Geotechnical Considerations for Offshore Wind Turbines

Geotechnical Considerations for Offshore Wind Turbines Zachary J. Westgate Jason T. DeJong

August 1, 2005

TABLE OF CONTENTS PREFACE ....................................................................................................................................................................6 GENERAL ...................................................................................................................................................................6 OBJECTIVES ...............................................................................................................................................................6 SCOPE & LIMITATIONS ..............................................................................................................................................6 RESOURCES ...............................................................................................................................................................7 1.0

THE GLOBAL OFFSHORE WIND INDUSTRY......................................................................................8

1.1 1.2 1.3 1.4 1.5 2.0

INTRODUCTION ............................................................................................................................................8 EUROPE .......................................................................................................................................................8 UNITED STATES ...........................................................................................................................................9 OTHER COUNTRIES ....................................................................................................................................10 FUTURE PROSPECTS IN OFFSHORE WIND ENERGY DEVELOPMENT............................................................10 TYPES OF FOUNDATIONS .....................................................................................................................12

2.1 2.2 2.2.1 2.2.2 2.2.3 2.2.4 2.3 2.3.1 2.3.2 2.4 2.4.1 2.4.2 2.4.3 2.5 2.5.1 2.5.2 2.5.3 3.0

DESIGN METHODOLOGIES ..................................................................................................................18

3.1 3.1.1 3.1.2 3.1.3 3.1.4 3.2 3.3 3.4 3.5 4.0

INTRODUCTION ..........................................................................................................................................12 PILED FOUNDATIONS .................................................................................................................................12 Introduction ........................................................................................................................................12 Monopiles ...........................................................................................................................................12 Multiple-Leg Foundations ..................................................................................................................13 Lattice Towers ....................................................................................................................................13 GRAVITY BASE FOUNDATIONS ..................................................................................................................14 General Description ...........................................................................................................................14 Material Considerations.....................................................................................................................15 SUCTION CAISSONS ...................................................................................................................................15 General Description ...........................................................................................................................15 Installation Principle..........................................................................................................................16 Site Considerations.............................................................................................................................16 FLOATING STRUCTURES ............................................................................................................................17 Introduction ........................................................................................................................................17 Tension-Leg Platforms .......................................................................................................................17 Low-roll Floaters................................................................................................................................17 LIMIT STATE DESIGN ................................................................................................................................18 Ultimate Limit State............................................................................................................................18 Fatigue Limit State .............................................................................................................................18 Accidental Limit State.........................................................................................................................18 Serviceability Limit State....................................................................................................................19 LOAD RESISTANCE FACTOR DESIGN (LRFD) METHOD ............................................................................19 DIRECT SIMULATION .................................................................................................................................19 TESTING BASED DESIGN ...........................................................................................................................20 PROBABILITY BASED DESIGN ....................................................................................................................20

SITE INVESTIGATIONS ..........................................................................................................................21

4.1 INTRODUCTION ..........................................................................................................................................21 4.2 PHASES OF THE SITE INVESTIGATION ........................................................................................................22 4.2.1 Introduction ........................................................................................................................................22 4.2.2 Geological Study.................................................................................................................................23 4.2.3 Geophysical Survey ............................................................................................................................24 4.2.4 Geotechnical Site Investigation ..........................................................................................................26 4.3 OFFSHORE GEOTECHNICAL SITE INVESTIGATION VESSELS .......................................................................27 4.3.1 Introduction ........................................................................................................................................27

2

4.3.2 Low Draft Barges ...............................................................................................................................28 4.3.3 Jack-up Rigs .......................................................................................................................................28 4.3.4 Specialty Geotechnical Drilling Vessels.............................................................................................28 4.3.5 Semi-submersible Drilling Rigs..........................................................................................................29 4.3.6 Supply Vessels ....................................................................................................................................29 4.4 DRILLING ..................................................................................................................................................29 4.5 SAMPLING .................................................................................................................................................31 4.5.1 Introduction ........................................................................................................................................31 4.5.2 Wheeldrive System..............................................................................................................................32 4.5.3 Grab Sampling....................................................................................................................................32 4.5.4 Push and Piston Sampling ..................................................................................................................33 4.5.5 Coring Techniques..............................................................................................................................33 4.5.6 Other Sampling Methods ....................................................................................................................35 4.5.7 Sample Recovery, Storage, and Transport .........................................................................................35 4.6 IN SITU TESTING .......................................................................................................................................36 4.6.1 Introduction ........................................................................................................................................36 4.6.2 Cone Penetration Testing ...................................................................................................................36 4.6.3 T-bar Testing ......................................................................................................................................37 4.6.4 Field Vane Testing..............................................................................................................................38 4.6.5 Other In Situ Tests ..............................................................................................................................38 4.7 CHARACTERISTIC SOIL PROPERTIES ..........................................................................................................38 4.8 LABORATORY TESTING .............................................................................................................................40 4.8.1 Offshore Testing Program ..................................................................................................................40 4.8.2 Onshore Testing Program ..................................................................................................................40 4.8.3 Soil Classification...............................................................................................................................41 4.8.4 Index Testing ......................................................................................................................................41 4.8.5 Consolidation Testing.........................................................................................................................42 4.8.6 Strength Testing..................................................................................................................................43 4.8.7 Other Laboratory Testing...................................................................................................................44 4.9 EFFECTS OF CYCLIC LOADING ...................................................................................................................46 4.10 SEAFLOOR STABILITY ...............................................................................................................................46 4.10.1 Introduction ........................................................................................................................................46 4.10.2 Slope Stability .....................................................................................................................................47 4.10.3 Hydraulic Stability..............................................................................................................................47 4.10.4 Earthquake Stability ...........................................................................................................................47 4.11 SCOUR .......................................................................................................................................................48 4.11.1 Scour Mechanism ...............................................................................................................................48 4.11.2 Types of Scour ....................................................................................................................................49 4.11.3 Prevention of Scour ............................................................................................................................49 4.11.4 Designing for Scour............................................................................................................................50 5.0

OTHER ENVIRONMENTAL CONDITIONS.........................................................................................52

5.1 5.2 5.2.1 5.2.2 5.2.3 5.3 5.3.1 5.3.2 5.3.3 5.3.4 5.3.5 5.3.6 5.4 5.4.1

INTRODUCTION ..........................................................................................................................................52 METEOROLOGICAL PARAMETERS ..............................................................................................................52 Wind Loading .....................................................................................................................................52 Wind Modeling ...................................................................................................................................53 Turbulence..........................................................................................................................................53 OCEANOGRAPHIC PARAMETERS ................................................................................................................54 Wave Loading.....................................................................................................................................54 Wave Modeling ...................................................................................................................................55 Current Loading .................................................................................................................................58 Current Modeling ...............................................................................................................................58 Ice Loading.........................................................................................................................................59 Ice Modeling.......................................................................................................................................59 COMBINED WIND AND WAVE LOADING ....................................................................................................60 Horizontal to Moment Load Ratio ......................................................................................................60

3

5.4.2 Combination Methods.........................................................................................................................61 5.4.3 Design Considerations .......................................................................................................................61 5.5 ENVIRONMENTAL CORROSION ..................................................................................................................61 5.5.1 Introduction ........................................................................................................................................61 5.5.2 Degradation Corrosion ......................................................................................................................62 5.5.3 Marine Growth ...................................................................................................................................62 5.6 OTHER LOADING CONDITIONS ..................................................................................................................63 5.6.1 Transportation and Installation Loading ...........................................................................................63 5.6.2 Vessel Collision ..................................................................................................................................63 5.6.3 Deformation Loading .........................................................................................................................64 6.0

FOUNDATION MODELING ....................................................................................................................65

6.1 INTRODUCTION ..........................................................................................................................................65 6.2 TYPES OF MODELS ....................................................................................................................................66 6.2.1 Plasticity Models ................................................................................................................................66 6.2.2 Finite Element Models........................................................................................................................67 6.2.3 Other Techniques................................................................................................................................68 6.3 DYNAMIC SENSITIVITY..............................................................................................................................69 7.0

FOUNDATION DESIGN ...........................................................................................................................71

7.1 7.2 7.3 7.3.1 7.3.2 7.3.3 7.3.4 7.3.5 7.4 7.4.1 7.4.2 7.4.3 7.4.4 7.4.5 7.5 7.5.1 7.5.2 8.0 8.1 8.2 9.0

INTRODUCTION ..........................................................................................................................................71 OFFSHORE TURBINE VS. OFFSHORE PLATFORM FOUNDATIONS ................................................................72 PILED FOUNDATIONS .................................................................................................................................73 Introduction ........................................................................................................................................73 General Design Considerations .........................................................................................................73 Grouting Operations...........................................................................................................................74 ULS Design.........................................................................................................................................75 SLS Design..........................................................................................................................................76 GRAVITY BASE FOUNDATIONS ..................................................................................................................77 General Principles..............................................................................................................................77 General Design Equations..................................................................................................................77 Stability of Foundation: ULS and ALS Design ...................................................................................78 Settlements and Displacements: SLS Considerations.........................................................................79 Grouting Operations...........................................................................................................................79 SUCTION CAISSONS ...................................................................................................................................80 General Principles..............................................................................................................................80 Design Studies ....................................................................................................................................80

FOUNDATION TRANSPORT AND INSTALLATION .........................................................................83 MARINE OPERATIONS................................................................................................................................83 PROJECT DURATION ..................................................................................................................................83 COSTS OF WIND ENERGY .....................................................................................................................85

9.1 RENEWABLE ENERGY MARKET INCENTIVES .............................................................................................85 9.2 GENERAL COST CONSIDERATIONS ............................................................................................................85 9.2.1 Introduction ........................................................................................................................................85 9.2.2 Cost of Energy Factors.......................................................................................................................86 9.2.3 Cost Optimization ...............................................................................................................................88 9.3 COSTS FOR UNITED STATES OFFSHORE WIND FARM DEVELOPMENT ........................................................89 9.3.1 Department of Energy Cost Estimates................................................................................................89 9.3.2 Other Estimates ..................................................................................................................................90 10.0

OFFSHORE WIND DEVELOPMENT IN NEW ENGLAND................................................................91

10.1 INTRODUCTION ..........................................................................................................................................91 10.2 ENVIRONMENTAL CONDITIONS AND CONSTRAINTS ..................................................................................91 10.2.1 Wind Speed .........................................................................................................................................91

4

10.2.2 Water Depth........................................................................................................................................91 10.2.3 Wave Loading.....................................................................................................................................92 10.2.4 Subsurface Conditions........................................................................................................................92 10.3 OTHER DESIGN CONSIDERATIONS .............................................................................................................92 10.3.1 Land Jurisdiction................................................................................................................................92 10.3.2 Electrical Grid Connections ...............................................................................................................93 10.3.3 Availability of Harbors for Construction and Maintenance...............................................................93 10.3.4 Local Public Acceptance and Conflicting Areas ................................................................................93 10.4 ADVANCEMENT OF DEEP WATER TECHNOLOGY .......................................................................................93 11.0

CONCLUSIONS..........................................................................................................................................95

5

Preface General Offshore wind farms are becoming increasingly popular in the quest for renewable sources of energy. The planning, design, inspection, and maintenance of offshore wind farms requires careful consideration of many variables, including local climate and site conditions, economic incentives, proximity to energy loads, environmental considerations, and legal issues. The various technologies involved in the design of offshore structures include oceanography, foundation engineering, structural engineering, marine civil engineering, and naval architecture, with specific aspects of each technology summarized in Table 1. The project certification process leading to successful development of an offshore wind farm includes six phases that incorporate verification of the design basis, the preliminary design, and the final design, as well as manufacturing surveys, transport and installation, and the final inservice state. The relationship of the wind farm development tasks to these certification phases are summarized in Table 2. This report will focus primarily on the site condition assessment and foundation modeling, design, and to a lesser extent, the installation aspects.

Objectives The objective of this report is fourfold: 1) to provide an updated account of the current state of the practice of the offshore wind industry and the technological direction in which the industry is headed, 2) to provide guidance in planning, designing, and constructing an offshore wind farm, specifically with respect to the site investigation program and foundation options, 3) to provide insight into the current research developments in offshore wind turbine foundations, and 4) to provide information useful for the planning and design of offshore wind turbines in the New England region of the United States.

Scope & Limitations This document is not meant to be a standalone resource for planning or design of offshore wind turbines. The guidelines included provide information on the activities associated with offshore site investigations and a design basis for the foundation support elements of the wind turbine

6

support structure. This document does not include recommendations for mechanical equipment design such as the rotor-nacelle assembly, nor does it include aspects of detailed structural design of the support towers. This document also does not account for local government codes, legal restrictions, environmental policy, or permitting restrictions.

Resources The referenced material contains additional resources that may specify other planning and design considerations not accounted for in this document, including industry standards (Table 3) for both offshore construction in general and offshore wind turbine construction, state of the industry reports and reviews, papers on both novel research concepts and proven offshore technologies, and past and present feasibility studies for global and local offshore wind energy development.

7

1.0 The Global Offshore Wind Industry 1.1

Introduction

The offshore wind industry has developed rapidly over the past 20 years with many industrialized countries now beginning to implement renewable energy requirements as part of their total energy production portfolio. Many of the areas with the greatest potential for offshore wind energy development are in coastal waters where shallow depths extend relatively far offshore (Table 4). While Europe has been the industry leader since the initiation of offshore wind turbine development, some of the other continents planning offshore wind energy development include Asia and North America, with countries such as China and the United States taking recent initiative.

1.2

Europe

In Europe, the offshore wind energy industry is gaining increased momentum as many countries are following the lead of the Nordic countries (e.g. Denmark, The Netherlands) and the United Kingdom. By the end of year 2003, there were 11 operational wind farms in Europe (Table 5), constituting 600 MW of energy generating potential, with individual projects of up to 160 MW each, enough to power 145,000 homes (EWEA 2004). Most of the projects have been limited to less than 30-m deep waters in the North and Baltic Seas, primarily in the 5 to 12-m water depth range (Musial and Butterfield 2004). The offshore potential in Europe is considered to be greater than its total electricity consumption (Beurskens and Jensen 2005). Based on sites within 30-km of shore and a generating density of 6 MW/km2, the offshore wind potential was estimated to be 596 TWh for water depths less than 10-m, and 3028 TWh for water depths less than 40-m (Gaudiosi 1996). Although this could be overestimated due to political restraints, the potential for further offshore development in deeper waters (e.g. North Sea) may substantially increase this estimate. By the end of year 2003, the European Wind Energy Association forecasted that by 2010, up to 10 GW offshore capacity would be operating in Europe. In the long term, a sea area of 150,000 square kilometers with water depth less than 35-m could be available, with the generation area capable of powering all of Europe (EWEA 2004).

8

Based on European wind atlas data for wind speed values through 10-km offshore, the United Kingdom is the leader for this potential, covering one third of the total potential (Table 6). This year 2004 estimate is a production level equivalent to three times the energy consumption in UK in 1998 (Sahin 2004).

The UK government’s Renewables Obligation established in 2002

mandated that 10% of the UK’s electricity must be generated through renewable sources by year 2010. In 2003, this was updated to require 20% from renewable sources by year 2020. Approximately 12,000 MW are required to meet the 10% requirement, which is equivalent to 1200-km2 of offshore generation area based on a typical 2 MW turbine (Byrne and Houlsby 2003).

1.3

United States

An exhaustive assessment of offshore wind energy potential for the United States was conducted (Kilar and Stiller 1980), resulting in a summary of the wind speed and water depth parameters for all potential wind farm water bodies in the US (Table 7). Significant portions of the US, namely the Bering Sea, Great Lakes, and parts of northern Alaska, are unsuitable for development due to deep water, floating ice, earthquake activity, or poor wind conditions. The most favorable regions found were the Northeast and Northwest Coasts and parts of southern Alaska. The study was based on a comprehensive evaluation for a 55-unit wind farm of 9 MW turbines on fixed and floating platforms in water depths from 3-m to 160-m. A tentative evaluation for water depth up to 50-m and wind speeds of 7 to 8-m/s resulted in an estimated 54 GW of total generating capacity at a wind power sea surface density of 6 MW/km2, with a total energy production of 102 TWh/year primarily in Northwest, Northeast, and Gulf of Mexico Coasts (Gaudiosi 1996). The Gulf of Mexico bordering the Louisiana coast has seen recent wind development interest in the concept of mounting wind turbines on existing oil platforms. The region has recorded wind speeds of 11-m/s at a height of 50-m above sea level. Wind Energy Systems Technologies LLC (WEST) has plans to develop 50 MW of power from several of the 5,200 existing platforms. WEST plans to purchase the platforms and liability for low cost, efficiently utilizing a structure which would cost up to $12M to disassemble and remove (Powers 2005).

9

Estimates of recently updated generating potential in the United States for shallow (i.e. less than 30-m water depth) and deep waters by region are summarized in Table 8. Distances of 0 to 5 nautical miles (nm) were completely excluded for environmental concerns, with distances from 5 to 20-nm having 67% exclusion, and 20 to 50-nm having a 33% exclusion of the potential developed area. In total, without considering the Gulf of Mexico and the Great Lakes, areas from 5 to 50-nm offshore in the U.S. contain 907 GW of energy potential, which is greater than current US electrical generation capacity. Only 10% of this 907 GW potential is in shallow waters, emphasizing the importance of the advancement of alternative technologies for deep water development (Musial and Butterfield 2004). A more detailed examination of offshore wind farm development in the New England region is provided later in the monograph.

1.4

Other Countries

In China, the extended shallow water depth of the Yellow and Eastern Chinese seas combined with predominant wind speeds between 7 to 8-m/s results in a total energy production of 254 TWh/year from 129-GW of installed offshore power (Gaudiosi 1996). Some of the issues with offshore wind turbines in this region of the world are their vulnerability to typhoons, resulting in extremely large wind and wave loading against the structures. As shown in Table 4, the countries bordering the Mediterranean and Black Seas have the potential to utilize over 12,000-km2 of seafloor area, resulting in an energy generation potential of 72 GW (Gaudiosi 1996).

1.5

Future Prospects in Offshore Wind Energy Development

Currently, plans for offshore wind farms across Europe and North America are in various stages of development. At present, the planning schedule is very dynamic and some projects have been put on hold or discarded altogether. Table 9 presents a list of the planned projects thus far. It is not considered to be comprehensive, and is based on year 2002 forecasts (Hendersen et. al. 2002). 10

In general, the trend of shallow water offshore wind energy development will likely show an increase in the number of turbines per wind farm, more efficient and cost-effective foundation installations, lighter component utilization for the wind turbine structure, and increased design lifetimes from 20 to 50 years (Gaudiosi 1996).

11

2.0 Types of Foundations 2.1

Introduction

The four main classes of offshore foundations consist of piled foundations, gravity base foundations, skirt and bucket foundations, and floating structures with moored foundations. The piled and gravity base foundations can be further classified into three structural configurations, namely, monopiles, which are designed as piled foundations and exhibit simplicity in fabrication and installation, tripod or quadruped configurations, which can be both piled or gravity based, and lattice configurations, which offer the most economical structural solution in terms of steel weight-to-capacity ratio. There are advantages and disadvantages fore each foundation class and structural subclass primarily based on site conditions and turbine size, as summarized in the following descriptions.

2.2

Piled Foundations 2.2.1 Introduction

Piled foundations are the most common form of offshore foundation, transferring both tensile and compressive loads from the foundation into the seabed. They have been installed since the 1940’s in water depths up to 150-m. They are simple to construct using large steel tubing and offer the most economical manufacturing option among the different types of foundations. The installation method involves lifting or floating the structure into position using equipment such as floating crane vessels, drilling jack-up units, and specially constructed installation vessels before driving the piles into the seabed. Installation depths are typically dependent on the environmental and soil conditions, and can range from 5-m to over 120-m below the seafloor for some offshore structures.

2.2.2 Monopiles Currently, the most popular design for offshore turbines is the monopile, a simple design that is made of large diameter (e.g. 4-m with tubing thickness of 5-cm) cylindrical steel tubing with a transitional sub-structure that connects the pile to the turbine tower, effectively extending the turbine tower below the water surface and into the seabed. The pre-fabricated transitional component is cast with the monopile, which can account for rotational and verticality errors that

12

occur during pile driving. The monopile is driven or hammered into the seafloor, and can be unsupported or supported (Figure 1).

Advantages include minimal seabed preparation

requirements, resistance to seabed movement, scour, ice flow damage, and inexpensive production costs. The disadvantages include sub-structure flexibility at greater depths, expensive and time-consuming installation that is dependent on turbine size and seafloor geology, and decreased stiffness relative to other foundation types, which can result in soil degradation or “potholing” around the pile (Irvine et. al 2003). Monopiles are also difficult to remove and as a result may pose problems for decommissioning activities. The limiting design condition for the monopile is the overall deflection and vibration during loading. Standard monopiles (i.e. no lateral support) are suitable for water depths up to 25-m. Supported monopiles (i.e. monopiles braced by lateral struts) are suitable for depths from 20 to 40-m, and are better suited to nonhomogeneous soils. Monopiles should generally be avoided in deep soft soils due to the required length of installation.

2.2.3 Multiple-Leg Foundations Multiple-leg foundations consist of tripod or quadruped structures made of cylindrical steel tubing with either angular or vertically-driven leg piles (Figure 2a), suitable for water depths from 25 to 50-m. They can also be built with a gravity base (Figure 2b), where piles driven below the base result in shared load between gravity and the pile, suitable for water depths up to 25-m. Advantages include their versatility, resistance to wave and current loading, and inexpensive fabrication cost. The disadvantages include expensive construction and installation, difficulty in multiple-pile removal, and their lack of resistance to ice flow damage (e.g. if slender sub-structure members are used).

2.2.4 Lattice Towers Lattice towers consist of a multi-leg foundation connected to a steel braced sub-structure (Figure 3). The piles are driven inside pile sleeves to depth for structural stability, suitable for water depths of 20 to 40-m. They are characterized by a stiff, dynamic response which makes them ideal for deeper waters in extreme environmental conditions, with the exception of ice load vulnerability due to the slender nature of the structural members. Another advantage is the ability

13

to be fully assembled before float-out installation. They generally require less scour protection than other piled foundation types.

2.3

Gravity Base Foundations 2.3.1 General Description

Gravity base foundations consist of a slender steel or concrete sub-structure mounted onto a single large square or circular foundation of reinforced concrete or a ballast-filled steel shell (Figure 4). They can comprise multiple foundations similar to the tripod or quadruped concept described in the previous section. Gravity base foundations can be skirted, which have the advantages of confining any soft soil layers and transferring the gravity load to the bearing soils, improving the hydraulic conditions, reducing the scour potential, and facilitating conditions for base grouting. Gravity base foundations are generally well suited to homogeneous soils due to settlement and bearing capacity distribution. However they can be used in virtually all soil conditions in water depths between 0 to 25-m. They require a flat base and scour protection requirements are dependent on local site conditions. The gravity foundation is designed to avoid tensile loads between the bottom of the support structure and the seafloor by providing sufficient self-weight dead loads to maintain stability. The overturning moment is resisted by a “push-pull” action where equal and opposite vertical loads (i.e. soil resistance on the downwind side and self-weight on the upwind side) act at the foundation level. An advantage to gravity base foundations includes simple transportation, as they can be installed partially in a fabrication yard and finished in the final position at sea, depending on the fabrication yard capacity, the available draft (i.e. water depth) during transport, and the availability of ballast materials. They can also be competitive when heavy lift vessels or other special installation vessels cannot be mobilized to the site. Gravity base foundations are also easily removed upon decommissioning.

14

2.3.2 Material Considerations Gravity base foundations are economically competitive when environmental loads are modest and when ballast for the foundation (e.g. pumped-in sand, concrete, rock, or iron ore) can be provided at low cost. Addition of ballast material can significantly reduce the size of the foundation required to achieve the conservative zero tension condition across the foundation base (Figure 5) (Houlsby et. al. 2001).

When considering the foundation material, although

lightweight, steel foundations can be filled with a very heavy mineral named olivine to stabilize the foundation (DNV-OS-J101 2004). In general, steel foundations are easier to transport due to lower weight and are easier to erect due to lightweight crane requirements. In either material case, gravity base foundations can be expensive due to the large weight involved, and will be subject to increased wave and current loads when the foundation is above the seafloor elevation.

2.4

Suction Caissons 2.4.1 General Description

Suction caissons are similar to gravity base foundations in shape and size but differ in the method of installation and primary mode of stability. They consist of a sub-structure column connected to an inverted steel bucket through flange-reinforced shear panels (Figure 6a. The shear panels distribute load from the column center to the edge of the bucket, which is comprised of vertical steel skirts extending down into the seabed from the horizontal base that rests on seafloor. The skirt length is approximately equal to the bucket width, where the volume of soil inside the caisson acts as a permanent gravity base foundation. Typical dimensions of suction caissons range from 2 to 4-m in diameter for water depths less than 5-m, and up to 12 to 15-m in deeper waters (Houlsby and Byrne 2000). Suction caissons can also be designed in a tripod or quadruped configuration, where the buckets replace piles or gravity base foundations in a conventional multi-leg structure, offering the advantages of applicability to deeper water, smaller required caissons, and easier ability to level the structure. The primary limiting design condition for the monopod suction caisson is the overturning moment, while for a multi-leg suction caisson configuration, the resistance to tensile loads is paramount.

Currently, there are no design

guidelines for suction caissons subject to large moment load to vertical load ratios, and should be designed on a site-specific basis (Byrne and Houlsby 2002).

15

2.4.2 Installation Principle The installation method of suction caissons is through a pressure differential, where water is pumped out through top of bucket once the rim of bucket seals with the seafloor (Figure 6b). This produces a net downward pressure, or suction, forcing the bucket into the seabed. In clays, the pressure is sufficient to bring the suction caisson to depth, but in sands, water inflow reduces the effective stresses in the sand, allowing the bucket to penetrate the seafloor. Once installed to sufficient depth, the pumps are removed and the valves are sealed, with the sand quickly regaining its bearing capacity. Suction caissons can easily be removed by reattaching the pumps and pumping water back into the bucket cavity, forcing it out of the seabed. The application principle of the suction caisson capacity is based on the relatively short wave periods to which the caissons are subjected, as time for the cavity within the bucket to fill during wave uplift is too short for significant inflow. The suctions that can develop are inversely proportional to the soil’s permeability, with the drainage path and wave periods to which the buckets can be subjected dependent on the depth of the bucket penetration into the seabed.

2.4.3 Site Considerations Suction caissons are most suitable in homogenous soils due to differential settlement and bearing capacity issues, but work well in sands and soft clays in a variety of water depths and tidal conditions. The advantages of this type of foundation include the ability of floating the structure to the site and the avoidance of heavy lifting equipment and pile driving requirements, making installation quick and inexpensive, as well as easy removal upon decommissioning. There is a large cost savings through the short installation time and minimal amount materials necessary for ballast. It is also not as weather dependent as a piled foundation installation is. A key requirement to suction caissons is the verification of installation ability and uplift capacity, where the seafloor soils must be penetrable and not prone to scour. A disadvantage to using suction caissons is the limited proven installation data for different types of soils, requiring detailed installation analyses prior to design. Suction caissons are also susceptible to scour in shallow waters, and piping below the bucket tip may occur in sandy soils.

16

2.5

Floating Structures 2.5.1 Introduction

There are two primary types of floating structures suitable for offshore wind turbines; the tension-leg platform and the low-roll floater. Both configurations can be used for single or multiple turbines, however the latter is far more cost-effective due to the installation cost of moorings and/or anchors. The feasibility of floating structures for wind farm design is still in its infancy, with many novel propositions currently under review. These two most common types are suitable for large water depths (greater than 50-m) and will be discussed briefly (DNV-OSJ101 2004).

2.5.2 Tension-Leg Platforms Tension-leg platforms, which are submerged using tensioned vertical anchor legs with or without ballast tanks (Figure 7), can be floated to a site in fully-commissioned condition and simply connected to the moorings or anchors. The base structure dampens the motion induced by wind and wave loading. Advantages of tension-leg platforms include easy disconnection for transport for repair or maintenance and installation depths in over 1000-m of water.

2.5.3 Low-roll Floaters Low-roll floaters are stabilized by mooring chains and anchors which dampen the motions of the platform (Figure 8). There is a stabilizer installed at bottom of floater to reduce roll. The installation is simple, similar to tension-leg platforms. The anchors may be fluke anchors, drag-in plate anchors, suction anchors, or pile anchors.

17

3.0 3.1

Design Methodologies Limit State Design

The limit states design methodology is the most typically used method for offshore wind turbine support structure analysis. A limit state is defined as the point when a structural element will no longer satisfy its design requirements. There are four types of limit states; the ultimate limit state (ULS), the fatigue limit state (FLS), the accidental limit state (ALS), and the serviceability limit state (SLS). For offshore foundation design, the ULS uses non-degraded soil parameters, while the other limit states use cyclically degraded soil properties for design calculations (Feld and Waegter 2002).

3.1.1 Ultimate Limit State The ULS corresponds to the maximum load carrying capacity, examining strength and overall stability aspects through elastic and plastic analyses by applying design loads and design material properties.

3.1.2 Fatigue Limit State The FLS is failure resulting from cyclic loading due to installation, operational, and nonoperational loads. Installation fatigue is based on a pile drivability analysis and should include the total fatigue damage accumulation.

3.1.3 Accidental Limit State The ALS corresponds to damage sustained through an accidental event or operational failure. It examines load conditions resulting typically from ship impact using very high load characteristic values applied to the structure for short durations, resulting in undrained soil conditions that may increase the potential plastic movement of the soil surrounding the foundation. Cyclic soil strength parameters are not considered due to the inability for soil degradation to occur during the loading duration.

18

3.1.4 Serviceability Limit State The SLS is a set tolerance criteria applicable to normal use or durability. It uses characteristic values to analyze deformations that result in settlements or structure rotations in which aesthetic or operational demands are compromised.

3.2

Load Resistance Factor Design (LRFD) Method

The LRDF method applies load factors to characteristic values of loads acting on the structure or load effects within the structure, and resistance factors to the resistance of the structure or strength of the structural materials. These factored values are compared to a specified design criterion to meet a specified target safety level.

The characteristic values of the load are

dependent on limit states accordingly; for ULS, the value is defined as the load effect with an annual probability of exceedance equal to or less than 0.01 (100 year recurrence period); for FLS, the characteristic load effect history is defined as the expected load effect history; for ALS, the value equals the most probable annual maximum load effect; and for SLS, the value is specified depending on certain requirements (DNV-OS-J101 2004). The basis for selection of the characteristic loads for operational design conditions is summarized in Table 10. The characteristic resistance is defined as the 5% quantile in the resistance distribution. Load factors for the ULS and ALS are summarized in Table 11. For permanent loads (G) and variable functional loads (Q), the load factor should be taken as ψ = 1.0, unless the load is such that a reduced value leads to an increased load effect in the structure. In such a case, a value less than ψ = 1.0 is required. The load factors for FLS and SLS are equal to 1.0 (DNV-OS-J101 2004). The structure should also be able to resist fatigue loading and meet serviceability requirements during temporary or operational design conditions.

3.3

Direct Simulation

The direct simulation method is similar to the LRFD method, except that it is based on direct simulation of the characteristic combined load effect from simultaneously applied load processes rather than a linear combination of individual load effects. It is generally more attractive when a structure is subject to two or more simultaneously acting loads, such as in the case of combined wind and wave loading. It can be inadequate when a load effect of an applied load process is 19

dependent on structural properties that are sensitive to other load processes. Since correct assumptions regarding structural damping during various loading regimes (e.g. wind flow, power production, wave load alignment) of a turbine may be difficult to determine, separate determination of the wave load effect alone may not be feasible. The characteristic combined load effect is obtained directly from the distribution of the annual maximum combined load effect from structural analysis based on the simultaneous application of two or more load processes (e.g. for ULS design, 99% quantile or 100 year recurrence period). The characteristic resistance is simply calculated as for the LRFD method (DNV-OS-J101 2004).

3.4

Testing Based Design

The testing-based design method consists of establishing load and resistance effects by testing or observation of performance of the full-scale structures. It is supported by analytical design methods and includes observation of the performance of the structure.

3.5

Probability Based Design

The probability-based design method defines the structural reliability as the probability that the structure will not exceed a specified failure criterion within a specified time period.

The

reliability analysis is based on Level 3 reliability methods, which are applicable to the calibration of Level 1 methods (e.g. deterministic analysis using one characteristic value to describe each uncertain variable), special case design problems, and novel designs for which limited experience exists (DNV-OS-J101 2004). The Level 3 methods use probability of failure as a measure, requiring knowledge of the distribution of all load and resistance variables. The target reliabilities should correspond to the consequences of failure, based on established cases known to have adequate safety. Otherwise, the target reliabilities can be based on transferable target reliabilities similar to existing design solutions, internationally recognized codes and standards, or other bases of transfer explained in “Classification note 30.6: Structural Reliability Analysis of Marine Structures” (DNV-OS-J101 2004).

20

4.0 Site Investigations 4.1

Introduction

The environmental conditions at a proposed wind farm location need to be determined in order to provide an accurate characterization of the environmental loads and subsurface soils to carry out the design process. The site conditions include data on the local geological, oceanographic, meteorological, human, and environmental characteristics of a wind farm site.

Table 12

summarizes the environmental parameters that affect offshore wind farm design, providing an outline of the various equipment and techniques used to determine these parameters throughout the different stages of an offshore wind farm site investigation. Although many of these parameters have a direct impact on the foundation design, detailed knowledge of the subsurface conditions is paramount for the safe and economical design of offshore foundations. The subsurface conditions are used to predict the overall response of the foundation for input into structural design models. The purpose of a subsurface investigation is to define various soil and rock data, their associated physical, mechanical, and engineering properties, and the potential geohazards to the structure. The extent of the site investigation is based on the depth and area of the seafloor that will affect or be affected by the foundation. The zone of interest to current offshore wind farm development lies within the continental shelf region of most of the world. The average width of the continental shelf is approximately 70-km, the widest areas occurring in the Arctic Ocean and the north and west sides of the Pacific Ocean, and the narrowest occurring in areas near young mountain ranges. The depths of these areas range from 10-m to 500-m with an average slope of less than 1 degree, before transitioning to the continental ridge and continental slope (Poulos 1988). The deepest areas occur near glaciated zones, and the shallowest near zones with extensive coral growth. The soil types occurring along the continental shelf and ridge are primarily fluvial marine deposits of sand and silt, transitioning to silt-clay mixtures and clay further offshore, as illustrated in Figure 9. Seabed soils are often normally consolidated and exhibit extremely compressive characteristics. However, they can vary considerably along coastal areas throughout the world. For example, in the UK, loose mobile sand banks are often found at shallow depths, underlain by soft clay in 21

some areas and glacial till in others. The northwest coast of Australia and parts of the Gulf of Mexico are characterized by deposits of soft calcareous sand-silt mixtures with localized areas of cementation that can be embedded through tens of meters of soil depth (Byrne and Houlsby 2003). Therefore, a foundation type that works in one area of a wind farm may not be appopriate at an adjacent location.

4.2

Phases of the Site Investigation 4.2.1 Introduction

Several major phases are necessary for an offshore site investigation (OSI) that may comprise up to a year or more of time investment depending on the available data for the proposed location and the extent of the project. Initially, a planning process must be completed, beginning with a definition of the scope of the work to be done and ultimately mobilizing the contractor to perform the work (Table 13). The initial regional desk study is completed during this process to identify an area of interest for a wind farm and to create a plan for the site investigation. The regional desk study incorporates all available data to define the probable site conditions that summarize the current state of knowledge of the site. It will also identify potential constraints and assist with further planning of the site investigation. The cost of this initial phase typically comprises approximately 2% of the site investigation costs (Poulos 1988). It should include investigation of geological databases, bathymetric data, geophysical and geotechnical data, metocean data, seismicity, performance of existing offshore wind farms, human activities in the area (e.g. pipelines, shipping routes, et cetera), and environmental concerns (e.g. bird migratory routes).

The planning process can take up to 6 months for large scale developments, and

generally takes a minimum of 3 months for small scale developments (Randolph and Kenkhuis 2001). The geotechnical risk involved in the development of a proposed offshore wind farm requires a concerted effort from developers and contractors alike to reduce the risk for all parties involved in the project. The three key elements in managing geotechnical risk are the initial regional desk study, the risk assessment, and the site investigation. The risk assessment is based on the regional desk study and the proposed development to define the range of geotechnical risks

22

involved, their potential consequences, and the options for reducing the impact of those consequences. This process should remain ongoing throughout the development process. The site investigation provides the site information necessary to carry out the proposed development. It should be optimized based on the results of the desk study, the risk assessment, and the geological and geophysical surveys. Integration of these phases is necessary to ensure that the soil conditions are determined to the fullest extent required to provide a sound geotechnical foundation design. Quality control and quality assurance measures should be implemented throughout the investigation to ensure that the data obtained are as accurate and as relevant to the proposed development as possible. The site investigation can then be divided into the geological study, the geophysical survey, and the geotechnical site investigation, which account for approximately 79% of the total cost. The remainder of the site investigation cost is divided between the laboratory testing and the engineering analysis and reporting, which account for 8% and 11% of the cost, respectively (Poulos 1988). The costs of a site investigation can be expensive, typically averaging $1M per week, primarily due to the mobilization of the vessel used (Randolph and Kenkhuis 2001). Using lump sums can help to control these high costs, but typically day rates are imposed by the contractor due to the high degree of risk assumed. These risks are primarily due to the variable nature of marine operations, specifically the need to halt investigation progress during inclement weather. Table 14 shows an offshore site investigation schedule representative of a typical offshore project. The general timeline will be tailored to the specific project, with phases such as the collection of oceanographic data reducing in duration if the environmental site conditions are already known.

4.2.2 Geological Study The geological study phase is based on the geologic history of the area, and when combined with the type, size, and importance of the wind turbine structure and the homogeneity of the soil deposit, it serves as a basis for the extent and method of the geotechnical site investigation. Geological maps and profiles are often available to provide general site information. Typically,

23

top soil deposits can help to identify zones of interest and aid in locating anchoring locations for subsequent investigation activities.

4.2.3 Geophysical Survey The geophysical survey establishes seafloor bathymetry, topography, stratigraphy, and identifies hazardous areas on the seafloor.

The upper stratigraphy should be assessed to determine

anchoring locations and potential problems with positioning jack-up rigs, as they may be used within close proximity to adjacent existing structures, affecting the surrounding soil. The topography should be assessed on a scale sufficient to identify outcrops, obstructions, and seafloor non-uniformities which can affect the relatively small footprint of typical wind turbine foundations. This is important in minimizing differential settlement of gravity base foundations over the lifetime of the structure. Sub-base grouting is typically used to account for these nonuniformities. The techniques used consist of bathymetry mapping with conventional single or multi-beam echo soundings or swathe bathymetry, seafloor mapping with side scan sonar or magnetometer readings, and seismic profiling methods using boomers, sparkers, pingers, chirp profilers, or high resolution digital surveys using airguns. A survey line plan is constructed for a proposed sight to create a grid over which the various characteristics identified in the geophysical survey can be plotted. Any existing data (e.g. boreholes) should be plotted on the survey line plan so that the geophysicists conducting the survey can calibrate survey data against borehole data to check for continuity between markers. Grid spacings are typically 75 to 100-m apart with cross lines spaced at 300 to 500-m, both of which can be adjusted for areas where known hazards exist (Randolph and Kenkhuis 2001). 4.2.3.1 Bathymetry Mapping The conventional echo sounder is a heave-compensated device which is calibrated to the seawater sound velocity. It produces spot measurements of depth along survey tracklines at distances of 25-m to 50-m apart that compile to form a contour map of seafloor. A Swathe device is a heave, pitch, and roll-compensated beam that records spot measurements as it sweeps from side to side over the area of the proposed site. The width of the coverage is approximately

24

four times the water depth. Because of the three dimensional capabilities of the system, it is ideal for areas of highly variable topography. The minimum required resolution for single and multi-beam echo sounders is approximately 1% of the maximum water depth (OSIC 2004). 4.2.3.2 Seafloor Mapping The side scan sonar system maps topographic features on the seafloor and identifies objects on or above the seafloor. It can help identify the type and distribution of sediment and surface forms such as sand waves. The system uses narrow beams of sound waves that are transmitted out to the sides and across the bottom of a hydro-dynamically stable towfish from transducers and reflected back to a receiver on the towfish. The system is dependent on frequencies of the transmitting waves, with a trade-off between resolution and distance traveled.

Higher

frequencies (500-kHz to 1-MHz) are high resolution, short distance waves, and low frequencies (50-kHz to 100-kHz) are the opposite. The intensity of the returning wave is dependent on both the topography and the material properties. The harder the object or material (e.g. gravel) on the seafloor, the darker the resulting image, as they reflect energy better than soft materials (e.g. muddy sands). Shadows show up as white areas due to the absence of reflected sound. When sand waves and ripples are large enough to produce shadows, there is an indication that currents are particularly high in that area and foundations would be subject to considerable scour. The resulting map of the variation in color, when combined with bottom samples or cores, produces a detailed map of seafloor features and interpreted sediment types. The side scan sonar system has a minimum required resolution of detecting a 10-cm3 object or a 10-cm wide linear object (OSIC 2004). The magnetometer device maps the locations of cables, pipelines, shipwrecks, and other ferrometallic objects, such as munitions dumps. The minimum required resolution is approximately one nanoTesla (i.e. one-billionth of the magnitude of a magnetic field vector necessary to produce a force of one Newton on a charge of one coulomb moving perpendicular to the direction of the magnetic field vector with a velocity of one meter per second) (OSIC 2004).

25

4.2.3.3 Seismic Profiling The seismic profiling methods use a sound source and receiver combination that produce a reflection record of the cross-section of an area of interest. The different methods available range in their ability to produce clean profiles (Table 15). Similar to side scan sonar, there is a tradeoff between the resolution and the depth of penetration, the latter of which is dependent on the seafloor composition and the underlying layers.

Typically, the minimum required vertical

resolution should be better than 0.5-m down to a depth of 5-m (OSIC 2004). The choice of the type of device is ultimately dependent on the specific application. However, the most precise is the high resolution digital survey (HRDS). Using an airgun as the sound source, the HRDS consists of multi-channel receivers aligned in streamers that vary in length, spacing, and grouping to target the resolution and depth of penetration of interest, which can reach up to 60-m (Jenner et al. 2002). The advantage of the HRDS is that both the sound source and the streamer are towed at shallower depths than for other equipment, resulting in increased ability to account for reflections from the water surface. However, the method is more affected by sea-state variability.

4.2.4 Geotechnical Site Investigation The geotechnical site investigation (GSI) is conducted following the geophysical survey to use the information obtained to target soil strata changes or specific seafloor features. The GSI includes in-situ testing for stratification identification and soil coring and sampling for material identification, characterization, and subsequent laboratory testing.

The geotechnical site

investigation characterizes the material properties of the soil in terms of five components: macroscopic (e.g. historical and regional characteristics), mechanical (e.g. strength and in situ stress conditions), microscopic (e.g. particle size, roundness, and shape), chemical (e.g. molecular structure), and water content properties. The GSI begins with in situ testing to obtain the stratigraphy of the area of interest, typically done with a cone penetrometer test. Once assessed, the stratigraphy will guide the development of the remaining in situ testing, sampling, and laboratory testing program. After completion, the geophysical survey and geotechnical site investigation data combine to form a geotechnical model of the area to develop the required design parameters.

26

The depth and aerial extent of the geotechnical site investigation are based on numerous factors, including the local geology, availability of previous investigations, site accessibility, soil variability, foundation structural configuration, the expected environmental loads, and limitations related to time and budget resources. The depth is primarily dependent of the presence of weak underlying layers which may influence foundation behavior. General rules of thumb have been established for typical foundation types to limit the required extent of the investigation. For pile foundations, borings and in situ testing should generally extend to at least one pile diameter below the level of the pile tip. For gravity base foundations, borings and in situ testing should extend to a depth equal to at least the largest horizontal dimension of the foundation structure. For single wind turbine structures, one boring to sufficient depth for recovery of a sample is recommended (DNV-OS-J101 2004). For wind farms, the soil investigation program should include combined techniques (e.g. one boring in each corner and center of farm combined with one CPT at each foundation location). The soil investigation program should be monitored by a geotechnical engineer to allow for program adjustments in case of soil non-homogeneities, the presence of problematic soils, or modifications to the second phase of testing if using a multiphased investigation program. The variety of equipment and techniques used in offshore site investigations is required to obtain quality in situ test results and quality samples for the array of soil conditions (e.g. various combinations of sand, silt, clay, rock, and shell deposits) in the offshore environment.

4.3

Offshore Geotechnical Site Investigation Vessels 4.3.1 Introduction

Site investigations can be performed from a variety of vessel types, the selection of which is controlled by sea conditions (e.g. water depth, sea state, anchoring conditions), the planned program of investigation, and the availability and cost of the equipment (Watson 2000). Vessel deployment and use can be expensive and usually constitutes the largest component of the site investigation cost. The options are grouped into four basic categories: in waters less than 20-m deep, a support vessel or low-draft barge using commercial divers is typically used; in waters greater than 20-m, specialized drilling vessels are used, or jack-up rigs if the water depth is less than 80-m; and for all water depths deployment of remote equipment from support vessels is typically included (Randolph and Kenkhuis 2001). Depending on the type of vessel used,

27

anchoring may be necessary and is limited to water depths less than 200-m; otherwise dynamically positioned vessels are used.

4.3.2 Low Draft Barges Low-draft barges are ideally suited to perform structure-to-shore pipeline and cabling investigations. Anchored using a multi-point mooring system, the minimum number of anchors is generally four unless high ocean currents cause instability.

4.3.3 Jack-up Rigs Jack-up rigs provide a stable working platform, however, expensive daily rates (e.g. up to $150K per day) and significant support requirements (e.g. manpower, transportation, and support vessels) can reduce their cost-effectiveness (Randolph and Kenkhuis 2001). They are typically used for oilfield activities and therefore tend to be unsuitable for geotechnical work unless landbased drilling equipment is provided. They also have space limitation issues and are not used for multi-boring investigations, as repositioning the rig is time-consuming.

4.3.4 Specialty Geotechnical Drilling Vessels Specialized geotechnical drilling vessels are ideal for multi-boring investigations. They typically have an on-deck crane with ample space, a moonpool for lowering equipment, and use a heave compensation system that can produce high quality geotechnical results. Their disadvantages include cost (e.g. up to $100K per day), limited endurance, and limited fresh water supply, which reduces the choice of drilling fluids to seawater-compatible types (Randolph and Kenkhuis 2001). The availability of these types of vessels is fairly limited. In the year 2002, there were less than 10 geotechnical drilling vessels in operation throughout the world (Jenner et. al. 2002). Most of these vessels are self-anchored with dynamic positioning systems (DPS) that allow deep water investigations to be conducted. The DPS also allow for large-scale investigations while minimizing the risk of losing expensive equipment. A wheeldrive drilling system is typically used that allows for a high production rate and affords the capability of continuous profiling of the subsurface stratigraphy.

28

4.3.5 Semi-submersible Drilling Rigs Semi-submersible drilling rigs are similar to jack-up rigs but more expensive with costs up to $275K per day (Randolph and Kenkhuis 2001). The anchoring process can take up to two days between installation and removal. However, multi-boring capabilities are possible with the long anchor lines that allow the rig to move over the footprint of a large foundation.

4.3.6 Supply Vessels Supply vessels are used for deploying remotely operated equipment using an A-frame or on-deck crane. They have a multi-point mooring system limited to less than 125-m of water depth without anchor handling assistance. They typically have a poor ability to maintain a stationary position and therefore should only used to transfer low-risk equipment. Costs are typically up to $25K per day (Randolph and Kenkhuis 2001).

4.4

Drilling

The drilling process should be conducted with the objectives of minimizing sample disturbance and providing borehole stability. The drilling process can contribute to the site condition insight as some of the drilling parameters give a rough idea of the soil conditions via interpolation/extrapolation of soil data, explanation of sample disturbance, and detection of soil stratigraphy. In certain soil types, such as soft rock or hard clays, rotary core drilling may be the best option. The sample quality is much less than for push or position systems, but recovery tends to be adequate. The most common types of rotary core drilling are Christensen coring and high speed diamond coring. Where more detailed investigation is required, geophysical borehole logging can be used through a wireline system. Some of the tools in use include: calipers, which measure the borehole diameter; gamma radiation, which measures the radioactive isotope concentration and identifies soil type and grain size distribution; neutron devices, which provide information on porosity; sonic devices, which measure sound velocity to identify soil type and character; and density devices, which provide information on the soil density.

29

Where shallow gas may be encountered, special precautions should be taken during drilling procedures, which include but are not limited to, no hot work or smoking on deck, keeping a supply of heavy mud available, having gas detectors on deck, and utilizing wind and current meters to provide information on gas dispersion should it escape from the borehole. A pilot hole without any downhole tool use should be drilled with a non-return valve in the drillstring. Using this technique, a borehole can be drilled next to the pilot hole with full use of downhole tools. A safety valve on top of the drillstring should be included when using wireline downhole tools, and an offset between the vessel and the borehole should be applied based on wind and current conditions. The drilling and coring process differs across the various types of vessels used. With barge vessels, subsea methods (i.e. divers on seafloor conducting the test) have several limitations including operating only during daylight hours, divers having little geotechnical drilling experience, short bottom times, shallow borehole depths (e.g. < 40-m), and drilling fluids are limited to seawater due to visibility requirements for the divers.

Video and audio

communication allows for the geotechnical supervisor aboard the barge to direct activity. However, complete oversight is uncommon as it also their responsibility to log and pack samples. From jack-up units, onshore drilling equipment can be mobilized on the rig, allowing experienced drillers to conduct the drilling operations. This technique usually produces the highest quality results provided that all necessary investigation equipment is available. A choice of drilling fluids is possible through ample fresh water supply, and operations can continue 24 hours per day. From specialty geotechnical drilling vessels or semi-submersible rigs, four main components comprise the operation: a seafloor reaction frame and utility guide frame, a heave compensation system, a bottom hole assembly, and umbilical downhole probes. The seafloor guide frame is lowered into position under tension, using ballast to provide reaction on the seafloor. During drilling operations, a hydraulic clamp mounted on the guide frame provides the reaction to the drill string. The drill string is typically 5-in in diameter to allow for the downhole tools to pass

30

through it. On specialty drilling vessels, a mud swivel on the top of the drilling unit allows insertion of the tool. However, on semi-submersible rigs, the drill string must be broken to allow insertion. The bottom hole assembly consists of an open drag bit, drill collars, and latching sections to provide weight to the setup and connect the downhole tool. The weight is necessary to provide reaction to the motion compensator to prevent disturbance to the testing and sampling activity.

4.5

Sampling 4.5.1 Introduction

There are three main types of testing, sampling, and coring methods: the wheeldrive system, push and piston sampling, and coring techniques. Within these three categories are the types of equipment used, with consideration in order of priority given to 1) thin-walled piston sampler, 2) thin-walled push sampler, 3) thick-walled push sampler, 4) hammer sampler, and 5) rotary core sampler. The techniques and equipment type should be chosen with consideration of the soil type, the type of testing to which the samples will be subject to, and of minimizing sample disturbance during both penetration of the device and extraction to the surface. Sampling can be done using one of two primary methods: the seafloor mode, where sampling is performed from the seafloor using a reaction frame with drilling; and the drilling mode, where sampling is performed through a drillstring at the bottom of the borehole. The seafloor mode is a umbilical technique that uses a electrical wires to send data to the surface in real time and hydraulic hoses to perform the test. The technique allows for alternative sampling and testing to be conducted at one borehole. The length of the sample obtained is dependent on several factors that include the geometry and characteristics of the sampler, the soil type, and the type of applied force (e.g. push, vibration, hammering). The cores removed from the sampling device should be photographed upon extraction and recorded in terms of color, texture, odor, and moisture characteristics. This should be performed immediately following removal as the appearance of the materials can change quickly when exposed to air. If rock coring is expected, is it possible to use a “piggyback” hard-tie (i.e. motionless) system of mounting a separate drill rig on top of a platform

31

which is mounted onto a separate riser that is secured to the guidebase on the seafloor. A separate drill string is then used to core the rock samples through the riser system. The speed of the sampling process and penetration rate is dependent on the phase of the site investigation. For the geotechnical phase, the penetration rate of sampling equipment will be tailored to achieving the highest sample quality for subsequent laboratory testing, while during the geological phase, the sampling penetration may be faster in order to cover as much area as possible for determining site stratigraphy. Any shock or vibration should be avoided unless samples are being retrieved for classification purposes only. Recovery of the sample on deck should also be conducted in a manner that minimizes sample disturbance.

4.5.2 Wheeldrive System The wheel drive system utilizes a heavy frame to provide thrust reaction, driving the rod into the seafloor using hydraulic motors mounted in series on top of the frame. It is primarily used for cone penetration testing, T-bar testing, and vane shear testing. However, it can be modified to perform rock coring and piston sampling. It is currently the most advanced equipment for deep water site investigations.

4.5.3 Grab Sampling Grab sampling is used to obtain large volumes of soil for classification purposes. Its applicability is suited for pipeline investigations where scour and erosion can be assessed for long lengths of the seafloor in a relatively short period of time. The grab sampler is essentially an articulated bucket ranging in size from a few liters to a cubic meter that closes upon contact with the seafloor. The method is simple and inexpensive, and can be equipped with hydraulics to obtain a higher recovery. It can also be used when other devices (e.g. vibro corers or gravity corers) cannot penetrate the seafloor sufficiently due to harder surface materials. The disadvantages with grab sampling are the high level of sample disturbance, the risk of washout during retrieval, and shallow depth of penetration. Non-representative samples may be obtained in cobbled soils and gravels and therefore require multiple grabs to achieve a large enough volume to adequately represent the soil type.

32

4.5.4 Push and Piston Sampling Piston samplers are a modification to drop corers via addition of a cantilevered weight attached to the top of a core barrel. The weight acts to penetrate the core barrel into the seabed at a predetermined height above the seafloor. The piston inside the plastic-lined barrel ensures adequate suction pressure to retain the sample in the barrel during extraction to the surface. For both piston and push sampling, thin-walled tubes (i.e. < 2-mm with diameter ≈ 75-mm) should be used whenever possible to minimize soil disturbance. In highly sensitive soils such as soft clays, shallow cutting angles (i.e. 5 degrees) on the shoe of the sampler should be used (Norsok 2004).

4.5.5 Coring Techniques 4.5.5.1 Remote Coring The remote coring system is an autonomous drilling unit that is hydraulically operated from a deck control system. The sampling is conducted via video monitoring in real time. The drilling fluid consists of seawater and the maximum drillable depth is approximately 5-m.

The

maximum water depth where this system can be used without weight compensation for the hydraulic umbilical line is approximately 130-m (Randolph and Kenkhuis 2001). It is ideally suited for pipeline investigations where numerous borings must be taken.

The subsurface

stratigraphy can be monitored during testing by mounting additional instrumentation onto the reaction frame resting on the seafloor. 4.5.5.2 Vibro Corer The vibro corer is used when sandy soils are prevalent and seafloor penetration resistance may be too high for the self-weight insertion of the piston corer. An electrical vibrating motor allows the plastic-lined steel tube corer to penetrate the seafloor via rotating eccentric counter weights, with barrel lengths typically reaching up to 8-m (OSIC 2004). To retain the sand sample upon extraction to the surface, a core catcher is fitted to the cutting shoe at the toe of barrel. Since the sample is considered highly disturbed due to vibration and densification in the barrel, their use is typically for classification purposes only. The water depths suitable for the vibro corer are generally less than 800-m (OSCI 2004). The advantages with vibro coring include its simple, lightweight, and inexpensive deployment. The disadvantages include lack of penetration ability in dense sands or hard clays, lack of stratification identification accuracy due to compaction,

33

plowing, sample loss issues, high level of disturbance applied to the soil, and difficulty with the handling of the larger vibrocore equipment. 4.5.5.3 Gravity Corer The gravity corer is typically used in place of the vibro corer in soft soils, as sample disturbance tends to be lower without vibration, especially when piston-type gravity corers are used. The principle of the device is to utilize the force of gravity during the last 5 to 10-m of deployment to allow the corer to penetrate the seafloor under its own weight. Quality samples can be obtained for depths up to 20-m below the seafloor with some of the larger piston-type gravity corers. The method is inexpensive, quick, and easy to conduct. The disadvantages with the gravity corer include lack of penetration is dense sands and stiff clays, and danger or difficulty in handling the trip-release mechanism on the piston-type corer. 4.5.5.4 Box Corer The box corer is a lightweight system used to obtain an undisturbed block of clay that can later be trimmed into multiple specimens for laboratory testing. The box corer is suitable primarily for soft clays and can retrieve samples from 10 to 50 liters in size (OSIC 2004). The primary disadvantage is the limited depth of penetration for the device. 4.5.5.5 Rock Corer The rock corer is a seafloor mounted device that can obtain continuous samples of rock from 3 to 9-m in length (OSIC 2004). It consists of a rotary drive mechanism with a diamond or carbide tipped drill bit with an inner core barrel. Remote systems exist that allow for autonomous handling of the equipment in deeper waters. Although the remote systems allow for deep water use, the reduction in the human factor of equipment operation tends to decrease the core quality. The main disadvantage to the rock corer is the lack of ability to obtain high quality cores. Lack of hole stability below granular soils, and size and weight of equipment are other drawbacks to its use.

34

4.5.6 Other Sampling Methods Besides soil sampling, pore water, gas, or ambient pressure sampling can be conducted to obtain other data important to site investigations and foundation behavior. For deep water sampling, drillstring-based equipment such as wireline percussion samplers, latch-in push samplers, hydraulic piston corers, pressurized core barrels, and samplers operated in a stabilized drillstring can also be used to obtain soil samples (Poulos 1988). Further details for soil investigations and sampling are explained in the Norsok Standard G-001 Marine Soil Investigations (Norsok 2004) and the DNV Classification notes No. 30.4 (DNV-OS-J101 2004).

4.5.7 Sample Recovery, Storage, and Transport Once the sample has been retrieved from the seafloor and brought to the deck of the vessel, care must be taken during handling and storage activities to minimize sample disturbance and maximize sample recovery. Samples can be stored in the tubes with which they were obtained in for push and piston samples. By placing a cap on the bottom of the tube immediately upon retrieval, soil and water escape will be prevented. The sample is then either extruded for offshore laboratory testing or sealed for transport to the onshore laboratory dependent on the strength and stiffness of the soil and the possibility of access to free water or mud. Extruded samples intended for onshore testing should be sealed with plastic and aluminum foil, followed by waxing to seal in the moisture content. It is important to note the spatial orientation of the sample in its in situ state to provide a correct reinstatement of the in situ stress state during laboratory testing. All samples being stored should be located in an area with a constant temperature (e.g. 7°C in onshore laboratory) and humidity, away from locations where excessive vibrations or heat transfer may occur, such as engine rooms. Samples should never be exposed to temperatures below freezing, as ice will form in place of the water causing swelling and characteristic soil properties to change. Transportation of samples should be conducted to minimize shock and impact loads to the samples. The samples should also be stored and transported in the same orientation from which they were retrieved. Horizontal storage and transport can be used for firm or stiff soils.

35

4.6

In Situ Testing 4.6.1 Introduction

In situ testing techniques include cone penetration testing (CPT), field vane testing, and various geophysical survey techniques such as downhole seismic testing. In situ testing is conducted using one of two methods similar to the sampling modes, namely the seafloor mode and the drilling mode. In both methods, the in situ tool is inserted into the seabed soils to either a predetermined depth or until refusal is met. With the seafloor mode, it is important to ensure that the rig does not affect the results of the in situ test. This can be reduced by considering whether the rig’s footprint is circular or square shaped with an open space where the in situ tool is placed, whether skirts could transfer the rig weight to stiffer surrounding soils, and whether the weight of the rig is sufficient to provide no more than what is required to maintain position during testing. Video cameras can be mounted to the rig to monitor testing and verify that these conditions are optimized. Remotely operated vehicles can also be used to perform shallow testing. The drilling mode is conducted to minimize soil disturbance below the drill bit. The in situ tool should be extended to a depth of 1-m below the drill bit if soil strength and density allow.

4.6.2 Cone Penetration Testing CPT testing typically uses a instrumented probe that allows for measurement of cone tip resistance, qc, sleeve friction, fs, and for the piezocone device, excess pore water pressures, u, that generate as the cone is pushed through the soil. The CPT is primarily a profiling tool, however, the measurements obtained from the two load cells can provide estimates of undrained shear strength with consideration for the degree of uncertainty, which can be rather high. The parameters that can be derived from the test include soil type, relative density, and stress history using empirical correlations. The advantage of the CPT over coring methods include a greater penetration depth, continuous measurement of the stratigraphy, real time data acquisition, reduction in the amount of laboratory testing that is typically required for samples, reliability of relative density determination for granular soils, and speed of testing. The main drawback of the test is the lack of retrieval of a soil sample, which is necessary to verify that the empirical correlations used are valid for the particular type of soil. The weight of the equipment required for reaction force is typically high, ranging from 5 to 10-tonnes, and the expertise required to conduct and interpret the test properly are other disadvantages of the CPT. 36

The testing procedure should be conducted at a penetration rate of 2-cm/s as continuous as possible dependent on the limitations of the equipment. Pore pressure dissipation tests are conducted during pauses in the penetration phase to assess the porosity of the soil. They should be carried out to greater than 50% consolidation, depending on the importance of the results and the time necessary to achieve the target level of consolidation.

The accuracy of the

measurements is based on soil type, distance between measurements, and the required degree of precision necessary (Table 16). Class 1 is for precise characterization in soft or loose soils. Class 2 is more appropriate for stiff clays and sands, while Class 3 is for stiff or dense soils. A seismic capability can be added to the CPT to measure the shear wave velocity (to calculate the shear modulus of the soil), in which sensors are located behind the friction sleeve of the penetrometer. The seismic source is installed on the seafloor frame and the data acquisition is connected through a umbilical line. The depth of shear wave penetration should reach down to 80-m below the seafloor under favorable conditions. The electrical conductivity of the soil can also be measured during the CPT using two electrodes placed on the penetrometer behind the friction sleeve.

4.6.3 T-bar Testing The T-bar tool is a recently developed device similar to the cone penetrometer that provides a more accurate means of estimating undrained shear strength in soft clays. Although the sleeve friction is not obtained, the resistance measured is subject to less uncertainty, and has the ability to measure the remolded strength (wheeldrive version only) of the soil as well. The surface of the T-bar is slightly textured through sandblasting, while the cone penetrometer tip surface is machined smooth. The T-bar is mounted on the cone penetrometer base to utilize the cone tip load cell to measure the T-bar resistance. There is no current standard operating procedure for conducting the T-bar test. However, the Norwegian Geotechnical Institute and the Center for Offshore Foundation Systems in Australia recommend that the T-bar resistance be measured during both penetration and extraction at a rate similar to the CPT. They also recommend that

37

for tests deeper than 5-m, the rod inclination be measured, and that for cyclic testing the vertical displacement limits are to be set at 0.5-m for at least 6 cycles (Norsok 2004).

4.6.4 Field Vane Testing The field vane shear test is conducted with a cross-shaped probe that is pushed into virgin soil at least 0.5-m ahead of the sampling device and rotated to measure the peak and residual shear strengths of the soil. The speed and number of rotations, direction of rotation, and the size of the vane (Table 17) can be adjusted to suit the soil type being tested and the type of strength desired (i.e. drained or undrained). The measured undrained shear strength is considered to be more accurate than the CPT measurements in soft clay soils down to penetration depths of 3-m below the seafloor. The main drawback is the slow operation speed and its limited applicability to other soil types.

4.6.5 Other In Situ Tests Other in situ tests that may be used in certain circumstances include pressuremeter testing (stress-strain characteristics), hydraulic fracture testing, dilatometer testing, permeability testing, in situ density testing, pile-model testing, seismo-acoustic techniques, plate loading tests, and pipeline trenching evaluation. The primary advantage to using some the model tests is the ability to directly model the predicted foundation behavior. Disadvantages include the size and cost of the larger-scale testing and their narrow scope of target parameters.

4.7

Characteristic Soil Properties

In situ testing and sampling are performed to gather the necessary geotechnical data for the soil deposits of interest. The exact information needed depends primarily on the type of foundation, but numerous parameters are needed that are independent of the type structure to be installed. The data should include, but are not limited to (Norsok 2004, OSIC 2004): 1) Summary of soil conditions: soil classification, description, and stratigraphy: total unit weight, solids unit weight, water content, void ratio, porosity, relative densities, liquid and plastic limits, and grain size distributions

38

2) Basic soil parameters: effective in situ overburden stress, σ’vo; in situ pore pressure, uo; preconsolidation stress, σ’p; overconsolidation ratio, OCR; coefficient of lateral earth pressure at rest, Ko; relative density of sand layers, Dr; 3) Deformation properties: undrained shear modulus, G; drained Young’s modulus, E; Poisson’s ratio, ν; constrained modulus, M; horizontal and vertical coefficients of consolidation, ch and cv; coefficient of permeability, k; creep parameters; cyclic loading parameters for settlement calculations; small strain shear modulus, Gmax; damping ratio, ξ 4) Shear strength parameters: friction angles for granular material, index, undisturbed, and remolded undrained shear strengths, su, and sensitivity for finegrained material; pore pressure parameters; parameters to describe excess pore pressure development and shear strength degradation due to cyclic loading; for drained clay analyses, cohesion and friction angles are needed 5) Parameters for specific applications: contour diagrams for cyclic effects; base contact stress parameters; skirt penetration resistance parameters; for piled and gravity base structures, mud mat stability and settlement parameters; for jack-up platforms, parameters for stability settlement and punching failure; for geohazard analysis, slope stability shear strength parameters; for earthquake analysis, dynamic soil parameters Along with the basic parameters, the appropriate types of in situ and laboratory tests to obtain specific parameters vary between soil type (Table 18). Additional soil parameters and their respective appropriate analysis for other offshore applications that may overlap with wind farm investigations in shown in Table 19. When characterizing soil parameters, consideration of the potential variability of the soil properties based on the volume of the tested sample should be addressed. A larger volume results in an averaging effect over the variability from point to point within the soil volume, while a smaller volume needs to account for the full variability of the local soil property. The characteristic value should be a cautious estimate of the property where the probability of a worse value (i.e. having a negative effect on the capacity of the soil property) is not greater then 5%, depending on the design limit state in question (DNV-OS-J101 2004).

39

4.8

Laboratory Testing

The laboratory testing phase of the offshore site investigation provides the majority of the geotechnical soil data. It is divided into two phases: the offshore testing program and the onshore testing program.

4.8.1 Offshore Testing Program The offshore testing program consists of routine classification tests following extrusion from sampling devices and undrained shear strength index tests for cohesive soils, and is necessary to update the in situ testing and sampling program based on the observed conditions. Wet and dry densities of the soil and the water content are also determined, as well as a determination of the sample disturbance. Select sections of the sample are then sealed with wax and stored in a humidity and temperature controlled environment for further onshore laboratory testing.

4.8.2 Onshore Testing Program The extent of the onshore laboratory testing phase is tied to the in situ test results and soil classification. Therefore, it is dependent on the degree of homogeneity of the soil deposit, possible stratification, the presence of critical and weak soil layers, and the design strategy. It should account for effects from thin problematic layers or pockets whose location relative to a foundation is unfavorable. Much of this information is based on the geophysical survey and the in situ testing results. The lab program should also account for actual stress conditions in the soil, including effects of cyclic loading caused by environmental loads. The duration of the laboratory testing phase can vary depending on the aforementioned considerations, however typical durations range from 3 weeks to 3 months (Randolph and Kenkhuis 2001). After this phase is complete, the results should be integrated with the in situ testing results to provide the design team with a comprehensive geotechnical design basis to begin analyses of foundation options. After a foundation design is finalized and construction begins, verification of the design basis should be conducted through monitoring foundation installation activities.

This may

include, but is not limited to, 1) blowcounts during pile driving to verify penetration resistance, 2) suction pressure measurements during caisson penetration to verify uplift capacity, 3) primary and secondary settlement measurements during and after gravity base foundation installation,

40

and 4) installing geotechnical instrumentation for long term verification of performance (e.g. storm event deformation, cyclic load deformation, differential settlement).

4.8.3 Soil Classification The soil description and classification should be based on an established classification system (e.g. Unified Soil Classification System) and should include the following characteristics: soil type, secondary or minor soil components important to the soil properties, clay strength or relative density and grading of sand, structure and texture of soil, color and odor of soil, photographs of each discernible layer, and any special characteristics such as the presence of cementation or calcareous ooze. The water content should be obtained for any samples that would subsequently be tested further. It should be obtained both prior to and following the advanced testing, and should comprise a series of continuous measurements to construct a water content profile of the subsurface. The liquid limit, wL, and plastic limit, wP, should be obtained based on standardized procedures for Atterberg limits.

The method should indicate if the soil was dried prior to testing and

whether coarse material was removed from the soil. Other parameters to include in this phase are the bulk density of soil or soil unit weight, specific gravity of soil, maximum and minimum density (for cohesionless soils), grain size distribution, sand grain angularity, and radiography of the soil to determine the quality of the samples.

4.8.4 Index Testing Index testing is performed on cohesive soil samples to estimate undrained shear strength. The types of index tests include fall cone tests, in which the depth of penetration of a freefalling small conical instrument is measured; pocket penetrometer tests, in which the force to penetrate a steel rod into the sample is measured; Torvane tests and miniature vane tests, in which the torque required to rotate a bladed cylinder embedded in the sample is measured; unconfined compression tests (UCT), in which the vertical force to fail a cylinder of soil is measured; and

41

unconsolidated-undrained (UU) triaxial tests, in which the sample is confined to the in situ stress for 10 minutes before being subject to a vertical force. Also, the remolded strength and sensitivity of cohesive soils should be measured during this phase using one of the aforementioned index tests or by using a ring shear device.

The

sensitivity, St, is the ratio of the undrained shear strength of the undisturbed soil to that of the remolded soil. Artificially low sensitivities can result from excessive sample disturbance and should be considered as such when interpreting the results.

4.8.5 Consolidation Testing Testing for determination of the deformation properties of fine-grained soils is primarily the oedometer test, in which the number of tests should be at least 2 per soil layer (DNV-OS-J101 2004). The results from laboratory and in situ testing must consider the inherent uncertainty and differences between results from the various testing methods. These differences may be from soil disturbance due to sampling, samples not reconstituted to in situ stress history, the presence of fissures, different loading rates between the test and the limit state, simplified laboratory test representation of complex load histories, and/or soil anisotropy effects. The installation effects should also be considered when implementing testing results in design. The selection of the type of consolidation test should be based on the stress history of the soil and the desired loading program including unload-reload cycles. The two primary types of tests are the incremental load test and the continuous load test (e.g. continuous rate of strain (CRS) test). The incremental load test is conducted by doubling the applied load at each increment, allowing the specimen to reach primary consolidation at each load step according to Taylor’s method. The CRS test compresses the specimen at a constant rate of vertical strain ranging from 0.2% to 5% per hour for a single drainage 20-mm specimen. Pore pressure is measured from the bottom of the specimen and drainage should be allowed at the top. The rate of strain should correspond to limiting the pore pressure to less than 10% of the total applied vertical load. Horizontal stress should also be measured during testing to obtain the coefficient of horizontal earth stress at rest, Ko.

42

Soil permeability should be measured using constant head permeability tests under constant stress, where the pore pressure should always be increased during testing, again limited to less than 10% of the total applied vertical load. Testing should continue until steady state conditions are attained. The coefficient of consolidation, cv, can be calculated from the coefficient of permeability, k, and the constrained modulus, M, of the stress-strain curve, according to the following:

cv =

Mk

γw

Equation 4-1

where:

γw

= unit weight of water

Sample quality evaluation can be performed to assess the confidence in the results of the laboratory testing. Using the initial void ratio, ei (ratio of pore space volume to soil particle volume), and the axial strain at the effective in situ overburden stress, a qualitative indication can be estimated using charts similar to those developed by Lunne et. al. (1998) in Table 20.

4.8.6 Strength Testing There are three primary test types used for determining strength properties. Static direct simple shear (DSS) and triaxial (TX) tests are used for sands and normally–consolidated clays, while cyclic DSS and TX tests are used for soils sensitive to cyclic loading (e.g. loose sands and overconsolidated clays). Ring shear tests are used for determining large displacement soil behavior along an interface between the soil and structural element. For one-way cyclic loading due to combined wind and wave loading, the cyclic strength may be higher for a normallyconsolidated (i.e. no stress history) clay than the static strength due to rate effects. Therefore, use of the conservative static strength tests for design can reduce testing time. For peat soils, ring shear tests with a low-humidified (dry) sample should be used. The typical number of tests conducted should be 4 to 5 per soil layer (DNV-OS-J101 2004). Extra soil from sampling tubes is generally used for classification purposes.

43

TX tests provide information on the stress-strain behavior of the soil. The various ways to perform TX test first consolidate the specimen and then either shear the specimen in drained or undrained conditions, using either an axial compression, axial extension, or constant volume condition. The type of test chosen is dependent on the predicted soil behavior at the location of interest (i.e. undrained compression below a footing in clay, or drained extension next to a bulkhead with a sand backfill). For the undrained conditions, pore pressure measurements are obtained through testing. Undisturbed clay samples are prepared using cutting tools while maintaining the in situ water content as well as possible, or remolded clay samples are prepared using standard remolding techniques. Silt and sand samples are reconstituted into the triaxial mold by hand tamping or wet/dry pluviating to the target relative density. The specimen is enclosed within a rubber membrane for both types of tests, and saturated with water (for sand specimens) before it is subject to a consolidation pressure of various methods (i.e. isotropic, anisotropic, Ko, SHANSEP, back pressure/saturation). After reaching primary consolidation, the specimen is sheared at a rate ranging from 1-µm/min to 1-mm/min. The TX test can be performed under a static load condition, or under cyclic loading where load control limits are imposed with a frequency dependent on the problem of interest. DSS tests are performed using a specimen that is consolidated to Ko conditions and subjected to a horizontal shearing load under drained or undrained conditions. The test provides stress-strain behavior and shear strength characteristics for the soil along a forced plane of failure. The specimen can be extruded and oriented in the device such that representative soil behavior can be obtained along any angle relative to the horizontal plane.

The specimen preparation and

consolidation phases of the test are similar to the TX test, however, only one-dimensional volumetric change in the vertical direction is allowed. Static or cyclic loading can be conducted, with displacement rates ranging from 1-µm/min to 2-mm/min, and cyclic load frequencies ranging from 0.01-Hz to 0.5-Hz.

4.8.7 Other Laboratory Testing Resonant column tests provide dynamic soil behavior characteristics such as shear modulus and soil damping. During cyclic loading, the soil can reduce or increase its shear modulus through volumetric contraction or dilation, respectively.

44

While this test is conducted at higher

frequencies than the cyclic DSS test and is used primarily for earthquake load investigations, it can provide useful information about the soil response not obtained through DSS testing alone. Piezoceramic bender element tests measure shear wave velocity through a soil specimen. This parameter can be used to calculate the maximum shear modulus, Gmax, and can be conducted at any phase of a TX, DSS, or consolidation test.

The test is performed by inserting small

piezoceramic probes at each end of a soil sample through which an electric charge is applied, resulting in a measurement of the time for a shear wave to travel over a fixed distance. The wave measurements are obtained through use of an oscilloscope. Thixotropy tests measure the increase in undrained shear strength of a remolded clay specimen over time. The clay is remolded under the in situ water content and tested using the fall cone apparatus to measure the shear strength at increasing time intervals. The test is useful for predicting the long-term internal shear strength characteristics of clay subjected to high disturbance during foundation installation. Heat conductivity tests provide information on the thermal characteristics of an undisturbed or remolded soil sample. Using electric current applied to a long needle inserted into the specimen, a time-temperature curve is generated to analyze the thermal conductivity, λ. Other tests that can be performed on samples to obtain pertinent data include plane strain tests, high pressure and variable temperature TX and consolidation tests, laboratory model tests, geological and geochemical tests to determine geological origin and sedimentary history, mineralogical or fossil analysis, amino acid or isotope analysis, gas analysis, carbon dating, organic content assessment, and assessment of corrosion potential.

For corrosion risk

assessment, offshore techniques include sulphide analysis, resistivity measurements, pH measurements, coupled with color and odor description and sulphur and organic content analysis. Onshore techniques include a more thorough sulphur-based compound analysis than the offshore technique.

45

4.9

Effects of Cyclic Loading

The effects of cyclic loading on the soil stiffness should be considered for soils prone to cyclic strength degradation (e.g. loose sands, calcareous soils, overconsolidated clays). Cyclic loading is caused by repeated wind and wave forces, and should be investigated for single storms or several successive storms, as relevant. Cyclic loading may lead to a gradual increase in excess pore pressure and a decrease in effective stress in undrained conditions (fine-grained soils), or the degradation of shear strength due to permanent volumetric strains in drained conditions (coarse-grained soils). For low permeability soil deposits, storm events causing excess pore pressure can be superimposed on another storm event, potentially doubling the excess pore pressures and subsequent strength reduction. However, such events can result in overconsolidation of the soil over time and hence a higher resistance to the excess pore pressure generation (Poulos 1988). In SLS design, the shear modulus G should be corrected when dynamic motions, settlements, and permanent horizontal displacements are expected (DNV-OSJ101 2004).

4.10 Seafloor Stability 4.10.1 Introduction There are generally three mechanisms in the continental shelf region that can produce seafloor instability and movement: gravity forces, hydraulic forces, and earthquake forces. Gravity forces primarily provide a mechanism for slope failure classified as either basic instability or creep phenomena. Basic instability is characterized by rapid, large displacement movements ranging in time from minutes to days. Creep phenomena is characterized by widely variable strain rates dependent on stress levels and environmental conditions, and occur over time periods ranging from hours to thousands of years. Hydraulic forces result from currents that cause erosion and scour in near surface deposits, tides that affect soft sediment stability in the intertidal zone, surface waves which induce pressure anomalies along the seafloor, and internal waves that result from density stratification of sea water from temperature and salinity variations (Poulos 1988). Earthquake forces are caused by a sudden release of strain energy from a fault, which is a function of source mechanism and regional seismicity.

46

4.10.2 Slope Stability Slope stability can be analyzed using limit equilibrium, continuum mechanics solutions, or finite element analyses, all of which usually employ plane strain conditions for simplification purposes. The slope stability should be calculated for natural slopes, slopes developed during and after installation of the structure, future anticipated changes of the existing slopes, the effect of continuous mudflows, and wave-induced and earthquake-induced soil movements (where applicable). Three soil conditions which should be investigated include undrained, fully drained, or partially drained conditions. Soils most susceptible to slope failure are soft, normally consolidated clays or loose sands. The safety against slope failure for ultimate limit state design uses the material factor γm equal to 1.2 for effective stress analysis and γm equal to 1.3 for total stress analysis. For accidental limit state design, γm is equal to 1.0 (DNV-OS-J101 2004).

4.10.3 Hydraulic Stability Soils susceptible to erosion or softening need risk assessment for the reduction of bearing capacity due to hydraulic gradients and seepage forces, formation of piping channels within internal soil erosion, and surface erosion in local areas beneath the foundation due to hydraulic pressure variations from environmental loads. Measures should be taken to prevent, control, or monitor erosion as relevant.

4.10.4 Earthquake Stability An earthquake releases energy in the form of stress waves, which are categorized into compression (P) waves, shear (S) waves, Rayleigh waves, and Love waves. The effects of these waves on onshore structures are similar to their offshore effects, with additional consideration given to the following: 1) the earthquake force will be acting in possible combinations with wave and wind forces, 2) the presence of water changes the ground motion characteristics, 3) flow slides occurring from earthquakes can travel further in offshore environments (e.g. up to several kilometers (Poulos 1988)), and 4) P-waves can cause increases in wave loading on an offshore structure in the form of a tsunami. Evaluation of earthquake stability should include limit equilibrium of the slope stability, in which earthquake loads are represented by equivalent vertical and horizontal loads, and liquefaction or cyclic mobility potential analysis, which

47

consists of comparison between the equivalent cyclic shear stresses incurred by the loading and the equivalent cyclic shear stress resistance of the soil. The assessment should be based on previous records of earthquake activity in terms of frequency of occurrence and magnitude if available. If no seismic information is available, a detailed investigation of geological history and regional seismic events should be conducted. If the region is seismically active, the following parameters should be determined: 1) Location and alignment of faults 2) Epicentral and focal distances 3) Source mechanism of energy release 4) Source-to-site attenuation characteristics 5) Local soil characteristics including stiffness and damping 6) Design earthquake or maximum credible earthquake 7) Tsunami (seismic sea wave) assessment (based on water depth) The seismic analysis can be conducted using pseudo-response spectra, specifically the response spectral acceleration, SA, the response spectral velocity, SV, and the response spectral displacement, SD. For lightly damped structures, SA ≈ ω2SD, and SV ≈ ωSD, where ω is the natural circular frequency of the structure. The earthquake-induced ground motions are then analyzed in two horizontal directions and one vertical direction. For dynamic analysis, the wind turbine can be represented by a concentrated mass on the top of a vertical rod, typically equal to the mass of the rotor-nacelle assembly and ¼ of the tower mass (DNV-OS-J101 2004). Sitespecific dynamic properties of soils subject to cyclic earthquake loads should also be evaluated. The ground motion characteristics should be evaluated at the structural base for free field and local soil conditions.

4.11 Scour 4.11.1 Scour Mechanism Scour is the movement of seafloor sediments due to currents or waves, occurring via natural geological processes or an interruption in the natural flow regime from structural elements. The 48

main driver of scour is the oncoming water flow in the form of a horseshoe-shaped vortex in front of the foundation structure, carrying sediment away from the foundation. The lower velocity flow behind the structure may allow sediment to accumulate, creating a zone of sediment higher than the original seafloor elevation.

4.11.2 Types of Scour There are three basic types of scour: local scour, global scour, and overall seabed movement (Figure 10). Local scour is characterized by conical scour pits around piles or pile groups similar to those observed in flume models (Gravesen 2004). Global, or dishpan, scour is characterized by large shallow basins around the foundation structure, resulting from overall structure effects, multiple structure interaction, or wave-soil-structure interaction. Overall seabed movement of sand waves, ridges, or shoals occurs naturally in the absence of a structure and can result in lowering or rising of the seafloor elevation. Most scour observed is usually a combination of the above characterizations. Similar to slope stability, the soils most susceptible to scour are loose sands and soft clays. Scour is most prevalent in areas with strong tidal streams or wave breaking zones. The affects of scour are specific to the foundation type around which it occurs. Piled foundations suffer from overburden loss and lateral resistance loss, while gravity base foundations may lose soil from beneath the foundation, resulting in bearing capacity reduction, settlement, or overstressing of the foundation elements (Watson 2000).

4.11.3 Prevention of Scour The extent of scour can be determined from previous records of sites with similar seafloor characteristics, from model tests, or from calculations based on prototype tests. There are two options for addressing scour. One option is to allow for scour in the foundation design, assuming all materials prone to scour are removed to a calculated depth below the original seafloor. In the absence of sufficient data, an assumed depth of scour for piled foundations can range from 1 to 2.5 times the pile diameter, which should be verified by regular monitoring (Gravesen 2004). The other option is to place scour protection around the foundation, typically consisting of a thick (e.g. 0.5-m), crushed rock protection bed or asphalt/concrete mattresses. Placement should

49

occur as soon after construction as possible, monitoring the seabed frequently and replacing lost materials as needed. The most inexpensive option is typically the crushed rock bed (van der Tempel 2002).

4.11.4 Designing for Scour Scour risk should be considered in design unless evidence can be demonstrated that scour will not occur for the expected range of water particle velocities. The design for scour must account for steady current, waves, and current or waves in combination as relevant, and design wind and wave loading with a recurrence period of 100 years, with inspection every 5 years (DNV-OSJ101 2004). The scour protection should provide both internal and external stability (i.e. transport of particles in underlying natural soil and excessive surface erosion, respectively). Both global scour and local scour should be considered in design, each having implications for construction of the p-y (i.e. lateral resistance versus depth) and t-z (i.e. shaft friction versus depth) curves. For global scour, effects of soil removal over a large area are estimated by constructing p-y and t-z curves on the basis of a lowered seafloor level equal to depth of general scour. For local scour, construction of the p-y and t-z curves are based on the original seafloor level, assuming no soil resistance within the depth of local scour on the curves (DNV-OS-J101 2004). Scour will lead to complete loss of lateral and axial resistance down to the depth of scour below the seafloor, as well as a reduction in the effective overburden stress, which reduces the resistance of the lower soil layers. The depth of overburden reduction is typically 6 times the pile diameter (van der Tempel 2002).

As the length of the pile will effectively lengthen, pile

flexibility increases resulting in decreases in the natural frequency of the pile. This effect is the driving factor in fatigue loading on the structure. An example flowchart for determining the occurrence of scour is shown in Figure 11. On the left-hand side of the flowchart, a maximum shear stress on the seafloor is obtained from the wave and current particle velocities. On the right-hand side, the seafloor allowable shear stress is determined. Accounting for amplification effects of the structure, the scour potential is

50

obtained from comparison of the stress loading to the stress resistance. Once the potential for scour is established, the depth of the scour can be calculated from numerous empirical equations (e.g. DNV-OS-J101 2004, van der Tempel 2002).

51

5.0 Other Environmental Conditions 5.1

Introduction

Other offshore environmental parameters impacting wind farm design, such as meteorological and oceanographic measurements, should be characterized for proposed wind farm sites for a period of approximately one year prior to development. There are several methods of assessing these parameters as outlined in Table 12. The oceanographic parameters to evaluate include wind speeds, wind directions, strength of wind gusts, and turbulence. Meteorological parameters of importance include the wave heights, tides, currents, and water turbidity.

Other site

conditions to evaluate include human factors such as fishing areas, shipping route proximity, and other restricted areas, and environmental considerations such as marine growth potential, migratory bird routes, or fauna reproduction areas.

5.2

Meteorological Parameters 5.2.1 Wind Loading

The wind force acting on an offshore wind turbine structure contributes approximately 25% of the horizontal load and 75% of the overturning moment to the foundation (Byrne and Houlsby 2003). Wind loading on the structure is represented by the 10-min mean wind speed, U10, and the standard deviation of the wind speed, σU. The short-term 10-minute stationary wind climate is represented by the power spectral density function of the wind speed process, S(ω), which describes how the energy of the wind speed is distributed between various frequencies. Local wind statistics are used as the basis for long and short-term wind condition representation where available, while empirical data from other sources must cover sufficient time period (e.g. at least 1 year).

The same averaging period should be used for the statistical basis of long-term

distributions of U10 and σU (e.g. 10 minutes). Otherwise, gust factors can be used. Hub height wind speeds should be used as the reference value; otherwise a wind speed profile above the still water level may be constructed. The characteristic values of the wind speed should be determined accounting for inherent uncertainties and wake effects due to other turbines upstream in a wind farm (DNV-OS-J101 2004).

52

5.2.2 Wind Modeling Existing wind farm data shows that offshore wind energy output is 20 to 30% larger than traditional wind modeling methods estimate (Krohn 2002). The first step in constructing a wind model is computing the spectral wind density, represented by the Kaimal spectrum: SU ( f ) = σ U

4 2

Lk U 10

 fL 1 + 6 k U 10 

  

5/3

Eq. 5-1

where:

f

= frequency

Lk

= integral scale parameter equal to 5.67z for z < 60m and 340.2m for z ≥ 60m where: z = height above sea water level

Full characterization of the wind speed probability distribution should be obtained for 0 to 200-m above sea level, referenced to a height of 10-m (Sahin 2004). This requires knowledge of upstream roughness length (zo), which is a function of the distance from shore and the height above sea level. Surface roughness is based on the rate of growth of internal boundary layers from surface discontinuities and assumptions of the shape of the wind profile. The wind profile can be described by the power law within layers, where growth of an internal boundary layer height is based on a function of zo and stability (Sahin 2004).

5.2.3 Turbulence Turbulence is created in a wind farm due to wake effects from other upstream turbines. Wake effects should be accounted for when turbines are spaced less than 20 rotor diameters apart (Frandsen and Thogersen 1999). This is done through a reduced U10 value relative to the ambient wind climate inside the wind farm perimeter, or the standard deviation σU of the mean wind speed.

Depending on the wind farm configuration, there may be multiple turbines

contributing to wake effects upon an adjacent turbine. Wake effects usually dominate fatigue loads on turbines, and fade out more slowly and over longer distances offshore than on land. The turbulence intensity, I, is defined as the standard deviation of wind speed divided by the 10 minute mean wind speed:

53

I=

σU U 10

Eq. 5-2

Wake effects are the main factor in decreasing the fatigue lifetime of the mechanical components of an offshore wind turbine due to the constant fluctuations in wind speed and direction. Offshore turbulence intensity is lower than it is onshore, and increases with the height above sea level. Therefore, offshore turbines should optimize design between lower hub heights and the associated smaller rotor length required. Based on the fatigue loading, the expected lifetime for an offshore wind turbine is estimated to be between 25 and 30 years compared to an onshore wind turbine lifetime of 15 to 20 years (Sahin 2004). A deterministic fatigue analysis should be applied to structures with natural periods different from the significant energy wave periods which also exhibit little dynamic amplification. Dynamically-sensitive structures should evaluate fatigue using a spectral analysis fatigue method, which provides a statistical assessment of the wave spectrum-based (e.g. PiersonMoskowitz or JONSWAP) structural response. For conceptual design studies, methods that combine time domain and spectral analyses into equivalent fatigue loads can be used for the acceptable level of accuracy necessary at that stage. The time domain simulation is conducted for a calm sea state to approximate aerodynamic fatigue loading, and a linear spectral analysis is used to approximate hydrodynamic fatigue loading. These two components are superimposed rather than added together due to the high non-linearity of the stress-induced fatigue damage (Watson 2000).

5.3

Oceanographic Parameters 5.3.1 Wave Loading

The wave loads acting on an offshore wind turbine structure affect both the sub-structure and the foundation. The wave loads can be represented by two main parameters, the significant wave height, Hs, and the spectral peak wave period, Tp. The significant wave height is a measure of the intensity of the wave climate accounting for wave height variability. It is traditionally defined as the mean height of the 1/3 highest wave, H1/3, but can also be defined as four times the standard deviation of the sea elevation process (i.e. four times the area under the wave spectrum,

54

Hm0) (Watson 2000, DNV-OS-J101 2004). The wave spectrum describes the frequency content of a sea state, typically based on a Pierson-Moskowitz spectrum for a well-developed sea state or a JONSWAP spectrum for a limited fetch and duration sea state. The spectral peak wave period, Tp, is related to the mean zero-crossing period, Tz, of the sea elevation process and is assumed to be constant (approximately 10-sec) over a short-term 3 to 6-hour sea state (Byrne and Houlsby 2003). The short-term 3 to 6-hour sea state is represented by a wave spectrum, or the power spectral density function of the sea elevation process, S(ω), which is a function of Hs and Tp and describes how the energy of the sea elevation is distributed between frequencies. Hydrodynamic loads should be determined by theoretical predictions and model or full scale measurements. The degree to which each is conducted is dependent on the uncertainty of the predictions. Wave statistics can be used to estimate these parameters given that they cover a sufficient period of time. If using data from a nearby area, consideration of the differences in water depth and seafloor topography should be taken into account. The Hs and Tp distributions should be based on the same averaging period for the waves as for the load determination; otherwise, use of appropriate adjustment factors is necessary. Simultaneous values of Hs and U10 should be obtained since waves are wind-generated and should be correlated in design. The directionality of each should be recorded as well.

5.3.2 Wave Modeling Wave load predictions should account for the size, shape, and type of the proposed structure. For piled foundations, Morison’s equation can be used to calculate the wave loads, consisting of a drag force component and an inertial force component (Watson 2000). The drag force is proportional to the overall combined water particle velocity, V, according to the following:

dFd = 12 C d ρDV 2 dL where:

dFd

= drag force component

ρ

= water mass density

D

= tower diameter

dL

= elemental length of tower

Cd

= drag coefficient 55

Eq. 5-3

The inertial force is proportional to the acceleration of the water particles due to Froude-Krylov forces and added mass forces according to the following:

dFi = ρAdL(a + C a a r )

Eq. 5-4

where:

dFi

= inertial force component

A

= cross-sectional area of tower

a

= water particle acceleration

Ca

= added mass coefficient

ar

= relative acceleration of water particles/tower

Diffraction effects may alter the wave pattern and loading for support towers that are large compared to the wavelength, typically significant when the tower diameter is greater than 20% of the wavelength (Watson 2000). At this point Morison’s equation is no longer valid, and the inertial force will dominate. For gravity base foundations, radiation analysis or diffraction analysis should determine wave loads (DNV-OS-J101 2004). Where steep wave crests are prevalent, the support tower and sub-structure may be subjected to highly localized impact loads, or slap forces, which are related to the rate of change of added mass. For monopile foundations, this can be important for structural design as the slap forces are not distributed as well as would be in a multiple-leg foundation. If wave steepness is assumed to be constant and the wave height is scaled to the water depth, then the drag and slap forces are proportional to the product of the squared water depth times the diameter, and the inertial forces are proportional to the water depth times the squared diameter (Watson 2000). Viscous and potential flow effects should also be considered. Waves that break against the structure are more prevalent in shallower waters where wind turbines will be located than typical water depths for larger offshore platforms, and therefore should be considered separately from non-breaking wave loads. There are three classifications to consider: surging, plunging, and spilling waves. These waves can result in large amplification factors depending on the wave frequency relative to the natural frequency of the structure.

56

A generic distribution, or scattergram, can represent the long-term probability distributions of the significant wave height, Hs, and the spectral peak wave period, Tp, consisting of a Weibull distribution for Hs and a log-normal distribution of Tp. In deeper waters, the short term probability distribution function of an arbitrary wave height H is assumed to follow a Rayleigh distribution in terms of Hs. The maximum wave height in a 3-hour sea state can be estimated to be equal to 1.89(Hs). In shallow water, the wave height is limited by depth, and the maximum height can be assumed to be equal to 78% of the water depth (DNVOS-J101 2004). The long-term probability distribution of an arbitrary wave height H is found by integration over all significant wave heights, which is used to calculate the distribution of the annual maximum wave height Hmax. Wave crest height Hc can be assumed to be equal to 0.65(H) (DNV-OS-J101 2004). The wavelength, λ, is given by the following equation:

λ=

g 2 2πd T λ 2π

Eq. 5-5

where: T

= period

g

= acceleration of gravity

d

= water depth

Analytical and numerical wave theories can represent the wave kinematics according to their ranges of validity. The linear wave theory that represents waves with a sine function is valid for d/λ ≥ 0.3. Stokes’ wave theory for high waves and the stream function theory are valid for 0.1 ≤ d/λ ≤ 0.3, and the solitary wave theory for very shallow water is valid for d/λ ≤ 0.1. The Airy theory is valid for all ratios of water depth to wavelength (DNV-OS-J101 2004). Recent studies indicate that traditionally-modeled wave loads (i.e. via Wheeler stretching of the pressure and velocity profile) are not adequate for extreme design cases, specifically in the range of 0.25 < Hs/h < 0.5, where h is equal to the water depth. A new Boussinesq model has been proposed that can generate a time series of the wave kinematics to define the input for the foundation, leading to calculation of the wave forces through integration of the inertial and drag

57

forces over the length structure. The results are scaled to the water depth through Froude’s scaling law, keeping the ratio of significant wave height to water depth constant. Because of this, both operational and fatigue extreme loads can be accounted for through accurate characterization of combined wind and wave forces (Gravesen and Bro 2003).

5.3.3 Current Loading The current load consists of two to four components, depending on water depth and geographical location: wind-generated current, tide-generated current, breaking waves (for shallow water), and ocean circulation (e.g. Gulf Stream currents). The waves and currents are assumed to be statistically independent (Watson 2000). The wind and tide-generated currents can be represented by current velocities, which vary with water depth (DNV-OS-J101 2004).

5.3.4 Current Modeling The current velocity can be estimated based on water depth according to the following: v( z ) = vtide ( z ) + v wind ( z )

Eq. 5-6

1

 h + z 7 vtide ( z ) = vtide 0    h 

Eq. 5-7

h +z  v wind (z ) = v wind 0  0  h0 

Eq. 5-8

where: v(z)

= total current velocity at level z

vtide(z) = tidal current velocity at level z vwind(z) = wind-generated current velocity at level z z

= distance from still water level, positive upwards

vtide0

= tidal current at still water level

vwind0 = wind-generated current at still water level h

= water depth from still water level (positive value)

h0

= reference depth for wind-generated current

Estimations should account for the variation in current profile due to the water depth variation from vertical wave action by stretching or compressing the profile, keeping the surface current

58

constant regardless of the wave action-induced instantaneous sea level (e.g. DNV-OS-J101 2004, Watson 2000).

5.3.5 Ice Loading Ice loading should be considered in areas where ice may develop or drift. Characteristics of ice loading include the geometry and nature of the ice, the concentration and distribution of the ice, type of ice (ridges, floes, or rafts), mechanical properties of the ice, direction and velocity of drifting ice, ice thickness, and the probability of an iceberg encounter. The build up and fracture of the moving ice should also be considered. Other ice loads that may need consideration include loads due to rigid ice coverages, masses of ice frozen to the structure, pack ice and ice wall pressures, thermal ice pressures, falling ice, and icing loads (DNV-OS-J101 2004). Ice build-up on the turbine may be due to sea spray, rainfall, snowfall, or air humidity, which can change the cross-sectional area of the structural elements or change the surface roughness. For floating structures, the uneven distribution of ice or snow accumulation should also be considered. The structure should be designed for both static and dynamic horizontal and vertical ice loads, assuming that horizontal ice loads act concurrent with the wind load direction. The water level that produces the most unfavorable reaction from the structure when calculating ice loads should be used for design (DNV-OS-J101 2004).

5.3.6 Ice Modeling The frost index, K, is used to determine the ice loading, defined as the sum of the daily mean temperature over all days whose mean temperature is below 0° C in one calendar year (DNVOS-J101 2004). It varies from year to year and can therefore be represented by a Weibull probability distribution function and a recurrence period. The ice compressive strength, bending strength, and thickness can be expressed as functions of the frost index or as probability distribution functions. The following ice parameter values may be used regardless of location (DNV-OS-J101 2004): 900 kg/m3

Density

59

Unit weight

8.84 kN/m3

Elastic Modulus

2 GPa

Poisson’s ratio

0.33

Ice-ice frictional coefficient

0.1

Ice-concrete dynamic frictional coefficient 0.2 Ice-steel dynamic frictional coefficient

0.1

The characteristic local ice pressure from moving ice can be determined as follows: 2

rlocal ,C = ru ,C

e 1+ 5 c Alocal

Eq. 5-9

where: ru,C

= compressive ice strength (≈ 1 to 2 MPa)

ec

= ice thickness

Alocal

= area of applied local ice pressure

During ice breakup, static and dynamic interactions between the ice and the structure occur, causing the breakup frequency of the ice and the natural frequency of the structure to become tuned together. Therefore, the structure should also be designed to withstand these load effects.

5.4

Combined Wind and Wave Loading 5.4.1 Horizontal to Moment Load Ratio

Load effects can combine to result in assumed intensities of multiple parameters acting during an environmental state (i.e. when an intensity parameters acts at an assumed constant value over a 10-minute to 1-hour period of time). Combined horizontal loads are generally equal to 1 to 5% of the resultant moment created from wind, wave, and current loading due to the typical height of the rotor-nacelle assembly (Byrne and Houlsby 2003).

However, the ratio of moment to

horizontal load fluctuates rapidly with time, dependent on water depth, sea state, and wind conditions. This loading scenario is atypical to that of other offshore structures due to their larger size, where the load ratios remain relatively constant regardless of environmental conditions. Because the wind and wave loading may not be coincident, the horizontal and moment loads may also not be coincident. However, the most unfavorable structural response is when wind and wave loads do act coincidently.

60

5.4.2 Combination Methods Two methods to combine wind and wave loads are the linear combination method and the combination by simulation. The linear method simply combines the calculated wind load effect and the calculated wave load effect by linear superposition. It works well as a preliminary combined load evaluation or when dynamic effects are demonstrated to be negligible (e.g. in shallow water) (DNV-OS-J101 2004). The equivalent damping resulting from the combined loads should be assessed in terms of turbine position. The simulation method is based on structural analysis in the time domain for the simultaneously-applied simulated time series of the wind and wave loads. The dynamic interaction between wind and wave loads also may be characterized by an amplification factor applied to the wind load that is proportional to the ratio of the dynamic wave load to the associated wind load and the relative wave load deformation to the 1st mode of the tower/foundation deformation (Gravesen and Bro 2003).

5.4.3 Design Considerations The load effects can be evaluated based on the ultimate and accidental limit states in terms of multiple combinations of environmental load type and its associated recurrence period. Each load combination should be evaluated for the operational and the non-operational states of the turbine, using the largest load effect for design. The characteristic load effect distribution for the fatigue limit state should be considered in terms of the expected load effect distribution over the design lifetime, which is a distribution of stress fluctuations from environmental loads that include all phases of turbine operation (e.g. start up, shutdown, idling, transport, etc) (DNV-OSJ101 2004).

5.5

Environmental Corrosion 5.5.1 Introduction

Corrosion for offshore structures can be divided into two categories: corrosion due to structural degradation, and corrosion in the form of marine growth.

There are three zones under

consideration for corrosion control; the lower splash zone, the buried zone, and the upper submerged zone. The lower splash zone is the part of the sub-structure and tower intermittently exposed to air and sea water, the lower limit of which is defined as one-third of the 100-year wave height (DNV-OS-J101 2004). The buried zone is the part of the support structure below 61

the seafloor (i.e. foundation), and the upper submerged zone is the part of structure in between the lower splash and the buried zones.

5.5.2 Degradation Corrosion Steel structures require corrosion control to ensure structural integrity throughout their design lifetime. Degradation corrosion in the lower splash zone is estimated to be approximately 0.5 mm per year, increasing with the age of the structure. It plays a significant role in fatigue analysis and approximately 3 to 5 mm of reduced steel thickness should be included in the analysis (DNV-OS-J101 2004). Corrosion protection systems in the splash zone should be based on the corrosion allowance with a surface protection layer such as glass fiber-reinforced epoxy coating. Steel structures in the submerged zone should be protected by a cathodic coating. Near the seafloor, piles should include a corrosion allowance of 3-mm in the fatigue life endurance limit, reduced by a factor of 2 in the buried zone to account for the anaerobic environment that reduces the cathodic corrosive ability. Further details for cathodic protection and protective coatings are given in DNV-RP-B401, and DNV-OS-C101, respectively (DNV-OS-J101 2004).

5.5.3 Marine Growth Corrosion in the upper submerged zone and lower part of splash zone consists of largely of marine growth, which can enhance or retard degradation corrosion attack. It can retard the corrosion process through corrosive metabolites, called Microbiologically Influenced Corrosion (MIC). MIC can occur in flooded steel members unless measures are taken to ensure water remains sterile via addition of a biocide. It can enhance degradation corrosion by interfering with corrosion control systems, such as coatings or linings or cathodic protection. Corrosion in the buried zone is predominately related to MIC. In undisturbed sediments, MIC is significant in uppermost layer only, but sediment disturbance through drill-cuttings or other effluents can enhance bacterial activity and MIC (Gravesen 2004).

Ice scoring in arctic

environments accelerates corrosion by removing rusting protection layers, corrosion protective coatings, or marine growth.

In deeper waters, oxygen content and sea currents are the

dominating parameters where minimal corrosion usually results (Gravesen 2004).

62

Marine growth can also increase the wave loads on the structure through an increase in the effective diameter of the support structure by 0.1 to 0.2-m, depending on the depth below mean sea level (DNV-OS-J101 2004).

Marine growth should be accounted for by assuming an

additional insulation layer with a density of 2200 kg/m3 surrounds the foundation without stiffness or damping capability (Zaaijer 2002).

5.6

Other Loading Conditions 5.6.1 Transportation and Installation Loading

There are several environmental parameters that may affect the transportation and installation of offshore wind turbines. These parameters include the expected wind speed, wave height and wave crest, water level, current, and ice for the duration of operations. Further details for characterizing these parameters are given in the DNV Rules for Marine Operations (DNV-OSJ101 2004). Other installation loading considerations include the damage sustained by the soil due to pile driving activities. The total fatigue damage in a piled foundation is derived from both the soil damage occurring during the installation phase and the cyclic soil damage occurring during operational phase. A pile drivability analysis should be conducted to verify that the foundation can be installed to the target depth given the local soil conditions, and to determine the effects on soil properties relative to measured values from in situ and laboratory testing results. Drivability analyses can also help determine the type of equipment required for installation. The analysis is conducted using a stress wave propagation dynamic analysis along the entire length of the pile, and should be verified against real data for accuracy. For foundation elements that cannot be inspected for long-term fatigue damage (i.e. below seafloor), a fatigue lifetime of the structure can be assumed to equal three times the economic lifetime of the structure (van der Tempel 2002).

5.6.2 Vessel Collision The loading and design of the wind turbine structure and foundation based on vessel collision should be evaluated for the ULS and the ALS design criteria. For the ULS, consideration should be given to the impact from the stem or stern of a characteristic vessel, while for the ALS, 63

unintended vessel collision by drifting vessels such as crane barges, working vessels, or unauthorized vessels should be considered. The drifting speed should be assessed for each case with a minimum speed of 2 m/s (DNV-OS-J101 2004). The structure and foundation should be designed so that impacts from any of the above cases will not reduce their capacities. This consideration is especially important in waters with high volume of commercial shipping traffic. The assessment should consider the volume of shipping traffic near the wind farm or turbine, identification of shipping lanes in the area, assessment of annual shipping traffic and of the types and sizes of ships in the shipping lanes. An environmental assessment of wind, wave, current, ice, and visibility that may affect collision frequency should also be considered. Calculation of the risk associated with vessel collision is defined as the product of the frequency or probability of a collision scenario (e.g. navigational error, disabled drifting ship, steering failure, etc.) and the consequence of a collision (structural safety, human safety, environmental safety). The total risk should be compared among wind farm locations for siting considerations, and verified against historical data of ship-to-ship collisions and groundings in the area. Risk reduction measures may need to be implemented such as additional structural reinforcement of the foundation and/or the support structure.

5.6.3 Deformation Loading Deformation loads should be considered when differential settlements impose strains on the support structure that may lead to increased fatigue effects or vulnerability to other environmental loads.

This is particularly important for gravity base and multiple-leg

foundations. In the case of gravity base foundations, a slight rotation from the vertical will induce an additional moment load onto the structure. In the case of multiple-leg foundations, the differential settlement may cause structural elements such as trusswork to assume additional stresses.

64

6.0 Foundation Modeling 6.1

Introduction

The dynamic excitation forces acting on an offshore structure consist primarily of waves and current on the tower, sub-structure, and foundation, wind on the rotor-nacelle assembly and tower, and the interaction between the structure and the rotor blades during operation. In design, the natural structural frequencies need to be avoided to minimize the chance of resonance with the excitation frequencies. To model dynamic loading, principles of fluid mechanics and ocean wave theories are applied using deterministic methods to simulate extreme waves in randomly generated sea states in shallow to deep water depths (Byrne and Houlsby 2003). In conjunction with dynamic loads within the turbine structure, load transfer into the seabed soils induce a dynamic response known as soil-structure interaction. Soil-structure interaction (SSI) evaluates the structural load effects on the integrated soil and structural system, using realistic assumptions for stiffness and damping of the soil and the structural members. To design and analyze offshore wind turbine structures, a coupled soilstructure analysis including extreme events and fatigue loading is recommended (Feld and Waegter 2002). Different foundation types have specific models on which to base analysis. For piles, axial and lateral resistance of the soil is typically modeled with non-linear springs, as shown in Figure 12. The non-linear behavior of axial and lateral pile-soil resistance can be accounted for by ensuring the load deflection compatibility between the structure and pile-soil system and accounting for the effects of geometrical and material non-linearities. Under the commonly-used Winkler assumption, empirically-derived nonlinear springs and dashpots each act independently of the springs and pile displacements at other locations (Kellezi and Hansen 2003). Although this assumption simplifies the behavior, it is established as a reliable method to model dynamic SSI. The lateral resistance stress-strain relationship is represented by the non-linear p-y curve, which varies for different soil types, pore pressure conditions, and static versus cyclic loading conditions. At higher displacements, cyclic loading produces lower resistance values than for static loading. The shaft friction stress-strain relationship is represented by the non-linear t-z curve, which uses dimensionless tabulated values of skin friction that also vary for different soil 65

types. The pile tip resistance relationship is represented by the non-linear q-z curve, which does not vary with soil type (Zaaijer 2002). For gravity base foundations, uncoupled springs, dashpots, and inertia are used to represent the soil-structure interaction model, as shown in Figure 13. Elastic half-space models are used to obtain the values of these elements, where radiation damping and spring and dashpot parameters are frequency-dependent. Modeling should account for the strain-dependency of the shear modulus and internal soil damping by using parametric studies including uncertainties in soil properties with upper and lower boundaries on shear moduli and damping ratios of the soil. For wind turbine structures, the dependency on frequency simplifies to allow for estimation of the model parameters through the equations shown in Table 21, where D is the diameter of a circular gravity base foundation, G is the shear modulus, and ρ is the soil density. For earthquake prone areas, frequency dependent reductions of the stiffness may be necessary.

6.2

Types of Models

Various modeling techniques have been used to model offshore foundations. In the following sections, different methods will be compared showing the effect of each on dynamic sensitivity for various foundation types typical of offshore wind turbines. Some of these models have been proven through industry implementation, while others are still in the research and development stage.

6.2.1 Plasticity Models Yield surfaces can be modeled in vertical:moment:horizontal load space (V:M/2R:H), in which the size of the surface is dependent on the vertical plastic displacement of a foundation element (Figure 14). Recent advances in strain-hardening plasticity theories have been proposed that provide more detailed information about displacements and the simulated loads that cause them (Byrne and Houlsby 2003). The foundation response can be expressed in terms of force resultants on the footing and the corresponding displacements, consistent with the time-domain approach used for structures which enables simultaneous modeling between soil behavior and structural analysis. Plasticity 66

models include four components consisting of the yield surface which defines allowable load combinations, a strain-hardening expression that defines how the yield surface expands or contracts, a flow rule that defines the plastic displacements at yield, and a model for the elastic response within the yield surface. The rule of behavior in the model is such that if a load combination is within the yield surface, an elastic response results; otherwise, plastic response results as defined by the flow rule. Results of a model study provided insight into cyclic soil-foundation interaction, such as reduction in the stiffness of a hysteretic response as tension is applied to the foundation. Multiple yield surfaces for the stiffness transition are each accompanied by a plastic potential to describe the plastic flow, which can then be used in constitutive soil models. A disadvantage to using this model is that the specific parameters, sometimes difficult to assess, must be specified for each surface. However, the concept of continuous hyperplasticity, based on thermodynamic principles, replaces plastic strain in conventional plasticity theory with a continuous field of an infinite number of yield surface-specific plastic strain components. It is based on the derivation of plasticity theory for dissipative material of two potentials, the Gibbs (Hemboltz) free energy and the dissipation function. As indicated in Figure 15, this theory closely matches laboratory behavior, and may prove to be a useful method of implementing plasticity models (Byrne and Houlsby 2003).

6.2.2 Finite Element Models Finite element models (FEM) that evaluate foundation behavior are composed of structural elements for the foundation and soil elements for the surrounding seafloor. FEM analysis accounts for initial conditions, nonlinear soil-structure interaction, and nonlinear soil behavior. Boundary conditions determine the constraints for coupling of the structural and soil elements. They are described using differential and integral operators of time and space developed through local schemes that are independent of the frequency of excitations, making them applicable for a time domain transient analysis. For static analysis, boundary elements are assumed to connect to a rigid surrounding, whereas in dynamic analysis, radiation damping at the soil interface needs consideration through multiple degree-of-freedom models. Under typical conditions, a Winkler assumption is preferred. This assumption is unsuitable, however, for flexible gravity base

67

foundations, geometrically-complex foundations, or foundations subjected to non-linear stressstrain cyclic loading (Zaaijer 2002). Many computer programs using FEM have been developed for the offshore industry. Examples include ABAQUS, which uses different models such as the Mohr-Coulomb theory with soil hardening/softening effects or a Drucker-Prager material model with a non-associated flow rule (Kellezi and Hansen 2003), or Ramboll’s multiple FEM programs (e.g. ROSAP, RONJA, ROSOIL) for the wind industry, which combine linear structures with nonlinear foundations. Most of these programs automatically generate the range of environmental and structural loads, in which any standard wave theory can be applied that comprise load situations in all limit states, incorporating both elastic and plastic behavior of the soil in the design (Feld and Waegter 2002).

6.2.3 Other Techniques The effective fixity length technique, which is based on the clamping effect of the soil surrounding piles, can be modeled using a rigid restraint located at an effective depth below the seafloor (Zaaijer 2002). Using approximate values of the effective fixity length (Table 22), preliminary dynamic analyses can be conducted for offshore structures. Due to the lack of bracing through a support frame as seen in typical offshore structures, monopile foundations exhibit different mode shapes of the effective fixity model. A stiffness matrix can be also used to represent the pile-soil stiffness at the seafloor, comprised of forces, moments, displacements, and rotations of the pile head (Zaaijer 2002). The advantages of a stiffness matrix include the consolidation of foundation properties helping facilitate information exchange between the geotechnical and structural engineers for frequency calculations. There are two methods for obtaining the stiffness matrix: a load-displacement analysis with p-y curves, or the Randolph elastic continuum model (Zaaijer 2002), which is based on the inverse of a matrix expression for pile head flexibility derived from dimensional and finite element analyses of piles in an elastic continuum.

The Randolph model is

parameterized in terms of both constant and linearly-increasing soil shear modulus, and therefore works well for sandy soils.

68

6.3

Dynamic Sensitivity

Due to the smaller size and loading nature of an offshore wind turbine structure, the dynamic sensitivity may be more significant than for traditional offshore structures (Watson 2000). The natural frequency of these structures lay within excitation frequencies that differ from traditional offshore structures in that the natural frequencies lay well above the wave excitation frequencies (Zaaijer 2002). The dynamic sensitivity of the structure to foundation stiffness should be determined through natural frequency analysis to obtain the important mode shapes and frequencies. If dynamically-sensitive, dynamic transient analysis using a detailed time history loading should be used to evaluate the structure. For increasing depths of foundation penetration, the natural frequency will achieve a maximum value where the lower portion of the pile does not move, similar to the effective fixity length principle. This value varies for foundation type between monopile, tripod, or lattice tower structures. Based on this analysis, a minimum dynamic sensitivity of the natural frequency can be expected for an offshore wind turbine structure, regardless of foundation penetration. A parametric sensitivity analysis for natural frequency dependency was conducted at the Delft University of Technology based on a 3 MW wind turbine in 21-m of water depth with uniform soil conditions. This analysis assessed the influence of parameter uncertainty, lifetime parameter change, and local parameter variation (e.g. within wind farms) (Zaaijer 2002).

Parameter

variations considered included soil, foundation, environmental, and structural variations. The sensitivities for the various foundation/structure configurations are summarized in Table 23. As indicated, soil parameters dominate the uncertainty of the natural frequency for all categories. The sensitivity of gravity base foundations to soil parameters is higher than for piled foundations, lending to the presumption that gravity base foundations cannot be uniform within a wind farm, but rather designed for local soil conditions at the location of the foundation within the limits of the wind farm. For structural considerations, the rotor-nacelle assembly had the largest effect on natural frequency, with other parameter variations resulting in frequency changes of less than 0.2%. The TUDelft study also examined the variation of natural frequency with foundation models (

69

Figure 16). As shown, the finite element and linear elastic models gave comparable results, resulting partially from the soil uniformity. Randolph’s method was less suitable for piles shorter than a critical pile length and for non-linear foundation deflections. The suitable value of the effective fixity depth for monopiles, tripod, and lattice structures differ due to dependencies on vibration mode shapes, pile stiffness, and soil properties. The stiffness matrix corresponded well with the finite element model, indicating that the stiffness matrix may replace the use of excessive computations from FEM in certain cases.

For the tripod and lattice tower

configurations, lateral flexibility was paramount in comparison to axial flexibility, mainly dependent on foundation stiffness, pile spacing, and the distribution of mass between the tower and the rotor-nacelle assembly. A comparison of these calculated natural frequencies were made to measured values for five turbines at the offshore wind farm Irene Vorrink. All values were within expected deviations form measured values, with the largest deviation being 5.3%. A comparison to the wind farm Lely resulted in a difference of 9% observed for one turbine, and a 30% difference for another turbine, which could not be explained without comprehensive re-assessment of model parameters. Comparisons of natural frequency and soil properties showed no changes over a 6 year period at the Lely farm (Zaaijer 2002).

70

7.0 Foundation Design 7.1

Introduction

The loading regime of the offshore wind turbine foundation is unique to offshore structures in that the weight of the turbine structure is low compared to the overturning moment and the horizontal load. It is critical that the foundation can sustain extreme loading conditions via any combination of environmental loads. When designing the foundation for an offshore wind turbine, it is important to keep in mind that since wind turbine farms contain numerous turbines, a single design for the entire farm is necessary to enable mass production and ensure expedient installation, both of which are necessary for the economic feasibility of a wind farm. The foundation should be designed in accordance with local, national, or international standards and regulations relevant to the site of the wind turbine. While there exist well-established design codes for offshore construction (e.g. ISO 19900 to 19903 Design of Offshore Steel and Concrete Structures, API Guidelines for Piles, etc.) that provide a good basis for offshore design, these codes generally do not cover the difficult design aspects such as layered soils or cyclic load effects, which should be investigated on a case-by-case basis.

In preliminary design

calculations, the number of load cases to consider can be reduced using load case lumping. This is a common offshore design technique that decreases the overall design effort by constructing a small number of lumped load cases to represent the full range of environmental characteristics at a site (Watson 2000). The choice of the foundation type should be based on site-specific information, including the adequate characterization of the soil conditions, water depth, scour and erosion potential, turbine capacity, foundation cost, and the environmental loading conditions. The design process must consider both the strength and the deformation characterization of the surrounding soils. The primary soil strength failures that can occur in an offshore environment include bearing capacity failure, sliding failure, and pile pull-out and punch-through failure. The primary deformation failures that can occur include large settlements or lateral displacements. Limit states categories are applied to foundation design where cyclic loading failure is treated as an ULS condition using partial and material load factors (DNV-OS-J101 2004). For effective 71

stress analysis, the tangent to the soil’s friction envelope is divided by material factor, γm, while for total stress analysis, the undrained shear strength is divided by material factor. Ultimately, the foundation design will depend primarily on the cost of installation due to the number of turbines in a typical wind farm, the in-service performance relative to repeated loading, large overturning moments, serviceability limit states, and considerations in the decommissioning and removal of the structure, and possibly the foundation. Because there is currently no economically feasible way to remove piles in offshore environment, foundations such as the suction caisson may soon become a reality in large scale wind farm development (Watson 2000).

7.2

Offshore Turbine vs. Offshore Platform Foundations

Much of the technology in use for offshore oil and gas structures can be applied towards the design of offshore wind turbine structures. However, there are a few major differences that must be considered, such water depth, loading condition, and the typical number of installations. Water depths for traditional fixed offshore structures have been in the range of 20 to 120-m, whereas wind turbines will be located in water depths in the 10 to 25-m range for the near future. The loading conditions for the three critical aspects, namely vertical, horizontal, and moment, all differ for offshore turbine versus offshore platform design (Figure 17). The vertical load for a traditional offshore platform varies between 5,000 and 30,000-tonnes, whereas the wind turbine structure load ranges from 100 to 300-tonnes. The horizontal load for traditional platforms is typically 10 to 20% of the vertical load, while it is much more variable for turbines, ranging from 70 to 150% of the vertical load (Watson 2000). The moment load imposed on the platform is calculated as the multiple of the water depth and the horizontal load, but for turbines it is equal to the multiple of the water depth plus 50-m (or more) and the horizontal load, due to the height of the rotor-nacelle assembly above mean sea level (Watson 2000). Typically, the number of installations of a traditional offshore platform ranges between one and four (depending on whether each pile group or gravity base is counted), but for an offshore wind farm, the number of installations can be greater than 200, and are usually at least 25. Because of the average number of installations occurring for a given project, the cost per structure must be

72

proportionately lower than single offshore installations to offset the high percentage of foundation installation in total cost (Watson 2000). Another aspect of offshore wind turbine design compared to offshore platform design is the importance of the structure. The offshore wind turbine is unmanned, is usually not a single supply source, has minor environmental impact in a failure state, and costs much less than the offshore platform. Therefore, a minimal factor of safety for the structure is required (Vugts and Harland 1997).

7.3

Piled Foundations 7.3.1 Introduction

Offshore piled foundations evolved from onshore designs using the large database of empirical evidence and experience to draw upon. However, the current knowledge base does not account for particulars for offshore piles including size, soil shear strength, and type of loading for offshore wind turbines. For these cases, the following parameters are needed for design: 1) axial capacity of piles in tension and compression, 2) load-deflection characteristics of axially and laterally loaded piles, 3) pile drivability characteristics, and 4) mudmat (i.e. temporary foundation that assists in alignment and leveling of pile foundation) bearing capacity (Gravesen 2004). Mudmat design is highly dependent on accurate slope characterization and assessment of shear strength of the soft surface soil layers. These parameters are needed to assess stability for both temporary and operational conditions (e.g. stability of drilled holes before pile placement).

7.3.2 General Design Considerations Piled foundations should be examined in the following contexts: 1) elastic ultimate limit state, where only one pile per foundation is allowed to reach yield point as a maximum, 2) plastic ultimate limit state (accounting for cyclic-load strength degradation), where piles are allowed to yield if still absorbing design loads, 3) fatigue in terms of both actual fatigue load on structure and damaging effects of pile driving, 4) pile driving analysis, 5) eigenfrequency analysis (i.e. characteristic representation of the dynamic response of the pile), and 6) soil damping estimate (Gravesen 2004).

73

Factors that need to be considered include the shear strength characteristics, deformation properties, in situ stress conditions, methods of installation, geometry and dimensions of the pile, and the types of load to which the pile will be subjected. For the design soil resistance for axial, lateral, and moment loading, the material factor for effective stress analysis shall be equal to 1.2 (ULS) or 1.0 (SLS), and for total stress analysis equal to 1.3 (ULS) or 1.0 (SLS). For axial loads in ULS design, a material factor equal to 1.3 should be applied to all characteristic values of soil resistance (e.g. skin friction, tip resistance) (DNV-OS-J101 2004). The design pile loads can be determined from structural analysis in which the pile is modeled with adequate equivalent elastic stiffness or with non-linear models reflective of the true stress-strain properties of the soil. For drilled piles, limit skin friction assumptions must be verified during installation due to drilling mud adhesion effects. Laterally loaded piles are analyzed based on realistic stress-strain curves for the soil and pile, which may be inelastic due to the combined effects from axial loading. The curves can be based on plasticity theory provided that characteristic resistance is in accordance with recognized plastic theorems to avoid non-conservative safety estimates, or based on calculations that assume that lateral deformations will completely plastify the soil.

7.3.3 Grouting Operations Grouting can be used for securing piled foundations into the seafloor where bedrock socketing is necessary, or where a monopile is connected to a transitional section for tower support. Advantages of grouting include the ability to account for pile installation tolerances, increased soil strength capacity, and higher fatigue resistance.

Some important characteristics of

competent offshore grout include its pumpability (i.e. sufficient viscosity and inner cohesion), low shrinkage potential, low thermal development during hardening, sufficient early strength development, and resistance to fatigue.

The grouting material used should be capable of

maintaining sufficient strength throughout lifetime of structure from all forms of deterioration, including chemical, mechanical, and/or mixing/dilution problems (DNV-OS-J101 2004). A Danish company Densit has developed Ducorit grout, applicable for offshore use. Ducorit is comprised of various quartz and bauxite aggregate that allows its application as an active structural component rather than simply filler material (J. Offshore Tech. 2002).

74

7.3.4 ULS Design 7.3.4.1 Monopile Foundations

For the ULS monopile design, soil strength values should be used which are equal to the characteristic soil strength values divided by the material factor. The design loads are calculated by multiplying the characteristic load by the specified load factor. The characteristic load is representative of extreme load conditions for both axial loading and combined lateral loading and moment loading. The axial resistance is calculated by integrating the design unit skin friction over the surface area of the pile, adding possible pile tip resistance and subtracting the weight of the pile. A general form of the equation is shown below: L

Pu =

∫ ( f C )dz + f s

b

Ab − W p

Eq. 7-1

z0

where: Pu

= ultimate load capacity

fs

= ultimate skin friction along length of pile Note: A static unit skin friction reduction factor of 0.9 for sands, 0.72 for clays is recommended (DNV-OS-J101 2004).

C

= circumference of pile

L

= embedded length of the pile

z0

= length of pile below the seafloor over which no skin friction is assumed to develop

fb

= ultimate resistance of the pile base

Ab

= gross area of the pile base

Wp

= weight of pile

The combined lateral and moment resistance should be ensured based on verification of lateral pile resistance through two requirements. 1) The first is ensuring that the theoretical design total lateral pile resistance, which can be found by vectorial integration of the design lateral resistance over the length of the pile, is not less than design lateral load applied at pile head. This verification can be done using FEM analysis with soil support springs in terms of p-y and t-z curves attached at the nodal points.

2) The lateral displacement of the pile head can be

calculated based on the soil resistance and soil stiffness, and it should not exceed limits specified based on the allowable displacement of the structure.

75

7.3.4.2 Multiple-Leg Foundations

For ULS multiple-leg foundation design, sufficient axial capacity and resistance to lateral and moment loading should be ensured for each pile by integrated analysis of the support structure and the foundation piles subject to design loads. P-y curves should be generated according to API RP2A code and DNV Classification Notes 30.4. T-z curves should be generated according to DNV/Riso: “Guidelines for Design of Wind Turbines”, dependent on the degradation potential of unit skin friction as mentioned previously (DNV-OS-J101 2004). The same reduction factors as for monopile foundations should be used for cyclic strength reduction. For grouped piles, additional requirements should be imposed as piles that are closely spaced may cause the grouped pile resistance to potentially be less than the sum of individual pile resistance, and the py and t-z curves should be adjusted accordingly. Also, the load transfer between closely spaced piles can cause soil deformations in soils supporting adjacent piles, resulting in a potentially softer pile group response.

This may be accounted for by elastic half-space solutions for

displacements in a soil volume due to applied point loads.

7.3.5 SLS Design 7.3.5.1 Monopile Foundations

For the SLS monopile design, the characteristic loads will be representative of loads causing permanent deformations of the foundation geometry (e.g. pile head tilt), and should account for the cumulative effects of long term cyclic loading on permanent deformations due to combined loads. It should be ensured that deformation tolerances, including tolerances for installation, are not exceeded in the form of permanent deformations, usually specified as a maximum allowable rotation of the pile head from the vertical plane based on visual and operational requirements of the wind turbine. 7.3.5.2 Multiple-Leg Foundations

For the SLS multiple-leg foundation design, the same procedures are used as for the monopile. Deformation tolerances should not exceed maximum specified pile rotations and horizontal displacements of the pile heads.

Separate tolerances may be specified to a pile group

immediately following installation versus the permanent cumulative displacements that occur throughout the design lifetime of the structure.

76

7.4

Gravity Base Foundations 7.4.1 General Principles

There exists a similar database and pool of experience for gravity base foundations as for piled foundations from the offshore oil and gas industry. For design purposes, the following factors should be accounted for: 1) total stability failure, 2) rupture in soil carrying capacity, 3) sliding ruptures, 4) combined ruptures in soil and structure, 5) ruptures due to foundation movements, 6) unacceptable movements and oscillations, 7) eigenfrequency analysis, 8) liquefaction risk analysis (when set upon sandy soils), 9) zones of local strong soil or rock, and 10) design of gravel bed (differential settlement analysis, requirement to grading curve, leveling of gravel base, base minimum thickness) (Gravesen 2004). For sliding rupture potential, calculation of the passive earth pressure should include the expected damage from scour and whether filling around the foundation is to be accounted for. This should include horizontal forces, overturning moment, and torsion moments about the vertical axis of the structure. Special attention should be given to any softening effects for residual top clay layers. The eigenfrequencies are modeled as springs attached to the foundation to demonstrate soil stiffness (see DNV Classification Notes 30.4) (DNV-OS-J101 2004). The analyses should be carried out with consideration that the wave load period is approximately 10 to 20-sec for resonance analysis, and that the soil will consolidate under the weight of the structure, creating initial excess pore pressure that may reduce strength parameters relative to design values (Poulos 1988). For gravity-based foundations with skirts, the hydraulic instability must be adequately guarded against for foundations which calculate tensile forces on differential water pressure. When tremie grouting, the stability of the structure and the surrounding soil is verified by comparing the volume of pumped concrete to the estimated volume of voids below the foundation (Gravesen 2004).

7.4.2 General Design Equations The two most common equations for design of gravity base foundations are Terzaghi’s general bearing capacity equation and Vesic’s bearing capacity equation, which accounts for many geometrical factors (Coduto 2001). Terzaghi’s equation for bearing capacity is as follows: 77

V peak = (12 BN γ + cN c + qN q )A foundation

Eq. 7-2

where: Vpeak

= peak vertical load that can be sustained by the foundation

B

= foundation width

c

= cohesion for soil below foundation

q

= overburden stress at depth of foundation

Afoundation

= area of the foundation

Nγ, Nc, Nq

= bearing capacity factors dependent on friction angle of the soil

Due to the complicated moment and horizontal loads acting on the foundation, the factors included in Vesic’s equation allow for a more precise design, and is currently the standard for bearing capacity calculation for homogeneous soil conditions in the offshore industry (Byrne and Houlsby 2003). qult

  c     = (c'+σ ' zD +0.5γ ' B )( Nsdibg )q  γ    

Eq. 7-3

where: qult

= ultimate bearing capacity of the soil

c’

= effective cohesion for soil below foundation

σ’zD

= effective overburden stress at depth of foundation

γ’

= effective unit weight of soil

N

= bearing capacity factor dependent on friction angle of soil

s,d,i,b,g

= factors accounting for shape (s), depth (d), load inclination (i), base inclination (b), and ground inclination (g)

7.4.3 Stability of Foundation: ULS and ALS Design The risk of shear failure below the foundation base should be investigated for all gravity base foundations, covering all potential shear surfaces, and accounting for foundation geometry. Special consideration should be given to soft soil layer and cyclic loading effects. For skirted gravity foundations, the base dimension is assumed to be at the skirt tip level. The design should

78

be analyzed for site-specific soil drainage conditions (i.e. fully, partially, or undrained conditions). Stability is evaluated based on either total stress analysis or effective stress analysis for laboratory shear strength and pore pressure measurements via stress paths. Effective stress analysis is based on the effective strength parameters of the soil and realistic pore pressure estimations dependent on the initial pore pressure, the build up of pore pressures due to cyclic load history, transient pore pressure through each load cycle, and any effects of energy dissipation. Total stress analysis is based on total shear strength parameters determined from tests on representative samples subject to appropriate as-built stress conditions.

Stability

analyses based on conventional bearing capacity formulae are only acceptable for uniform soil conditions and relatively small foundation bases. The material factors to be used are the same as previously used for ULS design, with ALS material factors equal to 1.0. Cyclic load effects should also be incorporated through load factors, and the overturning safety is evaluated in the ULS and ALS design. Both internal soil damping and radiation damping should also be considered in the design (DNV-OS-J101 2004).

7.4.4 Settlements and Displacements: SLS Considerations Analyses for settlements and displacements include initial consolidation, secondary settlements, differential settlements, long-term permanent horizontal displacements, and dynamic motions. The differential settlement analyses should account for lateral variations in soil conditions within the foundation area, non-symmetrical weight distributions, predominating environmental load directions, and liquefaction in seismically active areas.

7.4.5 Grouting Operations Base grouting is performed on gravity base foundations to avoid further seabed penetration, to retain vertical position, to ensure uniform stress distributions below the foundation, and to reduce the potential for piping that occurs below the base during environmental loading. It is particularly useful on sloping or uneven seafloors (Poulos 1988). Base grouting should not cause filling pressures to create channeling from one skirt to another or to the outside of the structural

79

periphery.

Grouting material properties should be similar to the requirements summarized

according to the piled foundations grouting operations.

7.5

Suction Caissons 7.5.1 General Principles

For suction caisson foundations, little experience and the lack of onshore precedent exist to base design procedures on. However, the design procedure is understood for relatively simple cases. More investigation is currently needed to describe the horizontal, overturning, and cyclic loading response of these foundations (Watson 2000).

Generally, the following factors should be

considered: 1) safe and even penetration of the caisson during installation, 2) plastic ultimate limit condition, 3) operational limit condition, 4) fatigue analysis, 5) eigenfrequency analysis, 6) and soil damping estimate. For suction-based installation, it must be demonstrated that the soil penetration resistance is lower than suction driving force, and that the soil inside the caisson does not dilate more than that induced by the displacement due to caisson penetration only. Particular consideration must be given to scour around the caisson perimeter, as this zone is critical to the uplift capacity of the caisson (Gravesen 2004).

7.5.2 Design Studies A simplified analysis for a multiple-leg foundation design with suction caissons was conducted by Byrne and Houlsby (2003) (Figure 18). There two geometrical aspects to the design that varied were the foundation separation and foundation dimension. The foundation separation established the critical case when the foundation rotated about the two downwind foundations. Considering an overturning moment M consisting of net horizontal load H acting through height h above the foundations, and a restoring moment consisting of vertical load V acting through structural center of gravity, the minimum required spacing between foundations was estimated accordingly: Hy M  (quadruped )  2 V =2V s= Hy M (tripod ) 2 3 =2 3 V V 

80

Eq. 7-4

The foundation capacity established the critical case when one foundation lay downwind. For a quadruped foundation, the tensile capacity will be larger than for an equivalent tripod foundation because there is always more tensile resistance from the upwind foundations. Given equivalent conditions, a comparison between quadruped and tripod configurations is shown in Figure 19. The increase in foundation size and mass for the tripod configuration was offset by the reduced amount of steelwork for the structure, which may also be subject to a reduced wave and current load. The capacity is calculated based on conventional bearing capacity theory as described previously. The results showed no difference in caisson size for the tripod versus the quadruped design. Figure 20a shows the relationship between vertical load, caisson diameter, and caisson spacing based on the quadruped design with a constant aspect ratio (i.e. length divided by diameter) of 0.5. According to the figure, as vertical load increases (e.g. through ballast addition) the required caisson size reduces.

As horizontal loading dominates, caisson size initially reduces,

transitioning to increasing caisson size as the vertical load becomes critical.

The length-to-

diameter ratios are typically restricted in certain soil types and in shallow water depths. In sands, an L/D ratio that exceeds 1 will typically cause piping failure to occur. In stiff clays, the maximum L/D ratio will be 2.5 (5 for soft clays) due to the occurrence of a reverse bearing capacity failure of the soil plug at the skirt tip from the suction pressure. For shallow water depths, the cavitation limit of the water will be the limiting factor for clay (Byrne and Houlsby 2002). From a design calculation standpoint, the multi-leg suction caisson foundations are a more straightforward design than for single caissons, as vertical loads on the seafloor can represent the moment loads applied to the structure. For mass production, however, the single suction caisson foundation has the advantage due to simplicity of installation and lower structural steelwork requirement.

81

A simplified design analysis for a single suction caisson foundation was also conducted by Byrne and Houlsby (2003). The vertical loads on the foundation remained constant while the moment loads varied with wave incidence. Without a standard design methodology, the results showed a linear relationship between the vertical load and the caisson diameter based on a function of the moment load to the horizontal load ratio and the weight of the soil plug within the caisson cavity (Figure 20b).

82

8.0 Foundation Transport and Installation 8.1

Marine Operations

Marine operations for offshore wind farm development can include the cover yard lift, load out, sea transportation, offshore lift, and installation operations. The costs of marine operations result in much more expensive per turbine installation costs than onshore wind turbine installation. Close integration of design and construction is needed due to the additional challenges of operating at sea. To appreciate the size of the equipment needed and why the costs of marine operations are high, the typical foundation sizes for turbines and construction aspects are shown in Table 24. The construction constraints associated with different foundation types also affect costs as summarized in Table 25. Offshore wind turbine foundations are usually installed by floating crane vessels or mobile jackup units, the choice of which is dependent on water depth, crane capability, and vessel availability (Figure 21). When using a crane vessel, it must be capable of lifting hook heights greater than the height of the rotor-nacelle assembly of the turbine. Some of the lift capacities along with other equipment specifications are summarized in Table 26. In shallow waters, conventional mobile jack-up rigs are typical, whereas for deeper waters, the floating crane vessels are usually deployed. Because the jack-up rigs establish their stability on the seafloor, they are susceptible to the same issues and risks that foundations are designed against.

8.2

Project Duration

Existing crane vessels have not been specifically designed for wind turbine installation, but large wind farms may benefit through time and cost incentives from purpose-built units. As an estimate, the total duration of installation for a multi-unit wind farm will take several months. All marine operations are subject to weather constraints and down-time, so scheduling during calm periods when wind and wave speeds are minimal is recommended. The actual construction time for piled versus gravity base foundations either from jack-up units or floating vessels are similar when considering weather-related down time.

However, there is a significant cost

savings for a two phase installation (i.e. foundation unit installation followed by sub-structure, tower, and rotor-nacelle assembly installation) versus a three phase installation (i.e. foundation,

83

sub-structure and tower, and rotor-nacelle assembly installed separately). To reduce marine operation costs in the future, there are three main objectives that should be addressed: 1) improved dissemination of offshore construction procedural and technical knowledge among designers and developers, 2) evaluation of resistance of existing offshore pile design techniques for low mass, fatigue-dominated applications, and 3) optimization of offshore wind structure installation cost-effectiveness using novel construction approaches (Watson 2000). More information on transport and installation aspects can be found in DNV Rules for Planning and Execution of Marine Operations (DNV-OS-J101 2004).

84

9.0 Costs of Wind Energy 9.1

Renewable Energy Market Incentives

There are two main types of market incentives established to increase the development of renewable energy production, namely fixed price systems and fixed quantity systems. Both incentive types are ways of creating a protected market and offsetting the competitive disadvantage that results when markets neglect environmental effects of conventional energy production. Their main purpose is to provide incentives for technological improvements and cost reductions. In fixed price systems, operators are paid a fixed price for every unit of output, and the extra cost is paid by taxpayers and electricity consumers. This system is highly effective in Denmark, Spain, and Germany. In fixed quantity systems, the national government decides the target level of renewable electricity to be achieved over a certain time period, while markets establish the price. This system is currently used in the UK, Belgium, Sweden, and Italy via tradable green certificate systems. It is still in its early stages, and long-term power purchase contracts are still challenging. An advantage to this type of incentive is that it introduces competition between the electricity producers (EWEA 2004). The current status of the wind industry shows that the majority of incentives in effect are financially-based, with governmental incentives outweighing the investment incentives. Tax incentives, including tax and duty discounts, and production incentives, such as renewable energy portfolios, production incentives, and fixed tariffs are less prevalent in the current global economy (Shaw et. al. 2002).

9.2

General Cost Considerations 9.2.1 Introduction

Offshore wind farm development is more expensive compared to onshore wind farm development due to foundation costs, access difficulties, water depths, weather delays, and corrosion protection requirements. However, the factors affecting the cost of energy are similar, as determined by the following eight factors: 1) total investment cost, 2) project preparation

85

costs, 3) operation and maintenance costs, 4) magnitude and availability of mean wind speed at turbine hub height, 5) technical design lifetime, 6) amortization period, and 7) real interest rate. Although these factors contribute to a higher investment cost, offshore wind farms can offer up to 40% more energy potential than onshore wind farms due to higher wind velocities and less turbulence (Irvine et. al. 2003). There are two primary economies of scale in offshore wind turbine development: wind farm size and wind turbine size. To be economical, wind farms generally have to be greater than 100 MW of generating capacity using turbines of at least a 2 MW capacity. Based on a study of existing European projects in year 2001, the cost of energy for offshore wind farm development ranged from $0.053 to $0.112 per kWh, increasing with distance from shore and decreasing with number of turbines per project. With further advances in technology, deep water wind turbines could reach $0.051 per kWh, while shallow water wind turbines are projected to reach $0.041 per kWh within the next seven years (Musial and Butterfield 2004).

9.2.2 Cost of Energy Factors 9.2.2.1 Total Investment Cost

The total investment cost includes the production, transportation, and erection cost of the turbine, which can range from $1200 to $2000 per kW, with production accounting for approximately half of this cost (Manwell 2001). Up to 40% of this cost can be the turbine itself, with the tower and foundation contributing from 28 to 34%, grid connection from 9 to 36%, and other capital cost accounting for 6 to 17% of the total cost. Considering the turbine size, installations costs for offshore foundations are 200 to 250% higher than for onshore foundations for medium (e.g. 0.5 to 0.8 MW) turbines in very shallow water (e.g. < 6-m). This reduces to 150 to 200% for installations of larger turbines (e.g. 1 to 1.5 MW) in 6 to 15-m water depths. Considering the foundation type, piled foundations versus gravity base foundations do not influence cost significantly for up to 15-m water depth (Gaudiosi 1996). 9.2.2.2 Project Preparation Cost

The project preparation costs include permits, land, and infrastructure. Electricity generation falls under this category, which can cost 30% higher offshore than onshore (Gaudiosi 1996).

86

Grid connection cost is proportional to the number of turbines, however the capacity of the cable does not affect cost (Krohn 2002). Typical costs for cabling can range from $500k to $1M per mile (Manwell 2001). 9.2.2.3 Operation and Maintenance Cost

The operation and maintenance costs (O&M) include service, consumables, repair, insurance, administration, and site leasing costs. Offshore O&M cost is approximately 100% higher than onshore (Gaudiosi 1996), specifically ranging from $0.01 to $0.02 per kWh (Manwell 2001). Generally, O&M cost is proportional to the number of turbines in a wind farm, distance form shore, and occurrence of inclement weather, decreasing with more reliable turbine design. 9.2.2.4 Wind Speed and Availability

The mean wind speed at hub height and its availability is also important to energy cost. The annual energy output can be estimated by following expression:

( )

E = 3.2 V 3 A

Eq. 9-1

where: E

= annual energy output (kWh/m2 swept rotor area)

V

= mean annual wind speed (m/s)

A

= swept rotor area (m2)

Generally, as the mean wind speed increases, the cost per kWh decreases exponentially to a stable value. The availability is defined as the time percentage that the wind speed falls between the turbine’s cut-in and cut-out wind speeds, which is typically higher than 98% of the time (Beurskens and Jensen 2005). 9.2.2.5 Technical Design Lifetime

The technical design lifetime is typically 20 years. Parts of the rotor-nacelle assembly system and other mechanical components usually need replacement every 1 to 5 years (Beurskens and Jensen 2005). Other parts subjected to fatigue loads may need replacement halfway through the design lifetime. Foundations are generally built to last 50 years. Therefore, foundations can theoretically go through two cycles of turbines, which can lower generating costs by 25 to 33%, falling within cost comparisons to onshore sites (Krohn 2002). Decommissioning costs also

87

should be considered in the overall estimate, which vary depending on the type of foundation used and the decommissioning policy. 9.2.2.6 Amortization Period

The amortization period depends on the economic model used, the type of investor, and their financial strategy. It is often equal to the technical design lifetime (Beurskens and Jensen 2005). An issue to consider when defining this period is the potential for technological advancement during the project preparation phase. 9.2.2.7 Real Interest Rate

The real interest rate influences the cost of wind energy according to the following: c=

aI tot +m AE

Eq. 9-2

where: c

= cost ($/kWh)

Itot

= total investment per m2 swept rotor area

A

= availability

E

= annual energy output per m2 swept rotor area (kWh/m2)

m

= operation and maintenance cost

a

= annuity factor calculated according to the following equation:

i (1 + i ) a= (1 + n )n − 1 n

Eq. 9-3

where: i

= real interest

n

= amortization period

9.2.3 Cost Optimization A study was conducted by the Riso National Laboratory in Denmark to determine the factors affecting cost optimization of an offshore wind turbine farm (Fuglsang and Thomsen 1998). The reference turbine was a 1.5 MW turbine with a rotor diameter of 60-m. Two separate wind farm configurations were examined based on turbine spacing as a multiple of rotor diameter. The cost

88

analysis was based on component weight in order to avoid subcontractor and market price fluctuations. The analysis showed that the offshore cost to onshore cost comparison was 42% higher primarily due to increased foundation and grid connection costs, with each cost component split into fixed and variable cost shares (e.g. transport and component weight, respectively). Table 27 shows the relative cost distribution for an onshore wind farm compared to an equivalent offshore wind farm based on the reference wind turbines. The results of the study indicate that the factors influencing the cost optimization of offshore wind energy compared to onshore cost are an increase in swept area due to rotor diameter increase, increase in rated power, a reduced rotor speed, and a lowered hub height. All factors combined resulted in an optimized cost of offshore wind energy of approximately 12% below the onshore wind energy costs, regardless of the spacing of the wind turbines.

9.3

Costs for United States Offshore Wind Farm Development 9.3.1 Department of Energy Cost Estimates

A 1979 study was conducted for United States offshore wind turbine development that resulted in a formula for fabricated cost based on loading and material parameters (Kilar and Stiller 1980). The formula was representative of piled foundations made of concrete (0 to 15-m water depth) or steel (15 to 250-m water depth), and gravity base foundations made of concrete (60 to 150-m water depth) or steel (15 to 150-m water depth). The resulting formula is represented as follows: C = x(h ) (Fe ) + 0.159 y

z

where: C

= Fabricated Cost (1979 $M)

h

= Water depth (ft)

Fe

= Effective free-end load (106 ft-lbs)

Mt

= Total tipping moment (106 ft-lbs)

x,y,z

= Depth and load coefficients based on foundation material

89

Eq. 9-4

The resulting costs for typical environmental conditions off the Atlantic Coast based on 100 unit plants of 7 to 10 MW output are shown in Table 28. The cost of energy ranged between $0.064 and $0.176/kWh, with the Pacific Northwest on the lower side of the scale and New England near $0.102/kWh, compared with land-based estimates of $0.03 to $0.05/kWh (Rogers et. al. 2003). Although the costs would have to scaled to the current economy and technological advances since 1979 would have to be considered, the table highlights the effects of water depth and mean wind speed on cost.

9.3.2 Other Estimates Other studies in the U.S. that were based on generic 5-MW turbines in a 500-MW wind farm founded on monopiles 15-miles from shore in wind speeds greater than 8-m/s (Musial and Butterfield 2004). Shallow water cost of energy estimates were based on design improvements and changes for offshore conditions, and a production learning curve for cost reduction per doubling of installed capacity. Table 29 shows a breakdown of the shallow water cost projections. The deep water estimate was based on 600-ft of water depth, with an initial marinization cost premium of 11% higher than the land-based value. Differences in cost were due to the floating structure and additional electric cabling. Baseline O&M costs were higher than for shallow water ($0.018/kWh) due to added platform hardware and distance to shore, all other assumptions and design remaining the same. Figure 22 shows COE projections graph for a novel deep water floating single turbine option (Musial and Butterfield 2004).

90

10.0 Offshore Wind Development in New England 10.1 Introduction The history of offshore wind energy potential in the US has focused on three primary areas: the Pacific Northwest, Great Lakes, and New England. Due to the deeper water depths of these areas compared to the developed wind farm areas of Europe, many initial concepts were based on floating structures. The greater prospect of wind farm development in New England compared to the other locations is due to two primary reasons: 1) there is a high wind resource in shallow waters close to major electrical loads, resulting in shorter electrical transmission lines, and 2) there are difficulties developing large onshore wind farms (Rogers et. al. 2003). Political factors that have contributed to the interest include implementation of utility deregulation legislation that mandated System Benefit Charges and Renewable Portfolio Standards (RPS), as well as the advocacy of wind and renewable energy professionals (Manwell et. al. 2001). RPS and the Systems Benefit Charge require development of up to 1100 MW of new renewable energy capacity by year 2009 (Rogers et. al. 2003).

10.2 Environmental Conditions and Constraints 10.2.1 Wind Speed Mesoscale modeling, developed by TrueWind Solutions, has provided estimates of wind resources out to 50 nautical miles (nm) offshore (Musial and Butterfield 2004). It is being used to validate previous observations at 12 monitoring stations, half of which are located on islands or along the coast, the other half located between 12 and 170-nm off of the New England coast. The average estimated wind speeds for the sites range from 7.0 to 8.4-m/s at 60-m height. The Measure Correlate Predict method (MCP) is being used to predict wind data at other locations based on reference sites in the general vicinity, implemented with Visual Basic computer code. It is combined with the log law to predict wind data at heights greater than those at which data was obtained (Manwell et. al. 2002). Results have indicated that there are immense areas of Class 5, 6, and 7 winds at 5 to 50-nm from shore (Table 30) (Musial and Butterfield 2004).

10.2.2 Water Depth The Massachusetts coast consists of shallow water depths (up to 18.3-m) that extend from 2 to 4km from shore. In the Cape Cod Bay, water depths less than 18.3-m extend out to 15-km from 91

shore. In Nantucket Sound and Buzzard’s bay, most of the water depth is shallower than 18.3-m. Based on these numbers, shallow water projects off coast of Massachusetts could produce up to 55,000 GWh, or 116% of the state’s energy needs (Rogers et. al. 2003).

10.2.3 Wave Loading As of yet, there is no well-defined wave data for New England. Wave, tide, and current conditions are less defined in the U.S than in the more shallow and sheltered European seas (Musial and Butterfield 2004). Wave current speeds of 1.3-m/s (2.5-knots) have been measured in Nantucket Sound, while the maximum measured waves heights at the mouth of Boston harbor were 9.1-m (Rogers et. al. 2003).

10.2.4 Subsurface Conditions The geology of Cape Cod and its surrounding areas results from glacial deposition, creating a mix of seafloor soil types. In the shallow waters around Cape Cod, sand is most prevalent, while there is a transition from a sand-clay mixture to only clay in the deepest waters. There are also scattered boulders, rocks, and outcrops throughout the seafloor. Nantucket Sound contains numerous shifting sand shoals, and significant slopes in these areas can create problematic installation issues (Rogers et. al. 2003).

10.3 Other Design Considerations 10.3.1 Land Jurisdiction The coastal waters in Massachusetts are divided into three categories of jurisdiction. The state has control for up to 4.8-km (3-miles) from shore. The Massachusetts Ocean Sanctuaries association has established that energy generation is prohibited within these limits, which may be subject to future amendments. Currently, there is no standardized procedure for permitting in Massachusetts, therefore siting a wind farm in waters further than 4.8-km from shore may significantly reduce the permitting process. The federal government has jurisdiction from 4.8 to 320-km (200-mile) from shore, headed by the Department of the Interior and the U.S. Army Corps of Engineers. Other entities of power include the U.S. Coast Guard, Federal Aviation Association, and the Department of Agriculture. Beyond 320-km, waters are considered to be under international jurisdiction (Rogers et. al. 2003). 92

10.3.2 Electrical Grid Connections Significant grid connections exist along the coast of Massachusetts. However, future power production could be limited by grid capacity on Cape Cod or from power lines crossing the Cape Cod Canal.

The primary option for transferring the power to shore is undersea cabling,

accomplished by washing the cables using high pressure water jets below the seafloor. This is the most economical method compared to trenching or plowing cables into the seafloor, however it is highly dependent on seafloor geology and soil conditions. As an example of grid connection sizes, current wind farms in Europe use 20 to 40-kV connections with a 30 to 150-kV transformer inside each wind farm, while links to the mainland have used 120 to 150-kV connections (Krohn 2002).

10.3.3 Availability of Harbors for Construction and Maintenance Massachusetts has significant dock, dry dock, and shipyard facilities, located in strategic towns such as New Bedford and Quincy. These areas are necessary for support craft, barge mounted cranes, cable-laying vessels, staging areas, and specialty-built vessels and floating systems.

10.3.4 Local Public Acceptance and Conflicting Areas There are certain public considerations that must be addressed when siting an offshore wind farm. A substantial consideration is the visual impact that the turbines will have on the horizon. Models have shown that a 64-m-tall turbine is visible above the horizon at 36.6-km away. At night, beacons may be required to alert airlines and ships. Current FAA regulations require anything taller than 61-m to have beacons (Rogers et. al. 2003). Another issue with siting an offshore wind farm in the New England area is the prevalence of shipwrecks. There are many archeological sites around coast of Massachusetts, especially off Nantucket Sound and east of Cape Cod coast.

10.4 Advancement of Deep Water Technology The expected progression in the U.S. towards deep water exploitation of wind farms will result from shallow water monopiles or gravity base foundations that provide application-specific experience towards the deeper water multiple-leg foundations and floating designs (Figure 23). 93

The economical step up from shallow fixed foundations to deep floating foundations is based on the additional floating requirements and grid connection system. These added costs can be based on oil and gas floating platforms with reductions made for the differences between oil and gas applications and wind turbine applications including: 1) higher safety margins for personnel, evacuation requirements, and spill prevention for oil and gas platforms, 2) the premise that depths up to 600-ft for floating wind turbine platforms would be adequate as opposed to 1500-ft and greater for oil and gas platforms, and 3) that wind platforms should be submergible to minimize wave loading as opposed to oil and gas platforms which minimize wave loading by constructing a above-water deck area (Musial and Butterfield 2004). A concerted research and development effort from the U.S. is necessary to address the critical issues in floating designs, including dynamic modeling of turbines and platforms, floating platform optimization, low-cost mooring and anchor development, floating wind turbine COE optimization strategies, deepwater erection and decommissioning, standards governing floating turbines, and deepwater resource assessment. Figure 24 shows a summary of the necessary steps in achieving this goal (Musial and Butterfield 2004).

94

11.0 Conclusions The global development of offshore wind energy is heading towards larger turbines, larger wind farms, and further distances offshore into deeper waters. The unique loading aspects of offshore wind turbine foundations require further development of the current state of practice in design of these structures. Although the costs of offshore wind energy can be competitive with onshore wind energy generation, refinement of the current foundation technology and offshore construction procedures are necessary to decrease the relatively higher foundation manufacture, transport, and installation costs. The technology required to achieve this advancement will require continued efforts from government bodies, private institutions, and the offshore energy industry. Although existing technology will be adequate to develop a substantial infrastructure for shallow water offshore wind energy generation, advances towards use of novel foundation concepts such as suction caissons and floating wind turbine platforms will be necessary to secure larger wind turbines in the ocean. The potential value in the investment of this technology can be seen by the current projections of wind energy generating capacity in the near future, as countries foresee the increased reliance on more renewable energy sources and less resource-depleting energy sources. Areas of the United States that are potentially high-output offshore wind energy producers include the shallow waters along the coasts of New England and the mid-Atlantic states, as well as parts of the West Coast, where deeper waters, and hence, a demand for technological investment, prevail. Areas of the Gulf of Mexico may also see increased interest due to the prevalence of existing oil and gas platforms that are nearing depletion and would otherwise face decommissioning and removal activities.

95

REFERENCES J. Offshore Tech., (2002). Securing Wind Turbines in the Water, Journal of Offshore Technology, Vol. 10, No. 3, pp.14-19. Beurskens, J., Jensen, P.H., Wind Energy, www.ewea.org/documents/20_Beurskens_Hjuler_ RD.pdf , Accessed May 2005, 30pp. Byrne, B.W., Houlsby, G.T., (2003). Foundations for Offshore Wind Turbines, Phil. Trans. Royal Society of London, Vol. 361, pp.2909-2930. Byrne, B.W., Houlsby, G.T., (2002). Investigating Novel Foundations for Offshore Windpower Generation, Proceedings of the 21st International Conference on Offshore Mechanics and Artic Engineering, Oslo, Norway, pp.567-576. Coduto, D.P., (2001). Foundation Design: Principles and Practices, 2nd Edition, Prentice Hall, New Jersey, 883pp. Det Norske Veritas DNV-OS-J101: Design of Offshore Wind Turbine Structures, June 2004. 138pp. European Wind Energy Association, (EWEA). (2004). The Current Status of the Wind Industry, 12pp. Feld, T., Waegter, J., (2002). Integrated Support Structure Design Analysis, Journal of Offshore Technology, Vol. 10, No. 3, pp.10-13. Frandsen, S., Thogersen, M.L., (1999). Integrated Fatigue Loading for Wind Turbines in Wind Farms by Combining Ambient Turbulence and Wakes, Wind Engineering, Vol. 23, No. 6, pp327-339. Fuglsang, P., Thomsen, K., (1998). Cost Optimization of Wind Turbines for Offshore Wind Farms, Riso National Laboratory, Roskilde, Denmark, 31pp. Gaudiosi, G., (1996). Offshore Wind Energy in the World Context, WREC, pp.899-903. Gravesen, H., (2004). #10105(12/13/02) and #10189(2/24/04): Foundation Design (Text Proposal for WG03), www.ecn.nl, Accessed March 2005. Gravesen, H., Bro, C., (2003). A Simple but Complete Design Procedure for Offshore Wind Turbine Foundations, #10148(10/30/03), www.ecn.nl, Accessed March 2005, 35pp. Henderson, A.R., Morgan, C., Barthelmie, R., Smith, B., Sorenson, H.C., Boesmans, B., (2002). Offshore Wind Energy – Review of the State-of-the-Art, Proceedings of the 12th International Offshore and Polar Engineering Conference, Kitakyushu, Japan, May 26-31, pp.494-498.

96

Houlsby, G.T., Byrne, B.W., (2000). Suction Caisson Foundations for Offshore Wind Turbines and Anemometer Masts, Wind Engineering, Vol. 24, No. 4, pp.249-255. Houlsby, G.T., Byrne, B.W., Martin, C.M., (2001). Novel Foundations for Offshore Wind Farms, A Research Proposal to EPSRC (August 2001), 9pp. Irvine, J.H., Allan, P.G., Clarke, B.G., Peng, J.R., (2003). Improving the Lateral Stability of Monopile Foundations, Proceedings of the BGA International Conference on Foundations innovations, observations, design and practice, Thomas Telford, London, pp.371-380. Jenner, C., Finch, M., Finlayson, K., Harker, G., (2002). Optimizing Integrated Site Investigations for Offshore Wind Farm Projects, Proceedings of the International Conference on Offshore Site Investigations: Diversity and Sustainability, November 26-28, London, United Kingdom, pp.141-149. Kellezi, L., Hansen, P.B., (2003). Static and Dynamic Analysis of an Offshore Monopile Windmill Foundation, Proceedings of the BGA International Conference on Foundations innovations, observations, design and practice, Thomas Telford, London, pp.401-410. Kilar, L.A., Stiller, P.H., (1980). An Assessment of Offshore Wind Energy, Energy Technology: Proceedings of the Energy Technology Conference, Vol. 2, pp.1471-1483. Krohn, S. (2002). Offshore Wind Energy: Full Speed Ahead, www.windpower.org/en/ articles/offshore.htm. Musial, W. and Butterfield, S., (2004). Future for Offshore Wind Energy in the United States, Proceedings of the EnergyOcean Conference, June 28-29, 2004, Palm Beach, Florida, 13pp. Manwell, J.F., (2001). An Overview of the Technology and Economics of Offshore Wind Farms, Presentation for the Cape and Islands Offshore Wind Public Outreach Initiative, Meeting No. 3, November 21, 2001. Manwell, J.F., McGowan, J.G., Rogers, A.L., (2002). Wind Energy Explained: Theory, Design, and Application, John Wiley & Sons, Ltd., England. 577pp. Manwell, J.F., Rogers, A.L., McGowan, J.G., (2001). Status of Offshore Wind Energy in the United States, Proceedings of the Second International Workshop on Transmission Networks for Offshore Wind Farms, March 29-30, 2001, Royal Institute of Technology, Stockholm, Sweden, pp10-13. Manwell, J.F., Rogers, A.L., McGowan, J.G., Elkinton, C.N., (2004). Characterization of External Conditions for the Design of Offshore Wind Energy Systems for the United States, Proceedings of the 2004 ASME Wind Energy Symposium at the 42nd AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, January 5-8, pp.487-497.

97

Manwell, J.F., Rogers, A.L., McGowan, J.G. and Bailey, B.H., (2002). An Offshore Wind Resource Assessment Study for New England, Renewable Energy, Vol. 27, pp.175-187. Norsok Standard G-0111, (2004). Marine Soil Investigations, Revision 2, October. 66pp. Petersen, P., Riber, H.J., Ronold, K., Det Norske Veritas (DNV), (2003). Rules for Design of Offshore Wind Turbine Structures, www.dnv.com/binaries/DNVrules_offshore_2003_tcm429389.pdf, Accessed March 2005, 8pp. Poulos, H.G., (1988). Marine Geotechnics, Unwin Hyman Ltd., London, 473pp. Powers, M.B., (2005). Platform-Based Wind Farm May Also Pump Up Oil Industry, Engineering News-Record, Vol. 254, No. 3, pp.31. Randolph, M.F., Kenkhuis, J., (2001). Offshore Foundation Systems 406: Site Investigation Planning – Course Notes, Center for Offshore Foundation Systems, University of Western Australia. Rogers, A.L., Manwell, J.F. and McGowan, J.G., (2003). A Year 2000 Summary of Offshore Wind Development in the United States, Energy Conversion and Management, Vol. 44, pp.215229. Sahin, A.D., (2004). Progress and Recent Trends in Wind Energy, Progress in Energy and Combustion Science, 43pp. Schneider, J., (2004). Nearshore and Offshore Site Investigations and Foundations, (Presented at the University of Massachusetts, Amherst), October. Shaw, S., Cremers, M.J., Palmers, G., (2002). Enabling Offshore Wind Developments, European Wind Energy Association, 133pp. Offshore Site Investigation Committee (OSIC): Allan, P., Cook, M., Power, P., (2004). Guidance Notes on Geotechnical Investigations for Marine Pipelines, Rev. 03, Society for Underwater Technology, 47.pp. van der Tempel, J., Zaaijer, M.B., Subroto, H., (2002+). The Effects of Scour on the Design of Offshore Wind Turbines, Delft University of Technology, The Netherlands, 9pp. Vugts, J.H., Harland, L.A., (1997). Optimization of the Design of Support Structures for Offshore Wind Energy Converters Using Advanced Offshore Wind Engineering Technology, Wind Engineering, Vol. 21, No. 5, pp.319-337. Watson, G., (2000). Structure and Foundation Design of Offshore Wind Installations, Final Report from the Offshore Wind Energy Network Workshop, March, CLRC Rutherford Appleton Laboratory. 27pp.

98

Zaaijer, M.B. (ed.), (2002). Design Methods for Offshore Wind Turbines at Exposed Sites (OWTES): Sensitivity Analysis for Foundations of Offshore Wind Turbines, Wind Energy Section, TUDelft. 49pp.

99

LIST OF TABLES Table

Page

Table 1. Technologies Involved in Offshore Structure Design. (Poulos 1988)......................... 102 Table 2. Phases of Project Certification and Associated Activities. (DNV-OS-J101 2004) ..... 102 Table 3. Design Standards for the Offshore Industry. ............................................................... 103 Table 4. Worldwide Offshore Wind Energy Potential. (Gaudiosi 1996) .................................. 104 Table 5. Operational Offshore Wind Farm Projects. ................................................................. 104 Table 6. European Offshore Energy Potential by Country. (Sahin 2004) ................................. 104 Table 7. Environmental Parameters for Coastal United States.................................................. 105 Table 8. Offshore Resource Estimates for U.S. (Musial and Butterfield 2004) ........................ 105 Table 9. Planned Offshore Wind Farms. ................................................................................... 106 Table 10. Characteristic Load Selection Basis. (DNV-OS-J101 2004)..................................... 107 Table 11. Load Factors for ULS and ALS. (DNV-OS-J101 2004) ........................................... 107 Table 12. Application of Site Condition Data for Project Phases of an Offshore Wind Farm. (Jenner et. al. 2002)............................................................................................................. 108 Table 13. Offshore Site Investigation Planning Process. (Randolph and Kenkhuis 2001) ...... 109 Table 14. Typical Offshore Site Investigation and Analysis Schedule. (Poulos 1988)............ 109 Table 15. Seismic Profiling Methods. (Randolph and Kenkhuis 2001) ................................... 110 Table 16. Accuracy Classes for Cone Penetration Testing. (Norsok 2004) ............................. 110 Table 17. Field Vane Blade Size. (Norsok 2004) ..................................................................... 110 Table 18. Suitability of Test Methods for Soil Parameters. (OSIC 2004)................................ 111 Table 19. Other Soil Parameters for Specific Applications. (OSIC 2004) ............................... 112 Table 20. Evaluation of Sample Quality. (Norsok 2004) ......................................................... 112 Table 21. Model Parameters for Gravity-based Foundations. (Zaaijer 2002) ........................... 113 Table 22. Estimations of Effective Fixity Length. (Zaaijer 2002)............................................. 113 Table 23. Important Natural Frequency Sensitivities for Various Structural Configurations. (Zaaijer 2002)...................................................................................................................... 114 Table 24. Typical Sizes and Construction Aspects. (Watson 2000)......................................... 115 Table 25. Foundation Construction Constraints. (Watson 2000) .............................................. 115 Table 26. Typical Installation Vessel Specifications. (Watson 2000) ....................................... 115 Table 27. Relative Cost Distribution for Onshore vs. Offshore Wind Farm Comparison. (Fuglsang and Thomsen 1998)............................................................................................ 116 Table 28. Costs of Offshore Wind Farms for Atlantic Coast Environmental Conditions. (Kilar and Stellar 1980) ................................................................................................................. 116 Table 29. Shallow Water COE Projections. (Musial and Butterfield 2004).............................. 117 Table 30. Wind Speeds Based on Class..................................................................................... 117

100

LIST OF FIGURES Figure

Page

Figure 1. a) Standard Monopile Structure, b) Supported Monopile Structure. (DNV-OS-J101 2004) ................................................................................................................................... 118 Figure 2. a) Tripod Structure, b) Gravity Pile Structure. (DNV-OS-J101 2004) ..................... 118 Figure 3. Lattice Tower. (DNV-OS-J101 2004)........................................................................ 119 Figure 4. Gravity Base Structure. (DNV-OS-J101 2004).......................................................... 119 Figure 5. Typical Relationship Between Ballast Component and Foundation Diameter. (Byrne and Houlsby 2002).............................................................................................................. 120 Figure 6. a) Suction Bucket Structure (DNV-OS-J101 2004), and b) Installation Principle. (Byrne and Houlsby 2003).................................................................................................. 120 Figure 7. Tension-Leg Platform. (DNV-OS-J101 2004) ........................................................... 121 Figure 8. Low-roll Floater. (DNV-OS-J101 2004).................................................................... 121 Figure 9. Ocean Sediment Distribution Throughout the Northern Hemisphere........................ 122 Figure 10. Scour Model. (Zaaijer 2002) .................................................................................... 123 Figure 11. Flowchart for Determining Scour Potential. (van der Tempel 2002)....................... 123 Figure 12. Spring Model of Pile-Soil Interaction. (Zaaijer 2002) ............................................. 124 Figure 13. Gravity-based Foundation Model. (Zaaijer 2002).................................................... 124 Figure 14. Example Yield Surface for Footings on Sand. (Byrne and Houlsby 2002) ............. 125 Figure 15. Comparison of a) Laboratory Test Data with b) Continuous Hyperplasticity Theory. ............................................................................................................................................. 125 Figure 16. Predicted 1st and 2nd Natural Frequency for Several Foundation Models. (Zaaijer 2002) ................................................................................................................................... 126 Figure 17. Loading Comparison of Offshore Platform to Offshore Turbine. (Schneider 2004) ............................................................................................................................................. 126 Figure 18. Multi-footing Suction Caisson Geometry. (Byrne and Houlsby 2003).................... 127 Figure 19. Foundation Size and Mass as a Function of Structural Configuration. (Byrne and Houlsby 2002)..................................................................................................................... 127 Figure 20. a) Design of Quadruped Suction Caisson, and b) Design of Monopod Suction Caisson Foundation. (Byrne & Houlsby 2003) ................................................................................ 128 Figure 21. Typical Foundation Installation Methods. (Watson 2000)....................................... 128 Figure 22. Deep Water COE Projections. (Musial and Butterfield 2004) ................................ 129 Figure 23. Expected Progression of Foundation Support Structures. (Musial and Butterfield 2004) ................................................................................................................................... 129 Figure 24. Deepwater Research and Development Strategy. (Musial and Butterfield 2004)... 130

101

Table 1. Technologies Involved in Offshore Structure Design. (Poulos 1988)

Table 2. Phases of Project Certification and Associated Activities. (DNV-OS-J101 2004)

102

Table 3. Design Standards for the Offshore Industry. STD. No. TITLE American Petroleum Industry (API) Recommended Practice for Planning, Designing and Constructing Fixed Offshore Platforms – RP-2A (WSD) Working Stress Design, 21st Edition 2000 Recommended Practice for Planning, Designing and Constructing Fixed Offshore Platforms – RP-2A (LRFD) Load and Resistance Factor Design, 1993, (suppl. 1997) Recommended Practice for Planning, Designing and Constructing Structures and Pipelines for RP-2N Arctic conditions, 2nd Edition 1995 Recommended Practice for Design and Installation of Electrical Systems for Offshore Det Norske Veritas (DNV) OS-J101 Design of Offshore Wind Turbine Structures, June 2004 OS-C101 Design of Offshore Steel Structures, General (LRFD Method), April 2004 OS-C201 Structural Design of Offshore Units (WSD Method), April 2002 OS-401 Fabrication and Testing of Offshore Structures, July 2004 RP-C203 Fatigue Strength Analysis of Offshore Steel Structures, October 2001 RP-B401 Cathodic Protection Design, January 2005 RP-C204 Design Against Accidental Loads, November 2004 RP-H102 Marine Operations during Removal of Offshore Structures, April 2004 CN-30.4 Foundations, February 1992 CN-30.5 Environmental Conditions and Environmental Loads, March 2000 CN-30.6 Structural Reliability Analysis of Marine Structures, July 1992 Rules for classification of fixed offshore installations, 1998 International Organization for Standardizations (ISO) 2394 General Principles on Reliability for Structures, 1998 Petroleum and natural gas industries -- Offshore structures -- Part 1: General requirements, 199513819-1 12, 1st edition. To be replaced , ISO TC 67. (ISO 19900) 13819-2

Petroleum and Natural Gas Industries – Offshore Structures – Part 2: Fixed steel structures, 1995

Petroleum and Natural Gas Industries – Specific Requirements for Offshore Structures – Part 1: Metocean design and operating conditions, 2003 19902 Piled Foundations: Fixed Steel Structures 19903 Gravity Foudnations: Fixed Concrete Structures 19904 Offshore Structures: Floating systems 3010 Basis for Design of Structures – Seismic actions on structures, 2001 Germanischer Lloyd (GL) Rules for Classification and Construction, III Offshore Technology, 2 Offshore Installations, Edition 1999 Rules for Classification and Construction, III Offshore Technology, 2 Offshore Installations, Guidelines for the Construction/Certification of Floating Production, Storage and Off-Loading Units, Edition 1999 International Electrotechnical Commission (IEC) 61892-3 Mobile and fixed offshore units - Electrical installations - Part 3: Equipment, (1999-02) 61892-6 Mobile and fixed offshore units - Electrical installations - Part 6: Installation,(1999-02) German Institute for Standardization (DIN EN) 12495 Cathodic protection for fixed steel offshore structures, 2000 10225 Weldable structural steels for fixed steel offshore structures, 1994 Norwegian Technology Center (NTC) NORSOK N-001 Structural Design, Rev. 3, Aug. 2000 Department of Energy (DOE) Offshore installations: guidance on design, construction and certification (fourth edition) HMSO 1990 ISBN 011 4129614 19901

103

Table 4. Worldwide Offshore Wind Energy Potential. (Gaudiosi 1996) Usable Wind Speed Offshore Area (m/s) (sq km) 9000 5.6-6.4 21500 5.6-6.4 162000 5.6-8.8 21000 5.6-7.0 5600 5.6-6.4 48000 5.6-8.8 1516 5.6-6.0 2278 5.6-6.0 8680 NA 1200 NA

Country/Region US China European Union Denmark The Netherlands UK Tunisia Egyptian Mediterranean Mediterranean Sea Black Sea

Power (GW)

Energy (TWh/yr)

54 129 NA NA NA NA 9 13 52 7

102 254 1853 260 75 626 16 24 100 14

Table 5. Operational Offshore Wind Farm Projects. Installed No. Type of Turbine Capacity Turbines (MW) Denmark Middelgrunden NA 2001 40 20 Bonus 2.0 MW Denmark Vindeby Baltic Sea 1991 4.95 11 Bonus 450 kW Denmark Tuno Knob Kattegat Sea 1995 5 10 Vestas 500 kW Denmark Horns Rev North Sea 2002 160 80 Vestas V80 2.0 MW Denmark Frederikshavn NA 2002 NA 4 NA Sweden Gotland-Bockstigen Baltic Sea 1997 2.75 5 Wind World 550 kW Sweden Utgrunden Baltic Sea 2001 10.5 7 Tacke TW1.5s Sweden Yttre Stengrund NA 2001 10 5 NEG Micon 2.0 MW The Netherlands Medemblik (Lely) Ijsselmeer Sea 1994 2 4 NedWind 500 kW The Netherlands Dronten Ijsselmeer Sea 1996 16.8 28 Nordtank 600 kW UK Blyth Offshore North Sea 2000 3.8 2 Vestas V66 2.0 MW Sources: Shaw et. Al. (2002), Beurskens and Jensen (2000), Gaudiosi (1996), Sahin (2004), Henderson et al. (2002), Manwell et al (2002) Country

Name of Park

Sea Location

Year Commisioned

Rotor Diameter (m) NA 35 39 NA NA NA NA NA 41 NA 43

Production (GWh/yr) 81 11.73 12.7 NA NA NA NA NA 3.5 NA NA

Distance Water from Coast Depth (m) (km) 1.7-3.5 3-5 1.5-3.0 3-5 6 3-5 14-20 NA NA NA 4.5 5.5-6.5 12.5 7.2-9.8 5 6-10 2 5-10 0.2 5 0.8 6-11

Foundation Details Gravity Foundation Box Caissons on Sandy Soil Box Caissons on Sandy Soil NA NA Drilled and Grouted Monopiles Monopiles Monopiles Steel Tower Driven in Sandy Soil Monopiles in Fresh Water Monopiles in bedrock

Table 6. European Offshore Energy Potential by Country. (Sahin 2004) Country UK Denmark France Germany Ireland Italy Spain The Netherlands Turkey Greece Portugal Belgium Total

104

Offshore Potential (TWh/yr) 986 550 477 237 183 154 140 136 130 92 49 24 3158

Table 7. Environmental Parameters for Coastal United States. Area

Mean Wind Speed (m/s)

Northeast Coast 4-8 Southeast Coast 4-8 Gulf Coast 4-7 Southwest Coast 3.5-5.5 West Coast 4-8 S. Alaskan Coast 5-9.5 Aleutian Islands 6-9 Hawaiian Islands 4-6 Puerto Rico and Virgin Islands 4-5 Notes: *Based on 100-year storm recurrence Source: Kilar and Stiller 1980

Max. Sustained Max. Wave Max. Surface Wind Speed* Height* Current (m/s) (m/s) (m) 45-65 30-50 0.2-1.1 55-85 30-50 0.25-1.8 45-70 30-40 0.2-0.9 35-45 20-30 0.15-1.0 35-50 30-35 0.2-0.7 50-60 35-40 0.2-0.8 50-60 30-40 0.15-0.5 40-50 20-35 0.2-0.7 NA NA 0.25-0.4

Table 8. Offshore Resource Estimates for U.S. (Musial and Butterfield 2004)

105

Table 9. Planned Offshore Wind Farms. Country

Name of Park

Sea Location

Year Commisioned

Installed Capacity (MW)

No. Turbines

Belgium Thorton Bank North Sea End 2005 NA 60-120 Belgium Vlakte van de Raan North Sea 2003 100 50 Belgium Wenduine NA 2003/2004 100 NA Belgium Raan NA 2003/2004 100 Denmark Roedsand Baltic Sea 2003 165.6 72 Denmark Omo Stalgrunde NA 2004 150 96 Denmark Samso NA 2003 23 10 Denmark Laeso NA 2003 150 78 Denmark Adlergrund NA 2004 480 Denmark Beltsee NA 2005 249 France Nord-pas-de-Calais NA NA 775 NA France Manche, Basse-Normandie NA NA 3500 NA France Bretagne NA NA 2050 NA France Languedoc-Roussilon NA NA 2800 NA Germany Pommersche Bucht Baltic Sea NA 1000 200 Germany Arkona Becken Sudost Baltic Sea NA 945 189 Germany Adlergrund Baltic Sea NA 790 158 Germany Kriegers Flak Baltic Sea NA 315 75 Germany Pilot Mecklenberg-Vorpommern Baltic Sea NA 40 21 Germany Beltsee Baltic Sea NA 415 83 Germany Sky 2000 Baltic Sea NA 100 50 Germany Dan Tysk North Sea NA 1500 300 Germany Weisse Bank North Sea NA 600 120-170 Germany Butendiek North Sea NA 240 80 Germany Offshore Helgoland North Sea NA 200 100 Germany Schleswig Holstein Nordsee North Sea NA 800-1000 200 Germany Amrumbank West North Sea NA up to 288 72 Germany Amrumbank/Nordsee-Ost AWZ North Sea NA 1250 250 Germany Meerwind North Sea NA 819 234 Germany Nordergrunde North Sea NA >200 76 Germany Wilhemshaven North Sea NA 4.5 1 Germany Dollart North Sea NA 9 5 Germany Juist North Sea NA 1400 280 Germany Borkum III North Sea NA 60 (later 1000) 12 (later up to 285) Germany Borkum Riffgrund North Sea NA 840 180 Germany Borkum IV North Sea NA 400 90-160 Germany Riffgat North Sea NA 135 30 Germany Borkum Riffgrund West North Sea NA 1800 458 Germany Sandbank 24 (Europe Pipe West) North Sea NA 2600 (1st phase 360) 120 Ireland Blackwater Bank 1 NA NA 260 NA Ireland Blackwater bank 2 NA NA NA NA Ireland Arklow Bank NA 2003-2006 (phased) 520 200 Ireland Codling and Greater Codling Bank NA NA NA NA Ireland Bray Bank and Kish Bank NA 2003 up to 300 MW NA Ireland Dundalk Bay NA NA NA NA Sweden Nogersund NA 1991 0.22 NA Sweden Rone Gotland NA NA 35 35 Sweden Klasarden NA 2003 42 21 Sweden Blekinge Oland Southern Skane NA NA 30 30 Sweden Blekinge Oland NA 1998 294 98 Sweden Ystad Skane NA NA 10 NA Sweden Lillgrund Bank NA 2002 86.4 48 Sweden Barsebank NA NA 750 NA The Netherlands Q7 North Sea 2003 120 60 The Netherlands Egmond aan Zee North Sea 2004 100 36 UK Kentish Flats NA NA NA 30 UK Gunfleet Sands NA NA 29.8 NA UK Scroby Sands NA NA 50 25 UK Cromer NA NA NA 30 UK Lynn NA NA NA 60 UK Inner Dowsling NA NA NA 60 UK Teesside NA NA NA 30 UK Solway Firth NA NA NA 60 UK Barrow NA NA NA 30 UK Shell Flat NA NA NA 90 UK Southport NA NA NA 30 UK Burbo NA NA NA 30 UK North Hoyle NA NA NA 60 UK Rhyl Flats NA NA NA 60 UK Scarweather Sands NA NA NA 30 UK NA NA 1987 20 50 Poland Bialogora NA 2004 61 NA Spain Cabo de Trafalgar NA 2005 200 100 Canada Nai Kun NA 2004/2008 700 NA USA Nantucket Sound, MA Nantucket Sound NA 425 NA Sources: Shaw et. Al. (2002), Beurskens and Jensen (2000), Gaudiosi (1996), Sahin (2004), Henderson et al. (2002), Manwell et al (2002)

106

Type of Turbine

Production (GWh/yr)

2.5-3.6 MW Vestas V80 2.0 MW NA

1000 NA NA

Distance from Coast (km) NA 12 NA

Bonus 2.3 MW NA Bonus 2.3 MW NA

480 NA NA NA

9-10 10 NA 40

NA NA NA NA 5 MW 4-5 MW 3-5 MW 3-5 MW Nordex, Neptun, Brand Elektro 3-5 MW 2 MW (1/3 Vestas, rest open) 5 MW 3.5-5 MW 3 MW 2 MW (Vestas) 4-5 MW 3-4 MW 5 MW (96x Repower NOK 5) 3.5 MW 2.5-5 MW 4.5 MW (Enercon E-112) 1.8 MW (Enercon) 5 MW 3.5-5 MW 3.5 MW 2.5-4.5 MW 4.5 MW (Enercon) 2.5-5 MW 3 MW NA NA NA NA NA NA NA Nordic 1 MW NEG Micon 2.0 MW 1 MW Nordic 3 MW NA Enercon E66/18.70 NA Vestas V80 2.0 MW NEG Micon 2.75 MW NA NA Vestas V80 2.0 MW NA NA NA NA NA NA NA NA NA NA NA NA 400 kW NA 2 MW NA NA

2400 10800 6300 10600 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 833 NA NA NA 350 NA NA NA 98 NA NA NA NA NA NA NA NA NA NA NA NA 40 NA NA NA NA

5-8 5-10 3-10 3.5-10 42 NA 40 35 15 30 19 60 60 30 13 15 35 17 NA 12-15 <10 NA NA 40 34 NA 15 45 120 NA NA 7-12 NA NA NA 0.25 NA NA NA 5 NA NA NA 23 NA 8 7 2.3 6.5 5.2 5.2 1.5 9 10 7 10 5.2 6 8 9.5 1-10 NA NA NA NA

Table 10. Characteristic Load Selection Basis. (DNV-OS-J101 2004)

Table 11. Load Factors for ULS and ALS. (DNV-OS-J101 2004)

107

Table 12. Application of Site Condition Data for Project Phases of an Offshore Wind Farm. (Jenner et. al. 2002)

108

Table 13. Offshore Site Investigation Planning Process. (Randolph and Kenkhuis 2001) Phase

3 - 6 Month Time Period

Activity

Pre-planning Period Work scope/budget definition Budget approval Pre-qualification Period Pre-qualify contractors Tenders list approval Tender Period Invitation to tender Assess tenders Approval to award Award tender Construction Period Contractor mobilization Work initiation

4-6 Weeks X X

3-5 Weeks

4-6 Weeks

6-8 Weeks

X X X X X X X X X

Table 14. Typical Offshore Site Investigation and Analysis Schedule. (Poulos 1988)

109

Table 15. Seismic Profiling Methods. (Randolph and Kenkhuis 2001)

Table 16. Accuracy Classes for Cone Penetration Testing. (Norsok 2004)

Table 17. Field Vane Blade Size. (Norsok 2004)

110

Table 18. Suitability of Test Methods for Soil Parameters. (OSIC 2004)

111

Table 19. Other Soil Parameters for Specific Applications. (OSIC 2004)

Table 20. Evaluation of Sample Quality. (Norsok 2004)

112

Table 21. Model Parameters for Gravity-based Foundations. (Zaaijer 2002)

Table 22. Estimations of Effective Fixity Length. (Zaaijer 2002)

113

Table 23. Important Natural Frequency Sensitivities for Various Structural Configurations. (Zaaijer 2002)

Structure

Parameter

Uncertainty

Location

Lifetime

% 4 0.06 0.02 % 19 0.01 0.2 % 0.9 <0.01 0.1 % 3 0.01 0.04 % 23 0.01 0.04

% 4 0.1 4 % 35 0.03 4 % 0.9 0.01 3.2 % 3 0.04 4 % 38 0.02 4

% 6 0.1 4 % 4 0.03 4 % 0.7 0.01 3.2 % 3 0.04 4 % 7.8 0.02 4

Tubular Tower on Monopile Soil Foundation Environment Structure Tubular Tower on Gravity Base Foundation Soil Foundation Environment Structure Tripod with Piles Soil Foundation Environment Structure Lattice Tower with Piles Soil Foundation Environment Structure Lattice Tower on Gravity Base Foundation Soil Foundation Environment Structure

114

Table 24. Typical Sizes and Construction Aspects. (Watson 2000)

Table 25. Foundation Construction Constraints. (Watson 2000)

Table 26. Typical Installation Vessel Specifications. (Watson 2000)

115

Table 27. Relative Cost Distribution for Onshore vs. Offshore Wind Farm Comparison. (Fuglsang and Thomsen 1998)

Table 28. Costs of Offshore Wind Farms for Atlantic Coast Environmental Conditions. (Kilar and Stellar 1980) Water Depth Mean Wind Speed @ 10-m a.s.l. (m/s) Plant Cost (1979 $M)

Shallow Water (0 to 15-m)

Deep Water (60 to 500-m)

5

7.5

10

5

7.5

10

7.3

5.7

5.7

7.5

7.7

5.7

116

Table 29. Shallow Water COE Projections. (Musial and Butterfield 2004)

Table 30. Wind Speeds Based on Class.

117

a)

b)

Figure 1. a) Standard Monopile Structure, b) Supported Monopile Structure. (DNV-OSJ101 2004)

a)

b)

Figure 2. a) Tripod Structure, b) Gravity Pile Structure. (DNV-OS-J101 2004)

118

Figure 3. Lattice Tower. (DNV-OS-J101 2004)

Figure 4. Gravity Base Structure. (DNV-OS-J101 2004)

119

Figure 5. Typical Relationship Between Ballast Component and Foundation Diameter. (Byrne and Houlsby 2002)

a)

b)

Figure 6. a) Suction Bucket Structure (DNV-OS-J101 2004), and b) Installation Principle. (Byrne and Houlsby 2003)

120

Figure 7. Tension-Leg Platform. (DNV-OS-J101 2004)

Figure 8. Low-roll Floater. (DNV-OS-J101 2004)

121

USA

Figure 9. Ocean Sediment Distribution Throughout the Northern Hemisphere. (Poulos 1988)

122

Figure 10. Scour Model. (Zaaijer 2002)

Figure 11. Flowchart for Determining Scour Potential. (van der Tempel 2002)

123

Figure 12. Spring Model of Pile-Soil Interaction. (Zaaijer 2002)

Figure 13. Gravity-based Foundation Model. (Zaaijer 2002)

124

Figure 14. Example Yield Surface for Footings on Sand. (Byrne and Houlsby 2002)

a)

b)

Figure 15. Comparison of a) Laboratory Test Data with b) Continuous Hyperplasticity Theory. (Byrne & Houlsby 2003)

125

Figure 16. Predicted 1st and 2nd Natural Frequency for Several Foundation Models. (Zaaijer 2002)

Figure 17. Loading Comparison of Offshore Platform to Offshore Turbine. (Schneider 2004)

126

Figure 18. Multi-footing Suction Caisson Geometry. (Byrne and Houlsby 2003)

Figure 19. Foundation Size and Mass as a Function of Structural Configuration. (Byrne and Houlsby 2002)

127

a)

b)

Figure 20. a) Design of Quadruped Suction Caisson, and b) Design of Monopod Suction Caisson Foundation. (Byrne & Houlsby 2003)

Figure 21. Typical Foundation Installation Methods. (Watson 2000)

128

Figure 22. Deep Water COE Projections. (Musial and Butterfield 2004)

Deep Water Wind Turbine Development

Current Technology Onshore Wind Turbine

Transitional Depths 30 M – 50 M

Shallow Water 0 M – 30 M

Deep Water 50 M – 200 M

Figure 23. Expected Progression of Foundation Support Structures. (Musial and Butterfield 2004)

129

Figure 24. Deepwater Research and Development Strategy. (Musial and Butterfield 2004)

130