Path: bloom-picayune.mit.edu!snorkelwacker.mit.edu!americast.com!americast.com\!americast-post Newsgroups: americast.mech From: americast-post@AmeriCast.Com Organization: American Cybercasting Approved: americast-post@AmeriCast.com Subject: Turbines Catch Their Second Wind Date: Fri, 13 Nov 92 21:37:04 EST Message-ID: Turbines Catch Their Second Wind Engineers have learned that a windmill is exposed to more extreme conditions than was first thought. This knowledge is helping designers create advanced wind turbines that may finally fulfill wind power's promise as a abundant, and low cost energy source for the future. By: Steven Ashley Associate Editor By the end of the 1980s, the prospects for large-scale wind power generation seemed dim. Only about one-third of the wind turbines installed in California the hub of wind farm activity during the decade proved successful. Poor load prediction, inadequate detailed design, and low quality control had led to failed turbine blades and overloaded gear boxes, brakes, and yaw drives. Meanwhile, many turbine manufacturers and operators failed. Far from being the clean cheap energy source that had long been promised, wind power seemed to many observers to be a mere tax write-off scheme that survived on government-mandated preferential power-purchase contracts with utilities. But veteran wind-energy researchers such as Sandy Butterfield often felt that wind energy had an undeservedly bad reputation. To compete successfully with fossil-fuel power plants in today's unsubsidized markets, utility wind turbines must be built inexpensively, but still must produce electricity day in and day out for up to three decades, noted Butterfield, who is senior engineer at the U.S. Department of Energy's National Renewable Energy Laboratory (NREL) in Golden, Colo. In a wind turbine, you have a large fiberglass or wood-composite rotor and a simple drivetrain sitting on a 100-foot tower that is expected to rotate at 60 rpm for 5000 to 6000 hours a year, he said. During its lifetime, that translates into something like 10-to-the-ninth cycles. If you do the calculations, that's the mechanical equivalent of a car going 300,000 miles a year for decades on end with only simple maintenence. Nobody has fatigue data out past, say, 10-to-the-sixth cycles. At the same time, we re requiring airplane-like performance with what's essentially boat technology. For example, the turbine rotor blades must retail for $6 to $10 per pound, as opposed to $100 to $200 per pound for a helicopter blade. That means no exotic high-performance aerospace materials such as filament-wound carbon composites, only hand-fabricated E- glass/ polyester and wood/epoxy. Meanwhile, wind researchers have found that a wind turbine's operating environment is considerably more severe than had been thought, said Bob Thresher, director of NREL's Utility Systems division. Continuous buffeting by the wind and array-induced turbulence subject turbine components to relatively extreme and random loads that can cause dynamic fatigue and cyclic stress, which often lead to failures, he said. Other hazards, such as performance-robbing blade soiling by insects and air pollutants and structural corrosion due to the elements, also add to costly system downtime. Next-Generation Turbines Because blade and rotor design generally drives the economics of wind turbines through both power production and turbine failures, greater understanding of the typical hazards has led to optimism that new advanced turbine designs may actually fulfill their decades-old promise despite the constant battle against stringent cost goals. Our growing ability to predict a turbine's operating environment is giving confidence to the wind industry to go through another design cycle and develop new, more efficient and cost-effective machines, Thresher said. Many manufacturers of turbines agree. The new technology and new understanding will bring a renaissance in the wind power business, said Robert Lynette, chief executive officer at Advanced Wind Turbines Inc. (Redmond, Wash.). For the first time in about eight years, new turbines are being developed, Butterfield said. The operators who stuck out the bad times by making the systems work mostly by maintaining the heck out of them gave the engineers enough breathing room to solve many of the earlier technical problems. Now they re taking advantage of the new learning by applying it to new machines. The prime example is the acknowledged leader of the U.S. wind industry, U.S. Windpower Inc. (Livermore, Calif.), a subsidiary of the privately held Kenetech Corp. U.S. Windpower researchers are now testing 22 prototypes of its new turbine, the 33M-VS, according to William E. Holley, director of research and development. Holley said that the 33M-VS will generate electricity at the rate of five cents per kilowatt-hour, two cents cheaper than current machines. The variable-speed machines use solid-state power electronic inverters and synchronous generators to condition the power output, a big plus for utilities, which are not fond of the reactive power characteristics of the widely used induction generators. Developed during a five-year $20 million project conducted in collaboration with Pacific Gas & Electric Co., Niagara Mohawk Power Corp., and the Electric Power Research Institute, the advanced 350-kilowatt turbines are slated to be installed in several wind farm operations. Four utilities in the Pacific Northwest plan to place almost 140 turbines in a 50-megawatt plant located in Washington State by 1996. Dutch utility EGD is building a 25-megawatt facility that is to be on-line by 1994. A 50- megawatt wind plant has been sold to the Sacramento (Calif.) Municipal Utility District. A joint venture has been established with Iowa-Illinois Gas and Electric Co. (Davenport, Iowa) to develop a $200 million 250-megawatt (700 turbines) plant somewhere in the northern Midwest. Meanwhile, the U.S. Department of Energy, in collaboration with utilities and the wind industry, has undertaken the Advanced Wind Turbine (AWT) program, a five-year $55 million cost-shared project to develop new wind turbine systems. The first part of the program calls for development of improved turbines that can produce power for five cents per kilowatt-hour in consistent wind regions by incorporating incremental refinements into existing designs by the mid-1990s. Advanced Wind Turbines Inc. (Redmond, Wash.), Northern Power Systems (Moretown, Vt.), and Atlantic Orient Corp. (Norwich, Vt.) are participating in this effort. The second part of the AWT program focuses on the development of next-generation technology that would generate power at four cents per kilowatt-hour by the turn of the century. Researchers in the AWT program are focusing on a range of technical improvements, including advanced airfoils and aerodynamic control (airfoils), flexible lightweight blades, optimized integrated drivetrains, smart controls and sophisticated power electronics, new tower designs and materials, and improved foundation materials. Experts said that the near-term machines will be sufficiently cost-effective to permit their use in many parts of the United States, especially in the Midwest where the greatest potential for wind power exists. Although it will be years before wind farms add much to the national energy grid, wind power's contribution to the energy supply could match that of hydroelectricity about 10 percent. Utilities view wind as a fuel saver, NREL's Thresher said, which means it allows them to pull back the throttle on their most expensive fuel-burning generator. Wind farms will never replace other forms of power plants, he emphasized, because of the intermittent nature of wind. But many utilities still want to put wind on their grids to save fuel costs, to lower the emissions of air pollutants and greenhouse gases, and for peak load leveling. Today, about 15,500 turbines are producing about 1620 megawatts of power annually in California. Most of those systems were designed in the United States, though the Danish company Vestas and Japan's Mitsubishi Electric Corp. have sold a number of their three-bladed turbines here. Foreign wind research is moving at a brisk pace, Thresher said. Projections indicate that European countries plan to install from 2000 to 4000 megawatts of wind turbine capacity in the 1990s. European government support for wind power research in 1993 could reach $150 million, he said. That figure contrasts with the roughly $24 million the U.S. government will spend on wind research, development, and demonstration next year, according to Thresher. Wind Turbine Basics Wind is the result of uneven solar heating of land and water. Wind machines extract the kinetic energy of moving air molecules by slowing them down and converting the energy into electricity. A typical horizontal-axis wind turbine consists of airfoil- shaped rotor blades mounted on top of a tower and connected by gears to a drive shaft that spins a generator, typically an induction-type. The amount of electric power it delivers depends on wind speed and blade length. Wind speed averaging at least 15 miles per hour is considered best for power generation. The electrical power available from the wind is proportional to the cube of the wind speed, so that even small velocity increases can mean substantially more power. Most designers are focusing on two- or three-blade horizontal- axis turbines, somewhat like the auto industry settled on four- wheel models with front-mounted gasoline engines, said Sandy Butterfield, who also manages the applied wind research program at NREL. This configuration has been chosen over vertical-axis Darrieus rotors, which look like oversize eggbeaters. Vertical- axis machines look simple and all the hardware is on the ground, which should intuitively lead to lower costs. But vertical-axis machines require a more massive structure and you can't get the rotor high enough into the air to produce a lot of power, Butterfield said. Turbine designers have several primary design choices to make in configuring the form and function of wind energy systems. One decision is whether to place the rotor blades upwind or downwind of the main frame, the structure that contains the drivetrain, Butterfield said. The downwind configuration allows the rotor to yaw freely without a yaw drive. It also makes it easier to cone the blades insert a preset deflection of the blades from the plane of rotation which relieves steady bending moments at the blade root. Placing the rotor upwind gets the blades out of the tower shadow, which reduces turbulence-induced loads. Another design divergence among wind turbines is whether to allow them to orient themselves into the wind around the vertical axis like a weather vane, which is called free yaw, or to employ a forced-yaw arrangement, in which the rotation is controlled by a yaw drive mechanism and wind-direction sensors. Free-yaw systems avoid the extra cost of a yaw drive, but can permit damage to the rotor if it swings around too quickly, said Jim Tangler, NREL project manager. Forced-yaw systems are safer but involve the added cost of yaw drive equipment. Determining the number of rotor blades is another design key. Most early wind turbines had three blades connected by a rigid hub. This sturdy arrangement, however, can transmit dynamic loads into the drivetrain because nonuniform winds cause differential loading. The newer and potentially lower-cost approach is to use a pair of blades that teeter back and forth like a seesaw. The trick is to make the rotor flexible so that rotor loads stay out of the drivetrain, said Clint Coleman, vice president for engineering at Northern Power Systems. Teetering turbines have a hinge at the hub, which allows the rotor to seesaw when it is asymmetrically loaded. This occurs with every rotor revolution because the wind at the top is faster than that at the bottom due to the boundary- layer shear effect. When the top blade is loaded, it teeters in the direction of the wind, flying away the differential air load while the bottom blade moves into the wind. Teetering is a passive way to change the apparent wind speed at the blade and so to reduce load transmission into the drivetrain. Some wind researchers are working on composite flexbeam structures to accomplish similar goals. Yet another major design issue is the method by which rotor overspeed and drivetrain torque are controlled. The goal is to keep the generator in a power excursion from cooking as a result of high winds or loss of electric load. Various types of brakes are used, but only in combination with other methods. Most existing wind turbine blades are stall-limited or stall- controlled, which means they go into stall (lose lift due to flow separation on the airfoil) at a certain maximum wind speed. Alternatively, some turbines use full-span pitch control, which involves a pitch bearing at the root of each blade and a hydraulic actuating system. This approach allows the angle of attack to be changed along the entire blade for peak power control. Other systems pitch the outer portion of each blade when high speeds induce high centrifugal forces. Another approach to this issue is the use of electromagnetically latched aerodynamic tip brakes, which are wing-tip plates that can be turned into the wind to create drag. Another method is the use of trailing-edge ailerons to control blade lift. The type of tower chosen is also important. Open-truss or lattice towers block less of the wind flow and are less expensive than tubular structures, which offer lower maintenance costs (fewer joints) and protection from the wind for wind farm technicians. Guyed systems will likely become more popular as towers get taller, in order to capture the faster winds at higher altitudes. Technical Challenges The turbine inflow environment in wind farm arrays must be better characterized, Thresher said. The higher the inflow turbulence level, the higher the dynamic and fatigue loads, he explained. If the design turbulence levels are set too low, the machines will fail or wear out early. If they are set at too high a level, the turbines will be overdesigned and expensive. Therefore, reliable predictive design methods for the unsteady stalled-flow regime of wind turbine operation have to be developed, Thresher said. Moreover, without accurate dynamics codes, the load predictions for new turbines will be off. Also needed is the development of improved system-dynamics models that account for three-dimensional stochastic wind inputs and unsteady stall and include the important structural motions of turbines, Thresher said. Finally, the implementation by turbine designers of comprehensive stress- and fatigue-analysis tools and techniques is necessary, as is a disciplined design process based on careful analysis, developmental testing, and full-scale verification field testing. Copyright 1992, Mechanical Engineering. For more information, send-email to American Cybercasting Corporation (usa@AmeriCast.COM)