Seawater-Based Hydraulics for Offshore Wind Turbines N.F.B. Diepeveen (TU-Delft)

(We@Sea project 2004-012)

WE@Sea Progress Report

Seawater-Based Hydraulics for Offshore Wind Turbines

August 2009

N.F.B. Diepeveen

Preface Purpose of This Document The purpose of this document is to report on the research which has done by the author as part of the We@Sea program. It covers the period from August 15th 2008 to August 15th 2009. This period coincides with the first year of PhD research performed by the author. During the first three months, the plan for the Delft Offshore Turbines (DOT) project was constructed. It was written to lay the foundation for several PhD topics. The research topic of the author concentrates on the hydraulic energy transmission using seawater. This has so far resulted in two conference papers, one focused on hydraulic transmission in wind turbines, the other specifically on the pump requirements.

Outline of This Document This document is contains the following four parts: - Executive Summary An extensive summary is given of the set up of the DOT project and the related research activities in the first year. Preliminary conclusions and an outline of the future research efforts are also provided. 1. DOT Project Plan Author: ir. N.F.B. Diepeveen, Supervisor: dr.ir. J. van der Tempel 2. Closed-Loop Fluid Pumping as a Means to Transfer Wind Energy Conference paper for the European Wind Energy Conference 2009, submitted on March 16th 2009. Author: ir. N.F.B. Diepeveen. 3. Pump Design Requirements for Seawater-Based Hydraulic Power Transmission for Offshore Wind Turbines Conference paper for the European Offshore Wind (EOW) Conference 2009, submitted on September 14th 2009. Author: ir. N.F.B. Diepeveen.

i

ii

Seawater-Based Hydraulics for Offshore Wind Turbines

WE@Sea Progress Report

Executive Summary Introduction Delft University of Technology is taking a radical step away from incremental development of offshore wind turbines. It has started a research project on using a 2 bladed, fixed pitch turbine (5 − 10M W ) to directly drive a water pump in the nacelle. By channeling the pressurized water of all turbines to one transformer platform, electricity generation is centralized. The design goal is to reduce the number of components in the offshore turbines drastically to come to the ultimate offshore turbine. Current offshore wind turbines are marinized land turbines with only a few add-on features to keep out the salty air. Improvements of turbine technology are only incremental and do not take full benefit of the offshore environment. The Delft University of Technology has a history in offshore wind research of over 25 years and has formulated a radical concept change of offshore wind energy conversion that helps develop a completely new system and spark revolutionary developments on sub-system and component level. Typically, offshore wind farms have a generator platform that gathers all electricity of the different turbines, steps up the voltage and feeds the power through shore connection cables to the onshore grid. The DOT takes boundary conditions from this existing configuration: horizontal axis turbine with blades and a platform where the combined electrical power is fed to the onshore grid. Everything in between can be changed. The DOTs focuses on radical technology changes. To facilitate this, a short list of design pointers has been defined to test all developments against and to keep as life line throughout the project execution. Offshore, one thing is abundant: water. The current turbine technology sees the nacelle weight increase steadily giving increasing challenges in support structure design and installation. Furthermore, power electronics help harness wind power slightly more efficiently, but also add weight and components (that can fail) to the turbine system. Offshore wind energy has high potential. Currently the price for placing turbines offshore is too high. Projects are not yet economically feasible without government subsidies. The overall goal of the Delft Offshore Turbines project is therefore to design a wind turbine infrastructure specifically for offshore purposes and thereby rendering offshore wind energy more economically attractive. This translates to a design goal of the overall project which is to reduce the number of components in the offshore turbines drastically to come to the ultimate offshore turbine, which is characterized by: - Very low maintenance - High availability; as a direct consequence of high reliability. - High efficiency iii

iv - Easy installation - Low production costs The Delft Offshore Turbines (DOTs) project aims to circumvent the need of the generator by using the rotor shaft torque to power a pump in the nacelle. This, along with the proposal to have a two-bladed rotor, will lead to a significant reduction in weight of the rotor-nacelle assembly. One of the fundamental parts of the DOT is the hydraulic transmission. The PhD research project in this report focuses on the design of a seawater-based high pressure hydraulic energy transmission system, from the rotor shaft to the generator platform. This PhD project was set up as follows. First a plan for the overall DOT project was drafted (part I) to lay the foundation on the basis of which up to 8 PhD students can select research topics. The next step was to select one of these topic for my own research project. Before any form of design could start a study was performed to look at the possibilities of using fluid power for wind energy transmission. This was the subject of the European Wind Energy Conference (EWEC) 2009 paper (part II). The purpose was to gain insight in how fluid power circuits operate. This meant mapping which type of systems exist, which is the most efficient and why and what the key performance indicators are. Gain insight in the applications and the potential of fluid power circuits, i.e. what has already been done with fluid power in similar applications, in general and hydraulic wind turbines in particular. The next step was to perform a research based selection of the pump type and a first look at seawater as hydraulic fluid. This was the subject of the European Offshore Wind (EOW) Conference 2009 paper (part III). The goal was to select the best suitable pumping principles and investigate their commercial availability. Next to that insight was gained into which challenges arise from using seawater as hydraulic fluid.

Fluid Power Applications High pressure fluid power has been applied for many years in many industries. The number of applications continues to grow. One of the earliest large scale projects were the Victorian age tap-water hydraulics. Pumping stations outside the center of cities like London, New York and Melbourne delivered water pressurized up to 60 bar through underground mains to power facilities like elevators, cranes and even theater curtains. Nowadays, fluid power is used in shredders, feeders, roll mills, cranes, bulldozers, jack-up systems, etcetera. These applications use electricity to efficiently acquire power in the form of high torque through high pressure fluid transmission. The DOT energy transmission concept is the exact opposite. High torque is converted into a high pressure flow. Classification of pumps Pumps can be divided in two general categories: kinetic (or hydrodynamic) and positive displacement pumps. In hydrodynamic pumps such as centrifugal pumps, the flow is continuous Seawater-Based Hydraulics for Offshore Wind Turbines

WE@Sea Progress Report

v

from inlet to outlet and results from kinetic impulse given to the fluid stream. The output is characterized by low pressure and high volume. Inefficiency and easy stalling as a result of back-pressure make these pumps unsuitable for control. In positive displacement pumps, fluid flows through an inlet into a chamber. As the pump shaft rotates, the (positive or definite) volume of fluid is sealed from the inlet and transported to the outlet where it is subsequently discharged. The essential difference between these two main categories is that kinetic pumps are for fluid transport systems and PD drive systems are for fluid power systems. The power-to-weight ratio of pd pumps is much higher than that of the generators used in wind turbines. This is without taking into account all the extra components required for the use of an electricity generator. Hydraulic Fluids Seawater is preferable in to hydraulic oil in terms of dynamic performance. This is due to the higher bulk modulus of seawater. Having an open-loop system means that the temperature of the water will remain well within its liquid range. However, low viscosity of seawater also means poor lubricity and high potential of wear due to erosion and corrosion.

Hydraulic Turbines Hydraulic turbines are not new. Using seawater as hydraulic fluid is. Currently ChapDrive AS in Norway, Artemis in Scotland and Voith Turbo (WinDrive) in Germany are all developing hydraulic gears for wind turbines. The most similar to the DOTs project is the ChapDrive, having its generator placed at the foot of the turbine tower. These systems all use hydraulic oil as medium.

DOT Concepts The Delft Offshore Turbines (DOTs) convert wind energy into a high pressure flow of water. A pump is connected to the rotor directly, generating a high pressure flow. The pressurized water is collected at a transformer platform, where generators are located comparable to a hydro plant. The platform can be fitted with limited water storage/accumulation capacity to smoothen energy variations. From this platform, an electricity cable connects to the onshore grid. So far, two concepts have been derived: 1. closed-loop + open-loop A high pressure fluid power circuit forms a closed loop between the pump connected to the rotor shaft in the nacelle and a motor just below sea-level. The motor is connected to a second pump which pumps seawater to the central generator platform. Having a closed-loop systems requires subsystems for cooling & pressurizing. This adds significantly to the total number of components. 2. open-loop The pump connected to the rotor shaft generates a high pressure flow. At the base, a WE@Sea Progress Report

vi small portion of this pressure is used to power a booster pump which ensures the flow of sea water to the pump in the nacelle. The rest of the pressure is used to generate electricity at the central platform.

Selection of Pumping Principle Candidate pump types for the Delft Offshore Turbines are the vane pump and the radial piston pump. Vane pumps can cope with low viscosity fluids like water but are limited in terms of pressure (< 100bar). The radial piston pump can generate high pressure (> 500bar) and can be designed to operate efficiently (> 95%) at rotation rates matching those of the wind turbines. However, in particular the clearance between the piston and its casing is a concern when using seawater. Corrosion and erosion of a pump will lead to a rapid decline in efficiency. A solution must be found to prevent these phenomena from occurring.

Conclusion The main challenge for the DOT hydraulic energy transmission is to have a robust, yet efficient system. Hydraulic drive systems have been applied in many industries for many years. High performance systems are characterized by high efficiencies and low maintenance needs. The power production of a Delft Offshore Turbine can increase beyond rated due to the characteristics of hydraulic drive systems. The cut-out wind speed is determined by the control characteristics of the rotor (blades). The current maximum size of suitable pumps is under 2 MW and requires the use of hydraulic oils as power fluid. Larger systems are technically feasible. They have not yet been produced due to lack in demand. Despite the additional mass of the hoses and power fluid, having only a pump in the nacelle significantly reduces the total mass of the wind turbine. For concept evaluation purposes, further research needs to be done on the reduction of the nacelle mass and subsequently the support structure mass. A 5M W turbine will require a volume flow of close to 10, 000 liters per minute. Since the idea behind the DOTs project is to design a turbine specifically for offshore purposes, oil will not be the preferred power fluid in the long run. Instead we aim to use the offshore environment to our advantage and use seawater. The research presented in part III shows that multi-MW high pressure seawater pumps do not yet exist. The main reason for this is that there has never been a real need for them. To make the DOTs a reality, such a pump will need to be designed. The main challenge is how to design for the use of seawater as hydraulic fluid. For this the radial piston pump appears to be the best suited. A pump with rated capacity of 5MW is not yet commercially available, let alone one capable of pumping seawater for long periods of time without maintenance. In terms of power production, the H¨ agglunds CBP 210, with a little over 2.3M W rated power at 350bar pressure is the state of the art. More powerful systems do already exist in the form of prototypes. However, all require hydraulic Seawater-Based Hydraulics for Offshore Wind Turbines

WE@Sea Progress Report

vii

fluids with high viscosity. The conclusion therefore is that a specific system for high pressure 5M W seawater pumping will have to be designed. Important design considerations are: - The H¨ agglunds CBP 210 (2.3M W ) operates with efficiency of 9096%. In these systems, the larger the piston, the higher the efficiency. Therefore ever more efficiency can be expected of a 5M W pump. - The low viscosity of seawater will lead to an increase leakage flow, reducing the efficiently slightly. - The high bulk modulus of seawater should significantly benefit the overall efficiency. - The pump has to operate efficiently also at very low rpm. Trends in radial piston pumps and commercial wind turbines indicate that designing for matching rpms is feasible. One of the first steps in the design process should be to find a solution for the poor lubricity and the corrosive and erosive characteristics of seawater. The logical first step is to look for a structural material that is sufficiently damage resistant for the fluid carrying parts of the pump. Finding this material is potentially crucial to the development of the project. The next phase of the project will be focused on finding a suitable structural materials for the high pressure seawater pump.

WE@Sea Progress Report

viii

Seawater-Based Hydraulics for Offshore Wind Turbines

WE@Sea Progress Report

Part I

DOTs Project Plan

roman

Contents 1 Introduction 1.1 The Future of Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 1.3

Current Wind Turbine Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . The Proposed Idea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 4

2 Design Rules

5

3 Design Options

7

4 How DOTs Work

9

4.1 4.2

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Power Transmission: Wind to Water to Electric . . . . . . . . . . . . . . . . . . . .

9 10

4.2.1 4.2.2

A: The Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B: The Closed-Loop System . . . . . . . . . . . . . . . . . . . . . . . . . . .

10 11

4.2.3 4.2.4

C: The Open-Loop System . . . . . . . . . . . . . . . . . . . . . . . . . . . D: The Generator Platform . . . . . . . . . . . . . . . . . . . . . . . . . . .

14 14

Simulation Results for Varying Wind Speeds . . . . . . . . . . . . . . . . . . . . .

15

5 Plan of Execution 5.1 DUWIND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

19 19

4.3

5.2 5.3

DOTs Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Time Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

19 20

5.4

Potential PhD Research Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . .

21

References

23

i

Contents

DOT Project Plan

ii

1 Introduction 1.1

The Future of Energy

Europe is becoming increasingly dependent on imported energy. The current level of 50% is expected to rise to 70% by 2030 by the European Wind Energy Association (EWEA) [2]. Meanwhile, energy prices are soaring and a future energy crisis is looming. Nearly all imported energy is derived from the burning of imported fossil fuels, releasing vast amounts of carbon-dioxide (CO2 ) into the atmosphere. The global consensus is that CO2 emissions lead to global warming. Of the energy consumed in Europe, only about 9% comes from renewable sources [1]. To become more independent and reduce CO2 emissions it is necessary for Europe and to produce more and cleaner energy. An increasingly popular, clean and renewable energy source is wind, and in particular offshore wind. Targets In 2007 the European Union and the EWEA set the following targets. Of the total energy consumed in Europe in 2020, 20% should come from renewable sources (EU). The goal is to let 5% of this renewable energy come from offshore wind farms. According to the EWEA offshore wind has the potential to deliver up to 25% by 2020. This translates to an estimated total target of 40GW from offshore wind farms. Assuming one wind turbine produces an average of 4M W , a total of 10,000 structures is therefore required. So far, around 530 offshore wind turbines have been placed, mainly around Denmark and Great Britain. In order to meet these requirements, the completion of these 10,000 structures needs to be realized within 10 years. Compare this to the oil and gas industry where worldwide 7,000 offshore structures were built in 70 years [3]. Challenges To achieve these targets many challenges have to be overcome. The essential technical design challenge is to improve performance, particularly in terms of reliability. Other design considerations include, - Wind energy economics. To make wind energy (more) profitable, the high costs of fabrication and installation must be reduced. - Impact on the local environment. The long term effects of offshore wind turbines on local flora and fauna are not yet known. A popular theory is that they can function as an artificial 1

Introduction

Current Wind Turbine Technology

Figure 1.1: Operational and under construction offshore wind farms around Europe. reefs, supporting marine life. However, for birds and their flight paths, offshore wind farms (OWFs) may be a nuisance. - Competition for space with other marine users. A popular location to place OWFs is on sand banks. Unfortunately, these relatively shallow sites are often also popular with fishermen. Once a site is officially assigned to a OWF developer, it becomes a restricted area and can thus not be accessed anymore for fishing.1 Other common objections to OWFs are the fear of obstruction of seaways and possible interference with surface radar. - Installation logistics. Currently there is a shortage of vessels which can be used for OWF installation. Hence, there is a need to optimize the use of cranes and other offshore construction vessels. - Compatibility with the European grid infrastructure.

1.2

Current Wind Turbine Technology

The conventional model for wind turbines both onshore and offshore is to install a generator in the nacelle. Supplying each turbine with its own generator has several disadvantages. 1 Organisations such as Greenpeace actually promote OWFs to create safe-havens for marine life and combat overfishing

DOT Project Plan

2

Introduction

Current Wind Turbine Technology

1. It demands large amounts of copper, making wind farms expensive. 2. The nacelle becomes heavy and thus requires a strong support structure. 3. Continuous efficient conversion from kinetic to electric energy requires huge amounts of switch gear. This severely complicates the installation and maintenance. 4. Many components are leading to many failures. To make wind energy truly beneficial requires more than incremental improvements.

Figure 1.2: Layout of the nacelle of the Vestas V90 wind turbine [8]

DOT Project Plan

3

Introduction

1.3

The Proposed Idea

The Proposed Idea

The idea proposed in this project plan is the following. Picture a wind turbine in an offshore wind farm. We take practically everything out of the nacelle. All that remains is the idea of rotating kinetic energy. This kinetic energy is at some stage transformed into electric energy. The whole process in between is for us to design.

Figure 1.3: Offshore wind farm at Egmond aan Zee, 10 km from the Dutch coast.

DOT Project Plan

4

2 Design Rules The main idea behind the DOTs is to get more energy from wind in a less labor-intensive manner. To realize this, the focus of the design process will not only be on turbine energy production, but also on important economical and practical issues such as the optimization of production, assembly and maintenance. The DOTs project aims to bring about a radical change in (offshore) wind turbine technology. It will not be performed for standard incremental improvements. However, research will not be limited to one specific grand design. Many of the developed concepts will also be applicable to current wind turbine technologies. For instance, one likely target for research is the concept of boundary layer suction along the rotor blades to reduce drag. This research will be part of the DOTs project, but the results can just as well be applied to other (common) wind turbines. Already in this early phase a number of initial design rules have been determined. 1. The wind turbine will have two blades. The current standard for onshore wind turbines is to have three blades. The main reason for this is that human spectators experience three blades as the more tranquil view. Offshore, where spectators do not have to be taken into account, the advantages of having two blades can be exploited. Such advantages are: - lower production costs - easier to assemble - higher rotational speed 2. The blades will have a fixed pitch. This simplifies the design of the rotor significantly. Theoretically, pitch regulation enables higher turbine efficiency at varying wind speeds. However, advanced aerodynamic profiling of the blades will be applied to minimize the difference. 3. The transport of energy from the nacelle to the base will be done using a closed-loop system of pumps and pipelines. 4. A single pipe will transport seawater from each turbine base to the generator platform. This makes the wind farm an open-loop system. 5. The conversion from kinetic to electric energy will happen on a central platform. Hence the cable to the shore can be plugged in directly to the power grid. 6. As few components as possible will be used to simplify the assembly. Figure 2.1 demonstrates how these rules can be translated to an initial design concept.

5

Design Rules

Figure 2.1: A functional flow diagram of a DOT proposal

DOT Project Plan

6

3 Design Options The design rules of chapter 2 form a platform from which to start the actual design process. During this process, many important choices will be made, such as: Turbine design - The actual size of the rotor. - Placing the rotor up- or downwind. - How to optimize flow around the blades with a minimum of moving parts. Support structure - Using a truss or a monopile as foundation. - If and how to apply the slip joint connections. - How to maximize the ease of installation. Pumping system/energy transmission - What kind of pumps to use. - If and how to install a superconductor pipeline. - The design of the transformer station. General design choices - If and how to include an option to store energy. - Whether to design the system as self-installing. - Which materials to use. - How to deal with marine growth. - How to minimize and ease maintenance requirements.

7

Design Options

DOT Project Plan

8

4 How DOTs Work 4.1

Introduction

In this chapter the theory behind the DOTs is presented. The basis for the theory is the initial concept design in figure 2.1. To review this design piece by piece, the entire system is split-up into the following subsystems: A The rotor, which turns as a result of wind blowing. B The closed-loop system, which transfers power from the rotor shaft in the nacelle, to the base of the structure. Its main components are: • pump A, which is directly driven by the rotor shaft. Its function is to induce a flow in pipe 1.

• pipe 1, where fluid flows from the nacelle to the base. • motor B extracts mechanical power from the pressure flow in pipe 1 at the base of the structure.

Figure 4.1: Identification of the subsystems of the DOT concept from figure 2.1 9

How DOTs Work

Power Transmission: Wind to Water to Electric

C The open-loop system, where seawater is pumped from the base of a structure to the generator platform through pipe 2. Pump C is directly powered by motor B. D The generator platform, where the seawater flow from all pipes 2 is collected and used to generate electricity. An overview of these subsystems is given in figure 4.1. Section 4.2 presents the theoretical power transmission throughout the entire system. The initial simulation results are presented in section 4.3. The (potential) initial conditions of this simulation are defined in table 4.2.

4.2

Power Transmission: Wind to Water to Electric

4.2.1

A: The Rotor

Figure 4.2: Subsystem A: the rotor Conventional turbines have three blades with variable pitch. The rotor of a Delft Offshore Turbine (DOT) will have two blades with fixed pitch. The power Protor extracted from the wind is a function of the wind speed v, the rotor radius r, the air density ρair and the induction factor cp . Protor = cp ·

1 · ρair · π · r2 · v 3 2

(4.1)

The induction factor cp is in fact a pressure coefficient. Its theoretical maximum value is 0.59. This is known as the Betz Limit [6]. The latest designs of turbine blades have inductions factors larger than 0.50. As part of the DOT project, research will be done on the application of boundary layer suction, to further increase the cp . The rotation of the rotor gives the rotor shaft a torque T at a rotation rate ω. Protor = T · ω DOT Project Plan

(4.2) 10

How DOTs Work

4.2.2

Power Transmission: Wind to Water to Electric

B: The Closed-Loop System

(a) layout sketch

(b) functional diagram

Figure 4.3: Subsystem B: the closed-loop system

Pump A For conventional turbines, electricity is already generated in the nacelle. A system of gears and a generator converts the power in the rotor shaft to electric power. The electricity is conducted through a cable, along the base of the turbine support structure, to a generator platform. The closed-loop system of the DOT (see figure 4.3) is in fact a hydraulic gear which transmits the power from the top of the wind turbine to the base. Pump A is directly linked to the rotor shaft. Its purpose is to generate, as efficiently as possible, high pressure in pipe 1. The specific pumping technique has yet to be determined. A likely candidate for pump A is a radial piston-design unit. Currently, the most power P any hydraulic unit is able to transfer is around 1.75M W (the H¨ agglunds MB 3200 [12]). However, these units can be applied in parallel to cope with higher torque/power.1 Despite their relative small size, such hydraulic units are designed to operate under pressures of up to 350bar with high efficiencies (ηpump > 95%). PpumpA = ηpumpA · Protor

(4.3)

Essentially, what pump A does is initiate a volume flow Q at high pressure p in the closed-loop pipe. At the end of the entire system is the generator. The harder it is to turn the rotor of a generator (the more torque is required), the more power it produces. The torque required to power the generator determines the pressure in the open-loop system. The pressure and the volume flow of the fluid in the open-loop system determine the torque that motor B is required to produce. The 1 The rotation rate of hydromotors is in the same order as the rotation rate of the rotor shaft. Hence, no gearbox is required. The MB 3200 operates at a maximum of 16rpm. The popular Vestas V90 turbine (3.0M W ) operates in the range of 8.6 − 18.4rpm.

DOT Project Plan

11

How DOTs Work

Power Transmission: Wind to Water to Electric

torque motor B is required to produce determines the pressures in the closed-loop system. Hence, the torque required to power the generator can be translated into flow resistance throughout the entire system. This flow resistance determines the pressure. Within pump A, the height difference ∆z (head) is assumed to be zero. So, the pressure change over pump A in the nacelle is, ∆p = phigh − plow

(4.4)

Pipe 1 Pump A generates a flow in pipe 1 at the nacelle. Assumptions with respect to flow through a pipe: • the fluid inside is incompressible • the inner surface of the pipe is smooth (roughness factor e = 0) From equation 4.5, it is evident that for constant power P , the volume flow Q decreases as pressure change ∆p increases. PpumpA = ∆p · Q

(4.5)

The volume flow Q in a pipe is a function of the pipe diameter D and the flow velocity v of the fluid. A=π·



D 2

Q=A·v

2

(4.6) (4.7)

An overview of how several flow characteristics change for different ranges of power is given in table 4.1. Here, Dmin is the minimal pipe diameter for laminar flow (< 6m/s). Power P 1.7 M W 3.0 M W 5.0 M W 10.0 M W

Pressure p 350 bar 350 bar 350 bar 350 bar

Volume flow Q 0.049 m3 /s 0.086 m3 /s 0.143 m3 /s 0.286 m3 /s

2910 l/min 5140 l/min 8570 l/min 17100 l/min

Dmin 0.102 m 0.135 m 0.174 m 0.246 m

4.00 5.31 6.85 9.69

in in in in

Table 4.1: An overview of flow characteristics for different turbine power capacities.

Losses Due To Friction To gain high efficiency, friction must be minimized wherever possible. The pressure loss due to friction in the pipeline is related to the value of the Reynolds number Re of the transported fluid. Re =

v·D ν

(4.8)

Here ν is the kinematic viscosity.

DOT Project Plan

12

How DOTs Work

Power Transmission: Wind to Water to Electric

The friction factor f for flow through a pipe is derived from references [5] and [7]. A distinction is made between disturbed (turbulent) and undisturbed (laminar) flow (see figure 4.4). • For laminar flow (Re < 2300 in pipes): f=

64 Re

(4.9)

• For turbulent flow (Re > 4000 in pipes): 1 √ = −2 · log f



2.51 e/D √ + 3.7 Re · f



(4.10)

Turbulent flow is undesirable because within a pipe the flow is subject to much more friction than in the case of laminar flow.

(a) Laminar flow

(b) Turbulent flow

Figure 4.4: The two types of flow through a pipe The loss of pressure due to friction is modeled as, ploss = f ·

L 1 · · ρf luid · v 2 D 2

(4.11)

Notice that the length of the pipe L is directly proportional to the pressure loss. Motor B At the base of a DOT, motor B uses the pressure difference in pipe 1 to generate mechanical power. The low pressure part of the closed-loop pipeline experiences the same head ∆z as the high pressure part. So, the pressure change over motor B at the base is the same as at pump A (equation 4.4) minus the losses in pipe 1 (equation 4.11). ∆p = ∆p − ploss

(4.12)

This pressure change is converted to mechanical power by motor B. PmotorB = ηmotorB · ∆p · Q

(4.13)

The extracted power PmotorB is used to drive pump C. DOT Project Plan

13

How DOTs Work

4.2.3

Power Transmission: Wind to Water to Electric

C: The Open-Loop System

Figure 4.5: Subsystem C: the open-loop system

Pump C At the base of the structure, power from the closed-loop system in pipe 1 is used by pump C to pump seawater into a second pipe. The open-loop system hence uses seawater as a means to transfer energy from the base of a DOT to the generator platform. The functioning of pump C is similar to pump A. It is mechanically powered by motor B and it generates pressure in the open-loop pipe 2. PpumpC = ηpumpC · PmotorB

(4.14)

Pipe 2 Pipe 2 connects the base of a DOT to the central generator platform (see figure 2.1). As with pipe 1, the power PpumpC from pump C can be expressed in terms of the pressure change ∆p and the volume flow Q. PpumpC = ∆p · Q

(4.15)

The radius r and hence the area of the cross section A is likely to be larger than that of pipe 1. However, the main difference between the two pipes is that the length of pipe 2 will be much greater. Using the exact same method as with pipe 1, the losses due to friction can be found.

4.2.4

D: The Generator Platform

Figure 4.6: Subsystem D: the generator platform In DOT farms, all (N ) pipes 2 come together at the generator platform. Here the pressurized seawater is distributed over several generator turbines. As with motor B, the power extracted by DOT Project Plan

14

How DOTs Work

Simulation Results for Varying Wind Speeds

a generator is a function of the pressure change ∆p over it, and the volume flow Q through it. Assuming no further losses before the generator turbines, the total power generated in the wind farm is expressed as: Pgen = ηgen · N · ∆p · Q

4.3

(4.16)

Simulation Results for Varying Wind Speeds

To better understand the (theoretic) performance of a DOT, a numeric input is given to the theory of the previous section. The wind speed is varied from 0 to 20m/s. All other initial conditions are derived from existing techniques. They are summed up in table 4.2. Parameter general νsw νf l ρsw ρf l rotor cp r pumps ppumpA ppumpC ηpumpA ηmotorB ηpumpC pipes D1 D2 L1 L2 pmin generator ηgen

Value

Units

Description

1.17E − 6 4.20E − 4 1030 1000

[m2 /s] [m2 /s] [kg/m3 ] [kg/m3 ]

0.50 63.0

[−] [m]

induction factor radius

35.0E6 10.0E6 0.95 0.90 0.80

[P a] [P a] [−] [−] [−]

pressure generated by pump A pressure generated by pump C efficiency pump A efficiency motor B efficiency pump C

0.30 0.50 100 500 2.0E6

[m] [m] [m] [m] [P a]

diameter pipe 1 diameter pipe 2 length pipe 1 length pipe 2 minimum pressure in pipe 1

0.90

[−]

efficiency generator

kin. viscosity seawater at 15◦ C kin. viscosity of fluid in pipe 1 density of seawater density of fluid in pipe 1

Table 4.2: Initial parameter values The two main output parameters of interest are: - the power production of a DOT - the volume flow of seawater into the generator platform from pipe(s) 2 The amount of water flowing into the generator station is dependent on the pressure in the openloop pipe 2. As section 4.2 explains, for the purpose of efficiency, it is beneficial to use high pressure and subsequent low velocity flow. For flow calculations (figure 4.7), the critical value of the Reynolds number Re is assumed as 2800.

DOT Project Plan

15

How DOTs Work

Simulation Results for Varying Wind Speeds

−1

10

Pipe No.1 Pipe No.2

Friction factor f [−]

Laminar

Turbulent

−2

10

3

4

10

5

10

6

10

10

Re = v⋅D/ν [−]

Figure 4.7: The friction factor versus the Reynolds number for flow regimes in both pipes. Figure 4.8 shows the power curves of a DOT (unrestricted) and the REpower 5M [13].2 Both turbines have the same rotor diameter.

7

P

rotor

P

pump

6

B

Power [MW]

P

gen

5

REpower 5MW

4 3 2 1 0

0

2

4

6

8

10 12 Windspeed [m/s]

14

16

18

20

Figure 4.8: Power curves of the DOT and the REpower 5M In figure 4.9 the volume discharge is plotted against the wind speed. For the (assumed) rated wind speed of 13m/s the volume flow of seawater to the generator is approximately 650liters/s. In every energy generating system, some energy is lost on its way to the consumer. A percentage of the DOT’s kinetic energy at the rotor is lost on its way to the generator platform. Most of these losses will be due to friction in the pipelines. The key to increasing efficiency in theory is to improve the pumping systems and minimize losses in the pipelines. The key to increasing efficiency in practice is to reduce maintenance and increase uptime.

2 The REpower 5M is the largest turbines currently in operation. So far these turbines have been installed at offshore locations near Scotland (Beatrice) and Belgium (Thornton Bank).

DOT Project Plan

16

How DOTs Work

Simulation Results for Varying Wind Speeds

2.5 Pipe No.1 Pipe No.2

3

Discharge Q [m /s]

2

1.5

1

0.5

0

0

2

4

6

8

10 12 Windspeed [m/s]

14

16

18

20

Figure 4.9: The volume flow in the open-loop and closed-loop pipelines of a DOT.

DOT Project Plan

17

How DOTs Work

DOT Project Plan

Simulation Results for Varying Wind Speeds

18

5 Plan of Execution 5.1

DUWIND

The Delft University of Technology has joined together the research groups involved in wind energy research into the Delft University Wind Energy Research Institute, DUWIND. This institute envelops sections of 5 of the 8 faculties of the DUT as shown in figure 5.1. DUWIND totals 60

Figure 5.1: Overview of participants in DUWIND FTE of whom 30 FTE are PhD students. The topics covered range from offshore engineering, aerodynamics, aeroelastics and material engineering to control systems, generators, grid and policy. DUWIND is one of the largest institutes on wind energy in the world with a particularly large focus on offshore wind energy. DUWIND drafted its first research plan in 2002 with a scope of 15 PhD students on all new topics. This resulted in the current double of that in 2008. At this moment, the new research plan is being finished with a goal for the next 5 years to reach 60 PhD students. Within this plan, the DOTs also has its place. It has been identified as a showcase of combining scientific knowledge and bring it to the market.

5.2

DOTs Organization

With the DOT project being an integrated part of DUWIND, the DOT team members will reside under the different sections within DUWIND. Each PhD student will have a specialist professor to act as coach and promotor. The PhD students will participate in their specific section in research and education. On top of that, the group of 8 will also have a shared work place to increase the team spirit and focus the group attention to reaching their common goal. The group will furthermore be supervised by 1 supervisor, who is responsible for the day-to-day co-ordination of the DOT project progress. Figure 5.2 shows the team structure organization chart. 19

Plan of Execution

Time Planning

Figure 5.2: DOT Team structure/organogram

5.3

Time Planning

The project started with the first PhD student, Niels Diepeveen, commencing his work on 1 August 2008. This detailed plan is the first work from his hand in shaping the DOT process. Over the next months, financing needs to be secured for the project and new PhD students are being recruited. The plan is to start full force in January 2009. The typical PhD study time has been set to 3 years, which is somewhat shorter than the ”normal” Delft duration. The PhD students will start consecutively over a 1.5 year period, stretching the DOT project from 1-8-2008 to 1-8-2012 as shown in figure 5.3.

Figure 5.3: Gantt chart The goal of the DOT project is to develop all components for the new offshore wind turbine technology. The combined outcome of the 8 PhD projects is a blueprint for the construction of a first demonstration offshore wind farm. The project therefore does not end with the last PhD. Early 2012, efforts will start to acquire funding for a demonstration farm in the order of magnitude of 50 - 100 M e, to be constructed in 2013/2014. Following this planning, the DOTs will be ready for commercial application in the second half of the next decade, exactly when the exponential increase in offshore wind turbine installations is anticipated. As the goal of the DOT is not only to create an entirely new wind turbine system, but also be unrestricted in component improvements DOT Project Plan

20

Plan of Execution

Potential PhD Research Projects

that can be re-applied in current turbine technology, the result of the project will give a boost to the wind energy industry as a whole.

5.4

Potential PhD Research Projects

The DOT project will be realized by 8 PhD students, each having their own research topic. These topics will be detailed further as all PhD students start their work during the next 6 months. For the moment, the following areas of research have been identified: Aerodynamics (2 PhDs) The design of the rotor is critical for the performance of a wind turbine. Possible topics include boundary layer suction (to reduce drag of the blades) and the effects of stall behavior. Hydraulics (2 PhDs) The central theme in the DOT project is the use of hydraulics to conduct energy to a central electricity generator. The main items in this system are the pumps and the pipes. Mechanics (2 PhDs) Of vital importance to the success of the DOT project are the required installation methods and the overall energy balance. Both will require extensive research. Support Structure Design (1 PhD) So far, the design of offshore support structures for wind turbines is largely based on norms set by the oil & gas industry. Possible topics include structural materials and soil mechanics. Electronic Engineering (1 PhD) Eventually, the mechanical power generated by the wind at the rotor of a DOT has to be converted to electricity. This requires the design of the generator platform and all its components.

DOT Project Plan

21

Plan of Execution

DOT Project Plan

Potential PhD Research Projects

22

References [1] European Wind Energy Association, Pure Power: Wind Energy Scenarios up to 2030, March 2008. [2] European Wind Energy Association, Delivering Offshore Windpower in Europe: Policy Recommendations For Large-Scale Deployment Of Offshore Wind Power In Europe By 2020, December 2007. [3] Veldman, H., Lagers, G., 50 Years Offshore, Foundation for Offshore Studies, Delft, 1997 [4] Anderson JR., J.D., Fundamentals of Aerodynamics, Second Edition, 1991 [5] Battjes, J.A., Fluid Mechanics, Lecture Notes, Delft, March 2000 [6] Betz, A., Das maximum der theoretisch m¨ oglichen Auswendung des Windes durch Windmotoren, Zeitschrift f¨ ur gesamte Turbinewesen, vol. 26, 1920 [7] Colebrook, C.F., Turbulent Flow in Pipes, Journal of the Inst. Civil Eng. (11), 1938 [8] www.vestas.com [9] www.randstad380kv.nl [10] www.ewea.org [11] www.efunda.com [12] www.hagglunds.com [13] www.repower.de

23

Delft Offshore Turbines Re-designing wind turbine technology for more power, better performance and better economy Wind energy is booming. The focus is

The DOTs project aims to bring about a

moving to offshore production. Current

radical change in (offshore) wind turbine

technology

technology. It will not be performed for

is

far

from

optimal.

standard incremental improvements. The main idea behind the DOTs is to get more energy from wind in a less labor-

Research will not be limited to one specific

intensive manner. The focus of the design

grand design. Many of the developed

process will be on wind farm energy

concepts will also be applicable to current

production, but also on important

wind turbine technologies.

economical and practical issues such as the optimization of production, assembly

For more information, visit us at:

and maintenance.

www.offshore.tudelft.nl/offshorewind

Design rules Two rotor blades - lower production cost easier assembly, higher rotational speed The blades have a fixed pitch angle and use boundary layer suction (Actiflow) Transport of energy from nacelle to base through closed-loop hydraulic system. A single pipe transports seawater from each turbine base to the generator Conversion to electricity happens on a central platform As few components as possible

rotor

Functional diagram Pumping fluid at high pressure in combination with low volume displacement allows for highly efficient energy transfer. S o - c a l l e d f l u i d p o w e r i s a p r ove n technology in many industrial sectors.

low pressure

high pressure

pump

generator platform

Civil Eng Mechanic Eng

power to shore

motor

Electrical Eng Tech Management Aerospace Eng

seawater

pump

high pressure

The DOT team members will all be PhD students, residing under different sections within DUWIND.

Part II

Conference Paper EWEC 2009

roman

CLOSED-LOOP FLUID PUMPING AS A MEANS TO TRANSFER WIND ENERGY N.F.B. Diepeveen DUWIND, Faculty of Civil Engineering and Geosciences, Delft University of Technology Stevinweg 1, 2628 CN Delft, The Netherlands Tel.: +31 15 27 88030, E-mail: [email protected] SUMMARY The current standard for wind turbines is to have a generator placed in the nacelle. The Delft Offshore Turbines project aims to circumvent the need of the generator by using the rotor shaft torque to power a pump in the nacelle. This pump adds pressure to the liquid in a closed-loop pipe circuit, creating a flow. At the base of the offshore wind turbine, the pressure (energy) is taken out of the flow by a motor. Most of the energy losses between the nacelle and the base of the OWT will occur due to friction in the closed loop pipeline and mechanical & volumetric losses in the hydraulic drive systems. The success of this new turbine concept depends in part on the efficiency of the energy transfer. Considering the environment they are placed in, the drive systems need to be efficient, robust and requiring low maintenance. This paper presents the modelling results of a 5MW DOT concept which applies closed-loop fluid pumping as a means to transfer wind energy. The key elements in this system are the hydraulic drive systems, the hoses between them and the power fluid. Since suitable drive systems for this sort of multi-MW application does not yet exist, modelling is done using extrapolated properties of systems that do. This paper identifies significant design challenges and required system properties of a 5MW offshore hydraulic wind turbine by modelling the concept of a closed-loop fluid power circuit. The resulting general characteristics are compared to those of traditional offshore wind turbines. Keywords: Fluid power, hydraulic drive systems, closed-loop

1 INTRODUCTION A new concept for the design of offshore wind turbines/farms incorporates the idea of using fluids to transfer energy. The current standard for wind turbines is to have a generator placed in the nacelle. The need for additional support systems results in a large nacelle mass. This and frequent component failures are not beneficial to the economics of offshore wind farms.

-

losses. The high availability should however lead to a large overall increase in relative power production. Easy installation Low production costs

One design concept in which this can be applied is to split the wind farm components in two types of systems. 1. The closed-loop hydraulic wind turbine. The pump in the nacelle adds pressure to the flow in the circuit, creating a power flow. At the base of the offshore wind turbine, the pressure is taken out of the flow by a motor and converted to mechanical energy. 2. The open-loop hydro-power system. The motor at the base drives a second pump which pumps freestream seawater to a central power hub where this open-loop power flow is converted to electricity.

The Delft Offshore Turbines (DOTs) project aims to circumvent the need of the generator by using the rotor shaft torque to power a pump in the nacelle. This, along with the proposal to have a two-bladed rotor, will lead to a significant reduction in weight of the rotor-nacelle assembly. The overall goal of the Delft Offshore Turbines project is to design a wind turbine infrastructure specifically for offshore purposes and thereby rendering offshore wind energy more economically attractive. This means: - Very low maintenance - High availability; as a direct consequence of high reliability. - Reasonable efficiency; hydraulic drive systems often experience small

Fluid power circuits have been applied successfully on different scales in many industries. The idea of applying this method to wind turbines is not new. Literature on the topic

1

can be found from as early as 1981 [2]. So far a successful launch has not yet occurred mainly because the required components are were not readily available. They still aren’t for multi-MW systems. The aim of this paper is to identify significant design challenges and required system properties of a 5MW offshore hydraulic wind turbine by modelling the concept of a closed1 loop fluid power circuit. The resulting general characteristics are compared to those of traditional offshore wind turbines.

Rotor

increased chord and blade thickness, same relative thickness. A thicker blade is a stronger blade, hence less structural material is required, making the rotor lighter. This reduction in weight means the rotor has a lower moment of inertia. This leads to relatively higher angular acceleration and consequently a higher rotational velocity. Number of blades Rotor diameter Rated wind speed Rated power Rated rotor speed

Pump

Repower 5M 3

DOT 2

126m 13.0 m/s

126m 13.0 m/s

5 MW 12.1 RPM

5 MW 18.15 RPM

Table 1: Turbine characteristics Fluid Power Circuit Characteristics The main components of the circuit are [1]: - a pump; - a high pressure hose - a motor to extract the power from the flow - a low pressure hose - a combined cooling/boosting system which o keeps the temperature of the fluid in the system below a predefined maximum o keeps the pressure at the entrance of the pump at the required level. For multi-MW (>2MW) turbines, suitable pumps/motors are not yet commercially available. There however appears to be no technical reason why they have not yet been produced.

Motor Pump Seawater

To central power hub

Figure 1: DOT functional diagram 2

SYSTEM DEFINITION & MODELING

Turbine Characteristics Initially, the modelling of the DOT is done with reference to an existing wind turbine, the Repower 5M. The DOT is given the same rotor diameter as the Repower, 63m. The main difference is that the DOT has 2 blades instead of three. They also share the same rated wind speed. The Delft Offshore Turbine will have two blades to improve ease of installation and reduce rotor mass. The size and shape of the two blades is such that together they produce the same amount of lift (& torque) as the three-bladed rotor of the Repower. The necessary blade area is distributed over two blades instead of three. This results in an

Drive systems: pumps & motors There are many types of hydraulic drives. For these types of systems, the positive displacement pump is most suitable. Criteria for selecting a pump: 1. General purpose - the general purpose of the pump is to transfer energy through high pressured flow as efficiently as possible whilst bringing a significant reduction in weight to the nacelle. 2. Amount of the fluid - this depends on the size of the rotor (wind turbine) , the nominal operating pressure and the length & diameter of the hoses. 3. Fluid properties – in this early stage hydraulic oil is used.

1

Note that this paper presents the modeling of one possible concept and by no means the final design.

2

4. Required head - again, this depends on the rotor size. 5. Specify type of flow – preferably laminar, to maximize efficiency. 6. Power supply - wind (the rotor) 7. Cost - not yet considered. 8. Efficiency - this is one of the most important properties. Slightly lower efficiencies are acceptable if required maintenance is significantly less. 9. Cost compared to efficiency – not relevant at this stage 10. Lifespan - minimum of 20 years 11. Noise level - for offshore application noise is not considered as a form of hindrance, only as a means of energy loss. 12. Operating pressure - as high as realistically possible. This is one of the most important properties. 13. RPM - this will be a dynamic signal, whose exact nature is to be determined at a later stage.

demonstration, the motor has a swept volume 5 times smaller than the pump. The result of this is that the motor will turn 5 times faster than the pump. Constant swept volume Previous research has been done on the application of variable stroke/swept volumes [6]. This way the pressure in the high pressure hose can be kept as high as possible to minimize flow velocity and thus maximize efficiency (higher pressure difference = lower volume flow). (1) P = ∆p ⋅ Q For example, at the start-up, when the rotor begins to turn, the swept volume of the pump and/or the motor is very small. Hereby a very low velocity, high pressure flow is initiated, which is beneficial in terms of efficiency. For the system described in this paper however, the drive systems will have constant swept volumes albeit of different magnitudes. The main reasons for this are that: - The drive systems will need less moving parts - The reduction in efficiency will only occur at lower wind speeds where it will be minimal. At start-up, the pressure throughout the system will be at the charge-level. Once the rotor begins to spin, the pressure in the system builds up very quickly. Figure 5 shows that the losses at start-up are minimal.

Displacement

Rated Speed

Max. speed

Max pressure

Max torque

Rated power

Max power

For the modelling of the drive systems, the characteristics of the Hägglunds’ Compact CB series have been used [4]. These systems are also known as radial pistondesign units. They can operate efficiently (>95%) at high pressure.

Vi

nrated

nmax

pmax

Tmax

Pra

Pm

ted

ax

l/rev

rpm

rpm

bar

kNm

Pump

472

18.1

27.2

350

2480

M W 5.0

M W 7.1

Motor

118

72.6

108.9

350

620

5.0

7.1

Pipe/Hose-flow modelling The standard formula for pressure loss in a pipe or hose is given by equation 2 [5].

ploss = f ⋅

L 1 ⋅ ⋅ ρ ⋅ v2 D 2

(2)

To minimize losses the velocity needs to be as low as possible. Using the distance between the hub and the base and the rated capacity, the optimal pipe/hose diameter can be determined. This optimal diameter is chosen for a flow velocity at rated power where the flow is in the laminar to turbulent region, where the friction factor is minimal (see Figure 6). As base case a hub height of 90 m and a hose diameter of 0.3 m are taken. The power fluid of choice in this phase is a type of hydraulic oil, which is assumed to be incompressible. The hose itself is assumed to be hydraulically smooth [3].

Table 2: Properties of modelled drive systems with a gear ratio of 5 Since no system of this series is (yet) available in the 5MW range, all relevant system properties have been extrapolated to meet requirements. The properties of these non-existent drive systems are listed in Table 2: Properties of modelled drive systems. The stroke volume of the pump is determined by dividing the rated power by the maximum pressure and the rated rotational velocity. The motor at the other end of the circuit is the same type of system, but used in reverse and has a different stroke volume. The effect of this is that the whole system functions as a gear for the rotor shaft. For the purpose of

3

3. Once the maximum RPM is reached the flow of the wind around the blades is manipulated to also keep the rotation rate of the rotor shaft and therefore the power output at a constant maximum. At the point where the wind speed is too high for the power output to remain at it’s maximum, the blades will stall and the system shuts off. 8

p

hydraulic turbine traditional turbine

7

max

,n

max

power [MW]

6 pmax ,nnom

5 4 3 2 1 0

M

Cooling & charge pressure Closed-loop fluid power circuits produce heat due to internal friction and thus require cooling. This problem is initially solved by including a drainage reservoir which (together with a small storage tank) is connected to an extra pump. This pump adds charge pressure and cooled fluid to the low pressure hose of the circuit. The charge pressure before the pump prevents cavitation and increases the pump’s efficiency.

0

5

10

15 wind speed [m/s]

20

25

30

Figure 3: The power curve of a traditional and a hydraulic turbine This three-step modulation leads to a different look of the wind turbine power curve, as demonstrated in Figure 3. The upper limit of the power curve is not determined by electrical components. Instead the maximum allowable pumping conditions (pmax,nmax) define the shape of the curve.

This necessary Figure 2: CL circuit extra system does however mean that with cooling/boosting the amount of required components increases significantly. It also reduces the overall efficiency of the system, but it is a necessary requirement to compensate for small leaks and over-heating. Whether the subsequent heat dissipation is sufficient requires further investigation.

400

pmax ,nnom

350

force [kN]

300 250 200 150 100

3 SIMULATION RESULTS Using the system definition, a simulation model was constructed in MATLAB to analyze general characteristics. Load modelling 1. For a constant angle of attack, the lift and drag forces of blades are directly proportional to the wind velocity squared. Hence the torque produced by the rotor is also directly proportional to the wind velocity squared. Since the power in the rotor is directly proportional to the third power wind speed, the rotational speed of the rotor shaft is directly proportional to the wind speed. 2. After vrated, the torque produced by the blades may not increase, only the RPM. This is done by changing the flow around the blades.

50 0

0

5

10

15 wind speed [m/s]

20

25

30

Figure 4: The rotor force The same maximum force (fFigure 4) occurs at the rated wind speed. This is where the pump performs nominally (pmax,nnom). The flow around the blades is now regulated to avoid exceeding the maximum pressure until the maximum rotation rate of the pump is reached. From this point (pmax,nmax) the power output is kept constant until the rotor blades stall. Figure 5 demonstrates: - The build-up of pressure difference between the input of the pump and the input of the motor. The pressure build-up ∆p is directly proportional to the rotor torque build-up.

4

The increase in volume flow Q increases as long as the rotor RPM increases. The efficiency of the system is high for all rotor speeds except instantly after start-up.

-

Q [m/s 3]

∆ p [bar]

-

number of moving components without adding significant benefits and is therefore not desired. Despite the additional mass of the hoses and power fluid, having only a pump in the nacelle significantly reduces the total mass of the wind turbine. For concept evaluation purposes, further research needs to be done on the reduction of the nacelle mass and subsequently the support structure mass. The necessity of a subsystem for cooling & pressurizing adds significantly to the total number of components. A 5 MW turbine will require a volume flow of close to 10,000 litres per minute. The closed-loop fluid power circuit as described here requires the use of hydraulic oil. For future concept designs the use of seawater as power fluid will be addressed. Since the idea behind the DOTs project is to design a turbine specifically for offshore purposes, oil will not be the preferred power fluid in the long run. Instead we aim to use the offshore environment to our advantage and use seawater. Research will therefore be done on the design requirements for seawater-based (wind turbine driven) fluid power circuits.

400 200 0

0

5

10

15

20

25

30

35

40

45

0

5

10

15

20

25

30

35

40

45

0

5

10

15

30

35

40

45

0.4 0.2 0

η [-]

1 0.5 0

20 25 nrotor [RPM]

Figure 5: The rotor RPM vs. pressure change, volume flow & efficiency 2

10

1

f [-]

10

0

10

-1

10

REFERENCES [1] Cundiff JS. Fluid Power Circuits and Controls, Fundamentals and Applications. Virginia Polytechnic Institute & State University Blacksburg. [2] Unknown. Hydraulic Wind Energy Conversion. Jacobs Energy Research, Audubon. July, 1981. [3] Albers PS, et al. Vademecum Hydrauliek. Koopman & Kraaijenbrink Publishing. September 2008. [4] Hägglunds. Compact CB. Product Manual. [5] Batjes JA. Fluid Mechanics. Lecture Notes, Delft University of Technology. [6] Rademakers, LWMM. Possibilities of Variable Transmissions in Wind Turbines. MSc Thesis. Laboratory of Power Transmission, Eindhoven University of Technology

-2

10

-1

10

0

10

1

2

10

10

3

10

4

10

Re [-]

Figure 6: The Reynolds number vs. the friction factor for flow in a straight DOT hose 4 PRELIMINARY CONCLUSIONS From researching and the modelling of this initial concept a number of conclusions can be drawn. Hydraulic drive systems have been applied in many industries for many years. High performance systems are characterized by high efficiencies and low maintenance needs. The power production of a Delft Offshore Turbine can increase beyond rated due to the characteristics of hydraulic drive systems. The cut-out wind speed is determined by the control characteristics of the rotor (blades). The current maximum size of suitable pumps is under 2 MW and requires the use of hydraulic oils as power fluid. Larger systems are technically feasible. They have not yet been produced due to lack in demand. The efficiency of the system can be slightly improved by using more sophisticated drive systems. However, this would increase the

5

Part III

Conference Paper EOW 2009 [DRAFT]

roman

Design Considerations for a Wind-Powered Seawater Pump N.F.B. Diepeveen DUWIND, Faculty of Civil Engineering and Geosciences, Delft University of Technology Stevinweg 1, 2628 CN Delft, The Netherlands Tel.: +31 15 27 88030, E-mail: [email protected] Summary Offshore, one thing is abundant: water. The current turbine technology sees the nacelle weight increase steadily giving increasing challenges in support structure design and installation. Furthermore, power electronics help harness wind power slightly more efficiently, but also add weight and components (that can fail) to the turbine system. The Delft Offshore Turbines (DOTs) convert wind energy into a high pressure flow of water. A pump is connected to the rotor directly, generating a high pressure flow. The pressurized water is collected at a transformer platform, where generators are located comparable to a hydro plant. The platform can be fitted with limited water storage/accumulation capacity to smoothen energy variations. From this platform, an electricity cable connects to the onshore grid. High pressure fluid power is used in shredders, feeders, roll mills, cranes, bulldozers, jack-up systems, etcetera. These applications efficiently acquire power in the form of high torque through high pressure (op to 500 bar) fluid transmission. The DOT energy transmission concept is the exact opposite. High torque is converted into a high pressure flow. Currently ChapDrive AS in Norway, Artemis in Scotland and Voith Turbo (WinDrive) in Germany are all developing hydraulic gears for wind turbines. The most similar to the DOTs project is the ChapDrive, having its generator placed at the foot of the turbine tower. These systems all use hydraulic oil as medium. Seawater is preferable in to hydraulic oil in terms of dynamic performance. This is due to the higher bulk modulus of seawater. However, low viscosity of seawater also means poor lubricity and high potential of wear due to erosion and corrosion. The research presented in this paper shows that multi-MW high pressure seawater pumps do not yet exist. The main reason for this is that there has never been a real need for them. To make the DOTs a reality, such a pump will need to be designed. The main challenge is how to design for the use of seawater as hydraulic fluid.

1

Contents 1 Introduction

2

2 Delft Offshore Turbines 2.1 General Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Energy Transmission System Requirements . . . . . . . . . . . . 2.3 Motivation for Using Seawater as Hydraulic Fluid . . . . . . . . .

4 4 5 6

3 Fluid Power Circuits 3.1 Introduction to Fluid Power . . . . . 3.2 Basic circuit components . . . . . . . 3.3 Hydraulic Fluids . . . . . . . . . . . 3.4 Classification of Pumps . . . . . . . 3.5 Applications of Fluid Power Systems

6 6 6 7 8 8

4 Hydraulic Wind Turbines 4.1 Early Ideas . . . . . . . . . 4.2 Current Developments . . . 4.3 Advantages & disadvantages wind turbines . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . . . . . . . . . . . . . . . . of hydraulic power . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . . . . . . . . . . . . . . transmission for . . . . . . . . . .

9 9 9 10

5 Seawater as Hydraulic Fluid 10 5.1 Basic Properties of Seawater . . . . . . . . . . . . . . . . . . . . . 10 5.2 Systems for seawater pumping . . . . . . . . . . . . . . . . . . . . 12 6 Selection of Pumping Principle 13 6.1 Design Criteria for Seawater-based Positive Displacement Pumps 13 6.2 Characteristics of PD Pumps . . . . . . . . . . . . . . . . . . . . 13 6.3 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 7 Conclusion

15

References

16

1

Introduction

aanleiding The EWEA has set the target for an installed capacity of 40 GW for offshore wind turbines in 2020 []. Some of the main obstacles in achieving this target are the costs of the offshore support structures, installation and maintenance. According to critics [], the root of the problem is that the turbines being installed offshore were not initially designed for this environment. The goal of the Delft Offshore Turbines (DOT) project is to design and build a wind turbine (5+ MW) specifically for the offshore situation, thereby stepping away from incremental improvements. This goal translates directly into the driving project requirements: easy installation, robustness/low maintenance and high efficiency energy conversion. Eventually the implementation of these requirements will lead to lower IRR for offshore wind farms. To achieve these demands, the working strategy is to minimize the total number of systems,

2

minimize the use of expensive materials and where possible use the elements to our advantage. One of the major differences in the DOT design with respect to current mainstream wind turbine technology is the energy conversion method. Instead of applying electricity generators and all related components, the DOT will use a pump to convert wind power to high pressure flow of seawater. In other words, the idea for the energy conversion is to transmit power through a seawater-based fluid power system to a central generator platform in the wind farm. probleem Fluid power systems are used in many industries throughout the world. What makes the application different for the DOT is the fact that it is a multi-MW wind-driven and that the power transmission medium is seawater. The basic components of any fluid power circuit are: a pump (mechanical to power flow, a motor/generator (power flow to mechanic/electric), conductors (pipes or hoses) and a hydraulic fluid as power transmission medium. The first and most critical component is the pump driven by the turbine rotor. In recent years, the number of industrial applications of fluid power has risen dramatically.1 Hydraulic transmissions have often been considered for wind turbines, mainly because of the low maintenance requirements. The part-load efficiencies of available hydraulic drive systems however have so far been too poor to make them commercially attractive. The main initial system requirements are high power (5 + M W ), robustness/low maintenance and high efficiency. Belang Offshore wind energy has high potential. Currently the price for placing turbines offshore is too high. Projects are not yet economically feasible without government subsidies. Doelstelling The overall goal of the project is to make offshore wind economically viable. This translates to a design goal of the overall project which is to reduce the number of components in the offshore turbines drastically to come to the ultimate offshore turbine. One way in which this can be achieved is by having only one component (other than the rotor shaft and its support) in the nacelle: a pump. Hydraulic turbines are not new. Using seawater as hydraulic fluid is. The goal of the research for this paper is therefore to find a pump type suitable to be used for a DOT. The main requirements for this pump are: Efficient performance at low wind speeds as well as high Very low maintenance Considering that the principle hydraulic fluid is seawater, an essentially robust yet high performance pump is required. The goal of this research for this paper is therefore to find a pump type suitable to be used for a DOT. Hoofdvraag Werkwijze From the analysis of the main DOT requirements, the main pump functions & requirements are derived. Through the investigation of fluid power applications in general and in wind turbines specifically, the pumping principle which allows for the highest efficiency is selected. One of the most challenging design requirements is that the hydraulic fluid is seawater. The approach for this research was to: • Gain insight in how fluid power circuits operate. This means mapping 1 Currently

it is even being considered for power transmission in cars [9].

3

which type of systems exist, which is the most efficient and why and what the fundamental performance indicators are. • Gain insight in the applications and the potential of fluid power circuits, i.e. what has already been done with fluid power in similar applications, in general and hydraulic wind turbines in particular. • Select the prime candidates for the best pumping principle and investigate their commercial availability. • Gain a general insight into which challenges arise from using seawater as hydraulic fluid. Randvoorwaarden Structuurbeschrijving After elaborating on the Delft Offshore Turbines Project, the basic characteristics, application and components of fluid power systems are discussed. The next step is to look at current developments of hydraulic wind turbines and analyze the pros and cons of hydraulic power transmission. The most unique feature of the DOTs energy transfer system is that it uses seawater as hydraulic fluid. With this in mind a selection of the preferred pumping principle is performed.

Figure 1: The research map

2 2.1

Delft Offshore Turbines General Purpose

Delft University of Technology is taking a radical step away from incremental development of offshore wind turbines. It has started a research project on using a 2 bladed, fixed pitch turbine (5-10MW) to directly drive a water pump in the nacelle. By channelling the pressurized water of all turbines to one transformer platform, electricity generation is centralised. The design goal is to reduce the number of components in the offshore turbines drastically to come to the ultimate offshore turbine. Current offshore wind turbines are marinized land turbines with only a few add-on features to keep out the salty air. Improvements of turbine technology are only incremental and do not take full benefit of the offshore environment. The Delft University of Technology has a history in offshore wind research of over 25 years and has formulated a radical concept change of offshore wind energy conversion that helps develop a completely new system and spark revolutionary developments on sub-system and component level. Typically, offshore wind farms have a generator platform that gathers all electricity of the

4

different turbines, steps up the voltage and feeds the power through shore connection cables to the onshore grid. The DOT takes boundary conditions from this existing configuration: horizontal axis turbine with blades and a platform where the combined electrical power is fed to the onshore grid. Everything in between can be changed. The DOTs focuses on radical technology changes. To facilitate this, a short list of design pointers has been defined to test all developments against and to keep as life line throughout the project execution. Offshore, one thing is abundant: water. The current turbine technology sees the nacelle weight increase steadily giving increasing challenges in support structure design and installation. Furthermore, power electronics help harness wind power more efficiently, but also add weight and components (that can fail) to the turbine system. The DOTs convert wind energy into flowing water. A pump is connected to the rotor directly, generating a high pressure flow. The pressurized water is collected at a transformer platform, where generators are located comparable to a hydro plant. The platform can be fitted with limited water storage capacity to smooth energy variations. From this platform, an electricity cable connects to the onshore grid.

2.2

Energy Transmission System Requirements

- Low cost - the ultimate purpose of the DOTs is to make offshore wind a competitive energy source. - Low maintenance is a fundamental requirement for lowering costs of operation - Long lifespan, is arbitrary. Sometimes it becomes more viable to upgrade/replace a system before it is written off. If the payback time of a DOT can be driven back to under 6 years, placing a new & improved system every 10 years could be beneficial. - High efficiency. This is both scientifically and economically also an arbitrary issue. How does one define efficiency? In this case the most straightforward answer is the rate at which wind energy is converted to electric energy, which is then transported to the shore. Looking at offshore wind turbines, a system can function very efficiently. However, once it breaks down, its overall efficiency reduces. If there is not a weather window which allows for repairs for some time, this overall efficiency drops significantly. Efficiency can also be measured economically. If a lot of maintenance is required, the time and energy (cost) that requires directly cuts into the overall efficiency. So, by stating the requirement for high efficiency, this refers to the entire system, including all related costs such as for installation, operation, maintenance and decommissioning. High efficiency thus translates to every component of the entire system. The hydraulic energy transmission system starts with the hydraulic pump in the nacelle. For high efficiency during normal operation, a pumping system is required that is also efficient in case of partial loading. • ”use the elements/environment” + minimize No. of systems + minimize expensive stuff like copper • apply basic fluid power system using seawater 5

The mayor advantage of using seawater as a medium is that in this energy transmission system can be applied to any kind of offshore project. The idea behind DOTs is to make wind energy offshore more economically viable. This can be achieved by: • reducing the number of components • (thereby) reducing the weight of the nacelle and subsequently the support structure • improving robustness Here, we make a case for choosing reliability as the number one priority. Efficiency is secondary. To make offshore wind a competitive energy source for the future. DOT Functional Concepts Motivation for the choice to pump seawater Mission pump seawater - OL ’ sw pump at rotor - CL+OL ’ sw pump at base In CL, if medium = oil, pumps already exist. [1] [2]

2.3

Motivation for Using Seawater as Hydraulic Fluid

Pumping seawater - Disadvantages - Advantages possible effects: - slight local sea temperature rise. effects?

3

Fluid Power Circuits

3.1

Introduction to Fluid Power

A fluid is any substance that flows or deforms under an applied shear stress []. Power can be defined as the manifestation of control. For mechanical applications power is expressed in terms of energy over time.The basic theory for fluid power is found in Pascal’s law: Pressure applied to a confined fluid in transmitted undiminished in all directions, acts with equal force on equal areas and at right angles to them. Fluid power is defined as the change in pressure of a volume of fluid times the flow rate of that volume over time (P = δp · Q) Fluid or hydraulic power circuits are found in a wide range of industrial machinery. The most common application is for motion control.

3.2

Basic circuit components

A fluid power circuit (FPC) is defined as a system in which pressure and/or flow speed are the primary forms of output control. Common components are: • Pump - To pump is to use pressure to displace a fluid. • Hydraulic fluid - the energy transmission medium 6

• Motor - a (positive displacement) pump creates an energy flow, a motor extracts energy from the flow. • Pipes/hoses - to contain and direct the flow • Extras - valves, filters, accumulators and the sorts all are fundamentally important to the functioning of a FPC. The characteristics of a fluid power system are predominantly determined by the characteristics of the pump the motor and the fluid medium.

3.3

Hydraulic Fluids

In fluid power circuits, the hydraulic fluid is used as a medium to transfer mechanical power. Liquids are virtually incompressible, yet flow with little frictional resistance. This makes them an ideal medium for the transmission of power. Convention hydraulic fluids can be split up in two categories: • petroleum base fluids (hydrocarbons): highly flammable, restricted operational range • synthetic fluids: chemically compounded or water base fluids, resistant to burning Merritt [3]: Water is a poor hydraulic fluid because of its restrictive liquid range, low viscosity and lubricity and rusting capability However, the use of freshwater for high pressure hydraulics has recently gained new interest. As discussed in [4], water is environmentally friendly (thus readily disposable), non-toxic, non-flammable, inexpensive, and readily obtained. With health, safety and environment becoming ever more important in modern industry, it is likely that water-based hydraulics will gradually replace oil hydraulics. An additional advantage is the high bulk modulus of water compared to oil, resulting in better performance. The main drawbacks of water are the need to use corrosion resistant materials and the low viscosity which leads to bad lubrication. The application of water as hydraulic fluid is mainly reserved for closed loop systems. The way in which water is used is either as • ”dead” water • water based - glycol or another agent which improves lubrication. • no open-loops, water needs to be clean, low viscosity required finer filters Hydrauvision So, what about seawater? probably the worst choice? The relation pressure, density (volume) and temperature is described by the equation of state [3]. increasing pressure leads to a higher boiling point According to Merrit: 7

the bulk modulus is the most important fluid property in determining the dynamic performance of hydraulic systems. This is because β is a measure for the stiffness of the fluid. It is the inverse of the compressibility. Viscosity is an important property. Positive displacement pumps all employ close-fitting surfaces. If viscosity is too low, leakage flows increase If it is too high, power loss due to fluid friction occurs.

3.4

Classification of Pumps

Figure 2: Schematic classification of pumps Pumps can be divided in two general categories: kinetic (or hydrodynamic) and positive displacement pumps. In hydrodynamic pumps such as centrifugal pumps, the flow is continuous from inlet to outlet and results from kinetic impulse given to the fluid stream. The output is characterized by low pressure and high volume. Inefficiency and easy stalling as a result of back-pressure make these pumps unsuitable for control. In positive displacement pumps, fluid flows through an inlet into a chamber. As the pump shaft rotates, the (positive or definite) volume of fluid is sealed from the inlet and transported to the outlet where it is subsequently discharged. The essential difference between these two main categories is that kinetic pumps are for fluid transport systems and PD drive systems are for fluid power systems. By far the most widely used type of pump is the centrifugal pump (kinetic). Centrifugal pumps are used for all kinds of flows including sludge and slurry. Positive displacement pumps only account for about 10% [7]. For the DOT, a pump is required that can cope with 5 + M W of power. To minimize the flow speed and the necessary size of the pipe or hose diameter, high pressure is required. This and the superior performance in terms of efficiency make the positive displacement pump the prime candidate.

3.5

Applications of Fluid Power Systems

Evidence of the use of water power dates back to 250 BC. The most common application up to well into the 20th century was in the form watermills, which were used to grind grains. The use of high pressure in hydraulics was introduced on a large scale in the second half of the 19th century. In major cities throughout the world, hydraulic mains (first cast-iron, later steel) were installed beneath the streets. Pressure was maintained by five hydraulic power stations, originally driven by coal-fired steam engines. Short-term energy storage was provided by 8

hydraulic accumulators, which were large vertical pistons loaded with heavy weights and tanks in high towers. Applications included cranes, elevators and even theater curtains [13]. At its peak in 1939, the pumping stations in London were supplying an average flow of around 14,000 liters of water per min at nearly 60 bar pressure. This translates to an average power production of around 1.35MW. Wartime bomb damage, the departure of manufacturing firms from the city center and the rise of power electronics gradually led to the shut down of the last pumping station in 1977. Find out main complications of Victorian age tap water hydraulics: which parts needed to be serviced most? Modern industrial applications Heavy Lifting Machines Cranes Bulldozers Hagglunds drives CB/M Offshore: Jack-up hydraulic systems. Fluid power for leverage. Special application Taipei’s 101 tower []. Mining Pressure regimes Power Regimes Rpm regimes (GRAPH) Power to weight ratio pump - see Hagglunds product manual generator ABB

4 4.1

Hydraulic Wind Turbines Early Ideas

Nasa paper [8] Sir Henrey Lawson-Tancred in Yorkshire. variable hydraulic drives Luc Rademakers 1.3MW Bendix/Schachle turbine in the USA. variable hydraulic drives

4.2

Current Developments

Currently ChapDrive AS in Norway, Artemis in Scotland and Voith Turbo (WinDrive) in Germany are all developing hydraulic gears for wind turbines. The most similar to the DOTs project is the ChapDrive, having its generator placed at the foot of the turbine tower. These systems all use hydraulic oil as medium. Seawater is preferable in to hydraulic oil in terms of dynamic performance. This is due to the higher bulk modulus of seawater. Having an open-loop system means that the temperature of the water is likely to remain well within its liquid range. However, low viscosity of seawater also means poor lubricity and high potential of wear due to erosion and corrosion. Artemis - Digital Displacement Wind Turbine Transmissions replacing mechanical gearbox by a hydraulic transmission Artemis Intelligent Power Ltd. [9] This system is being developed with the aim to replace the traditional gear- box in the conventional wind turbine layout. One of the main advantages of a hydraulic drive over a gearbox lies in the ability to handle large shocks. This directly relates to the ruggedness and reli- ability disadvantage of hydraulic drives is low efficiency at part-loading A prototype is 9

currently under development and scheduled to be ready in ... Pump properties nominal pressure/Q/rpm Voith - WinDrive The variable speed of the wind turbine is transformed to constant rotational speed by the hydraulic motor. The generator can therefor be connected directly to the AC grid. No power converters are required, since only the strength of the electric current increases as the motor builds up more torque at higher windspeeds. Pump properties nominal pressure/Q/rpm [10] ChapDrive Compare performances! Also to other types of pump applications What happens in the event of - failures - partial load - cut-out wind speed [11]

4.3

Advantages & disadvantages of hydraulic power transmission for wind turbines

Advantages • Heat generated by internal losses is a basic limitation of any machine. • Hydraulic fluid acts also as a lubricant. This translates to long component life. The choice of hydraulic fluid and structural materials of the component obviously play an important part here Disadvantages • It is not possible to keep the fluid free from contamination. Filtering is required. The level of sophistication of the filter depends on the robustness of the system components. •

5

Seawater as Hydraulic Fluid

5.1

Basic Properties of Seawater

About 97% of the water on Earth is sea water. Almost every natural substance known to man is found in the world’s oceans and seas, mostly in very small concentrations [5]. The most notable characteristic component of seawater with respect to freshwater is salt. Although the vast majority of seawater has a salinity of between 3.1% and 3.8%, this number can vary significantly, for instance in response to addition of freshwater from rain and runoff, and removal of freshwater through evaporation. Despite small compositional irregularities, seawater behaves as a Newtonian fluid, which is beneficial in terms of performance as a power fluid. For the use as hydraulic fluid, it is important to note that seawater contains • suspended solids, practically any form of debris 10

T vs. ρ at atmospheric pressure

Density ρ [kg/m3]

1035 salinity 40

1030

salinity 35

1025

salinity 30 1020 salinity 25 1015 1010

salinity 20

0

5

10 15 Temperature T [°C]

20

25

p vs. ρ at T = 0°C, salinity 35 1080

Density ρ [kg/m3]

1070 1060 1050 1040 1030 1020

0

200

400 600 Pressure p [bar]

800

1000

Figure 3: Nonlinear density changes of seawater with pressure [12] • organic substances, one effect being marine growth • dissolved gases, ... The density of surface seawater ranges from about 1020 to 1029kg/m3 , depending on the temperature and salinity (figure 3). The pH of Seawater falls in the range 7.5 to 8.4. Compare this to freshwater which has a pH of approximately 7, depending on temperature. Ocean acidification (like other changes in oceanic composition) is a serious concern, but the time scale for it to be influential is to grand for this research. Also left out of this research are the additional benefits of pressurizing seawater. They will be investigated in due time2 . Seawater density depends on temperature, salinity and pressure. Colder water is denser. Saltier water is denser. High pressure increases density. The nonlinearity of the equation of state is apparent in contours of constant density in the plane of temperature and salinity (at constant pressure) - they are curved. They are concave towards higher salinity and lower temperature. Cold water is more compressible than warm water. That is, it is easier to deform a cold parcel than a warm parcel. Therefore cold water becomes denser than warm water when they are both submerged to the same pressure. Therefore various reference pressures are necessary. We use a pressure which is relatively close to the depth we are interested in studying. The compressibility effect is apparent when we look at contours of density at say 4000 dbar compared with those at 0 dbar. The freezing point of seawater is lower than that of freshwater, at around 2 Magnesium, bromine and sodium chloride (table salt) are all extracted from the sea on a global scale. In theory desalted seawater can provide a limitless supply of drinking water. So far this has been restricted due to the high processing costs.[14]

11

Element Oxygen Hydrogen Chlorine Sodium Magnesium Sulfur Calcium Potassium Bromine Carbon

Percent [%] 85.84 10.82 1.94 1.08 0.1292 0.091 0.04 0.04 0.0067 0.0028

Table 1: Seawater composition (by mass) (salinity = 35) ?2?C? As sea water freezes, it forms pockets of salt. The salt (brine) leaches out of the bottom of the ice and the brine drips into the water below the ice.

5.2

Systems for seawater pumping

Check: triplex mudmotors Schaalvergroting van 2 tot 4 is max toegestaan Later kijken naar low maintenance Dredging/centrifugal pumps ? − > lifetime Useful Properties of Centrifugal Pumps General description Dredging is an underwater excavation technique. Its main working principle is the gathering up of bottom sediments and disposing of them at a different location. In essence, dredging is not a form of fluid power application but a mass transfer method. Fluid, usually in the form of sweat or salt water, is used as a lubricant to avoid cavitation and aeration. Centrifugal pumps. This is probably the most applied type of pump in the world. The fact that the The size of sediment particles in a flow is typically defined in microns (micrometers). What makes these pumps suitable? (Essential properties) Resistance to erosion - no Resistance to corrosion - to a certain extend Dredging/centrifugal pumps − > lifetime Useful Properties of Centrifugal Pumps In contrast to centrifugal pumps, pd pumps are able to build up high pressure. Centrifugal pumps stall when the pressure inside a system becomes too high. Since there are no tight fit with sealing in a centrifugal pump stall will occur at relatively low pressures. A pump suitable for a Delft Offshore Turbine does not yet exist. Therefor either one has to be designed or an existing design is adapted to cope with seawater as hydraulic fluid. The logical next step is to determine the type of positive displacement pumping which is optimal for the DOT application.

12

6

Selection of Pumping Principle

6.1

Design Criteria for Seawater-based Positive Displacement Pumps

The main requirements for this pump are: • Positive displacement. The only pumps which are suitable for fluid power are positive displacement pumps. • High pressure • Operate for relatively large range of rpm • Robust • High efficiency The selection of the positive displacement pump mechanism depends on characteristics like • the potential power that it can deliver. • the pressure regime • the flow characteristics • the applicability of seawater as medium • the response to partial loading

6.2

Characteristics of PD Pumps

Reciprocating piston plunger pump Triplex pumps for near-continuous flow PLAATJE Axial piston pumps are used for smaller scale wind powered osmosis plants [] Reciprocating pumps require a crankshaft. This is unfavorable in terms of load eccentricities due to asymmetrical loading. CHECK Reciprocating diaphragm pump PLAATJE Rotary vane pump low viscosity, non-lubricating liquids, restricted pressure PLAATJE Rotary helix pump PLAATJE Rotary piston pump PLAATJE 13

Lobe / gear pumps Although these are more basic (less expensive) mechanisms, they only function at a limited rpm range. At low rpm for instance, the leakage is to great to generate a pressure flow. PLAATJE Radial Piston Pumps (Triplex ’ Crankshaft) Compare efficiency curves. Compare partial loading efficiencies. Model efficiencies of different systems Find efficiency relations with torque and rotational speed. The main area of concern in terms of lubrication and wear due to corrosion and erosion is the clearance between the piston and the cylinder. Once this clearance begins to increase, the resulting volumetric losses will cause the overall efficiency to drop. next step: find a material suitable for the fabrication of hydraulic components resistant to seawater 7

Rated power [MW]

Hagglund CBP pumps Wind Turbines 6

E−126

5

Bard 5.0 Repower 5M

4 SWT−3.6 3

V90

2

E−70 V80

1

E−53 E−33

0

0

50

100

150 200 Rated rotation rate [RPM]

250

300

350

Figure 4: Power vs. rotor rpm of commercial wind turbines and H¨ agglunds CBP pumps

Figure 5: A H¨ agglunds pump/motor of the radial piston rotating case type

6.3

Analysis

The pump types best suited for the Delft Offshore Turbine are vane pump and the radial piston pump. Vane pumps can cope with low viscosity fluids 14

Figure 6: Performance efficiency of H¨ agglunds CA 210 (4 ports) pump/motor [15] but is limited in terms of pressure. The radial piston pump can generate high pressure and can be designed to operate efficiently at rotation rates matching those of the wind turbines. However, the clearance between the piston and its casing is a concern when using seawater. The higher pressure regimes of piston pumps result in a much larger power-to-weight ratio. This is the main reason for selection the radial piston pump as the preferred choice for the system connected to the rotor shaft of the Delft Offshore Turbines.

7

Conclusion

The main challenge for the DOT hydraulic energy transmission is to have a robust, yet efficient system. For this the radial piston pump is the best suited. A pump with rated capacity of 5MW is not yet commercially available, let alone one capable of pumping seawater for long periods of time without maintenance. In terms of power production, the Hgglunds CBP 210, with a little over 2.3M W rated power at 350bar pressure is the state of the art. More powerful systems do already exist in the form of prototypes. However, all require hydraulic fluids with high viscosity. The conclusion therefore is that a specific system for high pressure 5M W seawater pumping will have to be designed. Important design considerations are: - The Hgglunds CBP 210 (2.3 MW) operates with efficiency of 9096%. In these systems, the larger the piston, the higher the efficiency. Therefore ever more efficiency can be expected of a 5M W pump. - The power-to-weight ratio of these types of pumps is much higher than that of the generators in wind turbines. This is without taking into account all the extra components required for the use of an electricity generator. - The low viscosity of seawater will lead to an increase leakage flow, reducing the efficiently slightly. - The high bulk modulus of seawater should significantly benefit the overall efficiency.

15

- The pump has to operate efficiently also at very low rpm. Trends in radial piston pumps and commercial wind turbines indicate that designing for matching rotation rates is feasible. - One of the first steps in the design process should be to find a solution for the poor lubricity and the corrosive and erosive characteristics of seawater. The logical first step is to look for a structural material that is sufficiently damage resistant for the fluid carrying parts of the pump. Finding this material is crucial to the development of the project. [Turbine No. of components reduction and thereby reduce the turbine mass and subsequently the support structure mass.] A this moment a pump fitting these requirement is not commercially available. The main challenge is to have a robust, yet efficient system.

References [1] Diepeveen, NFB, Closed-Loop Fluid Pumping as a Means to Transfer Wind Energy, DUWIND, Delft University of Technology, Proceedings EWEC 2009 [2] Diepeveen, NFB, Van der Tempel, J, Delft Offshore Turbines, Project Plan, DUWIND, Delft University of Technology, www.offshore.tudelft. nl/offshorewind [3] Merrit, EH, Hydraulic Control Systems, 1967, John Wiley & Sons Inc. [4] Lim, GH, Chua, PSK, He, YB, Modern water hydraulicsthe new energytransmission technology in fluid power, Nanyang Technological University, School of Mechanical and Production Engineering, 1 February 2003 [5] Turekian, K, Oceans, 1976, Prentice-Hall [6] Anderson jr., JD, Fundamentals of Aerodynamics, Second edition, 1991 [7] Cundiff, JS, Fluid power Circuits and Controls, Fundamentals and Applications 2002 [8] Unknown, Hydraulic wind energy conversion system, NASA STI/Recon Technical Report, July 1981 [9] Rampen, W, Gearless Transmissions for Large Turbines - The History and Future of Hydraulic Drives Artemis IP Ltd, Scotland, www.artemisip.com [10] M¨ uller, H, P¨ oller, M, Basteck A, Tilscher, M, Pfister, J, Grid Compatibility of Variable Speed Wind Turbines with Directly Coupled Synchronous Generator and Hydro-Dynamically Controlled Gearbox, Proceedings Sixth International Workshop on Large-Scale Integration of Wind Power and Transmission Networks for Offshore Wind Farms, 26-28 October 2006, Delft, NL [11] ChapDrive AS, www.chapdrive.com date of access: August 24th 2009 [12] Physical properties of sea water, http://www.kayelaby.npl.co.uk/ general_physics/2_7/2_7_9.html date of access: August 24th 2009

16

[13] www.subbrit.org.uk/sb-sites/sites/h/hydraulic_power_in_ london/, date of access: August 24th 2009 [14] Encyclopedia Britannica (online), www.britannica.com, search term “seawater” [15] H¨ agglunds Product Manual for Compact CA Motors 2004 www.hagglunds. com [16] H¨ agglunds Installation and Maintenance Manual for Compact CBP Motors 2004 www.hagglunds.com

17