By Steve Meyer, Contributing Editor
Mechatronic challenges come in all sizes, but the new 15-MW Drivetrain Test Facility at Clemson University may be the biggest in the world.
Ever since the publication of the Department of Energy’s “20% by 2030” windpower roadmap, the question of equipment testing has been noted as a key ingredient to wind power market success. With turbine price tags in the millions of dollars, testing equipment in the field and waiting for results is not the best strategy—especially when the consequences of failure can be catastrophic.
The wind power industry has been progressively building larger and larger systems in an effort to capture more wind and make turbine hardware more cost effective. Previous designs were focused on generating capacities of 250 kW to 2 MW. Now, the market is dominated by systems that are 3 MW and larger with Vestas’ latest 8-MW turbine being demonstrated for deployment offshore.
The recent opening of the NREL’s National Wind Technology Center in Louisville, Colo., provided a major asset to the United States wind industry’s competitive capability—the ability to test how wind turbines will behave mechanically and electrically before they are deployed for service. However, the 5 MW capacity of the system in Boulder has a distinct limitation in the ability to serve the industry at large.
In 2009, Clemson University’s Dr. Nick Rigas put together a team and created an industrial advisory board of wind turbine manufacturers to gather input about important testing methods and future equipment size that would be needed to support the industry. The Clemson team concluded that a 15-MW capacity test facility would be necessary to support the emerging wind turbine industry.
Facility with internal view of structure
Having completed sophisticated dynamometer test systems for BMW, Clemson was able to compete for—and win—the DOE grant to build the largest dynamometer test system in the world. These systems, at hundreds of horsepower and thousands of rpms, were not so different in concept from the testing behaviors needed for low speed wind turbine rotors operating at thousands of horsepower. Rigas’ team also included utility partners Duke Energy, Santee Cooper and South Carolina Electric & Gas, which participated in support of building the test facility.
There are very few mechatronic challenges at the level of 15 MW. Typical lightning strikes are estimated at 1-10 billion Joules of electricity dissipated in around 50 milliseconds. Much of the power is converted to heat along the way, but if it were all pure electricity, a 5 billion Joule lightning strike would run a large home for over a year. The 15-MW dynamometer and electrical test facility can provide the equivalent of a 1.4 billion Joule lighting strike continuously. From a mechanical perspective, 15 MW is the equivalent of 20,000 hp at the hub of a turbine. The weight of a 4 MW wind turbine is like stacking 15 school buses in a large container on top of a steel tower 250 ft high. Even 8 MW machines require increasing the size and weight of all the components, pushing the boundaries of materials science in all areas—blade construction, bearing manufacture, tower design and foundations. And industry experts foresee larger machines coming in the future.
The sheer magnitude of the energy involved makes the notion of building a test facility of this kind daunting, not to mention the need for creating test conditions in the control system that correspond to conditions found in the real world.
The mechanical testing challenge
A wind turbine consists of a propeller driving a gearbox that increases the slow speed of the rotor, typically in the range of 25 rpm, to the high speed output necessary to run a generator. To test the drivetrain, a mechanical system had to be devised that would apply real mechanical loads asymmetrically to the rotor hub in a manner similar to what might be experienced in the real world. So a system was created using two 8.4 MW motors driving a 120:1 gear reducer, so that force could be directly applied to turning the rotor hub as if the equivalent amount of wind were pushing the propeller blades.
Since the wind doesn’t always blow symmetrically across the 300+ foot diameter of the turbine blades, unbalanced forces must also be presented at various angles radially and at offset locations parallel to the axis of the drivetrain. These off-axis forces are what account for accelerated wear and premature failure in the bearings. The solution for applying these forces to the turbine input is the “Load Application Unit,” or LAU. In the 15 MW LAU, a 150 ton disk of steel is configured with 72 hydraulic actuators positioned perpendicularly around the face and circumference. These actuators can apply a point load sufficient to displace the disk more than 3⁄4-in. at a frequency of 2 Hz. Consider this the equivalent of 7 school busses strapped together and capable of being vibrated 3/4-in. up and down twice per second. In the photo above as well as in the drawing on page 69, the 7.5 MW drivetrain is pictured with 16 hydraulic actuators, servovalves and position transducers.
Electrical testing can fall into a few categories: Conditions inside the turbine, site conditions outside the turbine, and electrical conditions on the grid. Internally, the generator and associated power conversion electronics have to accommodate changing wind speeds which cause changing generator speeds. By “exciting” the turbine rotor with an external motor and drive train, the Grid Simulation system can easily measure the electrical output of the turbine’s power electronics.
The Grid Simulation System can also help manufacturers anticipate the effects of transmission line distances on the power quality of the generator and power electronics at a particular site. Turbine interactions at a given location cause low frequency harmonics that will also affect output—and these conditions can be simulated as well. By simulating these characteristics, turbine manufacturers can significantly decrease commissioning time at a new construction site and reduce the risk of fault conditions that might result from a mismatch of the turbine electrical configuration.
The main electrical grid, as much as we may think of it as a constant 120-V, 60-Hz supply of power, is subject to many forms of electrical disturbance. From weather-induced faults, such as trees dropping across power lines, to induced variations based on large loads being connected and disconnected; all produce changing conditions that impact generator performance. With the precise waveform and power simulation controls available from the Grid Simulation System, turbine manufacturers can see how changing conditions will impact their equipment while it is still under development.
The beauty of the Grid Simulation System is its ability to simulate, apply and gather precise performance data for any desired electrical test condition that turbine engineers can come up with. All of this capability is available up through the full rated power of 15 MW. The data gathering and control system hardware required to make all this possible is an area of expertise of National Instruments.
National Instruments has built a new class of test capability called “hardware in the loop” testing where system performance is tested at full load under a variety of programmed conditions. This enables engineers to create test strategies that minimize the possibility of failures in the field. NI engineers assembled a seamless solution to meet the rigorous requirements of the Clemson Wind Turbine Drivetrain Test Facility. The control system is made up of an NI PXI chassis populated with NI R Series multifunction reconfigurable I/O FPGA modules, IEEE 1588 Precision Time Protocol timing cards, NI FlexRIO FPGA modules and an NI CompactRIO expansion chassis for temperature monitoring and control. Plus, 24 terabytes of streaming storage memory are provided for acquiring extensive test data and the entire system is monitored and controlled through personal computers acting as man-machine interface terminals in the control room.
TECO Westinghouse, with decades of experience building motors and drives in the thousands of horsepower range, provided the expertise needed to design and build a 15-MW power supply to meet the requirements of the Clemson Wind Turbine Drivetrain Test Facility. The system is modular and configurable to provide power from 4160 VAC to 13,800 VAC and output frequency from 45 to 65 Hz. This makes it possible to test any turbine from any vendor for use in any part of the world. In addition, the power supply system uses fiber-optic interconnect for superior electrical isolation of the controls and a 12-kHz synchronization clock to produce 0.1% waveform accuracy.
The 15 MW power supply has 133% continuous overvoltage capability with intermittent power as high as 35 MVA for 3 seconds in a 10 minute duty cycle. These capabilities allow the Clemson test facility to subject turbines to many possible electrical conditions that will stress the device under test (DUT) and enable testing to protocols consistent for commissioning of utility equipment in many areas of the globe.
All of the forgoing technology, the 7.5-, 15- and 15-MW Electric Grid Test Facility make the Clemson Wind Turbine Drive Train Facility a world class test facility, not only for size, but also for sophistication of the testing capability. Produced by a world class team of science, engineering, control system, and high power electronics expertise, The Clemson WTDTF is the world’s largest dynamometer. As the wind industry is currently developing and demonstrating hardware above 8 MW, Clemson has a truly unique test facility available to explore new hardware performance.
Not only can the dynamometer provide test capability for wind turbines, but the adjacent Electric Grid Test Facility can provide extensive load and fault testing of any large scale energy source. Multi-megawatt solar inverters, flywheel or battery storage technologies can all be tested in real time with an array of test conditions configurable. All of which brings the US wind industry to a leadership position in testing.
Dr. Ben Black, National Instruments
Drew Pierce, National Instruments
Dr. J. Curtiss Fox, Clemson University Restoration Institute
Dr. Ryan Schkoda, Postdoctoral Associate, Clemson University Restoration Institute
Dr. Nick Rigas, Executive Director, Clemson University Restoration Institute