP strain sensors are available in different configurations and layout for
similar mounting as for strain gauges, see Figure 5-30 and 5-35.
Figure 5-30 Different types of FBG strain gauges from HBM

Date: 2017-06-15
Rev.no.: 3
Page: 75
True distributed fiber optic sensor system based on Brillouin or Rayleigh scattering can
sense strain along the entire length of a fiber optic cable and can therefore contain thousands
of measuring points along a single fiber. However, the signal interrogation is more complex
and includes averaging of many interrogation cycles. The acquisition time for recording
strain along the fiber can be from 5 to 20 minutes (dependent on required resolution).
Distributed fiber optic strain systems are mainly used for static measurements, the long
response time is not suitable for SHM on OWT foundations.
The FBG sensors are interfaced by an

optical sensor interrogation module. The
unit illustrated in Figure 5-32 can operate
up to 4 FBG fiber cables, each with 20
FBG sensors (80 FBG sensors in total) at a
maximum scanning rate of 500 Hz. Data
from the interrogator is streamed to the
Data acquisition PC by Ethernet interface.
Figure 5-31 Dynamic FBG sensor interrogation

module from Micron Optics
5.6.3 Resistance wire vs fiber optic strain gauge

The mounting and bonding of the strain gauge to the structure is important in order to
obtain representative readings, similar techniques (micro spot welding or glued) are used
for both resistance wire and fiber optic strain gauges.
Figure 5-33 shows measurements of 6

FBG strain sensors installed in pairs with
welded and non-welded (glued)
mounting, By plotting the output from the
welded gauges (WFBG) against the glued
gauges (FBG), it can be concluded that in
general the welded strain gauges have a
reduced sensitivity compared to the glued
sensors. However, for long term
applications and field mounting the
welded strain gauges are recommended
due to a more robust fixture.
Figure 5-32 Comparison of sensitivity between welded and glued FBG SG's

Date: 2017-06-15
Rev.no.: 3
Page: 76
Sealing against water ingress, although the fiber optic strain gauges not fail immediately
if exposed to seawater, they will degrade with time according to all suppliers. Thus, the
requirements for sealing of the measurement points and cable connections are similar for
both optical and resistance wire strain gauges. The sealing of the measuring point is also
important in order to prevent corrosion advancing into the bare steel and strain gauge
bonding area, potting of the measuring points in the field requires good workmanship and
patience, see Figure 5-34.
Figure 5-33 Installation of strain gauges for use under water; 1) Fix the strain gauge and cable
terminal on the thoroughly prepared steel surface, 2) hook-up and fix the cable, 3) seal (pot) the
cable connection and 4) Seal and pot the complete assembly, 5) is showing the completed
installation (photos: Strainstall/NGI)
Cabling and hook-up, the obvious benefit with fiber optic strain gauges is that only one
fiber is required for hook-up, compared to 4-6 leads required for hook up of a full bridge
measuring point with resistance wire gauges. However, fiber optic cable junctions and
splicing is more complicated and sensitive compared to electrical leads and continuous
routing of the optical cable to the data acquisition system will make installation on the
structure very challenging. The fiber optic cable is also more sensitive for rough handling
at the yard. Thus, pro and cons with respect to cabling depends on the instrumentation plan.
For example cabling of string with strain gauges string along a mono pile may be simpler
using fiber optic sensors as they are daisy chained along one continuous cable, for a braced
structure it may be the opposite as the complicated routing may call for several cable
junctions.
Opsens use white-light polarization interferometry (WLPI) for interrogation of their fiber
optic sensors, this technique is less sensitive for dampening/losses in fiber connections and
allows for simpler hook-up. The "quick connect" adaptor allows for hook-up using ordinary
fittings (Swagelock) or similar, see Figure 5-35. The "quick connector" is robust and can
resist pressures up to 5000 psi and suitable for permanent subsea use. Opsens also claims
that their WLPI sensors are less sensitive to temperature changes and transverse strain.

Date: 2017-06-15
Rev.no.: 3
Page: 77
Figure 5-34 Weldable fiber optic strain gauge from Opsens based on Fabry Perot sensing principle
and white light interferometry, fully submersible assembly with quick connector
Mechanical protection, strain gauge installations are fragile and must be well protected to
prevent damage. This is especially important for instrumented piles or other structural
elements being embedded in the seabed sediments. Normally the strain gauges are mounted
inside channel protection elements with cover plates for closure of the channels after
completed installation of the strain gauges and cabling. The lower end of the channel must
be equipped with a driving shoe to reduce shear forces on the channel. Cables etc. must be
well secured during pile driving. Figure 5-36 shows examples of protection and rigging.
Figure 5-35 Strain gauge channels with driving shoes at the inside of a monopole and (right) cables
bundles secured by ropes during pile driving
Noise, the fiber optic signals are immune against EM noise which is an advantage. However,
the EM noise influence can be limited to negligible magnitudes for resistance wire gauges
by means of proper configuration and wiring (twisted pairs).

Date: 2017-06-15
Rev.no.: 3
Page: 78
Precision and drift, the overall precision in strain measurement

is comparable for both types of strain gauges. They are both
sensitive for temperature drift which can be compensated by using
dummy gauges and/or measuring temperature in the vicinity of
the gauge. They are both sensitive to transverse loads which can
be compensated for by adding a transverse strain gauge in the
configuration, see Figure 5-37.
Figure 5-36 T-rosette configuration of strain gauges for compensation of transverse strain
Zero point readings, after mounting and bonding both types of gauges will read strains
which to a large extent depends on the stresses introduced during mounting. Thus, both
resistance and optical strain gauges must be calibrated after installation. The “as installed”
offset must be determined based on comparing sensor readings in a situation when the strain
in the structure is known or can be assessed theoretically. Offset readings can for example
be made when the foundation is hanging vertical in a crane or vertically parked on a flat
surface at the yard or feeder port. If only dynamic strain/stress variations are of interest the
mounting offset does not matter.
5.6.4 Other solutions for stress and deformation monitoring

Strain gauges provide point measurements with high sensitivity. If the strain is measured
across a longer base line, the deformation between the end points becomes larger and can
be recorded by extensometers or other types of deformation sensors. For example, vibrating
wires or long gauge fiber optic strain sensors. The long gauge type of instruments is suitable
where the interpretation of point measurements is difficult due to complex strain fields or
when local strain or deformation can be large or non-representative because of cracks or
other inhomogeneity's, for example in concrete, see Figure 5-38.
Figure 5-37 Application set-up for long gauge fiber optic strain sensor for measurements on
concrete specimen. The gauge length between fixture points A and B can typical be from 0.2 to 2m

Date: 2017-06-15
Rev.no.: 3
Page: 79
The long gauge fiber optic strain sensors transform a distance variation into a change in the
balance between two optical fibers that can be measured with a read-out unit. The sensors
can be based on different principles but with similar performance. Two types of long gauge
fiber optic sensors are illustrated in Figure 5-39, namely:
SOFO (Surveillance d'Ouvranges par Fibers Optiques) which is a low coherence
interferometric sensor using two fibers one for measurement and one for temperature
compensation
MuST sensor is made up by two FBG sensors along the same fiber. The first FBG
is located in the strained section of the fiber between the attachment points to the
structure, the second FBG is located in an unstrained section of the fiber directly
after the attachment point and used for temperature compensation. The outlines of
the MuST sensor is similar to the SOFO sensors.
5-38 SOFO and MuST deformation sensor assemblies by Smartec, the sensors can be directly
embedded in concrete or clamped externally on the element to be monitored. The Must sensor with
200mm active zone can also be fully encapsulated in a composite profile for use under water.

Date: 2017-06-15
Rev.no.: 3
Page: 80
The sensors consist of two main parts: active and passive. The active part contains the he
measurement fiber and measures the deformations between its two ends, see Figure 5-39.
For the SOFO sensor the temperature compensating (reference fiber) is located in the active
part (parallel to the measuring fiber) and for the MuST sensor, the reference fiber is located
in the passive part (in line with the measuring fiber). Both types of sensors have similar lay
out however, a SOFO sensor is single ended while the MuST sensors can be hooked up in
series along the same fiber. The MuST FBG sensors also can be interrogated at high speed
and is therefore better suited for dynamic measuring applications.
Other types of long base strain gauges are based on the vibrating wire principle or use of
LVDT (Linear Variable Displacement Transducer) or Fibre optic extensometers based on
the Fabry-Perot measuring principle. These instruments can be bolted or clamped to the
structural elements or retrofitted to pre-installed mounting brackets, see Figure 5-40.
Extensometer type of instruments can also be used to monitor crack propagation.
Figure 5-39 Vibrating wire strain gauge (left) and LVDT type of extensometer (right) often used as
crack or joint meter (larger deformation range than a strain gauge)
For interfacing vibrating wire sensors special excitation modules must be used, this type of
sensor is more common within geotechnical monitoring than SHM. The frequency signal
(only two wires are required for hook-up) are however not sensitive for EM noise and
suitable for long cables.
Probably the most accurate extensometers with respect to long term drift are based on
magnetostrictive sensing (micro pulse transducers), see Figure 5-41. Another advantage
with this type of sensor is that the sensing element is hermetically sealed and can be
submerged (the moving part only consist of a magnet).

Date: 2017-06-15
Rev.no.: 3
Page: 81
Figure 5-40 Magnetostrictive (micropulse) linear displacement transducer
Retrofitted instruments for strain/stress measurements are usually of long base extensometer
type and in many cases pre-installed supports or brackets are not available. The fixture must
then be customized for the application, for steel structures strong magnets can be used, see
examples in Figure 5-42.
Figure 5-41 (left) Extensometers on magnet pads from Pulse and (right) combined magnetic pad
and mechanical clamp fixture on NGI's pile-sleeve extensometer for monitoring of relative
deformations between pile and grout sleeve for jackets

Date: 2017-06-15
Rev.no.: 3
Page: 82
Fatigue life monitoring of the Wind Turbine structures can be carried out in two ways;
firstly, by analyzing data from installed strain gauges at likely fatigue locations and generate
stress cycles for estimating remaining fatigue life. Secondly, by using fatigue "pre-crack"
or "weak link" sensors integrated into the monitoring system, accurate predictions of
cumulative damage on welded steel structures can be made. The fatigue pre-crack sensor
comprises a steel coupon attached adjacent to a critical joint. Stress cycles cause fatigue
crack growth in the coupon that is detected electronically by loss of conductivity. For a
typical fillet welded joint the sensor output gives the proportion of the fatigue design life
that has been used. See Figure 5-43.
Figure 5-42 "Crack-first" fatigue pre-crack sensor from Strainstall (not for subsea installation)
5.7 Pore pressure monitoring

The pore pressure in the soil beneath and around the
foundation plays a vital role with respect to strength and
stiffness of the soil. The pore pressure is defined as the
pressure differential between entrapped pore water in the
seabed and hydrostatic (ambient) pressure at the same depth,
see Figure 5-44.
Figure 5-43 Illustration of entrapped pore water in the sea bottom
sediments
A piezometer is the instrumentation system used to measure pore pressures in the soil. The
pore pressure can be measured in three different ways as illustrated in Figure 5-45, using
total or differential pressure sensors. At shallow water depths the ambient hydrostatic
pressure is limited and two Total Pressure (TP) sensors can be used instead of one
Differential Pressure (DP) sensor.

Date: 2017-06-15
Rev.no.: 3
Page: 83
Figure 5-44 Piezometer configurations for pore pressure monitoring in the seabed (P refers to total
Pressure and dP to differential Pressure sensors)
For subsea pore pressure monitoring NGI often use the third set-up with a filter inlet
embedded in the soil with an hydraulic tube or standpipe routed up to the pressure port on
a differential pressure (DP) sensor located in a monitoring head near the seabed. The
reference port is routed directly to sea (alternatively using two TP sensors). With the sensor
enclosure above ground, hook-up or replacement is possible by divers. For instrumented
piles the more fragile sensor heads can be mounted after pile driving, see Figure 5-46.

Date: 2017-06-15
Rev.no.: 3
Page: 84
Figure 5-45 NGI's standpipe piezometer system for piles, the plot shows the recorded pore pressure
dissipation after driving
Figure 5-46 NGI's standpipe piezometer system on a bucket foundation (DONG Energy's BKR01 SBJ)

Date: 2017-06-15
Rev.no.: 3
Page: 85
The filters and standpipes must be open to sea during installation of the foundation and
flooded by seawater by means of open drainage on top (if the sensor house already is hooked
up remote operated bypass valves to sea are used). By means of the bypass valve, the
pressure port can also be periodically be opened to the sea allowing for zero point check of
the sensor (differential pressure should then be zero) and possible de-airing of the hydraulic
line, see Figure 5-48. A standpipe/hydraulic tubing system includes a larger water volume
between the filter and the sensor compared to embedded sensors which can be compensated
by using larger filter area and the dynamic response is sufficient to monitor cyclic
foundation pore pressures generated by wave loads.
120
Tidal variations
Pressure build-up
after sealing hole
100
Excess pore pressure, kPa
80
60
40
20
Automatic zero
Automatic zeropoint
pointchecks
checks
0
570 620 670 720 770 820 870 920
Days since Jan. 1, 2000
Figure 5-47 Example of down-hole pore pressure measurements (200m below the seabed) done by
NGI using up-hole differential pressure sensor with automatic zero point check every month (using
a solenoid operated bypass valve)
For piles and bucket foundations the pore pressures of primary interest to monitor are at the
interface between the foundation and the soil and the piezometers can therefore be
integrated in the structural foundation. In some cases, such as monitoring pore pressure
dissipation (consolidation) beneath a GBS structure it may be important to also monitor the
pore pressure below or beside the foundation. In this case the piezometers must be installed
in drilled boreholes (expensive) or pushed down (similar to CPT testing). For down-hole
installation beneath the foundation, pre-installed jacks may be used or drilling performed
through pre-installed casings through the foundation. For installation beside the foundation,
soil investigation equipment such as CPT, Vibrocorer or sampling rigs can be used for
installation of the piezometers.

Date: 2017-06-15
Rev.no.: 3
Page: 86
One issue in when integrating piezometers in foundation structures (piles or buckets), is

welding or making slots in the steel walls. In order maintain the structural strength the filter
elements used by NGI simply are made up from a piece with identical thickness and steel
quality as the rest of the structure. By welding this piece back into the slot in the wall, the
structural strength is maintained and the pore pressure filters are fully integrated with the
structure and flush with the wall.
5.8 Earth pressure (effective stress)

As described in Section 4.2.3 great care must be taken for successful monitoring of earth
pressure (especially the lateral stress). In order to derive effective stresses also pore
pressures must be monitored in the vicinity of the earth pressure cell.
The stiffness of the earth pressure cell is very important when measuring total stress against
a stiff structural element. To prevent arching effects around the cell it should not be
significantly lower than the medium it is embedded in (in this case steel). It is also very
important that the total stress cell is mounted flush with the structural surface that shall be
monitored such that the sensor itself do not disturb the soil. Any protruding parts (conduits
etc.) must be above the cell such that the sensing surface not is disturbed. Correct
measurements can be difficult in granular and stiff soil, the recorded stresses are often under
estimated.
The earth pressure cell can be of strain gauged membrane type or fluid filled pad connected
to a pressure sensor, see Figure 5-49. In order to achieve representative earth pressure
readings (limit singular effects in the soil) the cell should have a relatively large sensing
surface, normally with a diameter between 90-160mm. The cells will be subjected to large
forces and must be of rugged design, this is especially important for piles as the instruments
must survive the driving forces (without significant changes in zero point). As for all total
pressure instrument it is important to check the zero point at known reference pressure for
example before the instruments are penetrating into the seabed.
Figure 5-48 Example of membrane type earth pressure cell from Kyowa (left) and hydraulic pressure
pad type from Geokon (right). Both cells have a rigid plate at the backside and are suitable for wall
mounting

Date: 2017-06-15
Rev.no.: 3
Page: 87
Examples of combined earth and pore pressure assemblies integrated in pile or skirt walls
are illustrated in Figure 5-50.
Figure 5-49 NGI Earth pressure cell for piles modified for flush fit to the external pile wall and acting
as an internal driving shoe (left). Combined piezometer and earth pressure panel for integration
with bucket walls
5.9 Static displacement and settlement

In general, the use of Global Navigation Satellite Systems (GNSS) offers good flexibility
with respect to static displacement monitoring of structures, see Figures 5-51 and 5-53. The
precision which obtained can be critical issue for fixed offshore structures and specially for
settlement monitoring. The precision often depends on the distance to the reference station.
Also dynamic motion and vibrations may affect the positioning accuracy (averaging may
be required). The fixed reference or base station is required for corrections of the satellite
positioning data which depends on atmospheric conditions such as temperature and
humidity. The corrections are made based on the assumption that identical conditions apply
for both the measuring (rover) and the reference stations. This assumption may become
more uncertain with increasing distance to the reference station. For static events the
corrections can be made during post processing.

Date: 2017-06-15
Rev.no.: 3
Page: 88
For real time corrections which is required during offshore construction work (foundation
installation) or tracking of dynamic events Real-Time-Kinematic (RTK) systems are used.
The RTK reference station is continuously transmitting correction data to the measuring
stations for automatic corrections in real time.
The fixed position and distance to the reference station are governing for the precision of
acquired GNSS data (applies for both post processed and RTK data). The reference station
can be located on a piled and stable sub-station or can be land based if the distance to shore
is limited. The differential GNSS systems can use one single reference station serving a
number of mobile (rover) measuring stations. The RTK reference station re-broadcasts the
phase of the carrier that it observes, and the mobile units compare their own phase
measurements with the one received from the reference station. Differences between the
phases can be attributed to satellite ephemeris and clock errors, but mostly to errors
associated with atmospheric delay significant correction of these errors can be achieved by
using the data from the reference station post processed or in real time (RTK). Better
precision is normally achieved for post processed corrections as more statistical data is used
as basis for the correction factors.
Figure 5-50 Principles of differential GNSS for offshore application with real time corrections (RTK)
GNSS based systems for deformation or motion monitoring are deployed on many major
bridges for displacement monitoring. Compared to offshore wind bridge monitoring has the
advantage of fixed reference stations close to the mobile monitoring stations on the bridge.
The reference stations are usually installed on both bridge embankments, see data read-out
example in Figure 5-52.

Date: 2017-06-15
Rev.no.: 3
Page: 89
Figure 5-51 Real-time motion data provided by RTK stations on the Forth Road Bridge in Scotland
Figure 5-52 Recommended position for satellite monitoring stations (9) and settlement monitoring
system (10). GNSS antenna (right) on top of Lotte World tower (Seoul) during construction, a free
line of sight is required to the satellites.

Date: 2017-06-15
Rev.no.: 3
Page: 90
It is expected that the accuracy that can be achieved using GNSS system in offshore
application is within a couple of centimeters. Usually the precision is better for
measurements of horizontal compared to vertical displacement. Satellite position
monitoring is probably most relevant for floating OWT foundations.
Satellite positioning usually do not have the required precision for monitoring settlement of
fixed OWT foundations. For settlement monitoring settlement NGI use a fixed seabed
reference (for example small suction anchor/pile or clump-weight). The elevation is derived
using hydraulic liquid lines connected between the seabed reference and the measuring
points on the foundation and measuring the difference in hydraulic head between the
foundations and the seabed reference. By means of fully isolating the hydrostatic system
from seawater and using compensating lines to obtain the same reference (gas) pressure in
the whole system an overall accuracy within a couple of millimeters can be obtained. The
system can be configured in many ways. However, the reservoir reference must always be
at the highest point, see Figure 5-54. For measurement of differential settlements within one
foundation, levelling systems with reference to the liquid horizon can be used. These
systems are more stable with better accuracy (+/1 mm) and less long term drift compared to
hydraulic head systems. However, the range in elevation differences which can be
monitored is limited (typical max. +/- 200mm deviation from the horizon), see Figure 5-55.
Figure 5-53 NGI’s Liquid head system for settlement monitoring, measuring unit with differential
pressure sensors (top right) and Reservoir reference (bottom right)

Date: 2017-06-15
Rev.no.: 3
Page: 91
Figure 5-54 NGI' liquid level system on a pre-piling template for precise under water levelling
5.10 Mooring loads

Monitoring the mooring loads is
important for understanding of the
dynamic response of a floating
structure. For larger floaters, the
mooring chains are usually operated
from deck using fairleads and
windlasses for tensioning, the
mooring chain are finally secured by
using chain stopper, see Example in
Figure 5-56.
Figure 5-55 Topside arrangement for

operation of mooring lines on a large
floater or station keeping vessel

Date: 2017-06-15
Rev.no.: 3
Page: 92
For OWT floaters the limited deck space (and economy) calls for a subsea termination of
the mooring lines directly to the hull of the floaters. For Statoil's Hywind floater, the three
mooring lines are terminated by Y-legs to the hull of the spar. One leg provides a fixed
mooring strong point. The other Delta leg provides yaw stiffness to the spar floater and can
be adjusted (tensioned) during installation, see Figure 5-57.
Figure 5-56 Arrangement for mooring line termination the hull of Statoil's Hywind spar floater
In order to monitor the load on mooring lines connected to the floater below, the water line
subsea load shackles or load pins can be used, see Figure 5-58. The force is monitored
measuring the shear forces in the strain gauged load pin, different sizes are available for
load ranges up to 2000 T tension. The load pins or shackles are normally located at the upper
termination of the mooring line (pad eye).
Figure 5-57 Subsea load monitoring shackles (left) and load pins (right) from Strainstall

Date: 2017-06-15
Rev.no.: 3
Page: 93
Figure 5-58 Load monitoring pins mounted on mooring quick release hooks from Mampaey
For tension leg floaters, load monitoring in the tendons may even be more important to
monitor. The tendons are usually hooked up to the TLP porches by means of ballasting the
hull and increasing the draught, after the tendons are hooked up the hull is de-ballasted and
the tendons are tensioned based on the monitored load.

Date: 2017-06-15
Rev.no.: 3
Page: 94
Figure 5-59 Pelastar TLP concept from Glosten Asc.

Underwater compression (column type) load cells can be used in groups of three on the
porches around each mooring tendon of the TLP for measurements of the load, see Figure
5-61.
Figure 5-60 Column type under water loads cells from Strainstall
Dependent on the hook-up configuration other arrangements for load measurements may be
required.

Date: 2017-06-15
Rev.no.: 3
Page: 95
Figure 5-61 Instead of ballasting the floater, hook-up can be done with a special barge which is
pushing down the TLP such that tendons can be connected. Concept: Pelastar/Geosea. Three
mooring arms are replaced by five arms in figure 5-60 (redundancy?). Three tie-down points
constitute a static determined system with known distribution of loads and less anchors to install
5.11 Monitoring TLP anchors

The TLP's can be anchored by piles, gravity and/or suction anchors, see Figure 5-63.
Figure 5-62 TLP suction anchors for a large floater

Date: 2017-06-15
Rev.no.: 3
Page: 96
The suction anchors are usually designed to utilize suction for cyclic (wave) loads while the
permanent tension is taken by friction and ballast. There are arguments either for further
design optimization or for more conservative design criteria. It is therefore of interest to
monitor the differential pressure in the headspace on top of the anchor as well monitoring
possible vertical motion during storm loading. Possible settlement or pull-out can also be
recorded comparing hydrostatic pressure on the anchor with a fixed reference station at the
seabed. Cabling and online read-out to the surface is challenging for floaters and self-
contained recording units for temporary monitoring may be a reasonable alternative. NGI
has developed mobile solutions for recording differential pressure and dynamic motion of
suction anchors, see Figure 5-64. These standalone units may have a battery life of 3-6
months dependent on the recording schedule. The units can be deployed and recovered by
ROV and re-used on different anchors as long as they are outfitted with the required
outfitting (pressure port with receptacle and docking plate)
Figure 5-63 Self-contained suction anchor monitoring unit from NGI. The unit can also be used to
monitor settlement or static pull-out against a reference station at the seabed.
For piles, probably only dynamic motion would be relevant to monitor. For gravity
anchors dynamic motion, settlement and tilt may be relevant to monitor. The relevant
sensor configuration depends on anchor type, the monitoring units may also be fixed with
strong magnetic footings, see example in Figure 5-65.

Date: 2017-06-15
Rev.no.: 3
Page: 97
Figure 5-64 NGI's motion monitoring unit fixed with strong magnets, in this case deployed on an
unstable submarine wreck to detect possible motion and tilt.

Date: 2017-06-15
Rev.no.: 3
Page: 98
5.12 Corrosion
Based on experience from installed OWT’s in Europe, the corrosion inside the mono piles
is the most serious problem affecting the operational life of the foundation and consequently
one of the most relevant application for remote monitoring and follow up (reducing the need
for field trips and visual inspection).
Figure 5-65 Corrosion is a severe problem for long term operation in the offshore environment, the
photo shows the interior of a mono pile (courtesy Force Technology)
Figure 5-66 Corrosion and coating repair in the field is extremely costly and difficult

Date: 2017-06-15
Rev.no.: 3
Page: 99
Corrosion can be monitored on the OWT foundation by means of three approaches (usually
combined):
1. Monitoring the corrosion potential (environment)
2. Monitoring of the cathodic protection system
3. Direct monitoring of corrosion rates on dummy elements
5.12.1 Corrosion potential (environment)

Many foundations are designed under the assumption that no replacement of seawater
occurs in the closed compartments. However, experience has shown that seawater may enter
the mono piles to a varying degree for example through the J-tube gasket. The renewal of
oxygen should preferably be monitored using a dissolved oxygen probe because oxygen
represents the driving force for corrosion. Other instruments for monitoring of
environmental parameters affecting the rate of corrosion may be:
• pH meter
• H2 and H2S sensors (sulfates)
• Salinity probe
• Air temperature sensor (reference)
• Water temperature sensor (reference)
• Water level sensor (reference)
Dissolved oxygen can be measured either by electrochemical or optical sensors.
Optical dissolved oxygen sensors tend to be more accurate than their
electrochemical counterparts and are ideal for long-term monitoring programs due
to their minimal maintenance requirements. They can hold a calibration for several
months and exhibit little calibration drift.
Figure 5-67 Optical dissolved oxygen probe from Anderaa
Many suppliers offers multi parameter instruments

with the ability to customize the parameters to be
monitored using different plug-in probes.
Figure 5-68 Multi parameter seawater probe from YSI

Date: 2017-06-15
Rev.no.: 3
Page: 100
5.12.2 Cathodic protection (CP) survey

Monitoring of the cathodic protection (CP) can be relevant due to uncertainties about
regional and tidal effects on the outside, and effects from site-dependent microbiology,
sediment composition and aeration of seawater on the inside. CP monitoring is based on
potential measurement with durable reference electrodes and/or probes measuring the actual
corrosion rate of the protected steel. The protection current running between the structure
and the sacrificial anodes is also recorded as this parameter provides valuable information
about development in corrosive conditions as well as anode consumption.
In the simplest form, cathodic protection surveying of fixed offshore platforms is achieved
by the so called "dipping/drop cell" technique, dipping a reference electrode (drop cell) into
the sea and measuring a steel/sea potential with respect to it via an indicating voltmeter and
a metallic connection to the topside steelwork, see Figure 5-70. During such surveys the
distance between the electrode and the structure should be limited and constant during
lowering of the reference cell from the surface this is difficult to achieve in open sea but
applicable inside a mono pile. Reference electrode systems for seawater are usually
equipped with Ag-AgCl (sensitive) or Zink (robust) electrodes.
Figure 5-69 Principles for CP survey using reference electrode (drop cell), Illustration (left):
Stoprust.com. Seawater Ag-Ag-Cl Reference electrode from Borin (right)

Date: 2017-06-15
Rev.no.: 3
Page: 101
For better precision the reference electrode can be positioned with divers or an ROV, see
Figure 5-71. The "Bathycorrometer" or "CP-gun", has a stainless steel tip for subsea ground
contact and a reference electrode 1"-2" from the ground, the potential is recorded by internal
voltmeter with readout. Most instruments have two reference electrodes such that faults can
be detected immediately (the same potentials should be read for both electrodes).
Figure 5-70 Diver operated CP gun from Polatrak (left) and ROV operated version (right). Remote
readout of the voltmeter is then usually arranged via the ROV's communication system
The Field Gradient Sensor (FiGS) developed by Force Technology is a non-contact CP
inspection tool that performs highly accurate measurements of electric currents in seawater,
see figure 5-72. The system is suitable for external CP survey with an excellent resolution
and detection level. The sensitivity enables the identification of corrosion problems and the
characterization of CP system status on pipelines and subsea structures, even when buried.
The electric fields (strength and direction) set up by the cathodic protection system are. It
can also quantify the current flow to buried parts of the mono pile. With this information,
the level of protection and remaining lifetime of the CP system can be evaluated. The survey
is quick and can be conducted by ROV's, divers or from the surface in calm weather.
Figure 5-71 Field Gradient Sensor from Force Technology

Date: 2017-06-15
Rev.no.: 3
Page: 102
Reference electrodes for permanent use must be fixed to the structure if located externally,
see Figure 5-73. Permanent reference electrodes should always be equipped with dual
element electrode for redundancy. If combined with current density monitoring, the
polarization of the structure can be monitored. The permanent units are hardwired to the
voltmeter and data acquisition unit topside, thus it is important that the cable is of adequate
quality for permanent immersion in seawater (nothing lasts longer than the weakest part).
An Ag/AgCl reference electrode provides very accurate potential data for monitoring
polarization of the structure. A heavy-duty zinc-electrode element provides a reliable, long-
term data source alongside the more accurate, but less stable, silver chloride.
Figure 5-72 Permanent reference electrodes for external use from Polatrak, the (right) version is
equipped with a 1m2 current density monitor.
For corrosion protection of larger steel structures, Impressed Current Cathodic Protection
(ICCP) are often used instead of passive CP with sacrificial anodes, see Figure 5-74. The
ICCP anodes are connected to a DC power source. An essential feature of ICCP systems is
that they constantly monitor the electrical potential at the seawater/hull interface and
carefully adjust the output to the anodes in relation to this. Therefore, such systems may be
more effective and reliable than sacrificial anode systems where the level of protection is
unknown and uncontrollable. The placement of the active anodes is however very important,
these should be located at some distance from the structure for better distribution of the
current field. Local concentrations of the current will reduce the overall protection of the
structure and may cause cracking of the coating (if present).

Date: 2017-06-15
Rev.no.: 3
Page: 103
Figure 5-73 Impressed Current Cathodic Protection (ICCP) system for active corrosion protection on
OWT foundations (Illustration Cathelco)
The thyristor based control panel supplies an impressed current which provides the driving
force ‘potential’ which ensures protection of the whole submerged structure and also
extends beneath the mudline.
Figure 5-74 Thyristor based ICCP control panel from Cathelco

Stand-off disc anodes are typically used for mono pile structures where the bracket can be
mounted on the transition piece. As the anode is at a distance from the structure the need
for a di-electric shield is eliminated. In other cases linear surface mounted anodes can be
used where the backing shield provides di-electric protection, see Figure 5-76. Ag-AgCl
reference electrodes are commonly used in ICCP systems on fixed foundations as they are
highly accurate and more stable at higher negative potentials. The reference system
monitors the potential gradient between two electrodes which are mounted on a bracket
fixed to the structure.

Date: 2017-06-15
Rev.no.: 3
Page: 104
Figure 5-75 Examples of ICCP anodes and reference electrodes from Cathelco
At any time the performance of the system can be checked and controlled from an on-shore
control room or on PC via internet link. If the environment below the waterline changes the
ICCP system can be immediately adjusted to take account of this. Secondly, a record can
be kept of the performance of the system demonstrating the integrity of the foundations.
This avoids the time and expense of carrying out offshore surveys.
ICCP systems have a much higher driving voltage than sacrificial anodes enabling the
current to reach a depth of up to 5 meters below the sea bed (mudline). This extends
corrosion protection to the buried foundations of the structure and combats microbiological
influenced corrosion (MIC).
5.12.3 Corrosion rate monitoring

Coupons (weight loss) are the direct technique providing reliable data including the option
of examining scale and corrosion attacks. The only drawback is the need for retrieval to
obtain data, slow response rate, and that only historical data are obtained, not real time data.
Corrosion rates vary in time so, in order to measure the actual corrosion rates and record
changes, techniques such as electrical resistance (ER) or electrochemical corrosion rate
measurements like linear polarization resistance (LPR) can be introduced. Even if remote
monitoring methods are applied, coupons should be installed as a reference and back-up
solution, see Figure 5-77.

Date: 2017-06-15
Rev.no.: 3
Page: 105
Figure 5-76 Coupons rigged inside a mono pile above and below the water line for direct monitoring
of corrosion rates
Remote monitoring of corrosion rates are based on measurements on electrodes exposed to
the corrosive media and not on the foundation structures.
Electric resistivity (ER) probes are equipped

with an element that is freely "exposed" to the
corrosive fluid, and a "reference" element sealed
within the probe body. Measurement of the
resistance ratio of the exposed to protected
elements is made as shown in Figure 5-78.
Reduction (metal loss) in the cross section of the
structural element due to corrosion will be
accompanied by a proportionate increase in the
element's electrical resistance.
Figure 5-77 Example and working principles for

ER probe

Date: 2017-06-15
Rev.no.: 3
Page: 106
The Linear Polarization Resistance (LPR)

measurement technique is based on the
fundamental concept that when a test electrode in
an aqueous environment is polarized by a small
voltage, the apparent resistance measured from
resulting current flow is inversely proportional to
the corrosion rate. The advantage with the LPR
probe compared to the ER probes is that the
corrosion rate is measured immediately as for the
ER probe metal loss must occur before the
corrosion rate can be determined. The LPR probes
have two polarization electrodes (anodic and
cathodic sides), by introducing a third electrode the
probe can be calibrated/tested, three element LPR
probes are therefore recommend for long term
operation. When immersed in water without Figure 5-78 Examples of 3 and 2 element
currents, for example inside a mono pile, protruding LPR probes
electrodes are recommended, see Figure 5-79.
If corrosion rates are low, ER and LPR probes will generally have a long functioning time
– whereas in the case of high corrosion rates or localized corrosion, the service life of the
sensor is shortened.
The galvanic probe (Figure 5-80) is an indirect

measurement very sensitive to oxygen ingress,
based on zero resistance amperemetry between a
steel probe and a noble copper or brass probe.
When installed just below the water level, it will be
sensitive to ingress of oxygen from both top and
bottom. The recorded output of this probe type is
galvanic current which can be transformed to an
approximate corrosion rate (mm/y). Rapid changes
in the oxygen level will be registered and the design
life of this type of probe can be long. Reduction
(metal loss) in the element's cross section due to
corrosion will be accompanied by a proportional
increase in the element's electrical resistance.
Figure 5-79 Galvanic probe from Emerson process

Date: 2017-06-15
Rev.no.: 3
Page: 107
The mud line inside mono pile foundations may represent a risk of highly localized
corrosion due to the combination of bacterial activity in the mud and macro galvanic
elements between the oxygen-containing bulk media and the oxygen-depleted mud zone.
There is no straight forward and standardized way of measuring corrosion in mud lines, and
exposure of simple coupons may overlook critical effects. Specially designed probes are
required such as full-length corrosion coupons or real-time monitoring devices that assess
the risk of macro galvanic effects, microbial corrosion (MIC) or Hydrogen Induced Stress
Cracking (HISC).
As discussed in the introduction to corrosion monitoring. One type of sensor or parameter

and one monitoring position or elevation are usually not sufficient in order to obtain a clear
picture about the corrosive state of the foundation. Therefore a set of sensors are usually
deployed and hooked up to a common data logger, see Figures 5-81 and 5-82. The most
common and important measurements/instruments in such systems are dissolved oxygen,
corrosion rate (ER or LPR probes) and reference electrodes.
Figure 5-80 Instrumentation system and data examples for corrosion monitoring inside a mono pile
from Zensor

Date: 2017-06-15
Rev.no.: 3
Page: 108
Figure 5-81 Permanent corrosion monitoring panel from MetriCorr containing different types of
corrosion sensors
5.12.4 Ultrasonic thickness gauging

The most common method for in-situ
measurements of thickness is by
Ultrasonic Thickness (UT) inspection. By
UT gauging determine the local thickness
of a solid element (typically made of metal
or concrete) is determined based on the
time taken by the ultrasound wave to
return to the surface. Dual element
transducers are recommended for
corrosion inspection. These transducers
generate sound waves with one element
and receive with another – in a ‘V-path’
orientation, which increases sensitivity
when examining corroded or pitted back Figure 5-82 Principles for Ultrasonic thickness
walls, see Figure 5-83. gauging (Illustration INTECH NDE)

Date: 2017-06-15
Rev.no.: 3
Page: 109
UT gauging requires good contact with the material (usually gel), however for practical
field work dry coupling is used by adding a soft material such as neoprene rubber to the face
of the transducers. The UT accuracy is good (0.1 mm and less) and can be achieved using
standard timing techniques. There is no need to remove the coating of the metal however
measurements are affected by rust and are therefore made from the intact or cleaned side
(usually the coated outside) to determine how much metal has been corroded at the back
wall.
UT spot checks can be done manually were access is possible, for scanning of larger sections
crawlers are normally used, see Figure 5-84. By means of robotic scanning inaccessible
sections can be inspected and the surface coordinates for the mapping are recorded
automatically.
Figure 5-83 UT Crawlers for UT scanning (inserted pictures). The outside of for example the TP is
not easily accessible, the large photo shows sand blasting at Scroby Sands wind farm with tie-back
to the structure by a magnet system from Abfad. UT crawlers can be used for external UT scanning
of TP's provided calm weather.

Date: 2017-06-15
Rev.no.: 3
Page: 110
Figure 5-84 UT Data from Crawler scans, the left example is a 2D presentation from two scanned
profiles along the TP waterline. A maximum pit depth of 2.8mm is observed on the inside of the left
scan (UT scans performed by Force Technology). The right example shows 3D UT mapping of
corrosion (C-scan) compared with observations at the back wall (Courtesy: James Fisher NDT)
UT measurement can be performed under water, diver or crawler operated, see Figure 5-86.
Figure 5-85 Diver operated ultrasonic Thickness Gauge from Cygnus instruments
Electromagnetic Acoustic Transducer (EMAT) is a transducer that employ a
magnetostrictive effect to transmit and receive ultrasonic waves. EMAT is an ultrasonic
testing method which does not require direct contact or coupling substance, because the
sound is directly generated within the material (steel to be tested). As an emerging ultrasonic
testing (UT) technique, EMAT can be used for thickness measurement, flaw detection, and
material property characterization. However, limitations may apply for thicker steel
elements.

Date: 2017-06-15
Rev.no.: 3
Page: 111
Figure 5-86 EMAT Transducer from Olympus (left) and comparison between traditional
piezoelectric and EMAT UT. Non-contact EMAT measurements can be done from the corroded side
5.12.5 Acoustic Emission Monitoring

An alternative and relatively new approach for corrosion monitoring may be acoustic
emission measurements (AEM) for corrosion fatigue monitoring. AEM can be utilized for
both active corrosion and active crack growth monitoring.
Discrete acoustic events can be located by time of flight, (similar to seismic activity monitor
for earthquakes) and clusters of high location densities can be found immediately, also low
level corrosion can cause non-locatable Acoustic Emission activity. Acoustic waves caused
by active corrosion and/or fatigue cracks can propagate either in the metal to an acoustic
emission sensor being directly mounted on the surface or through the liquid to an acoustic
emission sensor immersed into the liquid. The sensors are normally of low cost piezo-
electric type, see Figure 5-88.
Figure 5-87 Piezo-electric AEM transducers (left), array of transducers mounted on a tank wall and
3D plot showing intensity and location of recorded acoustic emissions

Date: 2017-06-15
Rev.no.: 3
Page: 112
6 Data acquisition system and interface electronics

The data acquisition system and interface electronics provides power and reads data from
all sensors. The data is stored on disc and/or transferred to shore by available links installed
for operating the wind turbine, see Figure 6-1.
Sensors with analogue output must be interfaced with AD’s and multiplexed, SHM
instrumentation usually includes a mix of analogue and digital sensors, special sensor
systems such as resistance strain gauges or vibrating wires may require special interfaces
(preamps, excitation and counting modules, etc.). In some cases, it may be cost effective to
use subsea AD/MUX in order to reduce the extent of cabling to the surface and obtained
better noise immunity.
Figure 6-1 Components and hierarchy for a SHM data acquisition system on an OWT foundation

Date: 2017-06-15
Rev.no.: 3
Page: 113
6.1 DAQ system

Low pass filtering, re-sampling and processing of statistical data is performed in the topside
Data Acquisition (DAQ) computer in order to condition and reduce the amount of data
transferred to shore. Note for dynamic SHM parameters a minimum re-sampling rate of
20Hz is recommended.
A local data storage and battery UPS (Uninterrupted Power Supply) should also be included
in the offshore Data acquisition system in case of power or data link failures. Sensor
configuration including settings of conversion factors such as sensitivity and offset can be
performed from the data acquisition computer. Other logic functions may be time
synchronization, statistical data reduction and control of re-sampling rate. For example,
more frequent logging of recorded sensor data and storage of complete time series in stormy
weather. SHM recording schedules may also be linked to turbine operation settings
(SCADA system).
If possible, all relevant data for SHM should be merged, this includes relevant turbine data
from the SCADA system (if possible) and Metocean parameters from other systems
onboard. These data can either be imported directly to the SHM data acquisition system in
order to provide common and synchronized data files or merged via land based servers. In
some cases, the opposite requirement may apply, namely exporting the SHM data to the
SCADA system such that the same data transfer link to shore can be used. Some SCADA
suppliers are aiming for integrating the SHM system with the SCADA system which is
challenging as the type of measurements and analyses required for complete SHM
(including the seabed foundation) is very different from condition monitoring of the drive
train.
Depending on the bandwidth of the data transmission link to shore, data reduction may be
required, especially if traditional GSM modems are used. If data is transmitted through
optical fiber cable usually no practical bandwidth limitations applies for transfer of SHM
data. An onshore server receives the data for further distribution and processing. How to
handle large amount of SHM data is a very important and big topic in itself, which may
include both neural networks and machine learning. The data processing part is however
not addressed in this study.
For remotly operated small to medium size SHM systems with standard sensors the
acquistion system can be based on smaller loggers such as the CompactRIO from National
Instruments, see Figure 6-2. This system has embedded controllers with two processing
targets: (1) a real-time processor for communication and signal processing and (2) a user-
programmable to implement high-speed control and custom timing and triggering directly
in hardware.

Date: 2017-06-15
Rev.no.: 3
Page: 114
Figure 6-2 Small scale DAQ system based on CompactRIO logger from National instruments
It should be noted that space is often limited on board the OWT foundation and handling
can be difficult. It is therefore important that the data acquisition cabinets are as compact as
possible or split into smaller modules
Figure 6-3 Example of compact wall mounted data acquisition cabinet inside TP serving small scale
SHM systems (Force technology)
Data acquisition for larger SHM system (100 + sensors) must be based on logging system
capable to interface and process large amount of sensor channels and small loggers (type
CompactRIO) become insufficient. Figure 6-4 shows a high-end data acquisition system
delivered by NGI based on HBM's QuantumX data acquisition modules and an industrial
grade computer. The system is fully remote operated and can handle hundreds of sensor
channels with different output signals and logging rates.

Date: 2017-06-15
Rev.no.: 3
Page: 115
Figure 6-4 Large scale and modular SHM data acquisition system from NGI. Right, termination of
strain gauges cables requires space and focus (Photo NGI archives)
Although the Data acquisition
systems are remotely operated, a
local operator station is usually
integrated in the on-board system for
checks during installation and
service.
Figure 6-5 Example of processed data

trends (NGI-VISMON) for local and
remote visualization of data and system
configuration
To allow for data conditioning, the sensor signals should be over sampled. The actual
sampling rate depends on the monitoring application. As a rule of thumb dynamic data
should be sampled at 20-25 times higher rate than the typical frequency of the measured
application in order to properly track the dynamic response and avoid aliasing errors, see
Figure 6-6. For a mono pile the dynamic response of interest covers frequencies up to 1 Hz,
dynamic data should then be logged with minimum 20 Hz sampling rate. As a rule of thumb
static data such as settlement and tilt can be logged at 1-2 Hz sampling rate, whilst
cyclic/dynamic data should logged at minimum 20 Hz. If vibrations are of interest logging
may be performed at even higher rates. Over sampling also provides better basis for
conditioning (filtering) of re-sampled data.

Date: 2017-06-15
Rev.no.: 3
Page: 116
Figure 6-6 Real-time trace of a cyclic signal based on a logging rate corresponding to 25 samples
per period
General considerations when planning a data acquisition system:
Parameter(s) to be measured (analog or digital sensor interface)
Duration of measurement and power requirements
Environmental (EM noise, IP protection)
Resolution, e.g. 8, 10, 12, 16 bits logger and number of channels
Interface type / communication
Logging rates
Embedded processing, data reduction
Storage capacity
Size
Programming interface
Costs
The Analog-Digital converter connects analog sensors to a data logger. There are several
analog sensors with voltage or current output. As the A/D converter only converts voltage
to digital numbers, sensors with current output have to be converted to a voltage signal by
using a resistor. High quality resistors with low thermal drift and good long-term stability
should be used in order to maintain the accuracy in the measurements.
Most A/D converters can be used as single ended (typical for 2 wire sensors) or fully
differential inputs (typical for full bridge sensors). A 24 bit A/D converter may increase the
sensor resolution but not the sensing accuracy. A rule of thumb is to choose an A/D
converter giving a resolution 10 times better than the sensors accuracy.

Date: 2017-06-15
Rev.no.: 3
Page: 117
6.2 Signal transmission

The following analogue sensor signals are relevant for SHM sensors:
Voltage, usually ±5 or 10 V, sensor with low voltage output are EM; noise sensitive
and should not be used)
Current loop, normally 4-20 mA
Frequency, vibrating wires or quarts resonators
The current loop and frequency signals are not affected by cable resistance and suitable for
transmission through longer cables. In some applications, the analog sensor signal has to be
amplified or reduced before logging. The main reason may be to adjust the sensors output
to the A/D input. Avoid analog sensor signals exceeding the range for the A/D as this may
fail the unit and the readings will be out of range (clipped).
Many sensors have serial I/O (Input/output) and communicate by using special protocols.
Serial data transmission has better immunity to EM noise than analog signals. The
traditional serial signal formats are RS232 (only for short cables) or RS485 lines (balanced
signal for transmission in longer cables). A range of different protocols exists for serial data
communication, the most common for sensor communication and interfaces are RS422,
CANbus, Profibus, HART and USB. For digital signals, the data logger must have a digital
interface and driver software installed fitting the used protocol. For rare protocols
implementing the driver software may be time consuming if the documentation from the
supplier is limited. It is becoming standard that sensors and/or interface modules have
Ethernet output with larger bandwidth allowing for faster sampling. A data acquisition
system based on network interfaces is flexible and enables easy to hook-up different
interface modules with Ethernet output (IP address) such as FBG interrogators or other
streaming units. As discussed earlier, fiber optic sensor signals are completely immune to
EM noise and suitable for longer cables. However, splicing and amount of connectors must
be limited as signal loss may occur.
In addition to the EM noise from the power grid (50-60 Hz), near-field electromagnetic
fields with higher frequencies is emitted by the generator and switching components in the
turbine. The electromagnetic interference (EMI) from the high frequency components may
disturb signals in the kHz band. For most SHM monitoring applications frequency
components above 50 Hz are not relevant and EM noise can be removed by Low pass
filtering, for examples with 25 Hz cut off frequency. It is however important that the
amplitudes of the noise frequencies above the filter cut off do not saturate the transmitted
signal as this will make signal conditioning useless. As long as shielding (for example
twisted pairs) and robust signal format is used NGI has not experienced any problems with
EMI for SHM systems installed on offshore wind turbines. Also note that the attenuation of
EM fields is strong in seawater. Consequently submerged instruments are subjected to very
low EMI, though the cable connection to topside may be exposed.

Date: 2017-06-15
Rev.no.: 3
Page: 118
7 Practical advices for SHM instrumentation of OWT

foundations
7.1 Introduction
A successful instrumentation project is when the monitoring system provides the data
required by the end user. This involves multidisciplinary coordination and understanding,
the main considerations and disciplines such as:
• Proper definition of objectives for monitoring (what data is needed?) and realistic
expectations of what can be achieved
• Identifying appropriate monitoring approach and linked parameters
• Type of sensors and data recording infrastructure (instrumentation hardware and
DAQ system)
• How and when to install the instrumentation
• End users and interpretation of data (Metocean, Geotechnical, Structural and
Environmental disciplines)
All links in this chain are equally important. The instruments can work perfectly but the
output can be worthless as the installation changed the in-situ conditions. The selected
hardware may not be adequate for the monitoring application or simply not mounted
properly on the structure or at a position which not is representative for the desired
monitoring purpose.
It is important to maintain focus on all components in the instrumentation system such as:
• Sensing mechanism (interface to the media to be measured)
• Sensor (type of instrument)
• Location and configuration of sensors
• Permanent, temporary or mobile system
• Power supply
• Signal transmission (analog, digital, fiber optic, wireless, required baud rate, EM
noise, environmental aspects dry/wet)
• Installation method and place (yard, in the field, retrofit, under water, etc.)
• Operational life (protection, waves, corrosion, biofouling....)
• Data acquisition system
• Data base and interpretation
The balance between costs, backup and reliability of the monitoring system usually depends
on how important the measurements are for the operator and how much risk the contractor
is prepared to undertake. Due to demanding environmental and operational conditions as
well as interface constraints, the instrumentation design must often be custom made for the
application.

Date: 2017-06-15
Rev.no.: 3
Page: 119
For SHM instrumentation the main factors affecting the design may not always be to obtain
measurements with the highest precision but rather:
• Limited time for design and manufacturing the monitoring system (delivery times
from sub- suppliers).
• Minimized time for installation in the yard or in the field (being at the critical path
can be expensive).
• Handling and installation constraints, such as deployment in the sea water.
• Accessibility and safety (especially under water and also at heights).
• Retrofit required to structures already installed at the field.
• Constraints for cable routing on the structure.
Offshore operations are costly. As a rule

of thumb the costs for offshore and
subsea operations are three orders more
expensive than on land. The offshore
installation work can cost more than the
instrumentation hardware itself. Thus,
as much as possible of the
instrumentation should be prepared and
installed at the yard, bearing in mind
that many times the instrumentation
(especially cables) might be exposed to
the highest risk of damage at the yard.
For critical monitoring system, the
instrumentation must include a robust
strategy for backup and redundancy.
Figure 7-1 The yard can be a dangerous place especially for unprotected cables (photo: NGI)
Although a successful installation of a permanent monitoring system many times depends

on the “Mood of Mother Nature” during offshore operations, the human factor is still
probably the most frequent reason for failures. Skilled and thorough instrument engineering
supporting during the installation is a paramount factor for a successful monitoring project,
comparable with the importance of a qualified pilot for a safe flight. The true skills of the
supplier are proven when things fail or do not happen according to the plan and judged by
the ability to rectify the problem in a stressful situation.

Date: 2017-06-15
Rev.no.: 3
Page: 120
7.2 “Ten Commandments” for offshore and subsea

instrumentation
Some of the most important aspects and recommendations for successful execution of a
monitoring program are summarized in the following sections.
7.2.1 Planning and Design

The first thing to be done is to perform a thorough analysis of sensitive/critical parameters:
Where-, what- and how to monitor; i.e. develop an instrumentation philosophy for the
project. Budget constraints may set limitations to the amount of hardware available, and it
is therefore very important to have a clear opinion with respect to priorities, i.e. where
money can be saved and where money shall not be saved.
It is almost impossible to do a “Fit for purpose” design without having first-hand experience
from offshore installation work and subsea operations. As the development time usually is
limited for working prototypes, the planning and considerations of all installation steps and
contingencies is very important. Many times the evaluation of “What will work and what
will not work” is purely based on experience. The measuring system “surviving” the
construction work and offshore installation, i.e. ensure it works offshore, has first priority
and many times an optimal setting for the best measurements must be sacrificed. Probably
the most demanding application with respect to system “survival” is pre-installed
instrumentation of driven piles. Not only the sensor integrity is important, cable protection
is often more critical.
For both yard and offshore installation works, proper planning is essential as the given time
slots for doing the job often are limited. This also includes taking into account the
Environment, Safety, Access and Training required. It is essential that the design of the
system allows the onsite installation and commissioning to be carried out as smoothly as
possible. Offshore retrofit of instruments is challenging (especially under water or at height)
therefore the instrumentation should if possible be installed in advance, preferable at the
yard were access and handling assistance can be provided (cranes, scaffolding etc.)
7.2.2 Harsh conditions and Robustness

The conditions at offshore wind farms are usually rough and the environment that the SHM
systems have to be installed and commissioned in is difficult and hazardous, each turbine is
subject to potential extremes of wind, wave and temperature parameters which can change
quickly. Instrument pre-installed on the foundation structure may imply that the instruments
will be subjected to large forces and vibrations for example during pile driving, see Figure
7-2.

Date: 2017-06-15
Rev.no.: 3
Page: 121
Figure 7-2 Instruments mounted on a pile driving template. The bolts on the mounting bracket
where eventually sheared of during pile driving, however the instrument survived. Loosening of
bolts and fittings were observed and special arrangement to secure all fixtures was required. If the
OWT foundation is towed in the water submerged instrument will be exposed to large drag forces
and vortex vibrations (source NGI archives)
For typical wind farm locations, the
waves may not be the critical factor
for subsea intervention by ROV but
rather the underwater currents (tidal)
and poor visibility. The experience
from many European sites is that
subsea intervention is only possible
when the tidal current is turning i.e.
two times a day during slack tide.
For safety reasons diver intervention
should be avoided. A diver usually
has better control than a ROV but the
working conditions and abilities
cannot be compared with installation
Figure 7-3 Do not trust this fellow for fine tuning your
work at the surface.
subsea instruments

Date: 2017-06-15
Rev.no.: 3
Page: 122
7.2.3 Backup and contingency

Sensor costs are usually small compared to infrastructure for the instrumentation (pressure
enclosures, cabling etc.) and the operational time for replacement is expensive. Therefore,
cutting costs by eliminating back-up sensors may not make any big difference in total costs
but regrettable if something goes wrong. Consequential failures must also be considered,
such as configuration of cables and sensors, how much of the system is affected by one
ruptured cable? Contingency is a popular word for offshore operation schemes, there must
always be a plan for what to do if things do not happen according to the plan.
7.2.4 Pressure integrity and corrosion

Pressure integrity of sealed components is obviously important for subsea applications. This
aspect can be divided into two categories:
1. Leakage (water ingress)
2. Structural integrity (pressure collapse)
Most failures related to pressure integrity occur due to leakage and the risk for water ingress
may even be bigger in shallow water compared to deep as many sealing parts (O-rings,
connectors, etc.) actually seal better at higher pressure. Structural collapse often occurs
because it was not considered. For example forgetting to drill vent holes in tubular members
for pressure compensation when deployed in deeper waters. However, for subsea
instrumentation related to offshore wind turbines, the water depth is limited and structural
collapse due to the external pressure is less of an issue.
Figure 7-4 Water ingress into a subsea instrument enclosure (left) and pressure integrity test (right)
in a tank filled with water normally at 1.5 x the ambient pressure at site (Photos: NGI archives)

Date: 2017-06-15
Rev.no.: 3
Page: 123
The offshore environment is very corrosive both above the sea level and in shallow water,
especially in the splash zone. Corrosion is therefore important to consider in design of the
instrumentation hardware, especially for long term deployment. Even if the rules of thumb
are followed such as using noble metals in the galvanic series, not mixing metals with
different galvanic potential or using cathodic protection, the results may be discouraging.
Scratches in the anodized aluminum surface or anodes with oxidized surface may
compromise the protection. Mixing of metals or fixture to larger steel structures without
galvanic isolation can result in rapid corrosion even if the metals themselves are corrosion
resistant. NGI’s practice is to keep it simple and only use for example stainless steel in
combination with thermoplastic materials such as Delrin, see Figure 7-5.
Figure 7-5 “Bright and shiny” instrument unit mainly in stainless steel prior to deployment (left) The
same unit heavily corroded after 2 years deployment in seawater (middle). Corrosion “safe”
instrument housing made of Delrin with outfitting in stainless (right)
7.2.5 Functional Testing and calibration checks

The importance of functional testing can never be exaggerated. If possible, hooking up the
instruments in a representative configuration can save time for fault seeking and prevents
uncertainties during field installation (especially for signal addressing and sign convention).
Many times the obvious is forgotten or wrong. These checks are therefore nicknamed as the
“idiot test” internally at NGI.
Other instruments such as Inclinometers are simply difficult to calibrate (offset) offshore
when things are moving on the vessel and must thus be checked at the quay before leaving
port.

Date: 2017-06-15
Rev.no.: 3
Page: 124
Figure 7-6 Ex works functional testing of NGI's level monitoring system mimicking the configuration
and use on a pre-piling template (Geosea)
7.2.6 Flexibility and Installation friendly

As offshore installation time is expensive it is important to simplify the operations as much
as possible minimizing the time used offshore. This is especially crucial when under water
intervention is required for retrofit applications. Flexibility implies options to adapt the
equipment for unexpected conditions or changed set up for installation as well as access for
easy replacement of damaged parts.
Figure 7-7 Accessibility is important for maintenance and replacement (photo: NGI archives)

Date: 2017-06-15
Rev.no.: 3
Page: 125
7.2.7 Safety during Onshore and offshore installation works

Safety is a paramount issue when working on wind turbine structures, this applies both on
the construction yard and offshore. There are many hazards around when working on or
accessing these structures. In Europe, the following training is required:
• Offshore survival training and medical (offshore works)
• Offshore access (offshore works)
• Confined space training (offshore and yard)
• Working at heights (offshore and yard )
Special training is also required for access to the

turbine. Due to many activities performed
simultaneously at the yard such as lifting
operations, grinding, welding etc. special
attention must be paid. As instrument
technicians/engineer normally only pays shorter
visits to the yards, familiarization is important.
The offshore wind turbines are remote in
location and have to be accessed by work boat,
the transfer from the boat to the turbine tower
can be difficult and there are vertical ladders to
climb both externally and internally to gain
access to the necessary parts of the structure.
Figure 7-8 (above) A dangerous and uncomfortable workplace. (right) Offshore access to the OWT
foundation requires physical fitness and high safety awareness.
Often the installation work has to be done inside confined spaces or at height, appropriate
safety equipment is then required, see Figures 7-9 to 7-11.

Date: 2017-06-15
Rev.no.: 3
Page: 126
Figure 7-9 Access for installing sensors at the yard: Left, inside mono pile (photo: Allnamics) and
right, on top of suction buckets (photo: NGI)
It is important to involve the construction yard and/or offshore contractor for review of and
input to the installation plan and interfaces. Usually, the rate of success can be raised
significantly if the important details are thoroughly explained and understood by the
contractor. Some requirements for installing the instruments can indeed appear as weird and
easily neglected for someone who is not familiar with the “Art of instrumentation”.
Figure 7-10 Left, darkness and confined space when installing extensometers inside the TP just
above the sea water level (Photo: Strainstall) or right, when strain gauging a jacket leg (photo: NGI)

Date: 2017-06-15
Rev.no.: 3
Page: 127
Figure 7-11 Great view when working at height installing a wave radar on TP deck (Photos: NGI)
In order to fulfil the safety requirements, significant and costly training is required (in
Europe) to be undertaken by personnel that are involved in the installation and
commissioning of the SHM systems on the Wind Turbine structures. Still the alertness and
judgement of the field personnel themselves are the most important factors preventing
accidents.
7.2.8 Delivery time and coordination of installation work

Installing instrumentation at the construction yard or offshore is usually a small part of a
bigger scheme and the logistics involved must fit the overall plans. Waiting must be
expected for access but do not expect the yard or installation vessel to be waiting for you.
As cables are usually fabricated in batches and many times is the first item required for
installation, delivery times are often on the critical path.
7.2.9 Cable routing

Cable routing on the structure is usually the trickiest part with respect to coordination and
interfaces, usually instrumentation and cables are not included in the structural fabrication
drawings and plans. Although many times the cable routing ends up with ad hoc solutions
the general cables routes must be defined such that sufficient cable lengths can be ordered.
If possible the cables should be installed after grinding, welding and painting is completed
at the yard. It is strongly recommended to visit the yard in advance and discuss routing
details, sketches on photos are useful but still (unless incorporated in the structural
fabrication drawings) expect that details must be adapted on site, see spFigure 7-12.

Date: 2017-06-15
Rev.no.: 3
Page: 128
Figure 7-12 Example of cable routing details indicated during yard visit (left) and final installation
(right), photo from NG's archives.
Challenges with cable routing can be avoided using wireless sensor systems, however
wireless communication do not work well between closed steel compartments and battery
power supplies would be required. For under water communication acoustics, EM or optic
(OLED) data transmission can be used, but such systems are either power hungry or have
limited transmission range in addition they are more complicated and expensive compared
to direct cabling. For retrofit under water it may however be the only option, unless
standalone systems with integrated data loggers ("black boxes") are used.
Figure 7-13 Long cables makes big bundles and even less work space before final routing (photo
from NGI archives)

Date: 2017-06-15
Rev.no.: 3
Page: 129
External cables must be properly fixed to the structure, this is especially important for cables
routed through the splash zone as the slamming forces from waves will impose violent
battering on the cables, see Figure 7.14. Normally the cables are routed through a solid steel
conduit through the splash zone.
Figure 7-14 Battering wave action which can be expected in the splash zone (in the North Sea)
7.2.10 The Devil is in the details and Murphy's Law

It is a fact that most problems or malfunctions are caused by small details rather than major
faults in design or hardware. Therefore, "Paying attention to the details” is necessary for all
instrumentation tasks and especially important for monitoring systems which not are
accessible after installation and consequently not possible to repair or rectify when in
operation. According to Murphy's law, "Anything that can go wrong, will go wrong".
Therefore, plan for the unexpected, double check and check again.
The following example is taken from the internet (http://www.fino3.de/en/research/pillar-

foundation). A comprehensive research program was executed for the mono pile Metmast
(FINO 3), the project included extensive monitoring of the embedded pile both during
driving and operation. The pile was extensively instrumented with Geotechnical
Measurement Stations for Offshore Ground Structures (GEMSOG’s), containing
inclinometers, piezometer and total stress cells. The GEMSOG’s were mounted along the
outer pile wall in rows above each other, see Figure 7-15.
The instrumentation was nicely done with a completion which should survive driving and
under water long term immersion. However, the readings of total pressure during driving
quickly revealed an important detail for successful measurements of radial earth pressures
(effective stress) which was not taken into account when designing the monitoring system,
see figure 7-16.

Date: 2017-06-15
Rev.no.: 3
Page: 130
Figure 7-15 FINO 3 Mono pile with "GEMSOGS" instrumentation
Figure 7-16 Recorded radial earth pressures during driving of FINO 3 mono pile

Date: 2017-06-15
Rev.no.: 3
Page: 131
As seen from the plotted earth pressure recordings along one instrumented section of the
pile (Figure 7-16), only the instrument close to the tip is reading increased earth pressure
when the pile is driven deeper into the ground. All sensors above show more or less a
constant value (not representative for the effective radial stresses around the pile).
The explanation is straightforward as the “GEMSOG’s” instrument units were fixed and
protruded outside the mono pile (acting as driving shoes) a slot was generated in the soil
above when the pile was driven. Consequently, only the deepest GEMSOG did measure
representative earth pressures as the units above were affected by the groove created in the
soil when the sensor unit at the tip was penetrating. Earth pressure cells must be mounted
flush with the pile wall.
More details about lateral earth pressure instrumentation are discussed in Section 5.8.

Date: 2017-06-15
Rev.no.: 3
Page: 132
7.3 Executing a larger SHM project

As discussed earlier there are many items to consider executing a monitoring project and
the solutions which work best are many times based on long term experience (including
failures). Thus, for larger projects competent expertise is required, usually organized as a
System integrator or Main instrumentation contractor.
The purpose with this role is to design and integrate a system which will provide an
integrated monitoring solution meeting the objectives and requirements for data as specified
by the client. The system integrator should be able to select the most suitable components
from different sub-suppliers, design a suitable configuration, test and install the system on
the structure.
Figure 7-17 Organization and flow chart for execution of a monitoring project
7.4 Guidelines for selection of sensors

7.4.1 Basic considerations
Parameter to be measured: Determine the best method to obtain the required
parameter(s) including how to get the desired accuracy.
Specifications: Range, resolution and accuracy required. Cost is usually a function
of the specifications - choose the specifications appropriate for the requirements and
not simply the very best sensor on the market.
Signal type: Which signal type (voltage, current, frequency, digital or optical) is
best suited for this particular application? This has implications also for choice of
cabling and interface electronics.

Date: 2017-06-15
Rev.no.: 3
Page: 133
Priority: Which priority do you give to this particular measurement? This may
govern the type of equipment you choose w.r.t. price and quality.
Environmental: The environmental conditions must be taken into account when
choosing materials, ruggedness of enclosures and mounts etc.
Duration: For how long shall the measurement program last? Type of equipment,
choice of materials, etc. will depend on this. Bear in mind, however, that a
successful monitoring program which gives interesting data is often prolonged.
Apply a reasonable safety factor.
Sensor materials: Requirements regarding corrosion, pressure, size, electrical
effects etc.
Sensor manufacturer: Previous experience with supplier.
Cost and delivery time: Always to be considered.
7.4.2 Sensor accuracy

Sensor accuracy is often specified as the sum of the following errors, see also Figure 7-18.
Non-Linearity
Hysteresis and repeatability
Drift
Note that the sensor resolution does not
give any information about the
accuracy. In many cases mounting and
how the sensor is used are more
important for the overall precision in
the measurements than the sensor
accuracy.
Figure 7-18
Definitions of sensor specifications
affecting the overall accuracy
The maximum error is stated as an Engineering value or as a percentage of sensor full

range/scale (FS). If only a portion of the sensor full range is utilized, better accuracy can
often be obtained by re-calibration for the utilized (limited) range or taking reference or
offset readings.

Date: 2017-06-15
Rev.no.: 3
Page: 134
In conjunction with deformation and position measurements it is very important with

minimum slack in the attachment point and/or that the reference position/or environment
does not change. For example, the use of high precision inclinometers is meaningless if the
alignment and fixture to the structure not is appropriate.
7.4.3 Sensor backup and redundancy

Back-up is defined as an identical extra system in case the main system should fail.
Redundant system is different (maybe simpler) but shall provide similar functions as the
main system.
Executing a long-term monitoring project in deep waters involves a significant amount of

money. The costs of for example extra sensors (back-up or redundant) is often minor
compared to total costs. It is therefore strongly advised to not compromise on back-
up/redundancy in order to save costs as that will increase the risk failing with the monitoring
task.
Identify which parts of the system which requires redundancy. This is usually related to risk
of damage and accessibility for repair. e.g.:
Sensors embedded in the seabed (not accessible): Need redundancy
Topside and subsea sensors accessible for replacements: May not need redundancy
Topside Logging cabinet: Does not need redundancy.
Also sensor redundancy or supporting instrumentation will enhance the quality of the data
readings. For example if one sensor shows a trend development the possibility of drift in
the sensor cannot be disregarded, see Figure 7-19. However, if two different sensors show
the same trend it is more or less certain that the response is real.
Figure 7-19 Redundant sensors (A and B) - increased data reliability

Date: 2017-06-15
Rev.no.: 3
Page: 135
7.5 Guidelines for purchase and calibration of sensor systems

7.5.1 Receipt of sensors
Most sensors that are commercially available come ready calibrated from the manufacturer.
However, errors by the manufacturer and transport damage may occur, and it is therefore
important to check the sensors on receipt.
Function test the following on receipt of equipment (do this immediately when receiving
equipment):
Visual signs of damage
Power up sensor and check signal
Sign checklist after completion for Quality Control (QC) documentation
Calibration checks:
Calibrate 1-2 steps within range, check signs and zero point
Verify this is in accordance with specifications from manufacturer
Sign checklist after completion for QC documentation
7.5.2 Customized calibration

For instrument systems where high accuracy is required and only a limited part of the total
range will be used, it is often advisable to perform a re-calibration of the sensor within the
limited range which is relevant for the application. Special calibration may also be advisable
when operating at for example low temperatures, usually it is sufficient to check temperature
offsets. In some cases, the fixture or measuring arrangement may be the critical factor for
the overall measuring accuracy, in such cases it is recommended that a calibration check is
done on the complete system is performed (if possible) on a dummy set-up.
7.5.3 Functional and in-situ calibration tests

For sensors built into a system, calibration simulating the use and function of the system
should be performed as this will be representative for the monitoring application and include
all aspects affecting the overall accuracy. This type of functional testing/calibration is often
the most important check of the measuring system. In some cases, field calibration is
required. This is normally limited to in-situ zero point check of for example systems based
on pressure measurements (piezometers) or offset readings after inclinometers,
extensometer or strain gauges have been mounted on the structure.

Date: 2017-06-15
Rev.no.: 3
Page: 136
If possible, the sensors should be in operation and data recorded during installation. The
sensor response should be carefully logged and as observed values may be used directly as
new offsets in the data acquisition configuration files. Many times the readings taken during
installation provides an important baseline for later monitoring and data interpretation. See
Figure 7-16.
Figure 7-20 (left) Real-time display showing piezometer response directly after seabed installation
(in this case negative pore pressures were expected as the instrument was installed in swelling
clays). From NGI archives. (right) Tracking of fiber optic strain gauge during pile driving, these
records provide important information about sensor integrity after driving and the recorded
stresses provides an essential baseline for subsequent strain recordings during loading, Example
from validation tests of FBG sensors performed by UCD Doherty et al.
For larger instrumentation systems, it is required that the sensors ID’s and line/cable routing
is properly described in hook-up diagrams with sensor, cable and channel ID’s. A master
list showing all ID’s, as well as conversion factors is essential for hook-up and configuration
of the data acquisition system. Sensors and cables should be properly tagged, it is useful to
show the sign convention on sensor enclosures (for example inclinometers) and tag long
cables in both ends. Even with careful documentation and marking it is strongly
recommended that the sensor response is checked during field hook-up making sure that
both signal address and sign convention are correct. This may save a lot of questions when
reviewing and interpreting data during later operation.

Date: 2017-06-15
Rev.no.: 3
Page: 137
Figure 7-21 Make sure that signal cables and routing is correct during field hook-up! (Photo and
hook-up diagram: NGI Archives)
Wrong sensor address, sign
convention done or missing offset
reading during hook-up
unfortunately happens to often and
may lead to a lot of confusion or
even wrong interpretation of the
recorded data. Especially
inclinometers are susceptible to such
errors as different sign conventions
are used (pitch/roll or tilt) from
different suppliers.
The example in Figure 7-22 shows

recorded pore pressures from a
seabed piezometer. The data was
difficult to explain by the
geotechnical engineers but made
sense when a sign error was
discovered.
Figure 7-22 Data initially recorded with

the wrong sign (Source: NGI archives)

Date: 2017-06-15
Rev.no.: 3
Page: 138
7.6 Instrument corrosion and preservation

Corrosion is important to consider especially for long term deployment. Even if the rules of
thumb are followed such as using noble metals in the galvanic series, not mixing metals
with different galvanic potential or using cathodic protection, the results may be
discouraging. Scratches in the anodized aluminum surface or anodes with oxidized surface
may compromise the protection. Mixing of metals or fixture to larger steel structures
without galvanic isolation can result in rapid corrosion even if the metals themselves are
corrosion resistant. NGI’s practice is to keep it simple and only use for example stainless
steel in combination with thermoplastic materials such as Delrin. The use of thermoplastic
materials is usually OK in shallow waters with limited ambient pressure thus the structural
strength of the enclosure is less important. See also section 7.2.4.
Figure 7-23 Galvanic corrosion on metal enclosure with connectors in different metal grade (left).
“Corrosion safe” synthetic enclosure in Delrin for shallow water with connectors/outfitting in
stainless steel (right)
As for discussed in section the marine environment at the windfarm is very corrosive this
applies for instrumentation under water, in the splash zone and above sea level.
The general recommendations to limit corrosion:

Don’t “mix” metals (except if cathodic protection I used)
Determine local environmental conditions
Isolate dissimilar metals
Support items such as brackets, cable trays etc. should either be of similar steel as the
structure, coated and protected by the structural CP system or galvanic isolated from the
structure.

Date: 2017-06-15
Rev.no.: 3
Page: 139
7.6.1 Suitable materials

Titanium or titanium alloys have both high strength and very good resistance to all types
of corrosion in sea water, but expensive and difficult to machine than steel.
Stainless steel alloys, SS 316L or Duplex are recommended with respect to corrosion
resistance. Note that it is important that the surface is not treated, e.g. by painting, as the
stainless steel must have access to oxygen to oxidize the surface. It is also important to
treat the surface after machining to start the oxidization process. More exotic alloys are
Inconel 625, Hastelloy C or 254 SMO which are close to Titanium w.r.t. corrosion
resistance. Again these materials are expensive and sometimes difficult to machine, some
suppliers of pressure sensors offer Hastelloy membranes which are recommended.
Anodized aluminum, has good corrosion resistance in seawater. However, this requires
that the anodizing layer is intact and does not become scratched or damaged. Try to avoid
if unless weight optimization is required. Pure aluminum should not be used in marine
applications.
Non-metallic materials eliminate the risk for corrosion risk, but the application/design
must be suitable, e.g. deformation-under-pressure characteristics of pressure vessels.
Design of pressure vessels may be different from metal vessels. Synthetic materials which
absorb water such as Nylon should be avoided for subsea application due to swelling. Delrin
(or POM C) is a cost-efficient choice. Some plastic materials are sensitive to UV exposure
and should not be used externally above sea level.
For larger items such as seabed

frames etc. ordinary St 52
carbon steel can be used if the
surface is coated by two-
component epoxy paint and
anodes are used, see Figure 7-
24.
Figure 7-24 Coated bottom mount

for ADCP sensor with zink anodes
(Photo Mooring Systems Inc.)

Date: 2017-06-15
Rev.no.: 3
Page: 140
7.6.2 Galvanic isolation

If, for some reason, dissimilar metals have to be used together, they must be galvanic
isolated from each other. Figures 7-25 and 7-26 show examples of isolation methods.
Figure 7-25 Dissimilar metal plate isolated from housing
Figure 7-26 Dissimilar metal structure isolated from housing

Date: 2017-06-15
Rev.no.: 3
Page: 141
7.7 Pressure vessels and pressure seals

As an example, for subsea instrumentation, the focus for leakage integrity is on the high
loads given to the actual effect of the high-water pressure itself. This is a rather straight
forward design issue which can be tested in advance by pressure testing. The truth is that a
leak often occurs already at low pressure and usually caused by human errors such as a
damaged O-ring. It is therefore recommended to use two independent barriers (O-rings) on
seals which must be opened after pressure testing.
The basic considerations for water

ingress integrity and pressure testing are
described in section 7.2.4. Creeping
leaks are difficult to detect during
pressure testing and for permanently
immersed system prolonged testing
period is recommended (at least a couple
of days). It is also a good practice to
avoid sensitive electronics in the bottom
of the enclosure and have an "umbrella"
on top such that droplets of seawater do
not fall directly into the electronics, see
Figure 7-27. The bulkhead connector
may be a weak point with respect to
leakage. Such simple arrangement may
prolong the operational life of the
instrument unit with several months if a
Figure 7-27 Instrument enclosure with internal
creeping leak is present.
umbrella and sensitive electronics elevated from the
bottom (photo NGI archives)
A water ingress detector (two electrodes in the bottom of the enclosure sensing the
difference in conductivity when immersed in salt water) can be used for early warning of
leakage. However, leakage detection is only useful if the instrument enclosure can be
recovered from the sea within short time and not relevant for permanent equipment.
7.7.1 Pressure sealing approaches

Insert lids with O-ring grooves is the most common and recommended type of seal for
external pressures enclosures. The cylinder housing contracts evenly around O-ring
grooves, and lid will have very little tendency for geometric deformation when external
pressure is applied. Use two O-rings for redundancy, see Figure 7-28.

Date: 2017-06-15
Rev.no.: 3
Page: 142
NOTE: include a means to grip the lid so it can be lifted off. The O-rings and generated
suction pressure difference may make the lid difficult to pull out. Screws passing through
the edge of the lid and acting on the edge of the container can be used (turning the screws
lifts off the lid). Another approach is to machine small slots in the lid to allow several
screwdrivers to be used to pry off the lid.
Figure 7-28 Insert lid container: Held in place by bolts (left) or by locking ring (right)
From mechanical strength a cylinder enclosure is preferred, for larger diameters the wall
thickness must be increased. Mechanical strength is however normally not an issue at the
shallow depths relevant for offshore wind farms. To simplify opening and closure of the
container, the inside electronics can be attached to the lid which is equipped with the
external bulkhead connectors. Thus, all cabling and electronics are hooked-up and attached
to only one part of the instrument enclosure. Care should be taken when closing the lid,
pinched cable leads is probably one of the most common reasons for failure.
Figure 7-29 (Left) Electronics fixed to lid and hooked up to bulkhead connectors on the lid. (Right)
Inclinometer housing with connectors on the body and no outfitting attached to the lid (photo: NGI
archives)

Date: 2017-06-15
Rev.no.: 3
Page: 143
Other alternatives for securing the enclosure lid(s) are shown in Figure 7-30.
Figure 7-30 From left to right, Lid flange plates with through bolts, external stress rods and
enclosure with threaded locking ring/collars
Note that a leaking subsea container may be pressurized with the hydrostatic pressure at the
installed depth. This can be a potential hazard, however for shallow water application the
possible over pressure will be limited and it is sufficient to open the containers with care.
7.7.2 General types of seals

Various types of container seals are available:
Flat gaskets for no pressure differential (Pressure compensated vessels or 'Splash-
proof' rating only).
O-rings for applications with pressure differential
Incorporate ‘double barrier’ philosophy when possible, e.g. use two O-rings to make the
seal. Recommendations and check list for O-ring seals:
Always use on metal or thermoplastic containers for pressure sealing applications
Many O-ring types & different materials:
 Hardness to suit pressure range – check!
 Material to suit medium (sea water, oil, etc.) – check!
O-ring dimensions (thickness) influence sensitivity to small particle impurities; “fat”
O-rings are less sensitive to small particles.
O-ring groove dimensions vs. O-ring dimensions. Specific rules apply for O-ring vs.
groove dimensions. Check manufacturers/suppliers specifications

Date: 2017-06-15
Rev.no.: 3
Page: 144
O-rings should be moderately greased with appropriate silicone grease, Aqualube,

Molycote 111 or 44. Avoid Vaseline in colder temperatures (temperature sensitive
and hardens when cold).
7.8 Connectors and cables

Leakage integrity check of subsea
connectors and cables is also very
important in order to ensure long
term functioning of the subsea
instrumentation system, see Figure
7-31.
Figure 7-31 Creeping leakage along one

lead in a subsea connector due to
improper priming and bonding of
neoprene pin seals (source: NGI
archives)
7.8.1 Connectors
The most common (and less costly) subsea connectors are neoprene connectors with over
molded pin seals, see Figure 7-32. There is wide range of this type of connectors with
different layout and configuration available in the market. Some of the biggest suppliers are
Seacon (Wetcon), Subcon, Impulse, Burton., etc. All brands are more or less of similar
design. The connection should be secured with locking sleeves. The sealing is increasing
with the hydrostatic pressure and the over molded pins seals must be treated as O-rings
(greased and protected from dust). Unplugged connectors should always be protected by
blind plug or caps.
Figure 7-32 Neoprene bulkhead and inline connectors with over molded pin seals. Locking sleeve
arrangement shown to the right. These types of connectors are not water blocked.

Date: 2017-06-15
Rev.no.: 3
Page: 145
Metal shell connectors are robust and normally water blocked, see Figure 7-33. They must
be dry mated, available both for electric and fiber optic connections. Suitable for long
term operation and for larger number of pins than the over-molded connector (also more
expensive)
Figure 7-33 Metal shell connectors from Seacon and Burton

Penetrators provide continuous connection and are normally used for pressure tight
routing of fiber optic lines through a bulkhead. PG cable glands are not recommended for
permanent subsea operation.
Figure 7-34 Fiber optic penetrator (left) and typical PG cable gland (right)
External connectors above sea level must as a minimum have IP67 ingress protection, MIL
type of connectors are commonly used, see Figure 7-35. Subsea connectors should be used
in the splash zone.

Date: 2017-06-15
Rev.no.: 3
Page: 146
Figure 7-35 MIL type connectors (common field connector based on US Military specifications)
7.8.2 Offshore instrumentation cables

Polyurethane (PUR) jacket is recommended for all cables routed externally (both subsea
and topside). PUR cable are easy to handle, robust against abrasion, both saltwater and UV
resistant. Kevlar reinforcement is recommend, if the cable is subjected to tension during
installation and/or operation. The PUR jacket is also suitable for over molding of neoprene
inline connectors or splices, see Figures 7-36 and 7-37.
Figure 7-36 Typical cross section and examples Kevlar reinforced PUR cables for offshore and subsea
use

Date: 2017-06-15
Rev.no.: 3
Page: 147
Figure 7-37 Example of inline connectors molded directly to PUR cable and molded cable split which
many times is more cost efficient than using subsea junction boxes.
For short jumpers (not exposed to any loads
neoprene cables may be used). The cables
should be properly routed and secured to the
structure. Especially the connectors must
not be subjected to strain, see Figure 7-38.
Cables traversing the splash zone will be
subjected to significant slamming forces
and should be protected inside steel Figure 7-38 Example of proper cable routing (Photo
conduits (pipes). from NGI archives)
7.9 Electric system design

An offshore wind turbine is a power generator, thus considered as noisy with respect to
electro-magnetic (EM) conditions. The noise conditions (especially EM) are significantly
attenuated below the sea surface. Some general electric system design aspects are discussed
in the following sections:
7.9.1 Signal and cable routing

See also section 6.2 Signal transmission.
For hook-up of sensor arrays the sensor signal can be:
Analog voltage (0-10 V) or Current loop (4-20 mA) 2 or 3 wire system
Vibrating wire (frequency)
Digital for example RS232, RS485, Ethernet, CAN/ProfiBus (RS232 is usually not
suited for data transmission through long lines, without extender Ethernet signals
also have a limitation of about 100m cable transmission )

Date: 2017-06-15
Rev.no.: 3
Page: 148
In general, a cable does three important things to the signal:

(1) it attenuates the signal,
(2) it contributes its own inductive (L) and capacitive (C) reactance, which can alter different
parts of the signal differently--of particular importance, attenuating some frequencies, such
as high audio notes, more than it attenuates others, and
(3) it exposes the signal to electromagnetic energy from other sources which can enter the
cable and pollute the signal with noise.
Figure 7-39 The cable in theory

For most signals, twisted pair wiring is a good solution and provides the best shielding
against EM noise. It is usually no problem to mix power and signal leads in one cable. For
ethernet signals the reactance must be considered.
With respect to power it is important to check the voltage drop in long cables (dependent
on cross section on power conductors). In some cases (if the cable cross section should be
optimized) it may be necessary to feed the instruments with a higher voltage through the
cable and transform it to the specified input voltage using a local DC-DC converter in the
sensor enclosure.
7.9.2 Grounding
One of the basic electronic laws tells us that each current flowing from a power supply has
to come back to the source. This “back wire” is often called the ground and one of the most
common failures is to mix the ground path of different electric equipment liker power
consuming motors, low power analog signal and high frequency digital signals. If all
different signals share the same ground path, there will for sure be some influence between
the signals which most often result in noise on the analog lines. To overcome this problem,
star grounding with different ground loops for different needs are recommended.
Star grounding means at there are different electric connections to each device from one
central ground point. For a balanced system, the connector from the power supply to the
electric instrument should be equal to the return path from the instrument to the power
supply.

Date: 2017-06-15
Rev.no.: 3
Page: 149
Other grounding problems can arise when metallic under water enclosures are used. There
are several sensor types where the ground pin from the connector is connected to the sensor
housing. If the sensor housing is in contact with sea water an additional ground loop is
introduced. This may result in additional noise and corrosion problems. For subsea
instrumentation, the ground should not be connected to sea.
7.9.3 Mains power supplies

There are two types of power supplies on the market: Linear power supplies and switch
power supplies.
Category Linear Switch mode Comment

Size   Switch typical 80% smaller
Weight   Switched typical 80% lighter
Input voltage range   Linear 10% versus 300% switched
Efficiency   Power save over long term
Reliability   Probably equal
Ripple / noise   Switched needs extra filter
Transient response   Linear up to 1000 times better
Leakage current   Linear has low leakage current
In most applications one power supply serves several units like data logger, sensors,
communication devices and more. When designing the power distribution care has to be
taken to the possibility of a failure on one of the connected units. A short circuit on one unit
should not influence the functions of the other connected units. A simple possibility is to
use a fuse for each unit such as the faulty devices in case of a short circuit is disconnected
from the power line. An example is shown in Figure 7-40.
Figure 7-40 Main power supply with individual fuses for each unit

Date: 2017-06-15
Rev.no.: 3
Page: 150
Isolating power supplies should be used for safety. Most linear and switched power supplies
are isolating power supplies, means at there is no electrical connection between the input
voltage and the output voltage.
With respect to current loops under water it may not be sufficient to only use fuses at one
side of the loop. There may be situations where a damaged cable could cause a short circuit
to the sea allowing stray currents in the water to form another loop.
The current may not be sufficient to trigger any fuses but could cause rapid corrosion at
unexpected locations, see Figure 7-41. Thus, it is recommended to use fuses on both sensors
sides of a current loop.
Figure 7-41 Burned batteries and galvanic corroded connector caused by stray currents in the water
probably originating from broken cable isolation and fuses with insufficient sensitivity (Photo NGI
archives)
7.10 Quality control and documentation

Performing tests or checks along the way to completion can usually prevent costly mistakes
and longer delays for installation.
IMPORTANT: A test which is not documented in written form has no value with respect
to the QA/QC system. The documentation does not need to be extensive. It can be as simple
as a handwritten note describing what is tested, how the testing is done, what the results are,
and signed + dated by the person doing the tests.
In general, the following milestones checks should be included:

Date: 2017-06-15
Rev.no.: 3
Page: 151
7.10.1 Technical design stage

Concept design. What is to be measured? What are the specifications and the design
constraints? Involve project QA responsible as well as appropriate technical
experts/specialists at NGI. Verify with the client that we are designing equipment
to meet their needs.
Multi-disciplinary check. Have others reviewed the design concept?
Final design drawings. Pre-production control.
7.10.2 Procurement and Manufacturing stage

Orders of components (Send written orders; get quotes from several suppliers if
possible)
Make sure the that the supplier has understood the order correctly (order
confirmation)
Follow up on delivery times, special cables are often long lead items and only
produced in batches. Sometimes this can lead to delays.
Control of externally manufactured hardware. Include visit to factory if applicable.
Receiving of components (including initial function testing)
Calibrations of sensors
Software or firmware to run the data loggers
Testing of assembled sub systems (if appropriate)
Leakage testing of pressure containers
Assembly checks: Visually check system that everything looks correct and not
damaged (particularly after transporting of equipment). Check all subsea connectors
Electrical test: Perform instrument loop test and test isolation against earth. Check
power consumption against available battery power
Overall system testing (functional testing) before FAT
7.10.3 Delivery stage: FAT and SAT/SIT

Executing FAT (Factory Acceptance Test) and SAT/SIT (Site Acceptance Test/System
Integration Test) is in interest for both the SHM contractor (instrumentation supplier) and
the Operator (end user). As the instrumentation system may be subjected to conditions
outside the suppliers control, given warrantees must be limited. Witnessed testing of the
equipment (FAT and SAT) constitute important milestones in an instrumentation project
and demonstrates that the delivery is compliant with the specifications. From the clients
perspective these tests are important verification of the purchased system. Usually the
ownership of the equipment and payment is linked to those tests.

Date: 2017-06-15
Rev.no.: 3
Page: 152
FAT testing may include:

Submittal of test documented tests performed during the manufacturing stage
Inspection of mechanical completion of the system
Functional and calibration testing of the system, usually limited to spot checks
During FAT the hardware and software is tested and functionality verified in accordance
with the procedures and acceptance criteria. The Client's acceptance of the FAT is usually
also the acceptance of the delivery in accordance with the contract. FAT testing is usually
performed at the supplier's premises. If the FAT not is accepted and punch list is compiled
describing the rectifications required for acceptance.
SAT/SIT is normally performed at the yard or/and in the field and is limited to
demonstration proper rigging/installation and normal operation of the system after
installation. As for the FAT test procedures to be agreed on in advance. System integration
testing is relevant if the SHM supply consists of sub systems which not have been tested
together at an earlier stage. This may for example be communication with turbine SCADA
system.
7.11 Offshore and Subsea installation considerations

As described earlier it is advised that parts of the instrumentation system are pre-rigged
(mechanical components, cables and some sensor systems) to save offshore installation
time. However, for retrofit and maintenance offshore/subsea installation operations are
expected.
7.11.1 Weight of equipment

In general one should try to keep weight of equipment as small as possible to make yard
and offshore handling easier. Consider the following factors which may limit the total
weight:
Manual lifting on deck
Crane handling
ROV/Diver handling
In some cases extra dead weight of the equipment may be necessary for stability purposes,
e.g. subsea templates. In other cases buoyancy may be added to simplify subsea handling
by ROV/Diver
NOTE: Equipment that has to be lifted by crane should have offshore certified lifting points,
or must be able to be lifted in a certified basket.

Date: 2017-06-15
Rev.no.: 3
Page: 153
7.11.2 Safety and access

Offshore work is demanding and safety must be taken serious with highest priority. Access
and safety for offshore installation work must therefore be taken into account when
designing the system solution. This also includes safe access for later inspection and
calibration checks.
Figure 7-42 Difficult access for instrument mounting (left). Mounting instruments inside installed
offshore pile is not a comfortable or safe work place (picture from internet)
Figure 7-43 No longer a safe workplace!

Date: 2017-06-15
Rev.no.: 3
Page: 154
7.11.3 System rigging and hook-up plans

Points to consider when designing the layout of the instrumentation system:
Compact layout for handling onboard/lifting through water column. Split into
manageable sections if possible
How is it installed - requirements for lifting, work at heights or ROV/Diver
intervention
Hook-up. Some jacket/tripods are assembled as building blocks, when can
instruments be installed (accessible), when can cables be routed…etc. Is subsea
hook-up required by ROV/Diver?
Protection and securing equipment for rough handling bot at the yard and offshore
Figure 7-44 Dong Energy's SBJ foundation assembled as building blocks. How and when to install
strain gauges required thorough planning and cooperation with the Yard (Source NGI archives)
7.11.4 ”As installed” protection

Instruments and especially cables installed on the structure should if possible be protected
against falling objects and environmental forces. External cables run through the splash
zone are exposed to very rough conditions (breaking waves) and recommended to be fully
protected inside conduits or channels. The splash zone can also by highly exposed to severe
corrosion and marine growth, see Figure 7-45.

Date: 2017-06-15
Rev.no.: 3
Page: 155
Figure 7-45 Examples of the severe conditions which can expected at the splash zone
7.11.5 Summary of factors affecting the Long term durability

The ability of the installation to withstand the duration of the deployment must be evaluated.
Several factors should be addressed in the concept and detail design phases:
Corrosion: Integrity of the instruments, protection enclosures, brackets and fixture
and if relevant, the ability to operate moving parts (extensometers). Corrosion of
lifting slings/eyes can be a problem for later recovery.
Biofouling: Marine growth covering the instruments can affect the measurements
(especially for chemical sensors, see Figure 7-46). May also overload the equipment
and make access difficult.
Leakage (water migration) through synthetic materials: Neoprene plugs may leak
over extended deployments (many years) due to water creeping along inside of the
over molded pins if not water blocked. Some materials absorb water (swelling).
Degradation: Some materials become brittle over long term exposure to salt water,
or in shallow water/splash zones. Especially important for cables.
Mechanical protection: External instruments are subjected to significant forces
both from waves and wind. The risk of human interference is also a factor which
cannot be neglected (dropped objects, removal of components, sandblasting,
paintwork, welding, etc.). The requirements for mechanical protection and proper
marking of the installation depends on the location at the structure.
Electrical, Lighting, mains power failures and human interference (cables and
junction boxes).
Long term processes: Sensor out of range may be caused by drift or measured
parameter larger than expected.

Date: 2017-06-15
Rev.no.: 3
Page: 156
Figure 7-46 Biofouling on a chemical sensor immersed one year in tropical waters
Although NGI have kept SHM systems on offshore platform alive for more than 27 years it
is unrealistic to expect that the initial SHM installation will have the same operational life
as the OWT foundation itself, sensors or cables being damaged or failing with time must
be expected. Much shorter operational life must be expected for some types of instruments
like strain gauges under water or instruments embedded in the seabed, 5 years
operational life is then a realistic goal. If life time monitoring is the aim, future
replacement monitoring solutions must be planned for. Future technology may also allow
for better and simpler monitoring.
In order to keep the system in operation some maintenance must be expected, this is
limited to accessible instrumentation and DAQ system. Many times financing and lack of
interest for continued operation of the SHM system also limits the operational life of the
system.
Figure 7-47 The cassette tapes and recorders more or less disappeared 20 years ago and was
replaced by other technology. What about SHM monitoring solutions 20 years in the future?

Date: 2017-06-15
Rev.no.: 3
Page: 157
8 Standards and guidelines

Presently the only known wind power specific standard for Structural Health Monitoring is
the new German VDI Guideline 4551 “Structure Monitoring and Evaluation of Wind
Turbines and Platforms”. All other standards are mainly related to the drive train and wind
turbine itself (machinery).
9 References
A significant share of the information and examples presented in this study is based on NGI
in-house experience and knowledge. Other examples are based on articles and presentations
at homepages from cited suppliers and academic institutions.
Other published references are listed below:
Design of monopiles for offshore wind turbines in 10 steps

Laszlo Aranya, S. Bhattacharyab, John Macdonalda,, S.J. Hogan
Soil Dynamics and Earthquake Engineering
Volume 92, January 2017, Pages 126–152
Condition monitoring of wind turbines: State of the art, user experience and
recommendations - Project report 2015
Diego Alejandro Coronado and Katharina Fischer,
Fraunhofer Institute for Wind Energy and Energy Systems IWES.
Guideline for the Certification of Offshore Wind Turbines (2012).

Germanischer Lloyd Industrial Services GmbH. Hamburg, Germany.
Revisiting mono pile design using p-y curves - Results from full scale measurements on
Horns Rev
T. Hald, C. LeBlanc, C. Mørch, L. Jensen, K. Ahle, EOW 2009
Proceedings Frontiers in Offshore Geotechnics III, Oslo 2015
Structural Health Monitoring Systems in Difficult Environments—Offshore Wind Turbines

P. Faulkner, P. Cutter and A. Owens
6th European Workshop on Structural Health Monitoring - Fr.2.A.3
Structural health monitoring of a monopile foundation structure of an offshore wind turbine

Michael Link, Matthias Weiland
Department of Civil and Environmental Engineering, Kassel, Germany
Proceedings of the 9th International Conference on Structural Dynamics, EURODYN 2014
Porto, Portugal, 30 June - 2 July 2014

Date: 2017-06-15
Rev.no.: 3
Page: 158
Inspection and monitoring of corrosion inside monopile foundations for offshore wind
turbines
L. R. Hilbert, A. R. Black, F. Andersen, T. Mathiesen, FORCE Technology,
Paper no. 4730 EUROCORR 2011 Conference Stockholm September 4th-8th 2011
Monitoring resonant frequencies and damping values of an offshore wind turbine in parked
conditions
Christof Devriendt, Wout Weijtjens, Mahmoud El-Kafafy, Gert De Sitter
Offshore Wind Infrastructure Lab, Brussels, Belgium
Published in IET Renewable Power Generation 2014
Foundation Monitoring Systems: Analysis of 2 Years of Monitoring at the North Sea

Gert De Sitter, Wout Weijtjens, Yves Van Ingelgem, Daan De Wilde, Kristof Verlinden,
Stefan Millis, Christof Devriendt
Vrije Universiteit Brussel, Parkwind NV, Sirris, Offshore Wind Infrastructure Lab, Zensor
Poster EWEA 2014, Barcelona, Spain
Long-gauge fibre optic sensors: Performance comparison and application

Carlos Rodrigues ,Daniele Inaudi, Branko Glisic
International Journal Lifecycle Performance Engineering Vol X, xxx
Structural health monitoring of offshore wind turbines: A review through the Statistical
Pattern Recognition Paradigm
Maria Martinez-Luengo, Athanasios Kolios n, Lin Wang
Cranfield Energy,Cranfield University,BedfordMK430AL,UK
Journal article Renewable and Sustainable Energy Reviews 64(2016) 91–105

Kontroll- og referanseside/
Review and reference page
Dokumentinformasjon/Document information
Dokumenttittel/Document title Dokumentnr./Document no.
Guidelines for Structural Health Monitoring for Offshore Wind Turbine Towers & 16-1036
Foundations
Dokumenttype/Type of document Oppdragsgiver/Client Dato/Date
Rapport / Report BSEE Environmental Enforcement 2017-05-15
Division
Rettigheter til dokumentet iht kontrakt/ Proprietary rights to the document Rev.nr.&dato/Rev.no.&date
according to contract 3 / 2017-06-15
Oppdragsgiver / Client
Distribusjon/Distribution
BEGRENSET: Distribueres til oppdragsgiver og er tilgjengelig for NGIs ansatte / LIMITED: Distributed to client and
available for NGI employees
Emneord/Keywords
Offshore wind turbine, foundation, towers, Structural health monitoring, instrumentation, guidelines
Stedfesting/Geographical information
Land, fylke/Country USA Havområde/Offshore area
USA
Kommune/Municipality Feltnavn/Field name
Sted/Location Sted/Location
Kartblad/Map Felt, blokknr./Field, Block No.
UTM-koordinater/UTM-coordinates Koordinater/Coordinates
Zone: Type here East: Type here North: Type here Projection, datum: Type here East: Type here North: Type
here
Dokumentkontroll/Document control
Kvalitetssikring i henhold til/Quality assurance according to NS-EN ISO9001
Sidemanns- Uavhengig Tverrfaglig
Rev/ Egenkontroll kontroll av/ kontroll av/ kontroll av/
Revisjonsgrunnlag/Reason for revision av/
Rev. Colleague Independent Interdisciplinary
Self review by: review by: review by: review by:
Original document DRAFT (Working 2016-10-25
0
document) Per Sparrevik
2017-05-15
1 Finalized DRAFT report for BSEE comments
Per Sparrevik
Internal QC (grammar and language 2017-06-09 2017-06-09 2017-06-09 2017-06-09
2
corrections) Per Sparrevik Amir Rahim Amir Rahim Amir Rahim
Final report with BSEE comments 2017-06-15 2017-06-27 2017-06-27 2017-06-27
3
implemented Per Sparrevik Amir Rahim Amir Rahim Amir Rahim
2015-10-16, 043 n/e, rev.03
Dokument godkjent for utsendelse/ Dato/Date Prosjektleder/Project Manager

Document approved for release
29 June 2017 Per Magnus Sparrevik
C:\Users\PSp\Documents\Presentasjon vindmøller NGI\BAA application\report\Report 16-1036_NGI Houston_BSEE_OWT_Guidelines_rev 3.docx

NGI (Norwegian Geotechnical Institute) is a leading international center
for research and consulting within the geosciences. NGI develops
optimum solutions for society and offers expertise on the behaviour of
soil, rock and snow and their interaction with the natural and built
environment.
NGI works within the following sectors: Offshore energy – Building,

Construction and Transportation – Natural Hazards – Environmental
Engineering.
NGI is a private foundation with office and laboratories in Oslo, a branch

office in Trondheim and daughter companies in Houston, Texas, USA and
in Perth, Western Australia
www.ngi.no
NGI (Norges Geotekniske Institutt) er et internasjonalt ledende senter for

forskning og rådgivning innen ingeniørrelaterte geofag. Vi tilbyr
ekspertise om jord, berg og snø og deres påvirkning på miljøet,
konstruksjoner og anlegg, og hvordan jord og berg kan benyttes som
byggegrunn og byggemateriale.
Vi arbeider i følgende markeder: Offshore energi – Bygg, anlegg og

samferdsel – Naturfare – Miljøteknologi.
NGI er en privat næringsdrivende stiftelse med kontor og laboratorier i

Oslo, avdelingskontor i Trondheim og datterselskaper i Houston, Texas,
USA og i Perth, Western Australia.
www.ngi.no

770 Aa

Uploaded by

Copyright:

Available Formats

770 Aa

Uploaded by

Document Information

Original Description:

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

770 Aa

Uploaded by

Copyright:

Available Formats

REPORT

BSEE Offshore Wind Recommendations

DOC.NO. 20160190 16-1036

This document shall not be used in parts, or for other purposes

Ved elektronisk overføring kan ikke konfidensialiteten eller

Dokumentet skal ikke benyttes i utdrag eller til andre formål

NGI Inc. T +1 281 752 4667 NGI Inc. NGI Oslo

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

Rough environmental conditions can be expected at many wind farm locations

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

7.1 Introduction 118

Review and reference page

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

Three main objectives are relevant for SHM:

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

Figure 1-2 Parts of an Offshore Wind Turbine (OWT)

2 Overview of areas relevant to monitor

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

2.1 Types of OWT foundations

These four foundation categories are illustrated in Figure 2-1.

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

2.3 Structural integrity and response

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

Further structural issues deserving special attention:

2.5 Correlated parameters

2.6 Installation Baseline

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

3.1 General considerations for an integrated SHM system

3.1.1 Instrumentation plan

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

3.1.3 Sensor interface

3.1.4 Data logging

3.1.5 Data transfer

3.1.6 Data storage

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

3.1.8 Alarms and Follow up

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

that only can be determined by implementing an integrated, intelligent continuous

3.2 The value of integrated monitoring

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

By mean of continuous monitoring, the risk related to the management of offshore

3.3 Limitations and failures in a monitoring program

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

3.4 Reasonable extent of SHM monitoring

3.5 An extreme event on tape

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

4 Specific SHM objectives for OWT foundations

4.1 Objectives for Foundation Response Monitoring in the field

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

Figure 4-2 Load transfer to the seabed for

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx

c:\users\psp\documents\presentasjon vindmøller ngi\baa application\report\report 16-1036_ngi houston_bsee_owt_guidelines_rev 3.docx