Final Report Summary - SUPRAPOWER (SUPerconducting, Reliable, lightweight, And more POWERful offshore wind turbine)
Executive Summary:
Offshore wind market demands higher power rating and more reliable turbines in order to optimize capital and operational costs. The state of the art shows that conventional generators are difficult to scale up to 10 MW and beyond due to their huge size and weight. Superconducting direct drive wind generators are considered a promising solution to achieve lighter weight machines.
SUPRAPOWER (www.suprapowe-fp7.eu) was a research project funded by the EU FP7 programme. It started in December 2012 and finished in May 2017. The project was conceived to design an innovative, lightweight, robust and reliable 10 MW class offshore wind turbine based on an MgB2 superconducting generator (SCG).
Tecnalia Research and Innovation lead a project consortium of eight outstanding European companies and research institutions: Columbus Superconductors SpA, IEE-Slovak Academy of Science, University of Southampton, Karlsruher Institut fuer Technologie, D2M Engineering SAS, Solute Ingenieros and Ingeteam Services SA.
As result of SUPRAPOWER a 10 MW 8.1 rpm direct drive 48 salient poles synchronous generator has been developed and patented (EP 2521252). It is a partially superconducting generator (SCG), MgB2 superconducting wires at cryogenic temperature (20 K) are used in the field coils, but copper wires at ambient temperature in the armature coils. The cooling system is a cryogen-free topology. It has a warm iron rotor configuration which consists of one modular cryostat per pole that encloses only the superconducting coil, while the iron of the pole remains at room temperature. Heat is extracted by conduction through a thermal collector inside a cryostat which links all the modules. This is cooled by conduction by two-stage Gifford-McMahon cryocoolers which rotate jointly with the rotor. A helium rotary joint connects the stationary helium compressors to the rotating cryocoolers.
The 10 MW SCG active parts show a 25.6% weight reduction with respect to a 10 MW permanent magnets generator (PMG) ones and the total reduction of the tower head mass (including nacelle and blades) is of 7.2%. The achieved head mass reduction permits an 11% reduction of the tower weight and 9% of the fixed foundations (monopile). Thus 0.5-1 M€ of cost reduction can be achieve just due to material savings. Installation, O&M, decommissioning and all other marine operations do not complicate due to the SCG and 0.9 M€ savings could be achieved in this field.
A scaled magnetic rotating machine has been designed and constructed for the validation the concept of the 10 MW SCG. Main innovative elements and features as superconducting and cryogenic implementation, modularity and quench detection system are equal to those of the 10 MW SCG ones. Several new solutions have had to be developed for this first-ever manufacturing of this scaled machine, which was successfully completed by project end. The superconducting coils and the cryogen free cooling concepts have been successfully tested and validated. However, the test results of the scaled machine have not permitted its complete validation and have revealed new requirements for proper functioning.
The project has demonstrated the SCG is cost-competitive with, and provides several technical advantages over direct drive PMGs. Test results and experimental activities have enhanced the value of the patent of the SCG. But the development of the SCG still requires further R&D, validation and demonstration activities.
Project Context and Objectives:
Context of the Project
-----------------------------
Offshore wind energy is rapidly developing motivated by stronger and more regular winds at sea, and increasing site restrictions on land. At the end of 2016 the European cumulative installed capacity was 12.6 GW offshore and 141.1 GW onshore. With almost 300 TWh generated in 2016, wind power covered 10.4 % of the EU’s electricity demand.
In 2010, according to the International Energy Agency, global offshore wind cumulative capacity could reach 100 GW in 2020 of which 40 GW in Europe. Recent regulatory and economic developments in the EU have changed the offshore wind energy perspective for the next 15 years. In light of uncertain governance towards achieving EU climate and energy binding targets, EWEA updated the European wind energy industry’s vision to 2030. EWEA’s new Central Scenario expects 66 GW of offshore wind capacity and it estimates that wind could produce 245 TWh, covering 7.7%.
However, severe cost reductions are demanded in order to be able to reach these market perspectives. In this sense, higher power rate wind turbines are pointed out as the innovation with higher Levelized Cost of Energy (LCoE) reduction potential. Indeed, there is already a trend towards more powerful turbines: the average capacity rating of the 361 offshore wind turbines (WT) under construction in 2016 has been 4.8 MW, 15.4% larger than in 2015 and more than the double respect to 10 years before. The average rated power of new installed WTs is expected to reach 6 MW in 2016 and 8 MW in 2020, and it is already remarkable that in 2016 the first 8 MW turbine has been grid-connected at sea. Main WT manufactures are developing new turbines of 5 to 8 MW, mainly based on medium speed (60-600 rpm) and low speed (8-20 rpm) drive trains with permanent magnets generators (PMG).
For the last years the dominant technology has been the geared drive train induction generator (DFIG), while direct drive (DD) trains still represent a small share of the offshore wind market. Nevertheless, DD trains are a promising solution, as they show higher reliability and lower maintenance costs than geared options, thereby this technology is called to have a much more important share of the offshore wind market. But, current slow rotating PMGs are huge and heavy machines and this put at risk the technical and economic feasibility of DD turbines in the range of 10 MW. Moreover one of the problems associated to PMGs is the high volatility in the price of the rare earth materials used for constructing the excitation magnets. For example in 2011 rare earth prices reached a top of more than 20 times its previous 5 year average. Additionally the geographical concentration of rare earths production in China worries to the industry.
The size of electric generators, as indicated in the expression below, is mainly determined by the torque (T), being this proportional to the rotor volume (V) and the airgap shear stress (σ). This stress is proportional to the average airgap magnetic flux (B) and to the linear current density (A).
P=T·ω ->TαV·σ=V·B·A
Conventional high power DD wind generators have huge torque due to their low rotational speed (ω). “B” is limited by material properties, with typical maximum values of 0.9- 1 T. “A” depends on factors as the stator slot depth, packing factor of copper is slots or cooling/ventilation system. As previously commented these limitations makes high power DD conventional wind generators heavy and bulky.
In this context the use of superconducting materials in DD generators arise as a promising solution for achieving reliable, lightweight and higher power rate WTs for the offshore market.
The use of superconducting materials in field winding permits to achieve much stronger magnetic fields (B). Additionally as the armature teeth can be removed more space for AC stator winding can be gained and thus higher linear current densities (A) can be reached. Therefore, for a given torque, shear stress higher than in conventional generators can be obtained, with the result of machines with lower volumes and weights of active parts.
Thus the use of superconducting materials in DD generators arise as a promising solution for achieving reliable, lightweight and higher power rate WTs for the offshore market.
Superconducting generators (SCG) for wind turbines are an active R&D field. Several wind generator concepts, currently at different development stages, have been proposed, as those of General Electric, AMSC, RISO-DTU or AML. It is more than remarkable the EcoSwing EC funded project that pursues to construct a 2 MW superconducting generator and install it in an already existing wind turbine.
Nevertheless, some of these concepts face certain technical and economic barriers that could complicate their industrial feasibility for the offshore wind sector. The feasibility of concepts based on conventional pool boiling cryogenic cooling systems is debateable for offshore locations. These systems require huge amounts of cryogenic fluids, as LN2, GHe or LNe, and very complex cryostats and cooling circuits. Complex rotary joints can be also required to exchange huge amounts of cryogen liquids between stationary and rotating parts. SCGs based on LTS wires, as NbTi, require very low operating temperatures, in the range of 1.8-4.2 K, thereby the efficiency of the cooling cycle working at such temperatures is very low, in the range of 0.3%. Beside this, some concepts are based on still expensive and with limited commercial availability materials, such as 2G HTS wires, or materials without attractive cost reduction perspectives such as 1G HTS wires.
SUPRAPOWER project outline and objectives
--------------------------------------------------------------
SUPRAPOWER (SUPerconducting, Reliable, lightweight, And more POWERful offshore wind turbine) is a research project funded by the EU FP7 programme that started in December 2012 and finished in May 2017, with a budget of 5,398,019.03€, being 3,891,058.45€ funded by the EU.
Tecnalia Research and Innovation has leaded a project consortium constituted by 8 top-class European companies and institutions. Industrial partners are a wind farms O&M services company (Ingeteam Services SA), a wind energy engineering (Solute Ingenieros), a superconducting wire developer (Columbus Superconductors SpA) and an offshore engineering company (D2M Engineering SAS). In addition to the coordinator, research partners are a laboratory with deep experience in superconductivity (IEE, Slovak Academy of Science), a university (University of Southampton) and a national institute expert in cryogenics (Karlsruher Institut fuer Technologie-KIT). They also took part as partners during part of the project Acciona Energía, Acciona Windpower and Oerlikon-Leyvold.
The project was conceived to provide an important breakthrough in offshore wind sector by designing an innovative, lightweight, robust and reliable 10 MW offshore wind turbine based on an MgB2 superconducting generator, taking into account all the essential aspects of electric conversion, integration and manufacturability.
One of the objectives was experimentally validating the superconducting generator concept with a scale machine designed and built specifically for this purpose. To keep the maximum similitude between the model and the full scale generator, the scaling down was obtained by reducing the number of poles, maintaining the size of the superconducting field coils identical both in full generator and in small scale machine. The most innovative full scale generator features (superconducting coils, cooling systems, cryostat and quench detection) are similar, too. This resulted in a scale machine which fulfils the basic performance parameters of the 10 MW machine, but with a substantial reduction of diameter, weight and power permitting to test it in actual size laboratory benches.
SUPRAPOWER project overall objectives were as follows:
• To reduce the head mass, size and cost of offshore wind turbines by means of a compact superconducting generator.
• To reduce operating, maintenance and transportation costs and to increase life cycle using an innovative direct drive system.
• To increase the reliability and efficiency of high power wind turbines through a drive-train specific integration in nacelle.
• To maximize the power conversion and wind response of the wind turbine by means of dedicated control systems/procedures.
• To facilitate the development of the offshore wind potential and support its drastic increase.
And the specific aimed impacts were the followings:
• Reduction of the LCoE of the offshore wind by means of a higher power rate lightweight superconducting generator. This cost reduction could permit to achieve the offshore wind sector market perspectives.
• 30% weight reduction with respect to a permanent magnet generator. This permits easier installation processes, reduces vessels and crane costs and decreases mechanical requirements for foundations and floating platforms.
• Elimination of the gearbox that permits a more reliable and efficient drivetrain.
• Low maintenance with respect to other superconducting solutions thanks to use of a cryogenics liquids free cooling system.
• High on site efficiency (95%).
• Provide a wind generator independent from the rare earth materials market, which recently have shown high price volatility.
Project Results:
Note: Please note that due to the nature of the project figures are essential for understanding the obtained results. The attached Final Report .pdf file contains all the required information but illustrated with figures. Please see "D7.7 SUPRAPOWER: Final Report", which will be published in the project website (www.suprapower-fp7.eu) as soon as it is approved by the EC
Introduction
**************
As it has been explained before the main aim of SUPRAPOWER is developing a novel 10 MW superconducting generator and integrating it in an offshore wind turbine taking into account O&M and marine operations. The validation of the generator concept is done through a scale machine, which innovative features as superconducting and cryogenic implementation, modularity and quench detection system are equal or similar in both 10 MW and scale machines
SUPRAPOWER was organized in five R&D work packages, and two more related to dissemination and exploitation (WP6) and management (WP7). The main objectives of the technical work packages are as follows:
• WP1 Validations of the superconducting scale machine: the objective of this WP is the design, construction and test of a scale machine intended to validate the concept of the 10 MW superconducting generator.
• WP2 Design and validation of the superconducting coil based on MgB2 wire: this WP is intended to the development and design of the superconducting coil of the 10 MW generator.
• WP3 Design and validation of the modular rotating cryostat: this WP aims to develop the cooling concept of the generator including the cryostats of the superconducting coils
• WP4 Integration of the full scale superconducting generator in the wind turbine: this WP is the starting point of the project with the design of a new concept of 10 MW generator, all the developments in the other WPs are related to the generator concept defined in WP4. This WP is also intended to the integration of the generator in a 10 MW wind turbine both fixed and floating.
• WP5 Next steps towards the industrialization of the new wind turbine: this WP includes the analysis of O&M and marine operations of wind turbine based on the developed superconducting generator. As part of this WP improvements in the critical components of the generator will be analysed and finally a business plan will be developed.
The main scientific and technical results obtained in SUPRAPOWER are highly interconnected, thus next summary has not been organized by work packages but by the technical explanation of the developed 10 MW SCG concept.
10 MW superconducting generator
***************************************
Superconducting generator concept
------------------------------------------------
One of the main objectives and outcomes of SUPRAPOWER is the design of an innovative 10 MW lightweight, robust and reliable superconducting generator that permits to scale up wind turbines up to 10 MW and beyond and that overcomes some of the presented barriers of other SCG concepts.
The patented generator concept (EP2521252 B1) is a low speed salient pole synchronous machine. It is a partially superconducting generator, superconducting MgB2 wires at cryogenic temperature are used in the field coils while copper wires at ambient temperature in the armature coils. The cooling system uses a cryogen free topology that does not use liquids at cryogenic temperatures. It has a warm iron rotor configuration, which consists on one modular cryostat per pole that encloses only the superconducting coil (at cryogenic temperature) while the iron of the pole remains at room temperature, thus minimizing the mass to be cooled down and facilitating the repair and maintenance works onsite. Heat is extracted by conduction through a thermal collector inside a cryostat which links all the modules. This is cooled by conduction by two-stage Gifford-McMahon (GM) cryocoolers which rotate jointly with the rotor.
This design is intended to achieve much higher magnetic field in the airgap than conventional generators, which involves volume and mass reduction. MgB2 wire has been selected as it shows a cost-performance ratio more competitive than other HTS wires. The cryogen free topology that avoids the use of cryogenic liquids, simplifies the installation in the nacelle of a wind turbine and it also increase the reliability of the system.
10 MW superconducting generator design
************************************************
On the basis of this patented concept a 10 MW, 8.1 rpm and 48 poles based on MgB2 field coils has been designed. Initially, an 11.9 m airgap diameter and 60 poles generator was designed and after several iterations and optimizations 10.1 m airgap diameter and 48 poles generator design has been selected. The design of the SCG requires putting special focus on the design of superconducting coils and cryostats, described in detail in sections 1.3.2 and 1.3.3 of this report.
The electromagnetic layout of the SCG is similar to the one of conventional synchronous salient pole generators. The position of the stator is external to the rotor. As it can be appreciated in one of the main geometrical constrains for the SGC is a polar pitch of 660 mm (distance between poles), this value is higher than in conventional salient poles machines because it is conditioned by the coil and modular cryostat dimensions. Polar pitch influences highly on the design of the generator and determines the diameter of the airgap.
Electromagnetic Finite Element Analyses (FEM) at no load, nominal load and in shortcircuit conditions have been done in order to study the generator steady state and transient performance. Due to the symmetry of the generator, the simulation model has been reduced to 1/24 of the machine, which covers 2 poles. The generator starts saturation with a field current of approximately 50 A, while around 70 A are required for obtaining the rated voltage of 2,280 V in the stator winding terminals.
The iron poles are the most saturated parts with a maximum value of 2.36 T (not taking in account the pole tips) and the peak magnetic field in the superconducting coils is 1.37 T. The interpolar leaked flux represents the 12.5% of the total flux.
The magnetic field in the airgap has been extensively studied as it is one of the most relevant parameters to design an electric machine. The magnetic field, along a polar pitch measured in the diameter of the middle radius of the airgap winding (10.16 m), show a peak value of 1.5 T.
In order to evaluate the characteristics and advantages of the SCG, a DD PMG of same power and similar characteristics has been defined, following when possible the same criteria used for the design of the SCG. The resulting PMG is 10 MW 8.1 rpm machine with 360 Nd2Fe14B poles, 11.9 m airgap diameter and 1.795 m stack length.
The armature, the static part of this generator, basically consists on a magnetic core, a copper winding and an external fixing frame. The armature doesn´t have any magnetic teeth, as these would be saturated due to the high magnetic fields generated by the superconducting coils. Hence an airgap winding configuration has been chosen. Unlike conventional armatures, the most important constraint is that the magnetic core has no slots to hold the air gap winding and to transmit electromagnetic forces from the winding to the magnetic core, therefore specific solution to fix the armature coils must be used. As an inherent consequence of the no use of teeth, the weight of the armature is reduced and there is more space for the winding. The magnetic core is made of a stack of 0.5 mm of M45 magnetic Silicon steel sheets packaged with pressure disks and isolated bolts, which show 5.31 W/kg losses at 1.5 T. One layer lap winding of 3 kV class isolation level copper coils with identical coil span has been chosen.
The rotor of the SCG basically consists on iron back yoke and poles, superconducting coils, cryostats and cryocoolers. The back yoke and poles are made of cast low carbon Steel AISI 1008. Each field coil is housed in a modular cryostat concentrically placed around the pole.
The calculated onsite efficiency of the SGC, taking into account the power consumption of all the auxiliary elements as compressors and cryocoolers, is around 95.2% at the rated power, while the PMG one is around 94.5%. It is remarkable that the efficiency curve of the SGC is slightly above the PMG for all the power range. Cost estimation has been done on the basis of MgB2 wire cost of 3 €/m. Armature Cu coil cost of 50 €/kg and iron cost of 5.5-8 €/kg have been considered, depending on the component and current commercial costs of the auxiliary components. An overall SCG active parts cost of 307 k€/MW has been obtained, while the cost estimated for the PMG with NdFeB magnets price of 39 €/kg (November 2013) is 313 k€/MW. The SCG has a huge margin for further cost reduction driven by the MgB2 wire performance improvement and the industrialization of the manufacturing process.
MgB2 superconducting coils
*********************************
Superconducting wire
-----------------------------
Columbus Superconductors SpA with the collaboration of IEE has designed and manufactured a specific MgB2 wire for magnet applications. It consists in a 3 x 0.5 mm2 superconducting tape with 19 filaments embedded in a Ni matrix. The tape is produced with the Powders In Tube (PIT) manufacturing process in single batch length between 1.5 and 3 km.
To enhance the thermal stability of the wire and of the devices in which it’s implemented, a high purity copper strip is soldered on the tape surface after the final heat treatment of the wire. All the issues related to the perfect alignment of the copper strip on the tape surface have been solved by optimizing the soldering process and now this strip can be perfectly aligned and soldered.
Strain tolerance of the tape subjected to stress and bending has been extensively studied. In particular, the practical limit of bending diameter is 150 mm with the copper outside. Cost-performance ratio is in the range of 20 €/kA m and Columbus anticipates an improvement of this ratio based on the enhancement of the performance mainly due to new precursors, better process implementation, and manufacturing costs reduction due to economies of scale. The main details of the MgB2 wire can be found in “D2.1 Improved MgB2 wire” SUPRAPOWER project public deliverable.
IEE has analysed the perspectives of improving the critical current at relatively low fields (1-1.5 T) by working on the precursors powder quality in the MgB2 composition and a different synthesis route. Several wire architectures have been also analysed implementing the results about the powders, increasing the filling factor of the wire the strength of the metal sheaths of the wires. Finally a new wire configuration has been proposed to be implemented for future intermediate magnetic field application, as the one of SUPRAPOWER.
The critical current of this wire (1.5 mm diameter) has been analysed at different operating temperatures. At 20K and 1.8T field the critical current is higher than 600 A, three times higher in comparison with the tape used in the scale machine. This improvement could permit to simplifying the SC coil by reducing the length of MgB2 wire need and in consequence the number of turns of the coil. Finally this would permit further reduction of the weight and volume of the SCG. All the details of this study have been reported in “D5.1 Improvements in the critical components of the 10 MW SCG” SUPRAPOWER project public deliverable.
In conjunction with the experimental work performed in the prototype pancake coils during the project, a series of simulations were performed at University of Southampton to understand better the behaviour of the winding under external magnetic fields. The influence of a ferromagnetic metal matrix in the magnetic field inside the filaments of an MgB2 tape was analysed for different filament configurations. The results showed that the reduction of the critical current would be 1% - 2% depending on the magnetic field applied. The analysis was extended to the case of a SC coil based on the scaled machine coil designs. The determination of the load line of the coil, considering it as a bulk ferromagnet, leads to an underestimation of the critical current of the coil. If the filaments are not considered ferromagnetic, a reduction of the average magnetic field in the coil is observed. For the stack of 9 double pancake coils the magnetic field values and load-line of the coil can be determined using a pure non-ferromagnetic model without obtaining significant differences which saves computational time.
Additionally, modelling analysis to study the mechanisms that change the MQE in superconducting coils was done using two models: a FULL 3D and a CM. It has been seen that for a given set of parameters the MQE of the coil can be in the order of magnitude of that predicted by 1D analytic equation. The MQE is strongly affected by the thermal conductivity of the insulation and can increase the MQE of the coil of few orders of magnitude from 1D to 3D, due to the change in the volume of the MPZ. However the change of the thermal conductivity of the coil on its own does not explain the change in trend in the MQE observed experimentally in a solenoid coil. Only a smooth transition from 1D to 3D is observed in the simulations. To obtain the transition observed experimentally, cooling needs to be added from a boundary layer. This induces the transition due to the movement of the hotspot to an adjacent turn of where the heat is deposited due to the effect of cooling. Only the FULL 3D model was able to produce the transition while the CM model overestimates the MQE values. Therefore, the FULL 3D model is preferable to study the quench processes even if it has a higher computational costs, since there are mechanisms that cannot be reproduced with the CM.
Superconducting coil design
--------------------------------------
Field coils are constituted by a stack of 9 racetrack MgB2 double pancake (DP) coils connected in series. Each DP is made of two identical layers of MgB2 wire wound in opposite direction without splices and with an oxygen free copper sheet placed between the two layers to assure the refrigeration of the wire. Once wound, the DP is vacuum impregnated with Araldit F resin. The DP coils are stacked between two 8 mm coppers caps and they are connected in series through specific electric joints. The Cu inserts of the 9 DPs are thermally connected through Cu laps to the upper and lower caps, which are connected to the thermal collector to extract the heat from the coils.
The design of the coil has been supported by electromagnetic and thermal simulations carried out by the University of Southampton and Tecnalia. One of the most important design aspects is thermal behaviour of the coil given that heat is extracted by conduction from one connection point to the cryocooler. Simulated temperature gradients are below 0.1 K showing a good thermal stabilization that the Cu insert provide inside each DP. The models have been later validated by the experimental results.
Initial Small test coils
----------------------------
Before any full size coils were constructed, two small test coils were produced to validate the design and manufacturing process. First, a circular, 1 layer, 7 turns coil was built to analyse the wire behaviour under specific winding diameter and cryogen free refrigeration. Second, a circular, 30 turns DP coil was built to validate the winding, impregnation and test procedures to be applied in the real scale DP. These were tested in a general purpose cryostat at same working conditions. The manufacturing process and test results of this last coil are available in the SUPRAPOWER’s public deliverable “D2.3 Report on Design and manufacturing of an MgB2 coil”.
First MgB2 real scale double pancake coil
--------------------------------------------------------
After this small size coil, a first real scale DP coil was manufactured by Tecnalia at CIEMAT’s facilities in Madrid.
The DP has been extensively tested in a laboratory cryostat at Tecnalia [26]. All the details of this work are available in “D2.4 SUPRAPOWER Test report of the MgB2 coil” public report.
The transition to superconducting state during the cooling down of the DP has been recorded with 0.9 A applied current. Similar test has been carried out switching off the cryocooler in order to register the transition to resistive state during warming. Through the analysis of the obtained results it is estimated that the average transition in the coil is at 35.7 K. Voltage measurements have been fitted by using a power law (E= E0 (I/Ic)^n) where E is the electric field measured between the two ends of the coil, E0 is the criteria for the critical current, I the transport current and the critical current Ic(B,T) and n(B,T) the fit parameters. Ic of 91.79 A, 133.59 A and 146.26 A have been obtained at 30 K, 28.5 K and 27.5 K respectively. The fitted n value is about 10 for the overall coil at the studied temperatures. These Ic values are consistent with the expected results derived from the calculated load lines.
Finally it is concluded that the electrical behaviour of the DP, the obtained critical current and thermal behaviour as expected and according to the electromagnetic and thermal simulations carried out.
Manufacturing and test of the scale machine MgB2 field coils
---------------------------------------------------------------------------------
Once validated the first DP, Columbus and TECNALIA has manufactured and tested the 2 superconducting coils of the magnetic rotating validator.
The process to manufacture the complete DP ready to be assembled in the coil is quite complex. Once the winding is finished (about 360 meter of tape in two layers with 72 turns each one, by using the same procedure as in deliverable D2.3. Then the tape 2 ends are electrically connect the copper pieces that perform as electrical exists, this is done by soldering the copper pieces to the tape of the external turns.
Throughout the project lifetime the manufacturing process and the design of the coils have been updated according to the obtained experimental results. Thus for example the electrical connections were improved, copper pieces for thermal connections were slightly redesigned to improve the manufacturability and the impregnations process was modified so that to impregnate 9 DPs at time, while the first DPs were impregnated one by one.
The two coils have been extensively tested in a laboratory cryostat before its assembly inside the modular cryostat of the scale machine. Test results in the first coil have shown a very homogenous thermal behaviour, with less than 0.6 K difference between the opposite ends of the. It has been found that one of the DPs is damaged while the other DPs perform as expected.
Finally, even though one DP performance has not been as desired due to some manufacturing defects, it has been concluded that the developed coils are suitable for operating in a 10 MW wind turbine. These coils have then assembled in the scale machine.
Quench detection and protection system
------------------------------------------------------
The quench detection and protection systems are essential elements to protect the SCG in case of quench. This aspect has been extensively studied for both the SCG and the scale machine.
The quench detection system is based on the resistive voltage measurement and is a redundant system with three levels of protections.
• First level: threshold voltage, measured in the power supply terminals, which will work only in case of failure of the other two protection levels.
• Second level: comparison of two coils voltage, only different in case of quench.
• Third level: threshold voltage of the resistive voltage of the inductor circuit. Inductor’s circuit current and voltage are measured and then the inductive voltage component is calculated and subtracted.
The quench detection systems have been implemented in the National Instrument CompactRIO hardware platform. A protection system has been developed by Tecnalia for the scale machine. The system is based on the fast discharge on a dump resistor as soon as the quench is detected and switching off at the same time the power supply that feeds the coils. The resistor bank is constituted by 2 group of resistances serial connect as shown in, one is mounted on the back yoke that rotates jointly with the rotor and the other one is a stationary resistors bank connected to the coils through the slip rings. This topology permits to protect the coils even in case of a fail in the slip rings.
The protection of the rotor circuit with a simple dump resistor connected in parallel is valid for the case of the scale machine. However, it is not totally satisfactory for the 10 MW SCG due to the large discharge time, which could yield in hot spot in case of poorly refrigerated segments, so that solutions for MW scale machines must be further studied.
Cooling system and cryostats
**********************************
On the basis of the initial concept developed in the patent, KIT in collaboration with Tecnalia has designed the SCG cooling system and cryostats. Warm rotor iron configuration with one modular cryostat per coil has been chosen as the best option although other alternatives have been also considered. This configuration needs lower cooling power as the rotor poles are kept at room temperature and the modularity facilitates the repair and maintenance works onsite. The main drawback of this solution is a higher polar pitch than in conventional generators.
Superconducting coils are contained in 48 modular cryogen-free cryostats, one per coil. Heat is extracted by conduction through a thermal collector which links all the modules. This is cooled by conduction by two-stage GM cryocoolers which rotate jointly with the rotor. In contrast to pool boiling cooling system, only a small quantity of He gas is required for the thermodynamic cycle of the cryocoolers. This He circulates in a close circuit between the rotating cold heads and the stationary He compressors, which are connected by mean of a rotary joint.
The thermal collector consists of two high conductivity thermal circuits enclosed by a cryostat (called non modular cryostat). One is connected to the cryocoolers first stage and to the active cooled shield (Temperature ~80 K), and the other to the second stage of the cryocoolers and to the superconducting coils, maintaining them to their operation temperature at about 20 K.
A rectangular shape modular cryostat has been chosen according to the warm iron pole configuration and in accordance with the coil and pole dimensions and space restrictions. It is mainly constituted by a stainless steel vacuum vessel, a copper active cooled shield with a multi-layer insulation (MLI) of approximately 20 layers and the support structure. The support structure is a key component of the modular cryostat, 8 Ti 6A1 4V titanium rods have been adopted for the connection from active cooled shield to the superconducting coil and another 8 titanium rods from the shield to the vacuum vessel. This support structure is replicated in four locations along the cryostat, resulting 64 titanium rods per module.
First modular cryostat
-----------------------------
As a first step KIT has constructed one modular cryostat to experimentally test the thermal behaviour of one dummy coil inside the cryostat and validate the cryostat design. This first dummy modular cryostat basically follows the same design criteria of the SCG cryostat. Design details and test results of this dummy cryostat are available in “D3.3 First modular cryostat” and “D3.4 Test results report of the first modular cryostat”, SUPRAPOWER’s public deliverables.
The main difference of this cryostat respect to the ones of the SCG or scale machine, is that it integrates the GM cryocooler. The cryocooler is connected with flexible copper bridges, which act as thermal anchor fitting the displacement caused by the thermal contractions. The support structure is also slightly different but based on the same concept of using sets of titanium rods. A copper dummy coil has beeen constructed to emulate the cold mass and heaters has been installed to emualte the coil AC looses.
It took about 20 hours for the first stage thermal anchor to reach the lowest temperature, which is around 33 K. The second stage thermal anchor together with the linked dummy coil required 56.5 hours to reach the lowest temperature, which was around 9.8 K and 9 K respectively. These test results permitted to validate the cryostat and cooling concepts before the final design and manufacturing of these component for the scale machine.
Design and manufacturing of the cryostats of the scale machine
-------------------------------------------------------------------------------------
On the basis of the modular cryostats design, which is exactly the same for the 10 MW SCG and the scale machine, KIT has designed and constructed the cryostats of the SM.
The cryostat of the thermal link, so called non-modular cryostat, adopts a pipe design and a copper bar is used to thermally connect the two modular cryostats and the cryocooler. Both, the vacuum vessel and the thermal shield have a concentric half ring shape in order to achieve an easy assembly. This pipe concept is also valid for the non-modular cryostat of the 10 MW superconducting wind turbine.
The thermal shield made of copper and vacuum jacket made of stainless steel also adopts semi-circle. The cold head of the cryocooler is located in the middle of the semi-circle pipes. On the basis of the design, KIT has constructed 2 modular cryostats, one per coil, and the so called non modular cryostat.
In order to save space, the modular cryostat adopts the welding approach instead of flange connections as proposed in the prototype modular cryostat. It includes several steel stiffeners in the inner and outer part of the chamber, intended to give mechanical rigidity to the cryostat to prevent deformations during the assembly and welding process. After the assembly of the coil in the cryostat and welding the upper and lower parts of the vacuum chamber, these stiffeners have been cut.
He Rotary joint
--------------------
The helium rotary union (RHU) is the element prepared to work in the G-M cryocooler to transfer the helium gas between the rotating cold head and the stationary helium compressor.
The main functionality of the RHU is to connect the stationary part of the machine with the rotary one. It has three parts: gas helium feed-through to connect the cold head with the compressor, slip rings to power electrically the cold heads and optical fibre feed-through to transmit measurement and control signal. The RHU contains two flows of GHe at approximately room temperature, one from the stationary He compressor to the cryocooler (at a higher pressure of about 25 bar) and the other one back from the cryocooler to the He compressor (at a lower pressure of about 8 bar).
Two versions of the RHU have been constructed. The first version showed insufficient sealing during the pressure tests, so that a second version has been constructed. The optimised second version of the RHU has successfully reached the specification and design objectives. The second version of the RHU includes a mid-span bearing to provide a more robust configuration than the first version. The sealing between the stationary and rotating part inside the joint is realized by means of ferrofluid. The rotating seal consists of a permanent magnet, pole pieces, and ferrofluid. The pole pieces and the shaft concentrate magnetic flux from the magnet into a narrow gap region surrounding the shaft. The ferrofluid is attracted to the region known as “stage” forming rings that generate a hermetic seal. Each ferrofluid ring or stage has a pressure capacity of a 70-340 mbar typically.
After manufacturing and assembling the prototype RHU, series of tests have been performed to examine its performance. As result, no noticeable pressure drop has been observed in both helium lines. The leak rate of the RHU has been tested under stationary conditions using a leak detector. The measured leak rate of both helium lines at specified pressure is smaller than the sensor limit of the detector, which is 1.0·10 9 mbar·l/s. The running torque of the RHU at a rotation speed of 150 rpm has been measured.
The results have been recorded after 24 hours and 48 hours, respectively. After starting the unit, the running torque first rapidly increases to a peak value and then gradually reduces to a stable state. Since the torque transducer has a measurement range up to 50 N·m, the actual peak torque is not illustrated. However, using curve fitting it has been estimated that the maximum starting torque is 57 N·m and 62.5 N·m after a time period of 24 hours and 48 hours, respectively. The normal running torque of the RHU was around 23 N·m. These torque characteristics are very relevant for the selection of the appropriate AC motor to drive the RHU.
After this test the rotary joint has been validated. The design and manufacturing details and test results are available in “D3.5 Rotary joint: specifications, construction and test results” SUPRAPOWER’s public deliverable.
Scale Machine: Rotating Magnetic Validator
*******************************
Scale machine design
--------------------------------
At first it was studies the possibility to design and construct a scale generator (SG) in the range of 500 kW. To keep the maximum similitude between the model and the full scale generator, the power reduction was obtained by reducing the number of poles from 48 to 4, maintaining the size of the SC rotor coils identical both in full and small scale generator. However this option was finally discarded due to the complexity and high cost of constructing 4 superconducting coils and cryostats and mainly due to the very high cost of the bench needed to test a 500 kW superconducting machine.
On the basis of the first version a second scale machine (SM) has been designed overcoming the above mentioned problems but without relaxing its objectives. The final design of the SM is rotating magnetic machine that basically consists on a rotating external rotor with 2 superconducting field coils, each one enclosed in a modular cryostat around an iron pole at room temperature. Heat is extracted by conduction through a thermal collector that links both coils. Modular cryostats, coils and poles have the same size and are operated at similar conditions as the 10 MW SCG ones. The cooling concept is the same of the SCG, with the same cryocoolers rotating with the rotor and a rotary joint that links a stationary He compressor with the cryocoolers.
It incorporates a magnetic mirror in order to reproduce with the highest possible accuracy the magnetic behaviour and mechanical stresses of the rotating coils and cryostats inside the iron armature that were calculated for the 10MW SCG and SG (1st version). The magnetic mirror is constituted by 2 pieces made of soft ferromagnetic material that operate in non-saturated conditions. Thus it is establish the same boundary conditions in the air-gap region and in the inter-pole space when the magnetic mirror is not saturated.
The armature is not wounded so that this machine can be tested in much simpler test bench of 15 kW. The armature incorporates 4 test coils wounded in the armature magnetic core in order to measure the polar and interpolar leakage flux. This SM can be tested in a 15 kW test bench, which highly facilitates de experimental activity.
Electromagnetic 2D and 3D FEM analysis has been applied to the SM in order to validate the magnetostatic performance of the generator concept. The most saturated parts are magnetic poles, with a maximum value of 2.36 T. The peak magnetic field value in the SC coils is 1.24 T, located in the lower part of the straight section of the coil. This latest aspect has been studied in the detail as the magnetic field determines the performance of the SC wire.
Scale machine construction and test
------------------------------------------------
First of all, the elements of the scale machine has been constructed, including the superconducting coils and cryostats that as previously described has been directly constructed by the project consortium.
It has also been implemented the quench detection and protection system that has been previously described. The quench detection is system is implemented in a control hardware that also monitors the main parameters of the scale machine as temperature, magnetic field, current and voltage signals. This hardware rotates jointly with the rotor and the signals are transmitted to a stationary PC through the fibre optics link of the rotary joint.
Once manufactured all the elements next assembly sequence has been followed:
• Assembly of the coils inside the modular cryostats. This is one of the critical steps of the mounting process as the parts tolerances are very tight and it is necessary to handle all the components to avoid damages and undesired heat entrance points.
• Assembly of Set 1 and Set 2:
o Set 1: formed by the shaft, stator assembly, 2 mirrors, 2 bearings and their supports.
o Set 2: formed by the rotor back yoke, 2 poles and coils inside the modular cryostats
• Assembly of both sets in the support bench and then mounting of other components as slip rings, balancing rings and measuring coils.
• Assembly of the rotary joint
• Assembly of non-modular cryostat and its balancing pieces
All the details of the constructions and assembly of the scale machine have been included in “D1.4 Superconducting Electromagnetic Scale Machine” SUPRAPOWER’s public deliverable. [34]
The SM allows the experimental validation of the main innovative aspects and components at full scale of the 10 MW SCG. First of all, the test of the cooling system permits to probe the mechanical design of the cryostats, the good operation of the rotary joint and cryocoolers in rotation, the behaviour and efficiency of the overall heat extraction circuit and the capability of achieving the 20 K operating temperature in the coils. SC coils can be tested in rotation and under similar working conditions (I, B, T) to the SCG ones. The quench detection and protection system can also be tested and validated. Finally, test coils in the stator allows to check if the magnetic flux generated by the field coils in the airgap and the interpolar zones correspond to the values calculated and obtained in 3D electromagnetic FEM simulations, thus permitting to validate the electromagnetic design approach.
The test protocol, test results and conclusions have been reported in “D1.5 Test results report of the scale machine” SUPRAPOWER’s public deliverable.
Static vacuum and cooling tests have been carried in the SM to check the cryostat vacuum performance, the correct functioning of the cooling system including rotary join (in stationary position) and the temperature distribution in the Cu thermal circuits. A residual pressure of 9.4·10-4 mbar has been obtained by means of a turbomecular pump and before switching on the cryocooler. Next, the cryocooler was connected and the temperature in the ration shield of the cryostat was registered. These results have permitted to conclude that the system, including the rotary joint, performed as expected.
Cooling systems based on cryogen-free GM cryocoolers has very limited heat extraction power. Due to this it is indispensable to have a good current lead design with almost no heat transport or generation. First design of current leads was based on four SC Sumitomo BSCCO tapes (DI-BSCCO) Type HT, twisted along stainless steel capillary tubes. This design resulted not valid because of potential damages during tape twisting due curvature restrictions. In the new design the SC tapes have been directly soldered to the Cu terminals. Tapes have been inserted inside a Cu film with tubular form, which acts as current alternative path in case of quench.
Current leads have been tested under liquid nitrogen (LN2) before assembling them in the scale machine. The obtained critical currents (Ic) are below the expected ones (around 600 A) according to the manufacturer data. Also the “n” values are low, with the exception of one of the current lead. These results show that 3 of the 4 current leads, despite the performance is not as good as expected, could be good enough to energize the SC coil. The forth current lead should be manufactured and tested again. As conclusion the development of this component requires further R&D activity.
At the time of writing this report, the SC coils and the cryogen free cooling concepts have been successfully tested. Due to the complexity of the system, several components have required redesigns during the tests. Finally, taking into account these difficulties, the test results have not permitted to completely validate the scale machine. Additional R&D work should be done for validating the scale machine.
It has been concluded that the developed generator shows technical benefits in comparison to direct drive PMGs and at a very competitive cost. However, the development of the SCG still requires further R&D, validation and demonstration activities.
Some future R&D working lines have been identified. The improvement of the MgB2 wire performance would permit to improve the SC coil by reducing the amount of wire needed. This would permit to ease the design and manufacturing process of the coil and would yield in more available space inside the cryostat, which is also considered as an important improvement. Some wire improvements have already been analysed under the scope of SUPRAPOWER (see public Deliverable D5.1). The cooling system is based on heat extraction through conduction, which avoids the use of cryogenic liquids. However, this implies complex and very tough tolerances heat extractions circuits, which are made of a material (OFE Cu) difficult to mechanize with precision. The improvement of the heat extraction circuits would permit to ease the assembly of the system and increase its reliability. The design and manufacturing of the current leads should be improved, as they are key component for the validation of the system.
10 MW superconducting generator wind turbine
******************************************************
Integration of the full scale SCG in both fixed and floating offshore wind turbines
----------------------------------------------------------------------------------------------------------
R&D studies of the integration of both PMG and SCG in both flxed and floating WTs have been carried out by SOLUTE to evaluate the effect of the generator weight reduction on the rest of the turbine structural components.
First of all, as there were none or few available references of blades for 10 MW turbines, a special blade design has been performed with the aim of having lower weight. This concept has been extrapolated form NREL-5MW blade with no changes in aerodynamical profiles but with an optimization in weight. As a result, the Solute Hybrid Blade of 92.5m (SHB-92.5) made of glass fibre and carbon fibre has been released.
On the basis of this blade and the SCG, the drive train has been dimensioned with the main aims of reducing manufacturing hours, reducing the overall weight and offering a clean load path. The mechanical integration of DD generators of very large diameter requires different solutions than geared drive trains. The mainframe of a geared drive nacelle is substituted by a component named goose neck that attaches the generator to the WT tower.
The pitch system is located out of the hub to reduce its weight and size and also to ease maintenance operations. All the electromagnetic elements and the support structure of the SCG are designed so that can be divided into smaller parts to facilitate transportation and assembly operations. The generator rotor support is directly connected to hub, while the stator support offers a connection to the goose neck. It supports the whole drive train and nacelle rear structures, while it offers a connection with the yaw system and tower, offering a load path for the loads coming from the drive train. The hub, rotor and stator supports and the goose neck have been simulated as cast iron.
On the basis of the Rotor nacelle assembly (RNA) design, the tower and the fixed offshore substructure monopile has been designed. The main aim of this design process is to avoid rotor frequency (1P), designing a structure (tower plus monopile) with a first FA (Fore Aft) natural frequency 20% above 1P, i.e. around 0.17Hz. With a hub height of 125m above MSL (Medium See Level) the tower is 110m height and is supported by a monopile that stands 10.8m above MSL. Final calculated 1st FA frequency of the system (tower plus monopile) is 0.184Hz.
With the meteocean analysis of the selected location, and considering a set of load cases, which, according to experience, are the usual driving ones, the offshore loads analysis, following IEC 61400-3 standard has been performed. As a result, extreme and fatigue loads along the wind turbine have been obtained. Comparing this final loads with the PMG ones, main conclusion is that at RNA loads doesn’t differs too much, but along the tower, with a lower THM, loads decrease significantly (-10%). A strength analysis has been conducted, taking into consideration the aforementioned loads for fatigue and extreme conditions. As result some little reinforcements have been applied at goose neck, shaft and shaft support.
The results of these studies bring 25.6% weight reduction of the active parts of the SGC. The weight reduction of the THM is of 7.2%, this drop in the percentage is because the aerodynamic forces in a 10 MW turbine have more influence on the support structure than the weight of the active parts of the generator. It has been also evaluated that the achieved THM reduction permits an 11% reduction of the tower weight and 9% of the monopile.
Finally it has been estimated that this weight reduction could lead to cost shaving of 0.5 1 M€ respect to a PMG, even though at higher cost of the SCG (unless for the first units introduced in the market).
It has also been analysed the integration of the SCG in a floating platform. Nautilus floating platform has been taken as reference, as this platform is intended for 5 MW wind turbines, Tecnalia has made the redesign and main calculation for adapting the platform to the 10 MW, SCG scaling the 5 MW design for a 10 MW turbine. Dimensions have been calculated so that the floater could withstand maximum overturning moment along turbine life time and basic constrains for manufacturing, load-out and assembly process have been also taken into account. Then, mooring and hydrodynamics have been checked. Mooring have been designed by quasi-static methodology for 100 years return period, most representative mooring parameters have been obtained, such as chain diameter, length and anchor weight. Hydrodynamic features have been set by diffraction/radiation software, obtaining added mass, damping and stiffness coefficients which are essential to feed the numerical model used for dynamic simulations.
Furthermore stability and local strength have also been analysed. Stability and structure have been checked according to DNV standards. Then the final design has been validated by coupling codes, analysing a wide range of design load cases.
The same wind turbine used for fixed structure has been considered to get a good comparison between floating and fixed offshore substructure. As floating substructures show higher inertial loads than fixed ones, the tower model has been reinforced accordingly. Obtained total tower mass is 898 tons, i.e comparing with fixed turbine tower, it shows a 33% increase of mass and a final 1st FA frequency (tower plus floater) of 0.225Hz. Taking into account the calculated floating platform mass of 12,991 tons, the overall mass, tower plus floater, is 13,889 tons. It is estimated that the mass of the floater is in the range of 3-5% less than for the case of PMG base turbine. Load calculation studies have been performed for the same metoceanic conditions and load cases as for fixed substructure, checking that loads in the mooring lines, base of the tower, floater movements, etc are in the allowable limits.
Marine operations analysis
------------------------------------
To analyse the feasibility of a 10 MW SCG offshore WT it is also necessary to investigate and analyse transport, assembly and installation procedures and asses the benefits at this respect of the obtained weight reduction. All the technical details and results of this study are available in “D5.2 Study of transport, assembly and installation of offshore 10 MW superconducting generator” SUPRAPOWER’s public deliverable.
On the basis of main Rules and Guidelines for marine operations, the previously presented fixed WT and floating WT configurations have been studied, including: marine operations, required installations vessels, preliminary hydrodynamic analysis and hazard identification.
Marine operations have been assessed for both configurations and several options have been considered defining the process step by step from the harbour to the offshore project area. Then, based on preliminary assumptions on foundations and on the characteristics of the WT and the SCG (mainly weights and dimensions), the installation vessels have been analysed, highlighting the ones able to install the new 10MW SCG WT.
In general terms, it has been concluded that the transportation and the installation of the SUPRAPOWER’s WT can be realised with present-day equipment and with no major differences compared to existing WTs. It is concluded that thanks to the previously analysed weight reduction, a slight cost reduction could be achieved respect to a PMG WT. It has been estimate that the cost range for the installation of a 100 MW wind farm in the selected locations, with fixed foundations, could be in the range of 4-4.8 M€ and 28 to 26 days of installation time for the case of SCG WTs. While for the case of PMG WTs it has been estimated a cost in the range of 4.5-5 M€ and 27-38 days installation time.
The general conclusion that these analysis shows is that even if the SUPRAPOWER’s 10MW wind turbine is more powerful and hence bigger than current WTs (4.8 MW mean installed power rating), all the transportation, installation and assembly operations on both fixed and floating foundations can be performed with classic and existing means. All the operations can be performed following existing methodologies, with existing wind farm offshore means (vessels, cranes, tugs), by crews already trained and familiar with these classic tasks.
The safety considerations of the installation of the fixed and floating foundation WT have been analysed considering the means and procedures previously defined. A Hazard Identification (HAZID) method has been applied following the marine operation Rules & Guidelines, assessing the Severity and Frequency of all possible risks, defining the possible preventive and protective measures that may reduce them. Finally each risk has been classified in 3 different critically categories ranging from acceptable or low risk, to unacceptable or high risk.
Main HAZID conclusions are that typical superconductive risks, mainly related to the high pressure Helium, can be well mitigated and prevented, reducing these risks to low levels. On the other hand marine operations are similar than for other same sizes wind turbines, well known and mastered, with no additional risks.
Operation, maintenance and reliability analysis
--------------------------------------------------------------
O&M represents an important cost in the life cycle of an offshore wind turbine, indeed scaling up the power of the wind turbines could contribute to reduce these cost per MWh. Due to relevance of this aspect and the novelty of the superconducting technology, Ingeteam has carried out a study of the reliability, availability and maintainability of the SCG 10 MW WT, intended to evaluate the impact of the SCG in the O&M strategy and costs. Then a risk assessment of the O&M operations has been completed and on the basis of these outputs a Criticality Analysis has been completed. Finally, once identified the most critical components the maintenance strategy has been defined
The risk Assessment has shown few added risks between a PMG and SCG, which are related with very low temperatures and higher radiation generated by the powerful magnetic fields generated by the SC field coils. On the basis of this data a Criticality Analysis has been carried out. The analysis has shown the need of taking special care with the SCG, the Pitch System, the Blade System, the Power Electronics and the Yaw System. Superconducting elements, the cooling and vacuum systems have been deeply studied. It is clear that introducing more elements into a system decreases its reliability. Although in this case the number of elements introduced is high, they are theoretically not very critical. An exception is the rotatory joint, but an alternative to eliminate this “Middle-High” critical element has been analysed (see public Deliverable D5.1)
All the elements introduced related with the superconducting System have never been used in a similar application and least of all in such a harsh environment. The forces and vibrations to endure are also much noteworthy inside a wind turbine than in the most common applications for these technologies as MRI or magnets for scientific facilities. With this in mind, it seems logical that in the first prototypes and commercial units of this potential product, an in-depth study of behavior should be performed, requiring much more often and detailed maintenance and check visits than the theoretically needed. The modular design of the SCG, avoiding the need of using the most expensive and with low availability vessels is definitely an advantage that reduces significantly the “Impact of the Failure”.
Finally, with all this data and the best practices on offshore wind power O&M strategies, some guidelines have been produced for the Maintenance Strategies of a wind farm equipped with SUPRAPOWER technology WTs. The most critical components identified in the Critically Analysis will need a Condition Monitoring Systems so that to apply a predictive maintenance strategy. Preventive Maintenance periods, required tasks and means have been assessed for each maintenance period. For each of these tasks, the logistics (boats, vessels, helicopters, forecasting...) needed to perform all the operations were studied to achieve the highest availability of the plant with the less Operational Expenditure.
It has been estimated that installation, O&M, decommissioning and all other marine operations do not complicate due to the use of a SCG and moreover 0.9 M€ savings could be achieved in this field. This cost reduction is achieved thanks to the modularity and lighter weights of the components of the WT, which permit to have faster marine operations.
Potential Impact:
SUPRAPOWER potential impact
**************************************
SUPRAPOWER (www.suprapowe-fp7.eu) is research project funded by the EU FP7 programme that started in December 2012 and finished in May 2017. The project was conceived to provide an important breakthrough in offshore wind industrial solutions by designing an innovative, lightweight, robust and reliable 10 MW class offshore wind turbine based on an MgB2 superconducting generator, taking into account all the essential aspects of electric conversion, integration and manufacturability.
New Superconducting wind generator
--------------------------------------------------
The main outcome of SUPRAPOWER project is a novel 10 MW superconducting generator. This generator has been patented (EP2521252 B1) both in Europe (Spain, Germany, UK and France) and in the USA. The patented concept has been developed in detail and validated through a scale machine.
This generator is a low speed salient poles synchronous machine. It is a partially superconducting generator (SCG), superconducting MgB2 wires at cryogenic temperature are used in the field coils while copper wires at ambient temperature in the armature coils. The cooling system uses a cryogen free topology that does not use liquids at cryogenic temperatures. It has a warm iron rotor configuration, which consists on one modular cryostat per pole that encloses only the superconducting coil while the iron of the pole remains at room temperature. Heat is extracted by conduction through a thermal collector inside a cryostat which links all the modules. This is cooled by conduction by two-stage Gifford-McMahon (GM) cryocoolers which rotate jointly with the rotor.
This generator concept gives answer to the need of cost effective and more power wind turbines, while overcomes some of the challenges faced by other previously developed superconducting generator concepts. On the one hand, the selected superconducting material, MgB2, shows a much lower cost performance ratio than other HTS materials. On the other hand, the cooling system uses a cryogen free topology that does not use liquids at cryogenic temperatures, which highly simplifies the required cooling installation and minimizes the maintenance operations. Additionally, the modular design of the superconducting and cryogenics developments makes more feasible and less costly the corrective maintenance operations, which could be needed in case of a major failure.
The design process has also taken into considerations marine operations and O&M aspects, as they both represent a significant fraction of the offshore wind turbines LCoE. It has been concluded that using a novel SCG, which uses a technology never used before in wind turbines and least of all offshore, do not bring significant additional risks.
The SCG onsite efficiency, taking into account the power consumption of all the auxiliary elements as compressors and cryocoolers, is around 95.2% at the rated power, while the PMG one is around 94.5%. It is remarkable that the efficiency curve of the SGC is slightly above the PMG for all the power range.
SCG active parts cost has been estimated in 307 k€/MW, with current MgB2 costs, while the cost estimated for the PMG with NdFeB magnets price of 39 €/kg is 313 k€/MW. Thus the developed SCG is already cost competitive and with a huge margin for further cost reduction driven by the MgB2 wire performance improvement and industrialization of the manufacturing process.
This generator concept has been experimentally validated with a scale magnetic rotating machine designed and built specifically for this purpose. Main innovative features as superconducting and cryogenic implementation, modularity and quench detection system are equal or similar in both 10 MW and scale machines. Superconducting coils and cryogen free cooling system have been extensively investigated and as result real scale coils, modular cryostat and cooling systems have been constructed and experimentally validated by the consortium. At the time of writing this report the test of the scale machine was in progress.
It has been concluded that the developed generator shows technical benefits in comparison to direct drive PMGs and at a very competitive cost. However, the development of the SCG still requires further R&D, validation and demonstration activities
Facilitate more power and reliable wind turbines for LCoE reduction
------------------------------------------------------------------------------------------
As it has been explained before higher power rate wind turbines are pointed out as the innovation with higher Levelized Cost of Energy (LCoE) reduction potential. The state-of-the-art shows that both geared and direct-drive wind generators are difficult to scale up to 10 MW and beyond due to their huge size and weight.
Indeed, the development of large scale offshore wind turbines in the range of 10-20 MW and the improvement of the reliability through use of new materials, advance rotor design and control and monitoring systems, were relevant objectives of the “Investing in the Development of Low Carbon Technologies (SET-Plan),Technology Roadmap published by the EC in 2009”.
The SUPRAPOWER 10 MW SCG highly contributes towards these objectives by facilitating the development of cost effective large and reliable wind turbines.
It has been studies the integration of the 10 MW SCG in both flxed and floating wind turbines to evaluate the effect of the generator weight reduction on the rest of the turbine structural components. The obtained results have been compared with a permanent magnets generator (PMG) wind turbine so that to be able to quantify the benefits. The developed SCG shows 25.6% weight reduction of the active parts respect PMG. The weight reduction of the tower head mass (including nacelle and blades) is of 7.2%. Finally it has been evaluated that the achieved head mass reduction permits an 11% reduction of the tower weight and 9% of the foundations (monopile). It has been calculated that cost savings in the range of 0.5-1 M€ could be reached thank to materials savings in the structural part of the wind turbine. For the case of floating wind turbines it estimated that a 3-5% weight reduction could be obtained in the floating platform thanks to the reduction of weight of the turbine.
Transportation, installation and assembly, O&M, decommissioning and all other marine operations, on both fixed and floating foundations do not complicate due to the use of SCG. All the operations can be performed with classic and existing means and with already applied methodologies, so that there are not drawbacks related to using superconducting technology in the generator. It has been estimate that the cost range for the installation of a 100 MW wind farm in a representative locations, with fixed foundations 10 MW SCG wind turbine, could be in the range of 4-4.8 M€ and 28 to 36 days of installation time, which is the range of 5-10% less than for the case of using PMGs. Moreover the direct drive topology contributes to increase the reliability of the turbine and O&M cost can also be reduced driven by faster marine operations thanks to the modularity and lighter parts of the generator. It has been estimated that 0.7 M€ savings could be achieved related to improved reliability and maintainability.
A critically analysis has shown that the most critical components are conventional ones as power electronics, blade system or yaw mechanism. An exception is the Helium rotary joint that has been identified as medium-high critical component, so that the reliability and performance of this component in offshore conditions should be demonstrated.
As conclusion the developed SCG can significantly contribute to the reduction of the LCoE of offshore wind. Nevertheless there is still R&D work to be done for the development and demonstration of a 10 MW SCG offshore wind turbine.
Finally facilitating the deployment and development of more cost effective offshore wind farms, contributes to EU targets of having at least 27% share of electricity produced by renewable sources, as indicated by the 2030 Climate Energy Package, “A policy framework for climate and energy in the period from 2020 to 2030”.
Exploitable foreground
------------------------------
As result of the project the project consortium has achieved the following exploitable foreground, that in certain cases have applications also in other sectors. The main exploitable results are the followings:
• 10 MW superconducting generator (main beneficiary: Tecnalia). This is the main results of the project and the other obtained results give additional value to the patent. The exploitation strategy is the IP commercialization after the project end.
• MgB2 wire (main beneficiary: Columbus). The strategy is direct sales to Coil and high current cable manufactures. The wire developed in the project can have further applications than those related to generators.
• MgB2 coils (main beneficiaries: Columbus, Tecnalia). On the one hand the strategy is manufacturing coils for generators manufacturers. On the other hand the strategy also considers providing to third parties services for the design and manufacturing of superconducting coils.
• Cryogen free cooling system & Rotary joint (main beneficiary: KIT). The strategy is the commercialization of the IP and/or provision of R&D services.
• Design of 10 MW turbines. Integration of superconducting machines (main beneficiary: Solute). The exploitation strategy is offering high quality and specific engineering services regarding wind turbines & offshore.
• O&M procedures for 10 MW wind turbines (main beneficiary: Ingeteam).The exploitation strategy is providing services for high power wind turbines offered under “Trade Secret”.
• Marine operations procedures for 10 MW wind turbines (main beneficiary: D2M). The strategy is the exploitation of the gained knowledge by providing services in the field of marine operations for offshore wind turbine or other marine renewables.
Dissemination activities
***************************
During SUPRAPOWER project dissemination and communications have been an essential activity and extensive effort has been made to disseminate the superconducting generator concept, its associated results and in general terms the results of the project. During the first part of the project dissemination material was mainly focus to the communication of the project objectives and background, while the second part has been focused to communicating the project results and to the exploitation of the foreground. After the end of the project some further activities will be carried out in order to disseminate the final results.
More dissemination activities than those initially planned have been carried out. As a summary 2 press release have been produced, the consortium has taken part in more than 22 conferences and 13 articles have been published in peer review magazines.
Two workshops have been organised. The first one was held in April 2015 in Bilbao, Spain as a side event of the Bilbao Marine Energy week (www.bilbaomarinenergy.com). This workshop was focused on High power electric generators for cost reduction of offshore wind. Its aim was to provide an overview about trends towards high power wind turbines based on both conventional and superconducting electric generators and get feedback from industry and academia.
The second has been held in May 2015 at TECNALIA’s facilities in Bilbao, Spain. The workshop has presented the main results achieved and lessons learnt during the project, including a visit to the scale machine. The event also brought main European research initiatives in the field of superconducting generators for wind energy. This event took place at TECNALIA’s facilities in Bilbao, Spain.
The project web site (www.suprapower-fp7.eu) has been a key tool for the dissemination of project objectives and results, and it has accumulated more than 15,000 visits. In the public are of the web site several articles and presentations are available, including all the presentations of the 2 workshops of the project. Moreover it also has 13 deliverables, which are made public as soon as approved by the European Commission. The project web site will be kept active unless one year after the end of the project so that keep of the results available.
All the dissemination activities have been reported in “D6.3 Dissemination Material” public deliverable”, which is available in the project website.
List of Websites:
www.suprapower-fp7.eu
Offshore wind market demands higher power rating and more reliable turbines in order to optimize capital and operational costs. The state of the art shows that conventional generators are difficult to scale up to 10 MW and beyond due to their huge size and weight. Superconducting direct drive wind generators are considered a promising solution to achieve lighter weight machines.
SUPRAPOWER (www.suprapowe-fp7.eu) was a research project funded by the EU FP7 programme. It started in December 2012 and finished in May 2017. The project was conceived to design an innovative, lightweight, robust and reliable 10 MW class offshore wind turbine based on an MgB2 superconducting generator (SCG).
Tecnalia Research and Innovation lead a project consortium of eight outstanding European companies and research institutions: Columbus Superconductors SpA, IEE-Slovak Academy of Science, University of Southampton, Karlsruher Institut fuer Technologie, D2M Engineering SAS, Solute Ingenieros and Ingeteam Services SA.
As result of SUPRAPOWER a 10 MW 8.1 rpm direct drive 48 salient poles synchronous generator has been developed and patented (EP 2521252). It is a partially superconducting generator (SCG), MgB2 superconducting wires at cryogenic temperature (20 K) are used in the field coils, but copper wires at ambient temperature in the armature coils. The cooling system is a cryogen-free topology. It has a warm iron rotor configuration which consists of one modular cryostat per pole that encloses only the superconducting coil, while the iron of the pole remains at room temperature. Heat is extracted by conduction through a thermal collector inside a cryostat which links all the modules. This is cooled by conduction by two-stage Gifford-McMahon cryocoolers which rotate jointly with the rotor. A helium rotary joint connects the stationary helium compressors to the rotating cryocoolers.
The 10 MW SCG active parts show a 25.6% weight reduction with respect to a 10 MW permanent magnets generator (PMG) ones and the total reduction of the tower head mass (including nacelle and blades) is of 7.2%. The achieved head mass reduction permits an 11% reduction of the tower weight and 9% of the fixed foundations (monopile). Thus 0.5-1 M€ of cost reduction can be achieve just due to material savings. Installation, O&M, decommissioning and all other marine operations do not complicate due to the SCG and 0.9 M€ savings could be achieved in this field.
A scaled magnetic rotating machine has been designed and constructed for the validation the concept of the 10 MW SCG. Main innovative elements and features as superconducting and cryogenic implementation, modularity and quench detection system are equal to those of the 10 MW SCG ones. Several new solutions have had to be developed for this first-ever manufacturing of this scaled machine, which was successfully completed by project end. The superconducting coils and the cryogen free cooling concepts have been successfully tested and validated. However, the test results of the scaled machine have not permitted its complete validation and have revealed new requirements for proper functioning.
The project has demonstrated the SCG is cost-competitive with, and provides several technical advantages over direct drive PMGs. Test results and experimental activities have enhanced the value of the patent of the SCG. But the development of the SCG still requires further R&D, validation and demonstration activities.
Project Context and Objectives:
Context of the Project
-----------------------------
Offshore wind energy is rapidly developing motivated by stronger and more regular winds at sea, and increasing site restrictions on land. At the end of 2016 the European cumulative installed capacity was 12.6 GW offshore and 141.1 GW onshore. With almost 300 TWh generated in 2016, wind power covered 10.4 % of the EU’s electricity demand.
In 2010, according to the International Energy Agency, global offshore wind cumulative capacity could reach 100 GW in 2020 of which 40 GW in Europe. Recent regulatory and economic developments in the EU have changed the offshore wind energy perspective for the next 15 years. In light of uncertain governance towards achieving EU climate and energy binding targets, EWEA updated the European wind energy industry’s vision to 2030. EWEA’s new Central Scenario expects 66 GW of offshore wind capacity and it estimates that wind could produce 245 TWh, covering 7.7%.
However, severe cost reductions are demanded in order to be able to reach these market perspectives. In this sense, higher power rate wind turbines are pointed out as the innovation with higher Levelized Cost of Energy (LCoE) reduction potential. Indeed, there is already a trend towards more powerful turbines: the average capacity rating of the 361 offshore wind turbines (WT) under construction in 2016 has been 4.8 MW, 15.4% larger than in 2015 and more than the double respect to 10 years before. The average rated power of new installed WTs is expected to reach 6 MW in 2016 and 8 MW in 2020, and it is already remarkable that in 2016 the first 8 MW turbine has been grid-connected at sea. Main WT manufactures are developing new turbines of 5 to 8 MW, mainly based on medium speed (60-600 rpm) and low speed (8-20 rpm) drive trains with permanent magnets generators (PMG).
For the last years the dominant technology has been the geared drive train induction generator (DFIG), while direct drive (DD) trains still represent a small share of the offshore wind market. Nevertheless, DD trains are a promising solution, as they show higher reliability and lower maintenance costs than geared options, thereby this technology is called to have a much more important share of the offshore wind market. But, current slow rotating PMGs are huge and heavy machines and this put at risk the technical and economic feasibility of DD turbines in the range of 10 MW. Moreover one of the problems associated to PMGs is the high volatility in the price of the rare earth materials used for constructing the excitation magnets. For example in 2011 rare earth prices reached a top of more than 20 times its previous 5 year average. Additionally the geographical concentration of rare earths production in China worries to the industry.
The size of electric generators, as indicated in the expression below, is mainly determined by the torque (T), being this proportional to the rotor volume (V) and the airgap shear stress (σ). This stress is proportional to the average airgap magnetic flux (B) and to the linear current density (A).
P=T·ω ->TαV·σ=V·B·A
Conventional high power DD wind generators have huge torque due to their low rotational speed (ω). “B” is limited by material properties, with typical maximum values of 0.9- 1 T. “A” depends on factors as the stator slot depth, packing factor of copper is slots or cooling/ventilation system. As previously commented these limitations makes high power DD conventional wind generators heavy and bulky.
In this context the use of superconducting materials in DD generators arise as a promising solution for achieving reliable, lightweight and higher power rate WTs for the offshore market.
The use of superconducting materials in field winding permits to achieve much stronger magnetic fields (B). Additionally as the armature teeth can be removed more space for AC stator winding can be gained and thus higher linear current densities (A) can be reached. Therefore, for a given torque, shear stress higher than in conventional generators can be obtained, with the result of machines with lower volumes and weights of active parts.
Thus the use of superconducting materials in DD generators arise as a promising solution for achieving reliable, lightweight and higher power rate WTs for the offshore market.
Superconducting generators (SCG) for wind turbines are an active R&D field. Several wind generator concepts, currently at different development stages, have been proposed, as those of General Electric, AMSC, RISO-DTU or AML. It is more than remarkable the EcoSwing EC funded project that pursues to construct a 2 MW superconducting generator and install it in an already existing wind turbine.
Nevertheless, some of these concepts face certain technical and economic barriers that could complicate their industrial feasibility for the offshore wind sector. The feasibility of concepts based on conventional pool boiling cryogenic cooling systems is debateable for offshore locations. These systems require huge amounts of cryogenic fluids, as LN2, GHe or LNe, and very complex cryostats and cooling circuits. Complex rotary joints can be also required to exchange huge amounts of cryogen liquids between stationary and rotating parts. SCGs based on LTS wires, as NbTi, require very low operating temperatures, in the range of 1.8-4.2 K, thereby the efficiency of the cooling cycle working at such temperatures is very low, in the range of 0.3%. Beside this, some concepts are based on still expensive and with limited commercial availability materials, such as 2G HTS wires, or materials without attractive cost reduction perspectives such as 1G HTS wires.
SUPRAPOWER project outline and objectives
--------------------------------------------------------------
SUPRAPOWER (SUPerconducting, Reliable, lightweight, And more POWERful offshore wind turbine) is a research project funded by the EU FP7 programme that started in December 2012 and finished in May 2017, with a budget of 5,398,019.03€, being 3,891,058.45€ funded by the EU.
Tecnalia Research and Innovation has leaded a project consortium constituted by 8 top-class European companies and institutions. Industrial partners are a wind farms O&M services company (Ingeteam Services SA), a wind energy engineering (Solute Ingenieros), a superconducting wire developer (Columbus Superconductors SpA) and an offshore engineering company (D2M Engineering SAS). In addition to the coordinator, research partners are a laboratory with deep experience in superconductivity (IEE, Slovak Academy of Science), a university (University of Southampton) and a national institute expert in cryogenics (Karlsruher Institut fuer Technologie-KIT). They also took part as partners during part of the project Acciona Energía, Acciona Windpower and Oerlikon-Leyvold.
The project was conceived to provide an important breakthrough in offshore wind sector by designing an innovative, lightweight, robust and reliable 10 MW offshore wind turbine based on an MgB2 superconducting generator, taking into account all the essential aspects of electric conversion, integration and manufacturability.
One of the objectives was experimentally validating the superconducting generator concept with a scale machine designed and built specifically for this purpose. To keep the maximum similitude between the model and the full scale generator, the scaling down was obtained by reducing the number of poles, maintaining the size of the superconducting field coils identical both in full generator and in small scale machine. The most innovative full scale generator features (superconducting coils, cooling systems, cryostat and quench detection) are similar, too. This resulted in a scale machine which fulfils the basic performance parameters of the 10 MW machine, but with a substantial reduction of diameter, weight and power permitting to test it in actual size laboratory benches.
SUPRAPOWER project overall objectives were as follows:
• To reduce the head mass, size and cost of offshore wind turbines by means of a compact superconducting generator.
• To reduce operating, maintenance and transportation costs and to increase life cycle using an innovative direct drive system.
• To increase the reliability and efficiency of high power wind turbines through a drive-train specific integration in nacelle.
• To maximize the power conversion and wind response of the wind turbine by means of dedicated control systems/procedures.
• To facilitate the development of the offshore wind potential and support its drastic increase.
And the specific aimed impacts were the followings:
• Reduction of the LCoE of the offshore wind by means of a higher power rate lightweight superconducting generator. This cost reduction could permit to achieve the offshore wind sector market perspectives.
• 30% weight reduction with respect to a permanent magnet generator. This permits easier installation processes, reduces vessels and crane costs and decreases mechanical requirements for foundations and floating platforms.
• Elimination of the gearbox that permits a more reliable and efficient drivetrain.
• Low maintenance with respect to other superconducting solutions thanks to use of a cryogenics liquids free cooling system.
• High on site efficiency (95%).
• Provide a wind generator independent from the rare earth materials market, which recently have shown high price volatility.
Project Results:
Note: Please note that due to the nature of the project figures are essential for understanding the obtained results. The attached Final Report .pdf file contains all the required information but illustrated with figures. Please see "D7.7 SUPRAPOWER: Final Report", which will be published in the project website (www.suprapower-fp7.eu) as soon as it is approved by the EC
Introduction
**************
As it has been explained before the main aim of SUPRAPOWER is developing a novel 10 MW superconducting generator and integrating it in an offshore wind turbine taking into account O&M and marine operations. The validation of the generator concept is done through a scale machine, which innovative features as superconducting and cryogenic implementation, modularity and quench detection system are equal or similar in both 10 MW and scale machines
SUPRAPOWER was organized in five R&D work packages, and two more related to dissemination and exploitation (WP6) and management (WP7). The main objectives of the technical work packages are as follows:
• WP1 Validations of the superconducting scale machine: the objective of this WP is the design, construction and test of a scale machine intended to validate the concept of the 10 MW superconducting generator.
• WP2 Design and validation of the superconducting coil based on MgB2 wire: this WP is intended to the development and design of the superconducting coil of the 10 MW generator.
• WP3 Design and validation of the modular rotating cryostat: this WP aims to develop the cooling concept of the generator including the cryostats of the superconducting coils
• WP4 Integration of the full scale superconducting generator in the wind turbine: this WP is the starting point of the project with the design of a new concept of 10 MW generator, all the developments in the other WPs are related to the generator concept defined in WP4. This WP is also intended to the integration of the generator in a 10 MW wind turbine both fixed and floating.
• WP5 Next steps towards the industrialization of the new wind turbine: this WP includes the analysis of O&M and marine operations of wind turbine based on the developed superconducting generator. As part of this WP improvements in the critical components of the generator will be analysed and finally a business plan will be developed.
The main scientific and technical results obtained in SUPRAPOWER are highly interconnected, thus next summary has not been organized by work packages but by the technical explanation of the developed 10 MW SCG concept.
10 MW superconducting generator
***************************************
Superconducting generator concept
------------------------------------------------
One of the main objectives and outcomes of SUPRAPOWER is the design of an innovative 10 MW lightweight, robust and reliable superconducting generator that permits to scale up wind turbines up to 10 MW and beyond and that overcomes some of the presented barriers of other SCG concepts.
The patented generator concept (EP2521252 B1) is a low speed salient pole synchronous machine. It is a partially superconducting generator, superconducting MgB2 wires at cryogenic temperature are used in the field coils while copper wires at ambient temperature in the armature coils. The cooling system uses a cryogen free topology that does not use liquids at cryogenic temperatures. It has a warm iron rotor configuration, which consists on one modular cryostat per pole that encloses only the superconducting coil (at cryogenic temperature) while the iron of the pole remains at room temperature, thus minimizing the mass to be cooled down and facilitating the repair and maintenance works onsite. Heat is extracted by conduction through a thermal collector inside a cryostat which links all the modules. This is cooled by conduction by two-stage Gifford-McMahon (GM) cryocoolers which rotate jointly with the rotor.
This design is intended to achieve much higher magnetic field in the airgap than conventional generators, which involves volume and mass reduction. MgB2 wire has been selected as it shows a cost-performance ratio more competitive than other HTS wires. The cryogen free topology that avoids the use of cryogenic liquids, simplifies the installation in the nacelle of a wind turbine and it also increase the reliability of the system.
10 MW superconducting generator design
************************************************
On the basis of this patented concept a 10 MW, 8.1 rpm and 48 poles based on MgB2 field coils has been designed. Initially, an 11.9 m airgap diameter and 60 poles generator was designed and after several iterations and optimizations 10.1 m airgap diameter and 48 poles generator design has been selected. The design of the SCG requires putting special focus on the design of superconducting coils and cryostats, described in detail in sections 1.3.2 and 1.3.3 of this report.
The electromagnetic layout of the SCG is similar to the one of conventional synchronous salient pole generators. The position of the stator is external to the rotor. As it can be appreciated in one of the main geometrical constrains for the SGC is a polar pitch of 660 mm (distance between poles), this value is higher than in conventional salient poles machines because it is conditioned by the coil and modular cryostat dimensions. Polar pitch influences highly on the design of the generator and determines the diameter of the airgap.
Electromagnetic Finite Element Analyses (FEM) at no load, nominal load and in shortcircuit conditions have been done in order to study the generator steady state and transient performance. Due to the symmetry of the generator, the simulation model has been reduced to 1/24 of the machine, which covers 2 poles. The generator starts saturation with a field current of approximately 50 A, while around 70 A are required for obtaining the rated voltage of 2,280 V in the stator winding terminals.
The iron poles are the most saturated parts with a maximum value of 2.36 T (not taking in account the pole tips) and the peak magnetic field in the superconducting coils is 1.37 T. The interpolar leaked flux represents the 12.5% of the total flux.
The magnetic field in the airgap has been extensively studied as it is one of the most relevant parameters to design an electric machine. The magnetic field, along a polar pitch measured in the diameter of the middle radius of the airgap winding (10.16 m), show a peak value of 1.5 T.
In order to evaluate the characteristics and advantages of the SCG, a DD PMG of same power and similar characteristics has been defined, following when possible the same criteria used for the design of the SCG. The resulting PMG is 10 MW 8.1 rpm machine with 360 Nd2Fe14B poles, 11.9 m airgap diameter and 1.795 m stack length.
The armature, the static part of this generator, basically consists on a magnetic core, a copper winding and an external fixing frame. The armature doesn´t have any magnetic teeth, as these would be saturated due to the high magnetic fields generated by the superconducting coils. Hence an airgap winding configuration has been chosen. Unlike conventional armatures, the most important constraint is that the magnetic core has no slots to hold the air gap winding and to transmit electromagnetic forces from the winding to the magnetic core, therefore specific solution to fix the armature coils must be used. As an inherent consequence of the no use of teeth, the weight of the armature is reduced and there is more space for the winding. The magnetic core is made of a stack of 0.5 mm of M45 magnetic Silicon steel sheets packaged with pressure disks and isolated bolts, which show 5.31 W/kg losses at 1.5 T. One layer lap winding of 3 kV class isolation level copper coils with identical coil span has been chosen.
The rotor of the SCG basically consists on iron back yoke and poles, superconducting coils, cryostats and cryocoolers. The back yoke and poles are made of cast low carbon Steel AISI 1008. Each field coil is housed in a modular cryostat concentrically placed around the pole.
The calculated onsite efficiency of the SGC, taking into account the power consumption of all the auxiliary elements as compressors and cryocoolers, is around 95.2% at the rated power, while the PMG one is around 94.5%. It is remarkable that the efficiency curve of the SGC is slightly above the PMG for all the power range. Cost estimation has been done on the basis of MgB2 wire cost of 3 €/m. Armature Cu coil cost of 50 €/kg and iron cost of 5.5-8 €/kg have been considered, depending on the component and current commercial costs of the auxiliary components. An overall SCG active parts cost of 307 k€/MW has been obtained, while the cost estimated for the PMG with NdFeB magnets price of 39 €/kg (November 2013) is 313 k€/MW. The SCG has a huge margin for further cost reduction driven by the MgB2 wire performance improvement and the industrialization of the manufacturing process.
MgB2 superconducting coils
*********************************
Superconducting wire
-----------------------------
Columbus Superconductors SpA with the collaboration of IEE has designed and manufactured a specific MgB2 wire for magnet applications. It consists in a 3 x 0.5 mm2 superconducting tape with 19 filaments embedded in a Ni matrix. The tape is produced with the Powders In Tube (PIT) manufacturing process in single batch length between 1.5 and 3 km.
To enhance the thermal stability of the wire and of the devices in which it’s implemented, a high purity copper strip is soldered on the tape surface after the final heat treatment of the wire. All the issues related to the perfect alignment of the copper strip on the tape surface have been solved by optimizing the soldering process and now this strip can be perfectly aligned and soldered.
Strain tolerance of the tape subjected to stress and bending has been extensively studied. In particular, the practical limit of bending diameter is 150 mm with the copper outside. Cost-performance ratio is in the range of 20 €/kA m and Columbus anticipates an improvement of this ratio based on the enhancement of the performance mainly due to new precursors, better process implementation, and manufacturing costs reduction due to economies of scale. The main details of the MgB2 wire can be found in “D2.1 Improved MgB2 wire” SUPRAPOWER project public deliverable.
IEE has analysed the perspectives of improving the critical current at relatively low fields (1-1.5 T) by working on the precursors powder quality in the MgB2 composition and a different synthesis route. Several wire architectures have been also analysed implementing the results about the powders, increasing the filling factor of the wire the strength of the metal sheaths of the wires. Finally a new wire configuration has been proposed to be implemented for future intermediate magnetic field application, as the one of SUPRAPOWER.
The critical current of this wire (1.5 mm diameter) has been analysed at different operating temperatures. At 20K and 1.8T field the critical current is higher than 600 A, three times higher in comparison with the tape used in the scale machine. This improvement could permit to simplifying the SC coil by reducing the length of MgB2 wire need and in consequence the number of turns of the coil. Finally this would permit further reduction of the weight and volume of the SCG. All the details of this study have been reported in “D5.1 Improvements in the critical components of the 10 MW SCG” SUPRAPOWER project public deliverable.
In conjunction with the experimental work performed in the prototype pancake coils during the project, a series of simulations were performed at University of Southampton to understand better the behaviour of the winding under external magnetic fields. The influence of a ferromagnetic metal matrix in the magnetic field inside the filaments of an MgB2 tape was analysed for different filament configurations. The results showed that the reduction of the critical current would be 1% - 2% depending on the magnetic field applied. The analysis was extended to the case of a SC coil based on the scaled machine coil designs. The determination of the load line of the coil, considering it as a bulk ferromagnet, leads to an underestimation of the critical current of the coil. If the filaments are not considered ferromagnetic, a reduction of the average magnetic field in the coil is observed. For the stack of 9 double pancake coils the magnetic field values and load-line of the coil can be determined using a pure non-ferromagnetic model without obtaining significant differences which saves computational time.
Additionally, modelling analysis to study the mechanisms that change the MQE in superconducting coils was done using two models: a FULL 3D and a CM. It has been seen that for a given set of parameters the MQE of the coil can be in the order of magnitude of that predicted by 1D analytic equation. The MQE is strongly affected by the thermal conductivity of the insulation and can increase the MQE of the coil of few orders of magnitude from 1D to 3D, due to the change in the volume of the MPZ. However the change of the thermal conductivity of the coil on its own does not explain the change in trend in the MQE observed experimentally in a solenoid coil. Only a smooth transition from 1D to 3D is observed in the simulations. To obtain the transition observed experimentally, cooling needs to be added from a boundary layer. This induces the transition due to the movement of the hotspot to an adjacent turn of where the heat is deposited due to the effect of cooling. Only the FULL 3D model was able to produce the transition while the CM model overestimates the MQE values. Therefore, the FULL 3D model is preferable to study the quench processes even if it has a higher computational costs, since there are mechanisms that cannot be reproduced with the CM.
Superconducting coil design
--------------------------------------
Field coils are constituted by a stack of 9 racetrack MgB2 double pancake (DP) coils connected in series. Each DP is made of two identical layers of MgB2 wire wound in opposite direction without splices and with an oxygen free copper sheet placed between the two layers to assure the refrigeration of the wire. Once wound, the DP is vacuum impregnated with Araldit F resin. The DP coils are stacked between two 8 mm coppers caps and they are connected in series through specific electric joints. The Cu inserts of the 9 DPs are thermally connected through Cu laps to the upper and lower caps, which are connected to the thermal collector to extract the heat from the coils.
The design of the coil has been supported by electromagnetic and thermal simulations carried out by the University of Southampton and Tecnalia. One of the most important design aspects is thermal behaviour of the coil given that heat is extracted by conduction from one connection point to the cryocooler. Simulated temperature gradients are below 0.1 K showing a good thermal stabilization that the Cu insert provide inside each DP. The models have been later validated by the experimental results.
Initial Small test coils
----------------------------
Before any full size coils were constructed, two small test coils were produced to validate the design and manufacturing process. First, a circular, 1 layer, 7 turns coil was built to analyse the wire behaviour under specific winding diameter and cryogen free refrigeration. Second, a circular, 30 turns DP coil was built to validate the winding, impregnation and test procedures to be applied in the real scale DP. These were tested in a general purpose cryostat at same working conditions. The manufacturing process and test results of this last coil are available in the SUPRAPOWER’s public deliverable “D2.3 Report on Design and manufacturing of an MgB2 coil”.
First MgB2 real scale double pancake coil
--------------------------------------------------------
After this small size coil, a first real scale DP coil was manufactured by Tecnalia at CIEMAT’s facilities in Madrid.
The DP has been extensively tested in a laboratory cryostat at Tecnalia [26]. All the details of this work are available in “D2.4 SUPRAPOWER Test report of the MgB2 coil” public report.
The transition to superconducting state during the cooling down of the DP has been recorded with 0.9 A applied current. Similar test has been carried out switching off the cryocooler in order to register the transition to resistive state during warming. Through the analysis of the obtained results it is estimated that the average transition in the coil is at 35.7 K. Voltage measurements have been fitted by using a power law (E= E0 (I/Ic)^n) where E is the electric field measured between the two ends of the coil, E0 is the criteria for the critical current, I the transport current and the critical current Ic(B,T) and n(B,T) the fit parameters. Ic of 91.79 A, 133.59 A and 146.26 A have been obtained at 30 K, 28.5 K and 27.5 K respectively. The fitted n value is about 10 for the overall coil at the studied temperatures. These Ic values are consistent with the expected results derived from the calculated load lines.
Finally it is concluded that the electrical behaviour of the DP, the obtained critical current and thermal behaviour as expected and according to the electromagnetic and thermal simulations carried out.
Manufacturing and test of the scale machine MgB2 field coils
---------------------------------------------------------------------------------
Once validated the first DP, Columbus and TECNALIA has manufactured and tested the 2 superconducting coils of the magnetic rotating validator.
The process to manufacture the complete DP ready to be assembled in the coil is quite complex. Once the winding is finished (about 360 meter of tape in two layers with 72 turns each one, by using the same procedure as in deliverable D2.3. Then the tape 2 ends are electrically connect the copper pieces that perform as electrical exists, this is done by soldering the copper pieces to the tape of the external turns.
Throughout the project lifetime the manufacturing process and the design of the coils have been updated according to the obtained experimental results. Thus for example the electrical connections were improved, copper pieces for thermal connections were slightly redesigned to improve the manufacturability and the impregnations process was modified so that to impregnate 9 DPs at time, while the first DPs were impregnated one by one.
The two coils have been extensively tested in a laboratory cryostat before its assembly inside the modular cryostat of the scale machine. Test results in the first coil have shown a very homogenous thermal behaviour, with less than 0.6 K difference between the opposite ends of the. It has been found that one of the DPs is damaged while the other DPs perform as expected.
Finally, even though one DP performance has not been as desired due to some manufacturing defects, it has been concluded that the developed coils are suitable for operating in a 10 MW wind turbine. These coils have then assembled in the scale machine.
Quench detection and protection system
------------------------------------------------------
The quench detection and protection systems are essential elements to protect the SCG in case of quench. This aspect has been extensively studied for both the SCG and the scale machine.
The quench detection system is based on the resistive voltage measurement and is a redundant system with three levels of protections.
• First level: threshold voltage, measured in the power supply terminals, which will work only in case of failure of the other two protection levels.
• Second level: comparison of two coils voltage, only different in case of quench.
• Third level: threshold voltage of the resistive voltage of the inductor circuit. Inductor’s circuit current and voltage are measured and then the inductive voltage component is calculated and subtracted.
The quench detection systems have been implemented in the National Instrument CompactRIO hardware platform. A protection system has been developed by Tecnalia for the scale machine. The system is based on the fast discharge on a dump resistor as soon as the quench is detected and switching off at the same time the power supply that feeds the coils. The resistor bank is constituted by 2 group of resistances serial connect as shown in, one is mounted on the back yoke that rotates jointly with the rotor and the other one is a stationary resistors bank connected to the coils through the slip rings. This topology permits to protect the coils even in case of a fail in the slip rings.
The protection of the rotor circuit with a simple dump resistor connected in parallel is valid for the case of the scale machine. However, it is not totally satisfactory for the 10 MW SCG due to the large discharge time, which could yield in hot spot in case of poorly refrigerated segments, so that solutions for MW scale machines must be further studied.
Cooling system and cryostats
**********************************
On the basis of the initial concept developed in the patent, KIT in collaboration with Tecnalia has designed the SCG cooling system and cryostats. Warm rotor iron configuration with one modular cryostat per coil has been chosen as the best option although other alternatives have been also considered. This configuration needs lower cooling power as the rotor poles are kept at room temperature and the modularity facilitates the repair and maintenance works onsite. The main drawback of this solution is a higher polar pitch than in conventional generators.
Superconducting coils are contained in 48 modular cryogen-free cryostats, one per coil. Heat is extracted by conduction through a thermal collector which links all the modules. This is cooled by conduction by two-stage GM cryocoolers which rotate jointly with the rotor. In contrast to pool boiling cooling system, only a small quantity of He gas is required for the thermodynamic cycle of the cryocoolers. This He circulates in a close circuit between the rotating cold heads and the stationary He compressors, which are connected by mean of a rotary joint.
The thermal collector consists of two high conductivity thermal circuits enclosed by a cryostat (called non modular cryostat). One is connected to the cryocoolers first stage and to the active cooled shield (Temperature ~80 K), and the other to the second stage of the cryocoolers and to the superconducting coils, maintaining them to their operation temperature at about 20 K.
A rectangular shape modular cryostat has been chosen according to the warm iron pole configuration and in accordance with the coil and pole dimensions and space restrictions. It is mainly constituted by a stainless steel vacuum vessel, a copper active cooled shield with a multi-layer insulation (MLI) of approximately 20 layers and the support structure. The support structure is a key component of the modular cryostat, 8 Ti 6A1 4V titanium rods have been adopted for the connection from active cooled shield to the superconducting coil and another 8 titanium rods from the shield to the vacuum vessel. This support structure is replicated in four locations along the cryostat, resulting 64 titanium rods per module.
First modular cryostat
-----------------------------
As a first step KIT has constructed one modular cryostat to experimentally test the thermal behaviour of one dummy coil inside the cryostat and validate the cryostat design. This first dummy modular cryostat basically follows the same design criteria of the SCG cryostat. Design details and test results of this dummy cryostat are available in “D3.3 First modular cryostat” and “D3.4 Test results report of the first modular cryostat”, SUPRAPOWER’s public deliverables.
The main difference of this cryostat respect to the ones of the SCG or scale machine, is that it integrates the GM cryocooler. The cryocooler is connected with flexible copper bridges, which act as thermal anchor fitting the displacement caused by the thermal contractions. The support structure is also slightly different but based on the same concept of using sets of titanium rods. A copper dummy coil has beeen constructed to emulate the cold mass and heaters has been installed to emualte the coil AC looses.
It took about 20 hours for the first stage thermal anchor to reach the lowest temperature, which is around 33 K. The second stage thermal anchor together with the linked dummy coil required 56.5 hours to reach the lowest temperature, which was around 9.8 K and 9 K respectively. These test results permitted to validate the cryostat and cooling concepts before the final design and manufacturing of these component for the scale machine.
Design and manufacturing of the cryostats of the scale machine
-------------------------------------------------------------------------------------
On the basis of the modular cryostats design, which is exactly the same for the 10 MW SCG and the scale machine, KIT has designed and constructed the cryostats of the SM.
The cryostat of the thermal link, so called non-modular cryostat, adopts a pipe design and a copper bar is used to thermally connect the two modular cryostats and the cryocooler. Both, the vacuum vessel and the thermal shield have a concentric half ring shape in order to achieve an easy assembly. This pipe concept is also valid for the non-modular cryostat of the 10 MW superconducting wind turbine.
The thermal shield made of copper and vacuum jacket made of stainless steel also adopts semi-circle. The cold head of the cryocooler is located in the middle of the semi-circle pipes. On the basis of the design, KIT has constructed 2 modular cryostats, one per coil, and the so called non modular cryostat.
In order to save space, the modular cryostat adopts the welding approach instead of flange connections as proposed in the prototype modular cryostat. It includes several steel stiffeners in the inner and outer part of the chamber, intended to give mechanical rigidity to the cryostat to prevent deformations during the assembly and welding process. After the assembly of the coil in the cryostat and welding the upper and lower parts of the vacuum chamber, these stiffeners have been cut.
He Rotary joint
--------------------
The helium rotary union (RHU) is the element prepared to work in the G-M cryocooler to transfer the helium gas between the rotating cold head and the stationary helium compressor.
The main functionality of the RHU is to connect the stationary part of the machine with the rotary one. It has three parts: gas helium feed-through to connect the cold head with the compressor, slip rings to power electrically the cold heads and optical fibre feed-through to transmit measurement and control signal. The RHU contains two flows of GHe at approximately room temperature, one from the stationary He compressor to the cryocooler (at a higher pressure of about 25 bar) and the other one back from the cryocooler to the He compressor (at a lower pressure of about 8 bar).
Two versions of the RHU have been constructed. The first version showed insufficient sealing during the pressure tests, so that a second version has been constructed. The optimised second version of the RHU has successfully reached the specification and design objectives. The second version of the RHU includes a mid-span bearing to provide a more robust configuration than the first version. The sealing between the stationary and rotating part inside the joint is realized by means of ferrofluid. The rotating seal consists of a permanent magnet, pole pieces, and ferrofluid. The pole pieces and the shaft concentrate magnetic flux from the magnet into a narrow gap region surrounding the shaft. The ferrofluid is attracted to the region known as “stage” forming rings that generate a hermetic seal. Each ferrofluid ring or stage has a pressure capacity of a 70-340 mbar typically.
After manufacturing and assembling the prototype RHU, series of tests have been performed to examine its performance. As result, no noticeable pressure drop has been observed in both helium lines. The leak rate of the RHU has been tested under stationary conditions using a leak detector. The measured leak rate of both helium lines at specified pressure is smaller than the sensor limit of the detector, which is 1.0·10 9 mbar·l/s. The running torque of the RHU at a rotation speed of 150 rpm has been measured.
The results have been recorded after 24 hours and 48 hours, respectively. After starting the unit, the running torque first rapidly increases to a peak value and then gradually reduces to a stable state. Since the torque transducer has a measurement range up to 50 N·m, the actual peak torque is not illustrated. However, using curve fitting it has been estimated that the maximum starting torque is 57 N·m and 62.5 N·m after a time period of 24 hours and 48 hours, respectively. The normal running torque of the RHU was around 23 N·m. These torque characteristics are very relevant for the selection of the appropriate AC motor to drive the RHU.
After this test the rotary joint has been validated. The design and manufacturing details and test results are available in “D3.5 Rotary joint: specifications, construction and test results” SUPRAPOWER’s public deliverable.
Scale Machine: Rotating Magnetic Validator
*******************************
Scale machine design
--------------------------------
At first it was studies the possibility to design and construct a scale generator (SG) in the range of 500 kW. To keep the maximum similitude between the model and the full scale generator, the power reduction was obtained by reducing the number of poles from 48 to 4, maintaining the size of the SC rotor coils identical both in full and small scale generator. However this option was finally discarded due to the complexity and high cost of constructing 4 superconducting coils and cryostats and mainly due to the very high cost of the bench needed to test a 500 kW superconducting machine.
On the basis of the first version a second scale machine (SM) has been designed overcoming the above mentioned problems but without relaxing its objectives. The final design of the SM is rotating magnetic machine that basically consists on a rotating external rotor with 2 superconducting field coils, each one enclosed in a modular cryostat around an iron pole at room temperature. Heat is extracted by conduction through a thermal collector that links both coils. Modular cryostats, coils and poles have the same size and are operated at similar conditions as the 10 MW SCG ones. The cooling concept is the same of the SCG, with the same cryocoolers rotating with the rotor and a rotary joint that links a stationary He compressor with the cryocoolers.
It incorporates a magnetic mirror in order to reproduce with the highest possible accuracy the magnetic behaviour and mechanical stresses of the rotating coils and cryostats inside the iron armature that were calculated for the 10MW SCG and SG (1st version). The magnetic mirror is constituted by 2 pieces made of soft ferromagnetic material that operate in non-saturated conditions. Thus it is establish the same boundary conditions in the air-gap region and in the inter-pole space when the magnetic mirror is not saturated.
The armature is not wounded so that this machine can be tested in much simpler test bench of 15 kW. The armature incorporates 4 test coils wounded in the armature magnetic core in order to measure the polar and interpolar leakage flux. This SM can be tested in a 15 kW test bench, which highly facilitates de experimental activity.
Electromagnetic 2D and 3D FEM analysis has been applied to the SM in order to validate the magnetostatic performance of the generator concept. The most saturated parts are magnetic poles, with a maximum value of 2.36 T. The peak magnetic field value in the SC coils is 1.24 T, located in the lower part of the straight section of the coil. This latest aspect has been studied in the detail as the magnetic field determines the performance of the SC wire.
Scale machine construction and test
------------------------------------------------
First of all, the elements of the scale machine has been constructed, including the superconducting coils and cryostats that as previously described has been directly constructed by the project consortium.
It has also been implemented the quench detection and protection system that has been previously described. The quench detection is system is implemented in a control hardware that also monitors the main parameters of the scale machine as temperature, magnetic field, current and voltage signals. This hardware rotates jointly with the rotor and the signals are transmitted to a stationary PC through the fibre optics link of the rotary joint.
Once manufactured all the elements next assembly sequence has been followed:
• Assembly of the coils inside the modular cryostats. This is one of the critical steps of the mounting process as the parts tolerances are very tight and it is necessary to handle all the components to avoid damages and undesired heat entrance points.
• Assembly of Set 1 and Set 2:
o Set 1: formed by the shaft, stator assembly, 2 mirrors, 2 bearings and their supports.
o Set 2: formed by the rotor back yoke, 2 poles and coils inside the modular cryostats
• Assembly of both sets in the support bench and then mounting of other components as slip rings, balancing rings and measuring coils.
• Assembly of the rotary joint
• Assembly of non-modular cryostat and its balancing pieces
All the details of the constructions and assembly of the scale machine have been included in “D1.4 Superconducting Electromagnetic Scale Machine” SUPRAPOWER’s public deliverable. [34]
The SM allows the experimental validation of the main innovative aspects and components at full scale of the 10 MW SCG. First of all, the test of the cooling system permits to probe the mechanical design of the cryostats, the good operation of the rotary joint and cryocoolers in rotation, the behaviour and efficiency of the overall heat extraction circuit and the capability of achieving the 20 K operating temperature in the coils. SC coils can be tested in rotation and under similar working conditions (I, B, T) to the SCG ones. The quench detection and protection system can also be tested and validated. Finally, test coils in the stator allows to check if the magnetic flux generated by the field coils in the airgap and the interpolar zones correspond to the values calculated and obtained in 3D electromagnetic FEM simulations, thus permitting to validate the electromagnetic design approach.
The test protocol, test results and conclusions have been reported in “D1.5 Test results report of the scale machine” SUPRAPOWER’s public deliverable.
Static vacuum and cooling tests have been carried in the SM to check the cryostat vacuum performance, the correct functioning of the cooling system including rotary join (in stationary position) and the temperature distribution in the Cu thermal circuits. A residual pressure of 9.4·10-4 mbar has been obtained by means of a turbomecular pump and before switching on the cryocooler. Next, the cryocooler was connected and the temperature in the ration shield of the cryostat was registered. These results have permitted to conclude that the system, including the rotary joint, performed as expected.
Cooling systems based on cryogen-free GM cryocoolers has very limited heat extraction power. Due to this it is indispensable to have a good current lead design with almost no heat transport or generation. First design of current leads was based on four SC Sumitomo BSCCO tapes (DI-BSCCO) Type HT, twisted along stainless steel capillary tubes. This design resulted not valid because of potential damages during tape twisting due curvature restrictions. In the new design the SC tapes have been directly soldered to the Cu terminals. Tapes have been inserted inside a Cu film with tubular form, which acts as current alternative path in case of quench.
Current leads have been tested under liquid nitrogen (LN2) before assembling them in the scale machine. The obtained critical currents (Ic) are below the expected ones (around 600 A) according to the manufacturer data. Also the “n” values are low, with the exception of one of the current lead. These results show that 3 of the 4 current leads, despite the performance is not as good as expected, could be good enough to energize the SC coil. The forth current lead should be manufactured and tested again. As conclusion the development of this component requires further R&D activity.
At the time of writing this report, the SC coils and the cryogen free cooling concepts have been successfully tested. Due to the complexity of the system, several components have required redesigns during the tests. Finally, taking into account these difficulties, the test results have not permitted to completely validate the scale machine. Additional R&D work should be done for validating the scale machine.
It has been concluded that the developed generator shows technical benefits in comparison to direct drive PMGs and at a very competitive cost. However, the development of the SCG still requires further R&D, validation and demonstration activities.
Some future R&D working lines have been identified. The improvement of the MgB2 wire performance would permit to improve the SC coil by reducing the amount of wire needed. This would permit to ease the design and manufacturing process of the coil and would yield in more available space inside the cryostat, which is also considered as an important improvement. Some wire improvements have already been analysed under the scope of SUPRAPOWER (see public Deliverable D5.1). The cooling system is based on heat extraction through conduction, which avoids the use of cryogenic liquids. However, this implies complex and very tough tolerances heat extractions circuits, which are made of a material (OFE Cu) difficult to mechanize with precision. The improvement of the heat extraction circuits would permit to ease the assembly of the system and increase its reliability. The design and manufacturing of the current leads should be improved, as they are key component for the validation of the system.
10 MW superconducting generator wind turbine
******************************************************
Integration of the full scale SCG in both fixed and floating offshore wind turbines
----------------------------------------------------------------------------------------------------------
R&D studies of the integration of both PMG and SCG in both flxed and floating WTs have been carried out by SOLUTE to evaluate the effect of the generator weight reduction on the rest of the turbine structural components.
First of all, as there were none or few available references of blades for 10 MW turbines, a special blade design has been performed with the aim of having lower weight. This concept has been extrapolated form NREL-5MW blade with no changes in aerodynamical profiles but with an optimization in weight. As a result, the Solute Hybrid Blade of 92.5m (SHB-92.5) made of glass fibre and carbon fibre has been released.
On the basis of this blade and the SCG, the drive train has been dimensioned with the main aims of reducing manufacturing hours, reducing the overall weight and offering a clean load path. The mechanical integration of DD generators of very large diameter requires different solutions than geared drive trains. The mainframe of a geared drive nacelle is substituted by a component named goose neck that attaches the generator to the WT tower.
The pitch system is located out of the hub to reduce its weight and size and also to ease maintenance operations. All the electromagnetic elements and the support structure of the SCG are designed so that can be divided into smaller parts to facilitate transportation and assembly operations. The generator rotor support is directly connected to hub, while the stator support offers a connection to the goose neck. It supports the whole drive train and nacelle rear structures, while it offers a connection with the yaw system and tower, offering a load path for the loads coming from the drive train. The hub, rotor and stator supports and the goose neck have been simulated as cast iron.
On the basis of the Rotor nacelle assembly (RNA) design, the tower and the fixed offshore substructure monopile has been designed. The main aim of this design process is to avoid rotor frequency (1P), designing a structure (tower plus monopile) with a first FA (Fore Aft) natural frequency 20% above 1P, i.e. around 0.17Hz. With a hub height of 125m above MSL (Medium See Level) the tower is 110m height and is supported by a monopile that stands 10.8m above MSL. Final calculated 1st FA frequency of the system (tower plus monopile) is 0.184Hz.
With the meteocean analysis of the selected location, and considering a set of load cases, which, according to experience, are the usual driving ones, the offshore loads analysis, following IEC 61400-3 standard has been performed. As a result, extreme and fatigue loads along the wind turbine have been obtained. Comparing this final loads with the PMG ones, main conclusion is that at RNA loads doesn’t differs too much, but along the tower, with a lower THM, loads decrease significantly (-10%). A strength analysis has been conducted, taking into consideration the aforementioned loads for fatigue and extreme conditions. As result some little reinforcements have been applied at goose neck, shaft and shaft support.
The results of these studies bring 25.6% weight reduction of the active parts of the SGC. The weight reduction of the THM is of 7.2%, this drop in the percentage is because the aerodynamic forces in a 10 MW turbine have more influence on the support structure than the weight of the active parts of the generator. It has been also evaluated that the achieved THM reduction permits an 11% reduction of the tower weight and 9% of the monopile.
Finally it has been estimated that this weight reduction could lead to cost shaving of 0.5 1 M€ respect to a PMG, even though at higher cost of the SCG (unless for the first units introduced in the market).
It has also been analysed the integration of the SCG in a floating platform. Nautilus floating platform has been taken as reference, as this platform is intended for 5 MW wind turbines, Tecnalia has made the redesign and main calculation for adapting the platform to the 10 MW, SCG scaling the 5 MW design for a 10 MW turbine. Dimensions have been calculated so that the floater could withstand maximum overturning moment along turbine life time and basic constrains for manufacturing, load-out and assembly process have been also taken into account. Then, mooring and hydrodynamics have been checked. Mooring have been designed by quasi-static methodology for 100 years return period, most representative mooring parameters have been obtained, such as chain diameter, length and anchor weight. Hydrodynamic features have been set by diffraction/radiation software, obtaining added mass, damping and stiffness coefficients which are essential to feed the numerical model used for dynamic simulations.
Furthermore stability and local strength have also been analysed. Stability and structure have been checked according to DNV standards. Then the final design has been validated by coupling codes, analysing a wide range of design load cases.
The same wind turbine used for fixed structure has been considered to get a good comparison between floating and fixed offshore substructure. As floating substructures show higher inertial loads than fixed ones, the tower model has been reinforced accordingly. Obtained total tower mass is 898 tons, i.e comparing with fixed turbine tower, it shows a 33% increase of mass and a final 1st FA frequency (tower plus floater) of 0.225Hz. Taking into account the calculated floating platform mass of 12,991 tons, the overall mass, tower plus floater, is 13,889 tons. It is estimated that the mass of the floater is in the range of 3-5% less than for the case of PMG base turbine. Load calculation studies have been performed for the same metoceanic conditions and load cases as for fixed substructure, checking that loads in the mooring lines, base of the tower, floater movements, etc are in the allowable limits.
Marine operations analysis
------------------------------------
To analyse the feasibility of a 10 MW SCG offshore WT it is also necessary to investigate and analyse transport, assembly and installation procedures and asses the benefits at this respect of the obtained weight reduction. All the technical details and results of this study are available in “D5.2 Study of transport, assembly and installation of offshore 10 MW superconducting generator” SUPRAPOWER’s public deliverable.
On the basis of main Rules and Guidelines for marine operations, the previously presented fixed WT and floating WT configurations have been studied, including: marine operations, required installations vessels, preliminary hydrodynamic analysis and hazard identification.
Marine operations have been assessed for both configurations and several options have been considered defining the process step by step from the harbour to the offshore project area. Then, based on preliminary assumptions on foundations and on the characteristics of the WT and the SCG (mainly weights and dimensions), the installation vessels have been analysed, highlighting the ones able to install the new 10MW SCG WT.
In general terms, it has been concluded that the transportation and the installation of the SUPRAPOWER’s WT can be realised with present-day equipment and with no major differences compared to existing WTs. It is concluded that thanks to the previously analysed weight reduction, a slight cost reduction could be achieved respect to a PMG WT. It has been estimate that the cost range for the installation of a 100 MW wind farm in the selected locations, with fixed foundations, could be in the range of 4-4.8 M€ and 28 to 26 days of installation time for the case of SCG WTs. While for the case of PMG WTs it has been estimated a cost in the range of 4.5-5 M€ and 27-38 days installation time.
The general conclusion that these analysis shows is that even if the SUPRAPOWER’s 10MW wind turbine is more powerful and hence bigger than current WTs (4.8 MW mean installed power rating), all the transportation, installation and assembly operations on both fixed and floating foundations can be performed with classic and existing means. All the operations can be performed following existing methodologies, with existing wind farm offshore means (vessels, cranes, tugs), by crews already trained and familiar with these classic tasks.
The safety considerations of the installation of the fixed and floating foundation WT have been analysed considering the means and procedures previously defined. A Hazard Identification (HAZID) method has been applied following the marine operation Rules & Guidelines, assessing the Severity and Frequency of all possible risks, defining the possible preventive and protective measures that may reduce them. Finally each risk has been classified in 3 different critically categories ranging from acceptable or low risk, to unacceptable or high risk.
Main HAZID conclusions are that typical superconductive risks, mainly related to the high pressure Helium, can be well mitigated and prevented, reducing these risks to low levels. On the other hand marine operations are similar than for other same sizes wind turbines, well known and mastered, with no additional risks.
Operation, maintenance and reliability analysis
--------------------------------------------------------------
O&M represents an important cost in the life cycle of an offshore wind turbine, indeed scaling up the power of the wind turbines could contribute to reduce these cost per MWh. Due to relevance of this aspect and the novelty of the superconducting technology, Ingeteam has carried out a study of the reliability, availability and maintainability of the SCG 10 MW WT, intended to evaluate the impact of the SCG in the O&M strategy and costs. Then a risk assessment of the O&M operations has been completed and on the basis of these outputs a Criticality Analysis has been completed. Finally, once identified the most critical components the maintenance strategy has been defined
The risk Assessment has shown few added risks between a PMG and SCG, which are related with very low temperatures and higher radiation generated by the powerful magnetic fields generated by the SC field coils. On the basis of this data a Criticality Analysis has been carried out. The analysis has shown the need of taking special care with the SCG, the Pitch System, the Blade System, the Power Electronics and the Yaw System. Superconducting elements, the cooling and vacuum systems have been deeply studied. It is clear that introducing more elements into a system decreases its reliability. Although in this case the number of elements introduced is high, they are theoretically not very critical. An exception is the rotatory joint, but an alternative to eliminate this “Middle-High” critical element has been analysed (see public Deliverable D5.1)
All the elements introduced related with the superconducting System have never been used in a similar application and least of all in such a harsh environment. The forces and vibrations to endure are also much noteworthy inside a wind turbine than in the most common applications for these technologies as MRI or magnets for scientific facilities. With this in mind, it seems logical that in the first prototypes and commercial units of this potential product, an in-depth study of behavior should be performed, requiring much more often and detailed maintenance and check visits than the theoretically needed. The modular design of the SCG, avoiding the need of using the most expensive and with low availability vessels is definitely an advantage that reduces significantly the “Impact of the Failure”.
Finally, with all this data and the best practices on offshore wind power O&M strategies, some guidelines have been produced for the Maintenance Strategies of a wind farm equipped with SUPRAPOWER technology WTs. The most critical components identified in the Critically Analysis will need a Condition Monitoring Systems so that to apply a predictive maintenance strategy. Preventive Maintenance periods, required tasks and means have been assessed for each maintenance period. For each of these tasks, the logistics (boats, vessels, helicopters, forecasting...) needed to perform all the operations were studied to achieve the highest availability of the plant with the less Operational Expenditure.
It has been estimated that installation, O&M, decommissioning and all other marine operations do not complicate due to the use of a SCG and moreover 0.9 M€ savings could be achieved in this field. This cost reduction is achieved thanks to the modularity and lighter weights of the components of the WT, which permit to have faster marine operations.
Potential Impact:
SUPRAPOWER potential impact
**************************************
SUPRAPOWER (www.suprapowe-fp7.eu) is research project funded by the EU FP7 programme that started in December 2012 and finished in May 2017. The project was conceived to provide an important breakthrough in offshore wind industrial solutions by designing an innovative, lightweight, robust and reliable 10 MW class offshore wind turbine based on an MgB2 superconducting generator, taking into account all the essential aspects of electric conversion, integration and manufacturability.
New Superconducting wind generator
--------------------------------------------------
The main outcome of SUPRAPOWER project is a novel 10 MW superconducting generator. This generator has been patented (EP2521252 B1) both in Europe (Spain, Germany, UK and France) and in the USA. The patented concept has been developed in detail and validated through a scale machine.
This generator is a low speed salient poles synchronous machine. It is a partially superconducting generator (SCG), superconducting MgB2 wires at cryogenic temperature are used in the field coils while copper wires at ambient temperature in the armature coils. The cooling system uses a cryogen free topology that does not use liquids at cryogenic temperatures. It has a warm iron rotor configuration, which consists on one modular cryostat per pole that encloses only the superconducting coil while the iron of the pole remains at room temperature. Heat is extracted by conduction through a thermal collector inside a cryostat which links all the modules. This is cooled by conduction by two-stage Gifford-McMahon (GM) cryocoolers which rotate jointly with the rotor.
This generator concept gives answer to the need of cost effective and more power wind turbines, while overcomes some of the challenges faced by other previously developed superconducting generator concepts. On the one hand, the selected superconducting material, MgB2, shows a much lower cost performance ratio than other HTS materials. On the other hand, the cooling system uses a cryogen free topology that does not use liquids at cryogenic temperatures, which highly simplifies the required cooling installation and minimizes the maintenance operations. Additionally, the modular design of the superconducting and cryogenics developments makes more feasible and less costly the corrective maintenance operations, which could be needed in case of a major failure.
The design process has also taken into considerations marine operations and O&M aspects, as they both represent a significant fraction of the offshore wind turbines LCoE. It has been concluded that using a novel SCG, which uses a technology never used before in wind turbines and least of all offshore, do not bring significant additional risks.
The SCG onsite efficiency, taking into account the power consumption of all the auxiliary elements as compressors and cryocoolers, is around 95.2% at the rated power, while the PMG one is around 94.5%. It is remarkable that the efficiency curve of the SGC is slightly above the PMG for all the power range.
SCG active parts cost has been estimated in 307 k€/MW, with current MgB2 costs, while the cost estimated for the PMG with NdFeB magnets price of 39 €/kg is 313 k€/MW. Thus the developed SCG is already cost competitive and with a huge margin for further cost reduction driven by the MgB2 wire performance improvement and industrialization of the manufacturing process.
This generator concept has been experimentally validated with a scale magnetic rotating machine designed and built specifically for this purpose. Main innovative features as superconducting and cryogenic implementation, modularity and quench detection system are equal or similar in both 10 MW and scale machines. Superconducting coils and cryogen free cooling system have been extensively investigated and as result real scale coils, modular cryostat and cooling systems have been constructed and experimentally validated by the consortium. At the time of writing this report the test of the scale machine was in progress.
It has been concluded that the developed generator shows technical benefits in comparison to direct drive PMGs and at a very competitive cost. However, the development of the SCG still requires further R&D, validation and demonstration activities
Facilitate more power and reliable wind turbines for LCoE reduction
------------------------------------------------------------------------------------------
As it has been explained before higher power rate wind turbines are pointed out as the innovation with higher Levelized Cost of Energy (LCoE) reduction potential. The state-of-the-art shows that both geared and direct-drive wind generators are difficult to scale up to 10 MW and beyond due to their huge size and weight.
Indeed, the development of large scale offshore wind turbines in the range of 10-20 MW and the improvement of the reliability through use of new materials, advance rotor design and control and monitoring systems, were relevant objectives of the “Investing in the Development of Low Carbon Technologies (SET-Plan),Technology Roadmap published by the EC in 2009”.
The SUPRAPOWER 10 MW SCG highly contributes towards these objectives by facilitating the development of cost effective large and reliable wind turbines.
It has been studies the integration of the 10 MW SCG in both flxed and floating wind turbines to evaluate the effect of the generator weight reduction on the rest of the turbine structural components. The obtained results have been compared with a permanent magnets generator (PMG) wind turbine so that to be able to quantify the benefits. The developed SCG shows 25.6% weight reduction of the active parts respect PMG. The weight reduction of the tower head mass (including nacelle and blades) is of 7.2%. Finally it has been evaluated that the achieved head mass reduction permits an 11% reduction of the tower weight and 9% of the foundations (monopile). It has been calculated that cost savings in the range of 0.5-1 M€ could be reached thank to materials savings in the structural part of the wind turbine. For the case of floating wind turbines it estimated that a 3-5% weight reduction could be obtained in the floating platform thanks to the reduction of weight of the turbine.
Transportation, installation and assembly, O&M, decommissioning and all other marine operations, on both fixed and floating foundations do not complicate due to the use of SCG. All the operations can be performed with classic and existing means and with already applied methodologies, so that there are not drawbacks related to using superconducting technology in the generator. It has been estimate that the cost range for the installation of a 100 MW wind farm in a representative locations, with fixed foundations 10 MW SCG wind turbine, could be in the range of 4-4.8 M€ and 28 to 36 days of installation time, which is the range of 5-10% less than for the case of using PMGs. Moreover the direct drive topology contributes to increase the reliability of the turbine and O&M cost can also be reduced driven by faster marine operations thanks to the modularity and lighter parts of the generator. It has been estimated that 0.7 M€ savings could be achieved related to improved reliability and maintainability.
A critically analysis has shown that the most critical components are conventional ones as power electronics, blade system or yaw mechanism. An exception is the Helium rotary joint that has been identified as medium-high critical component, so that the reliability and performance of this component in offshore conditions should be demonstrated.
As conclusion the developed SCG can significantly contribute to the reduction of the LCoE of offshore wind. Nevertheless there is still R&D work to be done for the development and demonstration of a 10 MW SCG offshore wind turbine.
Finally facilitating the deployment and development of more cost effective offshore wind farms, contributes to EU targets of having at least 27% share of electricity produced by renewable sources, as indicated by the 2030 Climate Energy Package, “A policy framework for climate and energy in the period from 2020 to 2030”.
Exploitable foreground
------------------------------
As result of the project the project consortium has achieved the following exploitable foreground, that in certain cases have applications also in other sectors. The main exploitable results are the followings:
• 10 MW superconducting generator (main beneficiary: Tecnalia). This is the main results of the project and the other obtained results give additional value to the patent. The exploitation strategy is the IP commercialization after the project end.
• MgB2 wire (main beneficiary: Columbus). The strategy is direct sales to Coil and high current cable manufactures. The wire developed in the project can have further applications than those related to generators.
• MgB2 coils (main beneficiaries: Columbus, Tecnalia). On the one hand the strategy is manufacturing coils for generators manufacturers. On the other hand the strategy also considers providing to third parties services for the design and manufacturing of superconducting coils.
• Cryogen free cooling system & Rotary joint (main beneficiary: KIT). The strategy is the commercialization of the IP and/or provision of R&D services.
• Design of 10 MW turbines. Integration of superconducting machines (main beneficiary: Solute). The exploitation strategy is offering high quality and specific engineering services regarding wind turbines & offshore.
• O&M procedures for 10 MW wind turbines (main beneficiary: Ingeteam).The exploitation strategy is providing services for high power wind turbines offered under “Trade Secret”.
• Marine operations procedures for 10 MW wind turbines (main beneficiary: D2M). The strategy is the exploitation of the gained knowledge by providing services in the field of marine operations for offshore wind turbine or other marine renewables.
Dissemination activities
***************************
During SUPRAPOWER project dissemination and communications have been an essential activity and extensive effort has been made to disseminate the superconducting generator concept, its associated results and in general terms the results of the project. During the first part of the project dissemination material was mainly focus to the communication of the project objectives and background, while the second part has been focused to communicating the project results and to the exploitation of the foreground. After the end of the project some further activities will be carried out in order to disseminate the final results.
More dissemination activities than those initially planned have been carried out. As a summary 2 press release have been produced, the consortium has taken part in more than 22 conferences and 13 articles have been published in peer review magazines.
Two workshops have been organised. The first one was held in April 2015 in Bilbao, Spain as a side event of the Bilbao Marine Energy week (www.bilbaomarinenergy.com). This workshop was focused on High power electric generators for cost reduction of offshore wind. Its aim was to provide an overview about trends towards high power wind turbines based on both conventional and superconducting electric generators and get feedback from industry and academia.
The second has been held in May 2015 at TECNALIA’s facilities in Bilbao, Spain. The workshop has presented the main results achieved and lessons learnt during the project, including a visit to the scale machine. The event also brought main European research initiatives in the field of superconducting generators for wind energy. This event took place at TECNALIA’s facilities in Bilbao, Spain.
The project web site (www.suprapower-fp7.eu) has been a key tool for the dissemination of project objectives and results, and it has accumulated more than 15,000 visits. In the public are of the web site several articles and presentations are available, including all the presentations of the 2 workshops of the project. Moreover it also has 13 deliverables, which are made public as soon as approved by the European Commission. The project web site will be kept active unless one year after the end of the project so that keep of the results available.
All the dissemination activities have been reported in “D6.3 Dissemination Material” public deliverable”, which is available in the project website.
List of Websites:
www.suprapower-fp7.eu
