Siemens Tests its Direct Drive: Will Direct Drive Bring Down the Cost of Energy from Large-Scale Turbines?

This summer, Siemens Energy installed a highly innovative 3.6 MW direct drive ‘concept’ turbine near Ringkøbing in Denmark, where it will undergo a two-year period of testing. REW’s Eize de Vries was invited to Denmark to take a look at progress and to hear first-hand about Siemens’ key strategies behind this important project.

It was just four years ago — September 2004, shortly before its acquisition by Siemens of Germany — that Bonus Energy erected its geared 3.6 MW wind turbine prototype at a coastal multi-megawatt test site at Høvsøre (Denmark)

Featuring a 107-metre rotor, that SWT-3.6 machine has become one of the company’s two top-selling products, with an operational track record of about 40 turbines by June 2008. This flagship turbine has also emerged as one of the world’s most successful offshore machines. The first 25 offshore turbines became operational during 2007 at the UK’s Burbo Banks wind farm, and currently 54 of the 3.6 MW turbines are being assembled for the UK’s 180 MW Lynn & Inner Dowsing project. The order book stands in the region of 240 units.

And yet despite this success, Siemens has decided to see how a new, gearless, design would fare. Thus, during the second week of July, 2008, Siemens Energy erected an innovative 3.6 MW direct drive concept turbine for a planned two-year testing period at a coastal location near Ringkøbing, Denmark. Erection of a second, sister turbine featuring a generator of different design is planned before the end of 2008.

Siemens’ chief technology officer Henrik Stiesdal explains: ‘We have been producing geared wind turbines since 1980 and despite occasional issues over the years we have always been — and remain — happy with the technology. However, we also realize that direct drive wind turbines may become competitive with geared turbines for large turbine sizes, and we want to get hands-on experience with direct drive technology in order to conclude whether it can be made competitive with geared technology and, if so, from what power level. At the end of the day it can be boiled down to the question whether direct drive technology offers any substantial added value to Siemens’ customers.’

Expanding capacity

Following its late 2004 acquisition of the Danish wind pioneer, annual output of wind turbines from the former Bonus rose quickly from an initial 330 MW (2004) to nearly 1400 MW in 2007. The next goal is to increase production to 4500 MW by 2011, a plan that includes starting up wind turbine assembly in China for serving local and other Asia-Pacific markets. The enormous expansion of the company, now renamed Siemens Wind Power (SWP) in Denmark, is highly visible. Since September 2004 total floor space for offices as well as assembly and storage at the Brande site has been expanded almost beyond recognition. And, across the road, a new Siemens building is under construction, which from early 2009 will house office staff currently accommodated in temporary, containerized, offices.

A huge parts storage department stocks about 25,000 different components destined for turbine assembly, as well as providing spares supply for the upkeep of different generations of Siemens turbine models. When Siemens took over, it maintained the company’s highly valued components measuring department. Here, in a climate conditioned hall, full measurement checks are performed to all main shafts with a key focus on checking correct size tolerances of crucial machined areas like bearing seats.

In addition, all incoming gearboxes undergo extensive internal testing before being moved to the drive system assembly halls. In the assembly process Siemens works with a so-called ‘best match’ strategy, whereby components for certain assemblies are selected on the basis of optimizing size differences (tolerances) between matching parts.

Flagship turbine

Each Siemens turbine consists of a steel nacelle housing that accommodates the drivetrain and other main components, and systems like the AC-DC frequency converter, power cabinets and internal crane. The nacelle roof typically comprises two sections split along the centre line that can hinge open. The ‘open on demand’ roof feature eases operation and maintenance activities and allows fresh air to come in during favourable weather conditions. In addition, it enables large component exchange when necessary with hoisting operations from above. Another standard feature integrated in both the 2.3 MW and 3.6 MW models is a Turbine Condition Monitoring (TCM) system. This TCM system, among other functions, monitors drivetrain vibration levels, temperature development in critical components, and changes in acoustic levels. All these accumulated data are stored in each individual wind turbine database as well as a central database. Data analysis provides essential information on individual turbine condition and equally important condition changes. A second option is to use the statistical data as a monitoring tool for the operational behaviour of turbine fleets, specific wind farms and such like.

The Concept turbine

A ‘conventional’ fast-speed geared drivetrain always formed the distinct heart of all former Bonus, and now Siemens, turbines up to 3.6 MW. These geared wind turbines typically comprise a rotor connected to a (low-speed) main shaft that in turn is attached to a multi-stage gearbox and generator in a non-integrated line arrangement. The gearbox thereby converts the low rotor speed (13 rpm nominal for the 3.6 MW Siemens SWT-3.6-107) into the required generator speed. A majority of geared wind turbines are equipped with a 4-pole generator. Traditionally this type of generator would be required to run at a nominal speed of about 1500 rpm to produce European grid-compliant current at 50 Hz. With today’s variable speed systems using power electronics, the speed may vary over a wide range, but nominal power is still often produced at 1500-1800 rpm. An alternative is fitting a 6-pole (~1000 rpm) or an 8-pole generator (~750 rpm), where the higher the number of poles, the lower the nominal revs required. Choosing a generator with a higher pole number reduces the required gearbox step-up ratio, but with a 6-pole generator, for instance, there is a penalty, in that it is more bulky and heavier than a 4-pole generator of the same power.

The Multibrid concept used by Multibrid and WinWind, among others, has a single- or two-stage gearbox and a multi-pole generator operating at 100-150 rpm. The required (fixed) gearbox ratio in all cases has to match the difference between rotor and generator nominal speeds.

In contrast, a direct drive turbine features a low-speed multi-pole generator, which has a ‘fixed’ connection to the wind turbine rotor and therefore rotates at the same speed as the rotor. A direct drive generator eliminates the need for the ‘conventional’ combination of a multi-stage gearbox and fast-speed generator. However, in order to be able to convert the full rotor torque, the air gap diameter, and therefore the outer dimensions, are large when compared to a fast-speed generator with a similar power rating.

Stiesdal has been working on direct drive design options since 1999, but project development efforts remained low-key until 2005. The idea behind the concept turbine is to change only what is necessary. It therefore features a standard SWT-3.6-107 rotor and a functionally comparable twin-bearing main shaft support. Also unchanged are the 690 volt generator, full-power converter, control systems and the tower.

For the concept turbine an integrated main bearing housing was developed, which differs from the two separate main shaft bearing blocks applied in the SWT-3.6-107 both visually and mechanically. Another major mechanical modification the Siemens design engineers had to make was finding an alternative for the fail-safe disc brake originally located on the SWT-3.6-107 high-speed gearbox output shaft. In the concept turbine this brake is now fitted on the low-speed main shaft and located just behind the rotor hub. A key difference is that the new low-speed fail-safe brake features a disc with a large diameter and greater thickness compared to the original component it replaces. The difference in dimensioning can be explained by the fact that low-speed shaft input torque of a 3.6 MW Siemens turbine is about 115 times larger compared to gearbox output torque.

Integrating the direct drive generator into a proven wind turbine system substantially simplifies the testing arrangement. ‘It enables our engineers to focus on generator performance, system reliability and cost-effectiveness during the test period’, says Stiesdal, explaining why he prefers to speak of a ‘concept’ turbine rather than a prototype: ‘This is not a prototype of a new product; it is the physical manifestation of a technology project. If we decide to turn this into a commercial product then we will run as a second stage a classical R&D project, with design optimizations, etc. But I would also like to stress that such a decision has not yet been made, and if it is made it may be for a larger machine.’

Recognizable design heritage

The 1.3 MW and 2.3 MW Siemens turbine models are all clearly recognizable by a characteristic cylindrical cigar-shaped nacelle, and these familiar lines now reappear in the 3.6 MW ‘concept’. But unlike several other direct drive turbines that have the annular generator located in front of the tower, Siemens decided on a different option. The fully enclosed cylinder-shaped generator unit is supported by the main shaft rear end and is positioned behind the mainframe and tower. The generator is now located in what is normally the gearbox position in a Siemens turbine. A potential advantage is that a certain degree of load balancing can be achieved with the overhanging rotor mass located at the tower front side. All direct drive key components fit neatly into the cylinder-shaped nacelle, which has a 6-metre outer diameter, and a length of ‘only’ 13 metres. This outer dimension is about four metres shorter than the geared SWT-3.6-107 turbine nacelle length.

The fully enclosed air-cooled permanent magnet type generator does not have much free play inside the nacelle, as its outer diameter is of the order of 5.5 metres. The two different generators have weights in the 70-75 tonnes range, and the length is about 2.5 metres. The generator stator is fitted to a left and right-hand torque support attached to the integrated bearing housing. The generator cooling systems and the AC-DC power converter are all located in the nacelle rear.

The Concept turbine is fitted with a helicopter-landing platform located on top of the nacelle. The technology feature aims at facilitating personnel rescue operations. But the outer platform also enables service technicians working inside the nacelle to travel from the front section to the rear section and vice versa. Internal passage is not possible due to the fact that the circular gap between generator outer dimension and nacelle inner wall is insufficient.

A second ‘concept’ turbine, featuring a different design of generator, is planned for the test site before the end of the year. Converteam (UK) and the Large Drives Business Unit of the Siemens Industry Sector have each delivered a fully interchangeable generator for one of the two concept turbines.

Generator test rig

The Concept turbine was assembled in its own hall at the back of which a huge test rig has been built. The rig consists of two distinct parts in a set-up known as a back-to-back test arrangement. The first part comprises a complete SWT-3.6-107 drive system including cast mainframe and integrated fabricated welded steel sub-frame accommodating the main shaft, gearbox and generator. When used as a wind turbine, the rotor hub is bolted to the main shaft flange. However, in the testing arrangement the main shaft flange is connected to the direct drive generator input shaft. And instead of being driven by rotor power, the original fast-speed squirrel-cage asynchronous generator now operates in electric motor mode, driving the direct drive generator via the gearbox and main shaft. The test engineers have already exposed one of the two generators at the test rig to a three-month testing and optimizing period. Besides performance and generator efficiency tests, the engineers have optimized generator and frequency converter combination operational settings, among other things.

The Concept top head mass is about 265 tonnes (nacelle 165 tonnes; rotor 100 tonnes) assembled, as compared to a 235 tonnes THM for the SWT-3.6-107. One of Stiesdal’s future goals, provided direct drive is to stay, will be to push down the THM to that of their geared drive turbine equivalents through product optimization. Such optimization effort will pay off, both in terms of overall investment cost and dynamic load reduction, and will certainly pose a major challenge. Nonetheless, the impressive concept turbine already offers lots of potential to become a formidable wind technology asset for Siemens in the years to come, reducing the cost of energy over a turbine lifetime of 20 years plus.

Eize de Vries is Wind Technology Correspondent for Renewable Energy World magazine

Expanding production

Rotor hubs for the concept machine are assembled in Ølgod, located about 50 km away from Brande, at a new facility where SWP introduced a moving line assembly system, a production line first for the company. The new method aims to optimize overall production efficiency and make better use of available factory floor space.

In-house rotor blade manufacture commenced at pilot level in 1998, and in the summer of 2002 a full-blown manufacturing facility was established in Aalborg, an industrial city about 150 km north of the Brande main works. The 52-metre long B52 blade is currently the company’s largest product. All in-house manufactured rotor blade types are composed of state-of-the-art glass fibre reinforced epoxy (GFRE) laminate.

And, as a first major investment outside Denmark, Siemens also commenced production of rotor blades in the US, at Fort Madison in Iowa, in the first half of 2007. Due the site’s close proximity to water, rail and road transportation links, and its central location in the United States, Siemens says it is ideal for wind turbine blade manufacturing where logistics are critically important due to the massive size of the blades. Backed by several large orders from the US for its turbines, Siemens recently decided to expand rotor blade production at the Fort Madison plant.

For its rotor blade production process, SWP has developed a so-called monolithic manufacturing process or ‘One-shot’ technology, covered by several patents.

A great deal of effort has been taken to create optimal ‘closed environment’ conditions, totally preventing skin contact with harmful epoxy resin, which can cause allergic skin reactions. A key feature of the innovative technology is that the blades are manufactured in a single mould. A major advantage over ‘conventional’ blade manufacture — which involves two opposing blade shell halves — is the absence of a seam. This is the ‘critical’ joint where the two separate shell halves are normally bonded together; in most cases various spars are also joined in the bonding process. The One-shot manufacturing method begins with building a sandwich-type ‘glass-balsa wood-glass laminate’ structure in the bottom mould. The next step characterizing the One-shot method is that long overlapping glass fibre sections are put into the mould perpendicular to the blade’s long axis. The ‘surplus’ glass fibre sheeting is draped outwards on an adjoining platform. After putting in a single web (longitudinal reinforcement) and additional reinforcements, the glass fibre sections are draped back individually over the mould. This creates the familiar three-dimensional blade structure. The next step is to apply balsa wood ‘stiffener’ and the closing glass fibre layer. After the upper mould section is closed, the inner hollow structure is put under controlled vacuum. The blade manufacturing process is finalized by feeding liquid epoxy resin into the mould over its entire length. This ‘creeps’ into the glass fibre layers and all other air pockets, bonding the three-dimensional structure into a finished laminate.

Direct drive summary

In 2007 the cumulative market share of direct drive turbines was around 14%, a percentage that has varied little (between about 13%-15%) during past few years.

Since 1993, Enercon of Germany has dominated the direct drive wind market segment with a range of turbines from 100 kW to 6 MW, and the company has installed over 12,600 units since 1992. Enercon turbines are all fitted with in-house developed and manufactured generators with external field excitation. The largest direct drive turbine type in the world is the 6 MW Enercon E-126, introduced in 2007. This turbine features a 127-metre rotor diameter and a disc-shaped annular generator with a record 12-metre diameter. The power rating is likely to be raised to 7-8 MW or more in future.

However, the combination of large dimensions and simultaneous exposure to mechanical, electrical and thermal loads generally puts high demands on direct drive generator shape retention and air gap control.

Many international direct drive newcomers have attempted to enter the market segment since the early 1990s, some with highly innovative concepts, but so far with little overall commercial success. Examples of companies that have not succeeded in making a commercial impact include Jeumont Industrie of France (750 kW), Seewind of Germany (750 kW), and Heidelberg of Germany (30 to 300 kW).

For several years the former Lagerwey of the Netherlands took second place ranking in the direct drive segment with its 750 kW turbine series, but it erected only about 200 of these machines until filing for bankruptcy in 2003. Emergya Wind Technology (EWT) of the Netherlands bought the technology from the official receiver in 2004 and now markets a scaled-up 900 kW DirectWind 900*54 model.

A Zephyros consortium, led by the former Lagerwey, manufactured one 2 MW prototype, after selling the technology to a third party that again did not succeed. The latest owner of the Zephyros direct drive technology is Harokasan of Japan, which operates an assembly plant in the Netherlands.

However, the substantial number of new direct drive concepts introduced in recent years indicates that interest in the technology is growing. New contenders either developing or marketing commercial products include DarWinD (Netherlands; 4.7 MW), LeitWind (Italy, 1.2/1.5 MW), MTorres (Spain, 1.65 MW), Vensys (Germany; 1.2, 1.5, and 2.5 MW), Scanwind (Norway; 3.5 MW) and Unison (South Korea; 750 kW). Vensys turbines are also released under license, with Goldwind of China a prominent licensee. Goldwind now owns 70% of Vensys’ shares.

A majority of the new direct drive turbine concepts, including the Siemens 3.6 MW Concept, feature a permanent magnet (PM) type generator. PM generators are believed to have a superior partial-load efficiency and are compact compared to external field excitation generators. However, potential disadvantages are the loss of a field strength control variable and more stringent demands on manufacturing

Previous articleStandard Solar Raises US $8.5M
Next articleMorocco Backs Asola’s African Plans

No posts to display