IN DEPTH: Supersize turbine testing
As manufacturers race to build the next generation of offshore giants, the tools for testing are reaching an unprecedented scale.
Vestas’ director of global testing, Christian Fenselau, swings open the door with a dramatic flourish: “Here it is, the big beast!”
It is an apt description. At the centre of a hall the size of an aeroplane hangar in Aarhus, Denmark, sits the biggest wind turbine transmission system test bench ever built, a 20MW giant measuring 25 metres end to end, now being prepared to test the power train for Vestas’ flagship 8MW V164 turbine.
The rig will bring together the machine’s 8MW drivetrain, permanent-magnet generator (PMG), full-scale converter and control system — all components that have already been individually bench-tested — as part of system integration trials, the last step before Vestas erects prototypes on- and offshore next year.
The V164, unveiled in March 2011, features a Bosch Rexroth three-stage, medium-speed gearbox and a rear-mounted PMG from The Switch at the heart of its 375-tonne nacelle.
Its record-setting 80-metre blades, being readied at Vestas’ R&D facility on the UK’s Isle of Wight, will crank a rotational force from a swept area of 22,000 square metres — roughly three football pitches — into a transmission system designed for gusty high-speed winds averaging 11 metres per second.
Load levels that can be exerted by the 20MW test bench, built by GE-Converteam, are appropriately huge. Bolted in place, the drivetrain, wired up with 300 load-measuring sensors, will be shouldering torques of up to a wrenching 18MNm — almost ten times that of an industrial rock crusher — through a variety of extreme tilt, yaw and thrust moments, while powering a PMG specially designed for sharp jumps in torque from 60kNm to 300kNm to imitate the most turbulent offshore conditions.
“Testing is about trying to move the wind farm into the test centre,” says Vestas chief technology officer Anders Vedel.
“It also ensures we are taking all required considerations into the design process — and building these into our test and reliability protocols prior to launch a testing programme, just as we have done for the V164.”
The 20MW might of the test bench is meant to speed the V164 to market by 2016. On top of far-reaching function and reliability programmes, the Vestas rig can put the turbine’s 150-tonne transmission system through the make-or-break Halt tests at a rate of knots — putting ten years’ wear and tear on a gearbox for each calendar month — to gauge the performance of “every aspect of the machine, from mechanical input to electrical output”.
Vestas runs 50 test benches for drivetrains, generators, main bearings, blades and other components, as well as a climate chamber and materials lab, with a combined testing capacity of over 1,500 hours a day across its global R&D operation. “We used to have a seven-hour shift for testing, now it’s 24/7,” says Fenselau.
Its strategy has shifted somewhat too in recent years, with a move away from part-by-part testing towards a “critical component” focus with more time and money spent on systems integration testing.
The V164 drivetrain, generator and converter have each begun a six-week warm-up lap, after which they undergo parallel forensic exams, before being mounted together for an integrated run.
The first stretch of trials using the rig, which is grid-connected and churning out enough power to cover all but 10% of its own electricity bill, is expected to run six months, with Vestas angling to get the first fully assembled prototype up and turning in early 2014.
“Testing is increasingly a very significant part of the value chain,” says Fenselau. “It is helping reduce the CoE [cost of energy] ultimately, which is the main driver for everything. It is important that we do everything we can [to test a turbine’s components] before we think about erecting a prototype — especially one that is for offshore. With this machine it will be the real, final, proven product before it is installed at sea.”
Vedel adds: “We have made a significant investment in the testing technology and the people working in the test centres. The payback — for both Vestas and our customers — will be improved LPF [lost production factor] and so lower CoE.”
The push to bring down the CoE of ultra-large offshore wind turbines led Vestas — like its rivals Mitsubishi and Siemens — to put its flagship turbine through its paces in a closed shop first.
The test rigs built by manufacturers for these huge machines are a case of necessity being the mother of invention
Testing is fast entering a new era. In the late 1990s, when Henrik Stiesdal, chief technology officer at Siemens Wind Power, decided to develop testing technology that was as state of the art as the turbines, he was starting from scratch, looking to patch together the hardware and software from what could be scavenged from other industries.
But in the past five years that area has started coming into its own — although he remains of a mind to keep it in-house at the company’s R&D centre in Brande, Denmark.
“Testing now is so very important [for validating a new turbine] — we wouldn’t dream of not doing it ourselves because half of the learning is in the process and half is in the results,” says Stiesdal.
“And this doesn’t come cheap: beyond the cost of setting up a testing facility such as we have [at Brande], you basically destroy a multi-million-euro machine every time you test. It’s money out of the window, but worth every penny.”
The drivetrain test bench being employed to prove up the direct-drive system chosen for Siemens’ 6MW SWT-6.0 machine is only 8MW — more than enough overcapacity for a direct-drive transmission — but can apply a massive 20MNm of offshore wind-mimicking torque to simulate the loads brought to bear by a 154-metre-diameter rotor.
Stiesdal describes the rig, which can run a 6MW Halt test in three to six months, as a “cheap and cheerful” solution. “We have found that we have been able to build a full test bench and carry out a first full test for the same price as a third party would charge for a test alone — and we own the benches afterward. Much of what we use for testing is not sexy, but it does the job and only cost a couple of million euros to build.”
Siemens Wind’s head of advanced technology, SØren Oemann Lind, adds: “Because we design all our test rigs internally, we own the knowledge that we get from them — this has been the objective since we started formal testing programmes.
“In moving from testing the [3MW] 3.0 to the 6.0, we learned how to be much better at controlling the torque, pitch and yaw forces applied to the prototype to get a truer picture of the turbine’s [likely] performance offshore.” Siemens’ SWT-6.0 was first fitted with a 120-metre-diameter rotor and installed as a pair off southeast England at Dong’s Gunfleet Sands 3 pilot project earlier this year.
Plans are in chrysalis at Siemens for a 10MW turbine — a size of machine that would need a purpose-built test bench — although no location has been settled on. For the new model, Stiesdal expects to scale up from the bench devised for the 6MW, “with few real changes, except capacity”.
“The Halt testing we are doing today is as I think it should be done. It would take a very hard sell to convince me that we should be deviating from a testing methodology that has given us such a good track record for our turbines,” he says.
“When I look back at many of the smaller 1MW, 2MW models we produced in the past [which had no formal integrated testing regime], it would have really paid to do the testing we do now. The returns [for thorough testing of ultra-large offshore turbines] will be much, much greater.”