Báo cáo " WINDPOWER: A Turn of the Century Review " docx

P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 Annu. Rev. Energy Environ. 2000. 25:147–97 Copyright c  2000 by Annual Reviews. All rights reserved WINDPOWER: A Turn of the Century Review 1 Jon G. McGowan and 2 Stephen R. Connors 1 Department of Mechanical and Industrial Engineering, University of Massachusetts, Amherst, Massachusetts 01003; e-mail: jgmcgowa@ecs.umass.edu, jgmcgowa@aol.com 2 The Energy Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts 02137-4307; e-mail: connorsr@mit.edu Key Words wind energy, renewable energy, offshore, electricity, electricity competition ■ Abstract The 1990s saw a resurgence in the windpower industry, with installed grid-connected capacity expanding more than five-fold between 1990 and 2000. Most of this increase occurred in Europe, where governmental policies aimed at developing domestic energy supplies and reducing pollutant emissions provided a sheltered mar- ket for renewable energy generation. The 1990s were also marked by a return to large, megawatt-sized wind turbines, a reduction and consolidation of wind turbine manufac- turers, and increased interest in offshore windpower. This article reviews recent trends in the windpower industry, including some of the fundamental engineering principles of wind turbine design. Technological impediments and advances are discussed in the context of changes in the global electricity markets and environmental performance. CONTENTS INTRODUCTION 148 RECENT TRENDS 149 WIND ENERGY APPLICATIONS AND ECONOMICS 151 WIND TURBINE DESIGN CONSIDERATIONS 155 Rotor Axis 156 Orientation 157 Rotational Speed 158 Rotor Characteristics 158 Aerodynamic Power Control 159 Dynamic Load Management at the Hub 160 Tower Structure 160 Other Design Constraints 161 Maintenance Issues 162 Standards and Certification 163 ENVIRONMENTAL DESIGN CONSIDERATIONS 165 Land Use 165 Avian Interaction 166 Local Opposition 167 1056-3466/00/1129-0147$14.00 147 P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 148 MCGOWAN  CONNORS WIND RESOURCE CONSIDERATIONS 170 RECENT ADVANCES IN WIND TECHNOLOGY 173 Rotor and Blades: Aerodynamics 174 Blades: Materials and Testing 178 Drive Train and Generators 178 Controls and Conditioning 179 Towers and Construction-Erection Issues 181 Resource Trends 181 FUTURE WINDPOWER APPLICATIONS AND DEPLOYMENT 182 Development of Large Wind Turbines 183 Offshore Windpower 183 Small Wind Turbine Systems 187 INDUSTRY TRENDS 187 CONCLUSIONS 191 INTRODUCTION In 1990 there were roughly 2200 MW of grid-connected wind generating capacity in the world, mostly in California (1). After the end of the OPEC oil shock, and the end of U.S. investment tax credits for wind, the industry entered a period of slow growth. In the early1990s, withconcernsoverclimatechangeand an over-reliance on fossil fuels reemerging, governmental policies in Europe, the United States, and elsewhere were re-instituted to help renewable power generation. This, along with technology improvements and lower installed costs, has led to a remarkable resurgence in the industry. Denmark and Germany introduced rules that ensured that wind farms received payments of up to 85% to 90% of the retail price of electricity (2). In the United States, the Energy Policy Act of 1992 instituted a production tax credit for wind and other renewables of 1.5¢ per kWh. However, with the introduction of competition for electricity in nearly every industrialized country, the long-term planning function of vertically integrated electric utilities has all but disappeared. In the place of utilities’ integrated resource planning has arisen renewable portfolio standards and the potential to sell “value priced” green power. Against this background of liberalized electricity markets, wind turbine developers have continued to work, improving the technology and bringing out bigger and bigger machines. In Europe especially, issues regarding land use have wind farm developers looking to the sea, a very suitable place for large wind turbines and smoother, faster winds. To bring the reader up to date, this article covers three main topics. First are the recent changesinthe wind industry itself, withparticularattentionpaid to therange and types of wind turbines—or wind energy conversion systems (WECS)—that are now being installed in onshore and offshore wind farms. Second is a review of the key wind turbine design issues upon which the continued development of the wind industry depends. Third is a discussion of where the industry is going. Of particular interest is how increased competition, or liberalization, in the electric P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 WINDPOWER 149 sector will effect the market for windpower and, of course, how this impacts the increasing need to reduce pollutant emissions and mitigate global climate change. RECENT TRENDS At the end of 1999, it was estimated that there was more than 12 GW of grid- connected windpower in the world. This is more than five and a half times the amount of installed capacity in 1990 (1, 3). Figure 1 (see color insert) shows how installed capacity has grown from 1995 through 1999, broken down by geographic region (4–6). Here the influence of European renewable energy policies is appar- ent. Table 1 provides details for 1996, 1998, and 2000. In the mid-1990s, North America and Europe had roughly the same amount of installed capacity (at 46% of the world’s total each). However, by the end of 1999 Europe’s share of total installed capacity had risen to over two thirds. From 1997 to 2000, Europe installed new wind generating capacity at the rate of 1600 MW per year, and from 1995 to 2000 wind generating capacity grew at an average annual rate of 37%. Over the sameperiod, windcapacityin Asia hasquintupled,primarilydue to efforts inIndia. By the late 1980s, commercial grid-connected wind turbines were in the 150 to 450 kW range. By the late 1990s, most manufacturers had roughly doubled the size of wind turbines, offering600to750kW machines. 1000 to 1600 kW machines are now commercially available. The latest models being developed range well above 2 MW, primarily for offshore applications. Figure 2 shows the comparative height andsweptareaforvariousmachines. Althoughrotor diameterand tower/hub height varies among manufactures, the variation is not overly large. Tower height is the most variable, as site characteristics such as uniformity of the wind’s flow field, surface roughness, and visual impacts must be considered. However, towers are commonly one to one-and-a-half the rotor’s diameter in height. A good overview TABLE 1 Installed wind generating capacity (4, 5, 6) Region Jan. 1996 Jan. 1998 Jan. 2000 Europe 2518 52.0 46.1 4766 62.8 35.9 8349 67.0 29.1 North America 1676 34.6 −2.7 1615 21.3 0.2 2617 21.0 30.2 Asia and Pacific 626 12.9 157.6 1149 15.1 24.5 1363 10.9 8.4 Latin America 7 0.1 −30.0 34 0.4 21.4 87 0.7 67.3 Middle East 12 0.2 −50.0 21 0.3 0.0 36 0.3 38.5 Africa 0 0.0 3 0.0 0.0 3 0.0 0.0 Total 4839 30.0 7588 24.5 12455 26.9 (MW) (%) (D%) (MW) (%) (D%) (MW) (%) (D%) (D%—Percent change from previous year) P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 150 MCGOWAN  CONNORS 5000 kW 112 m 100 m : Capacity : Rotor Dia. : Tower Hgt. 750 kW 48 m 60 m 50 kW 15 m 25 m 1000 kW 60 m 70 m 300 kW 34 m 40 m 2000 kW 72 m 80 m 160 m 120 m 80 m 40 m Figure 2 Representative size, height and diameter of wind turbines. of how the size and performance of Danish wind turbines has changed over time can be found in References 7–9. Whereas most new wind farm installations remain onshore, The Netherlands, Denmark, and Sweden have begun to develop their expertise in offshore appli- cations. Table 2 lists current offshore wind farms (10). Most of these represent near-shore, sea floor mounted WECS installations. As is discussed below, off- shore applications present a tradeoff between installed costs and maintenance and superiorwind resourcesandlowerland-useandcommunity acceptanceconstraints. TABLE 2 Existing offshore wind installations (10) Location Country Year Capacity No. Size Manufacturer Vindeby Denmark 1991 4.95 11 450 Bonus Lely (Ijsselmeer) Netherlands 1994 2.00 4 500 NedWind Tunø Knob Denmark 1995 5.00 10 500 Vestas Dronten I (Ijsselmeer) Netherlands 1996 11.40 19 600 Nordtank Bockstigen Sweden 1997 2.75 5 550 Wind World Total 26.10 49 (MW) (kW) P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 WINDPOWER 151 WIND ENERGY APPLICATIONS AND ECONOMICS How individual wind turbines are bundled into wind farms depends upon the wind resource, topography, economics, and the sensitivity of local populations. Figure 3 (see color insert) shows some of the potential configurations a wind farm can take, includingsome prospectivearrangements foroffshorepower. Thelargewindfarms in California range from ridge-top arrays in the Altamont pass to large rectilinear arrays near Palm Springs. Europe, due in part to population density, has deployed its wind turbines in smaller groupings, as linear arrays or clusters of perhaps a dozen machines each (7). Another important factor is the regulatory treatment of grid interconnections. At what voltage level is the local utility comfortable with the insertion of a variable power source? Furthermore, there may be economies of scale for larger wind farms, especially if they are connecting to higher voltage transmission lines for delivery to distant population centers. Interest in offshore applications has increased because large high quality wind regimes are relatively close to population and load centers. As maintenance requirements drop and remote control and operation capabili- ties expand, the economics of co-location will diminish. In areas where there are more people, or existing agricultural land-uses, the “European model” of smaller groups of WECS allows better integration and synergy of windpower generation with existing land uses. Although the close proximity of wind may invite local opposition, if there is good community buy-in, owing in part to local economic and employment benefits, wind deployment can continue (11). Of course, interest in windpower is not limited to grid-onnected power. As illustrated in Figure 4 (see color insert), large- and smaller-scale grid-connected windpoweris only part of the picture. For veryrural areas, including village power in developing countries, there is considerable interest in hybrid systems, or mini- grids. Recent experience in wind-diesel applications in Alaska and Canada focus on the delivery of reliable power, especially when already expensive fuel deliveries are interrupted for part of the year due to harsh weather (12, 13). These smaller kW systems are driven by a different economic equation. Rather than competing against the grid price of power, they are measured by the value of the service they provide. In developing countries this can be measured in terms of improved medical services, equivalent cents per lumen from a kerosene lamp, or clean and reliablewatersupplies.Onthefarendofthisspectrumaresmall single-usesystems, generally associated withtelecommunicationsandnavigational applications. Here the electricaldemandis generally for acontinuouspowersource, ratherthana large demand for electrical energy. Such battery-charging systems are rarely judged on a cost per unit power basis. The primary tradeoff effecting the economics of windpower is the capital cost of the machine or farm and the quality of the wind resource. Currently, to be cost-competitive, wind farms must be sited in high quality wind regimes, nor- mally a Wind Power Class of 4 or higher, preferably 5 or higher. Figure 5 shows P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 152 MCGOWAN  CONNORS 0 10 20 30 40 50 45678910 Capacity Factor (%) 0 200 300 400 500 600 800 200 300 400 500 600 800 2000 - - - - - - - Wind Power Density (W m-1) 5.6 6.4 7.0 7.5 8.0 8.8 11.9 0.0 5.6 6.4 7.0 7.5 8.0 8.8 - - - - - - - Wind Speed Range (m s-1) ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) ( 7 ) Wind Power Class (at 50m height) (1) (2) (3) (4) (5) (6) (7) Average Wind Speed (m s -1 at hub height) Figure 5 Comparison of average wind speed and wind power class to capacity factor (14, 15). a plot of the annual generation from a Vestas 600 kW machine, expressed as a capacity factor—the percent of a year it would need to run at rated power to pro- duce its annual output (14). For reference purposes the equivalent Wind Power Classes have been included on the graph (15). As power output, and therefore generation, is related to the cube of the wind speed, slightly higher average wind speeds, or wind regimes with a higher variability in the high velocity range, can produce significantly more power. The very best wind sites tend to be Class 6. A Class 4 site is considered marginal by economic standards, especially when the wake effects of other wind turbines within a wind farm are taken into ac- count. Therefore, in today’s market, a capacity factor of about 25% can be con- sidered a lower bound, unless the combined capital and operating costs of wind turbines drop. Thecostofwind-generatedelectricityisinfluencedbynumerousfactors.Table 3 shows how the cost of windpower changes as assumptions regarding capacity fac- tor, capital cost, financing, and operation and maintenance change. Using conser- vative mid-range assumptions for costs and performance, six cents per kWh is in line with recent experiences. Costsarecontinuing to drop for windpower, and with turbine costs approaching $800 per kW, wind generated electricity costs of 4–5¢ P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 WINDPOWER 153 TABLE 3 Parametric evaluation of electricity cost from wind Best Mid- Worst range range range Unit Capacity/plant factor 40.0% 25.0% 20.0% % of year at rated output Greenfield overnight cost $750 $1,000 $1,500 $/kW Fixed O&M costs $10.00 $15.00 $30.00 $/kW-yr Variable O&M costs $2.00 $8.00 $12.00 $/MWh (mils/kWh) Cost of generation 2.63 6.05 11.47 ¢/kWh All three calculations use a levelized carrying charge of 10% per kWh are expected in the near future. The cents per kWh number is a simple calculation of annual fixed and variable costs divided by the expected generation suppliedto thegrid. The“BestRange” and“WorstRange”columns inTable3show how this number changes if the combined best/optimistic and worst/pessimistic assumptions are used from a recent literature review (16). Greenfield overnight costs represent the “all-in” cost of the generation facility including grid inter- connections and access roads, as well as wind turbine costs. Fixed operation and maintenance costs (O&M) refer to regularly scheduled servicing, while variable O&M includes utilization-based service and repairs. Payments to landowners and taxes can be either fixed or variable O&M, based upon contractual and other legal arrangements. The four to six cents number is currently one and a half to three times the average spot price of electricity in the United States, depending on region, so to be competitive on a head-to-head basis with other sources of wholesale electricity these factors have to change, or some sort of subsidization or credit calculation must occur. By comparison, the total cost of generation for a new natural gas–fired unit can range from two to four cents per kWh, based upon technology and fuel cost assumptions (16). In the last Annual Reviews chapter on windpower, Sørensen (17) discussed the avoided environmental costs of choosing windpower over other options with pollutant emissions, as well as the life cycle impacts associated with mining, refining, fuel transportation, and combustion. Also important in addressing the social costs of various generation alternatives are the risks of severe accidents and longer-term fuel and solid waste issues. The external costs of windpower are not included in Table 3’s calculations. Nor are there any credits given in the calculationforsubsidiessuch asthe ProductionTaxCredit, oravoidedexpenditures such as the cost of sulfur emissions allowances that U.S. fossil units must now consider. Portfolio benefits, as demonstrated in Reference 18, can reduce system- wide variability in costs and emissions, and have some synergistic benefits when coupled with end-use efficiency efforts. Such estimates of avoided environmental and other costs are difficult to make without detailed analyses that incorporate the P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 154 MCGOWAN  CONNORS ● ● ● 2.0 4.0 6.0 8.0 10.0 12.0 14.0 0.0 0.5 1.0 1.5 2.0 Cents per kWh Ratio to Mid-Range Assumptions Best Range Worst Range Var. O&M Fixed O&M Carrying Charge Overnight Cost Mid-Range Capacity Factor (Site Wind Speed) Figure 6 Parametric evaluation of cost of electricity from wind. composition of the regional power system, as well as other regional demographics, air quality, and other environmental criteria. With these factors in mind, what opportunities are there to bring down the cost of wind-generated electricity? Figure 6 shows how the cost of wind generated electricity changes as cost and performance assumptions are varied about the mid-range assumptions in Table 3. The lines showing variations in fixed and variable O&M and the lines showing capital costs and carrying charges overlay one another. Although still significant, changesin O&Massumptionsdonoteffecttheresultingcostas muchasdocarrying charge, capital cost, or capacity factor. As with other large capital projects, project finance (represented by the levelized carrying charge) can be as important to the success of the project as the technology cost itself. Not surprisingly, capacity factor also plays a large role. It must also be recognized that capacity factor is not just the wind resource alone. The amount of scheduled maintenance a particular technology requires and the amount of time a unit is unavailable due to unforeseen outages also effects capacity factor. How will the costs of windpower technologies change in the coming decade and beyond? Arecentstudyexamining the possible impacts of introducing 10 GW of windpower in the United States by 2006 assumed installed windpower capital costs would drop to $600 per kW in 2006, owing largely to the economies of mass P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 WINDPOWER 155 production (19). A U.S. Department of Energy (DOE)/Electric Power Research Institute report the previous year had costs dropping to $740 per kW by 2005 (20). The factors at work in these anticipated reductions are discussed by Neij (21), and include not only the economies of mass production, but the increasing expertise of the industry as it designs, builds, installs, and operates greater numbers of wind turbines. Using data from Denmark, Neij calculated experience curves and rates of technology improvement, and predicted that if a growth rate of 15%–20% can be maintained, the cost of wind-generated electricity can drop by 45% over the next 2 decades (21). For such significant cost reductions to occur, the application of experience will certainly be needed, not only in the installation and operation of wind turbines, but also in their design, materials selection, construction, and siting. Although such technology forecasting is a tricky business, it remains a valuable exercise. Anotherfactor toconsiderin estimating thefuturecost ofwind-generated electricity is the availablewind resource. Thereisa finite amount of land with high quality Class 5 and 6 winds. How much of this land can be used for windpower, owing to ecological, local acceptance, and other factors such as access to the high voltage grid are always a topic of debate. Therefore, over the long run, capital costs must drop such that more readily available Class 4 wind regimes can be utilized. It is estimated that in the United States alone there are 232,000 km 2 of Class 4 land within 10 miles of transmission facilities, nearly 8 times more land area than there is for Class 5 and 6 wind regimes combined (20). Therefore, a combination of capital cost drops and operating performance improvements are required if the predicted cost of wind-generated electricity predictions are to occur assuming Class 4 wind regimes. Larger land area Class 4 wind regimes also allow greater siting flexibility, and may avoid some of the siting problems past wind projects have experienced because they required wind ridge sites in order to be economically viable. With this ultimate tradeoff between cost reductions and the finite nature of high quality onshore wind regimes, the following sections look at some of the more fundamental aspects of wind energy engineering, beginning with wind turbine de- sign and environmental considerations, the effects of site selection (offshore versus onshore in particular), and recent technical advances and how they are effecting the industry in the development and deployment of windpower. WIND TURBINE DESIGN CONSIDERATIONS The design of a wind turbine involves the integration of a large number of mechan- ical and electrical systems. This process is subject to a variety of constraints that directly effect the performance and economic viability of wind-generated elec- tricity. As discussed above, the cost of electrical energy from a wind turbine is a function of many factors, but the primary ones are the cost of turbine itself and its annual energy productivity or capacity factor. These and other factors are directly P1: NRM October 21, 2000 10:48 Annual Reviews AR118-06 156 MCGOWAN  CONNORS influenced by turbine design and necessarily must be considered in the design. The productivity of the turbine is a function both of the turbine’s design and the wind resource. Whereas designers cannot control the wind resource, development of wind turbines that maximize performance given the variability of the wind and other meteorological factors is of paramount importance. Therefore, a fundamen- tal tradeoff exists between low capital costs and robust operating performance. Minimizing initial capital costs has far-reaching implications. It compels the designer to minimize the cost of the individual components, which in turn pushes him to consider the use of inexpensive materials. The impetus is also to keep the weight of the components low, for a variety of reasons. On the other hand, the resulting turbine must be strong enough to survive any likely extreme events and operate reliably with a minimum of maintenance for a long time. Wind turbine components, because they are kept light and flexible, tend to experience relatively high, variable stresses. These periodic stresses result in fatigue damage, which eventually leads to failure of the component, requiring its repair or replacement. The need to balance the cost of the wind turbine with the requirement that the turbine have a long, fatigue-resistant life is therefore a fundamental concern of the designer. Over the past decade, the general design of larger grid-connected machines has converged, at least to some degree. The overwhelming majority are horizontal axis machines, usually with three blades. Nearly all now utilize asynchronous generators that, although they require power conditioning to match the generator’s output to the grid, provide greater operational flexibility and energy capture from the wind. Asynchronous generators are now even employed on fixed-speed wind turbines. It should be noted that within the wind community there are proponents of par- ticular aspects of design, such as rotor orientation, number of blades, etc. A good overview of these disparate design philosophies can be found in Doerner (22). This debate is centered around the issue of how light a wind turbine can be and still withstand operational and environmental stresses it will experience during its intended service life. Similar issues are also discussed by Geraets et al (23). As such, there are a wide variety of possible layouts or “topologies” for a wind tur- bine. Most of these relate to the rotor. Below we discuss the design considerations related to rotoraxis, orientation, rotational speed, and other general characteristics, as well as aerodynamic power control and load management. Design considera- tions regarding choice of tower structure, meteorological and other environmen- tal factors, and issues related to maintenance and design certification are also addressed. Rotor Axis A fundamental decision in the design of a wind turbine is the orientation of the rotor axis—horizontal or vertical. In most modern wind turbines the rotor axis is horizontal (parallel to the ground), or nearly so. The turbine is then referred to as [...]... MCGOWAN AR118-06 CONNORS Evaluation of Control and Protection Systems Evaluation of Loads and Load Cases Evaluation of Structural Components Evaluation of Mech and Electrical Components Evaluation of Component Tests Evaluation of Foundation Design Requirements Evaluation of Design Control Evaluation of Manufacturing Plan Evaluation of Installation Plan Evaluation of Maintenance Plan Evaluation of Personnel... was the initiation of a program to generate long-term (10 year equivalent), highcycle fatigue data for candidate structural materials Following this recommendation, NREL has sponsored approximately 10 years of fatigue testing at Montana State University, Bozeman The result of this program has been the compilation of a database containing a broad range of glass-fiber-based materials parameters encompassing... cost-effective It is also important for a wind engineer to realize that the visual appearance of a wind turbine or a wind farm must be considered at an early stage in the design process For example, the degree of visual impact is influenced by such factors as the type of landscape, the number and design of turbines, the pattern of their arrangement, their color, and the number of blades Visual or aesthetic resources... spite of some promising advantages of the vertical axis rotor, the design has not met with widespread acceptance Many machines built in the 1970s and 1980s suffered fatigue damage of the blades, especially at connection points to the rest of the rotor This was an outcome of the cyclic aerodynamic stresses on the blades as they rotate and the fatigue properties of the aluminum from which the blades were... utilizes a small portion of the land In the United States wind farm facilities may occupy only 3% to 5% of the wind farm’s total acreage, leaving the rest available for other uses In Europe it has been found that the percentage of land use by actual facilities is even less than the California wind farms For example, U.K wind farm developers have found that typically only 1% of the land covered by a wind farm... 10:48 Annual Reviews AR118-06 WINDPOWER 157 a horizontal axis wind turbine (HAWT) There are two main advantages to having the rotor axis horizontal First, the rotor solidity of a HAWT (the total blade area relative to its swept blade area) is lower when the rotor axis is horizontal (at a given design tip speed ratio) This reduces capital costs on a per kilowatt basis Second, the rotor of a HAWT is... direction, at all times Another advantage is that in most VAWTs, the blades can have a constant chord or cross-section, and no twist These characteristics should enable the blades to be manufactured relatively simply and cheaply (e.g by aluminum extrusion) A third advantage is that much of the drive train (gearbox, generator, brake) can be located on a stationary tower, relatively close to the ground... 80 acres (4 to 32 hectares) per megawatt of installed capacity The dense arrays of the California wind farms have occupied from about 15 to 18 acres (6 to 7 hectares) per megawatt of installed capacity Typical European wind farms have the wind turbines spread out more and generally occupy 30 to 50 acres (13 to 20 hectares) per megawatt of installed capacity (7) Because wind generation is limited to areas... (the area of the blades relative to the swept area of the rotor) A higher speed rotor with longer blades will have less blade area, or solidity, than the rotor of a slower machine For a constant number of blades, the chord and thickness of the blades will decrease as the turbine’s solidity decreases Owing to structural limitations, there is a lower limit to how thin the blades may be Thus, as the solidity... to the natural and cultural features of an environmental setting that are of visual interest to the public An assessment of a wind project’s visual compatibility with the character of the project setting is based on a comparison of the setting and surrounding features with simulated views of the proposed project To address the potential impacts, the National Wind Coordinating Committee developed a list . generating capacity grew at an average annual rate of 37%. Over the sameperiod, windcapacityin Asia hasquintupled,primarilydue to efforts inIndia. By the late. the choice of the generator. Using a low speed generator can eliminate the need for a gearbox and have a dramatic effect on the layout of the entire machine.