Large wind-turbine projects in the United States wind energy program

Journal of Industrial Aerodynamics 5 (1980) 323--335

323

© Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands

LARGE WINI~TURBINE PROJECTS IN THE UNITED STATES WIND ENERGY PROGRAM

RONALD L. THOMAS and WILLIAM H. ROBBINS

National Aeronautics and Space Administration, Lewis Research Center, Cleveland, Ohio (U.S.A.)

Summary The technology for large, horizontal-axis wind turbines (100 kW--2500 kW) has been under development since 1973 as a major part of the United States Wind Energy Program which is sponsored by the U.S. Department of Energy. Specific projects are being managed for the Department of Energy by the Lewis Research Center of the National Aeronautics and Space Administration. The objective of the United States Wind Energy Program is to accelerate the development of reliable and economically viable wind-energy systems and enable the earliest possible commercialization of wind power. To achieve this objective requires advancing the technology, developing a sound industrial technology base, and addressing the non-technological issues which could deter the use of wind energy. Significant advances have been made in the technology of large, horizontal-axis wind turbines since 1973. Technical feasibility has been demonstrated in utility service for systems with a rated power of up to 200 kW and a rotor diameter of 125 ft. (Mod-0A). There appear to be no major feasibility issues to be resolved. The activation of larger prototype units in utility service in 1979 (Mod-1; 2000 kW, 200-ft. diameter rotor) and in 1980 (Mod-2; 2500 kW, 300-ft. diameter rotor) are expected to confirm this assessment. The long-term reliability of wind-turbine systems is yet to be demonstrated. This will require time to accumulate service experience. In addition, machine capital costs must be further reduced through a combination of continued research and technology development and quantity production. The "second-unit" capital costs* for large, horizontal-axis wind turbines currently range from about $8000/kW for operational prototype units in the 200-kW class down to approximately $1000/kW for advanced design prototype units in the multi-megawatt range. Current designs of large wind turbines such as the 2500-kW Mod-2 are projected to be cost competitive for utility applications when produced in quantity, with capital costs of $600 to $700/kW. At this level, wind turbines will produce electricity in the range of $0.03/kWh to $0.04/kWh when operated at sites with a mean annual wind speed of 14 mph (at 30 ft. ). These machines will be attractive to some utilities, but for widespread use by many utilities, COEs of $0.02/kW to $0.03/kW will have to be achieved. We expect the advanced technology machines to achieve these COEs when produced in quantity and operated at sites with moderate wind speeds (14 mph yearly average). In addition to the projected favorable economics, wind turbines appear to have no significant environmental impacts and use a replenishable, non-polluting source of energy. These features make wind turbines today one of the most attractive potential solar options for widespread utility application. *All costs in this paper are in 1977 dollars.

324

Introduction

Since 1973, The United States Government has sponsored an expanding research and development program in wind energy in order to make wind turbines a viable technological alternative to existing electrical generating capacity. The current U.S. Wind Energy Program, under the sponsorship of the Department of Energy, is directed toward the development and production of safe, reliable, cost-effective machines which will generate significant amounts of electricity. One element of the U.S. Wind Energy Program is Large Horizontal Axis Wind Turbine Development, which is being managed by the NASA Lewis Research Center. This activity consists of several ongoing wind-system developments oriented primarily toward utility application. Four wind turbine projects designated the Mod-0, Mod-0A, Mod-1, and Mod-2 are part of the current development program for large, horizontal-axis wind turbines in the U.S. The machine configurations are illustrated in Fig. 1. In addition to the configurations currently under development for testing, efforts aimed at achieving lower machine costs will be initiated in 1979. These will include an advanced MWclass wind-turbine project and an advanced 200 to 500 kW wind-turbine project. The machine design and technology development projects have been supported by substantial analysis and hardware/material testing. These include

1 WIND 555'

r"

MAX. POWER OUTPUT:

MOD-O MOD-OA

MOD-1

MOD-2

lO0-200kW

20OOkW

250OkW

Fig.1. Large wind turbines.

WAI ;HINGTON MONUMENT

325

efforts to improve the methods of structural dynamic analysis, assessment of utility interface problems, testing of component materials and evaluation of new blade concepts by analysis, laboratory testing of blade sections and operational testing of full-scale blades. This paper presents an overview of the NASA wind-turbine activities. More detailed descriptions of the projects are presented in refs. [1--4]. Development status of large wind turbines Mod-O The current program of research and technology development on large, horizontal-axis wind turbines was initiated with the Mod-0. The Mod-0 is a 2bladed, 125-ft. diameter, research wind turbine rated at 100 kW. This machine was designed by the Lewis Research Center. The Mod-0 has aluminum blades and the rotor is located downwind of the tower. However, Mod-0 has also been operated with the rotor upwind of the tower to assess the effects on system structural loads and machine control requirements. The rotor speed is maintained at a constant RPM by rotating or pitching the blades about their lengthwise (spanwise) axes to control the aerodynamic torque imparted to the rotor as the wind speed varies. This type of speed control is referred to as fullspan pitch control. The nominal rotational speed of the Mod-0 is 40 RPM, but a belt drive incorporated in the drive train system (see Fig.2) has permitted the machine to be run at several different speeds for test purposes. Power is transmitted from the rotor through a speed-increasing gearbox to a synchronous generator operating at 1800 RPM to produce 60-hertz power.

/BLADE /I

/, u,r,u

¢

o~=n

GEAR

HUB ~

~

,/

;ER

YAW

DRIVES -~ . . . . .

--r

Fig.2. Mod-0/0A drive train assembly and yaw system.

326

The entire assembly illustrated in Fig.2 is mounted on a steel, open-truss tower. This assembly is oriented to the wind by a yaw control mechanism. With a changein wind direction, the yaw control system orients the entire assembly using a hydraulic yaw drive connected to a large-diameter ring gear. The Mod-0 is installed at NASA's Plum Brook facility near Sandusky, Ohio and became operational in the Fall of 1975. It is being run in an automatic, unattended mode and synchronizes routinely with the Ohio Edison utility net~ work. It has proved to be a valuable engineering test bed for evaluating advanced design concepts and validating the analytical methods and computer codes which are being used to design advanced machines. Mod-OA The Mod-0A Project will place four prototype units of the Mod-0 class into utilities to gain early in-service experience. The Mod0A is essentially the same design as the Mod-0 except for a larger generator (200 kW) and larger gearbox. The Westinghouse Electric Corporation of Pittsburgh, Pennsylvania is the prime

Fig.3. Operating Mod-0A wind turbines.

327

contractor responsible for assembly and installation. The blades are built by the Lockheed California Company. The first Mod-0A was installed at Clayton, New Mexico; first rotation occurred in November of 1977. Following a checkout period, Lewis turned the machine over to the City of Clayton in March of 1978 to operate as an integral part of their utility system. The machine has operated successfully; it is operationally compatible with the utility grid and has generated 2--3 percent of the energy at Clayton since the machine was activated. As expected of the first machine in service, machine hardware problems have been encountered and have been corrected as they occur. A second Mod-0A was installed at Culebra, Puerto Rico for the Puerto Rico Water Resources Authority and was activated in July 1978. A third Mod-0A is installed at Block Island, Rhode Island and first operated in May 1979 for the Block Island Power Company. Photographs of these three operating ModOA wind turbines are shown in Fig.3. A fourth Mod-0A is planned for Hawaii and will be activated in 1980 on the Island of Oahu for the Hawaiian Electric Company. The Mod-0A project has demonstrated the technical feasibility of wind turbines in utility applications. It has also provided valuable in-service testing of hardware and operations to help guide technology development. Mod-1 The Mod-1 project was started in 1974. The Mod-1 is a 2-bladed, 200 ft diameter wind turbine with a rated power of 2000 kW. The blades are steel and the rotor is located downwind of the tower. Full span pitch is used to control the rotor speed at a constant 35 RPM. The gearbox and generator are similar in design to the Mod-0A but, of course, are much larger. The tower is a steel, tubular truss design. The General Electric Company, Space Division of Philadelphia, Pennsylvania is the prime contractor for designing, fabricating and installing the Mod-1. The Boeing Engineering and Construction Company of Seattle, Washington, manufactured the two steel blades. A single protytype is installed at Boone, North Carolina (Fig.4) and will supply power to the Blue Ridge Electrical Meml~ersi~ipCooperative. The Mod-1 first operated in May 1979. Mod-2 The Mod-2 project was initiated in 1976. Three machines are currently being fabricated. These are 2-bladed, 300-ft. diameter wind turbines with a 2500-kW rating. These machines are being designed with a new technology base developed as a result of research and development efforts on Mod-0, Mod-0A, and Mod-1. Because of this, the Mod-2 is referred to as a second generation machine. The rotor will be upwind of the tower. Rotor speed will be controlled at a constant 17.5 RPM. In order to simplify the configuration and achieve a lower weight and cost (the cost of these machines is closely tied to weight and complexity), the use of partial span pitch control is being incorporated rather

328

Fig.4. Mod-1 wind turbine at Boone, North Carolina. than full span pitch. In this concept, only a portion of the blade near the tip (outer 30% o f the span) is rotated or pitched to control rotor speed and power. To reduce the loads on the system caused by wind gusts and wind shear, the rotor is designed to allow teeter o f up to 5 degrees in and o u t of the plane of rotation. This reduction in loads saves weight and, therefore, cost in the rotor, nacelle and tower. The Mod-2 tower is designed to be " s o f t " (flexible) rather than "stiff" (rigid). The softness o f the t o w e r refers to the first-mode natural frequency of the t o w e r in bending relative to the operating frequency of the system. For a two-bladed rotor the tower is " e x c i t e d " twice per revolution (2P) of the r o t o r If the resonant frequency of the tower is greater than 2P it is referred to as "stiff". Between 1P and 2P it is generally characterized as " s o f t " and below 1P as "very-soft". The stiffer the tower, the heavier and more costly it will be. The tower's first-mode natural frequency must be selected to be sufficiently displaced from the primary forcing frequency (2P) so as n o t to resonate. Care must also be taken to avoid higher-mode resonances.

329

PLANETARY

GEARBOX--X\

ROTOR r'NNER- \ SHAFT ~' "--',

17.5rpm ~. OUTER-~ J)C'~ TEETER ,~C'~[ ] HUB--~\ ~ ~ ..,~-'~..--<'NF~

\

~ . ~

~-- HIGH-SPEEDSHAFT

I \, ILBEDPLATEI I- BLADE

1~OOrpm

~-- YAW DRIVE

ENCLOSURE

"- HYDRAULICSUPPLY FORPITCHCHANGE

Fig.5. Mod-2 nacelle.

Fig.6. Mod-2 wind turbine (artist concept).

330

The tower is a welded-steel, cylindrical-shell design. This design is more cost~effective than the stiff, open-truss tower. The gearbox is a compact, epicyclic design which is lighter weight than a parallel-shaft gearbox such as used on Mod-1. The nacelle configuration of the Mod-2 is illustrated in Fig.5. The Boeing Engineering and Construction Company is the prime contractor for the designing, fabricating, and installing the Mod-2. Current plans call for 3 prototype units to be installed at a single site during 1980--81. The site has not yet been selected. An artist concept of the Mod-2 is shown in Fig.6. Wind turbine cost-of-electricity The cost of electricity (COE) in cents per kilowatt hour is plotted as a function of site mean wind speed (at 30 ft.) in Fig.7 for the NASA wind turbines either operational or under development. This COE plot reflects the machine capital costs for the second units that are built and assumes that the machines will operate and generate electricity 90 percent of the time that the wind is in the proper speed range.

4045 ~ ~IoEC0scN0:L ~UNIT ~,

MOD-OA

30 ~

COST

25 b

w

8

is

5

_ TECHNOLOGYIMPROVED~MOD-1 -- ~'-~.. MOD-2 MATURE ]-"'''")PRODUCT 1 12 SITMEAN E WINSPEED, DlP mph

(3

~ O D - 2

18

Fig.7. Cost of electricity. The cost of electricity (COE) produced by wind turbines is computed as follows: (Capital Cost, $) (Fixed Charge Rate, %) COE (cents/kWh) = (Annual Energy, kWh) + (Annual O&M cost, $) (Levelizing Factor) (100) (Annual Energy, kWh)

331 The cost of electricity is taken to be at the output of the installation's stepup transformer. Capital cost, O&M costs, annual energy production, fixed charge rate and levelizing factor, are briefly discussed in the following paragraphs.

Capital costs The installed equipment costs of the prototypes of the large, horizontalaxis wind turbines currently under development are (19775): Mod-0A Mod-1 Mod-2 Cost ($M) ($/kW)

1.61 8050

5.40 2700

3.37 1350

Second-unit costs are quoted so as not to include the nonrecurring costs associated with the first prototype unit. The Mod-0A second-unit prototype cost represents Lewis' estimate of the cost that would be required to build a second unit identical to the first unit at the Clayton, New Mexico site and reflects the knowledge and experience gained as well as the actual costs incurred in that first installation. For Mod-1, the second-unit prototype cost is based upon an estimate made by the General Electric Space Division after they completed the fabrication and testing of the first Mod-1 prototype. The Mod-2 second-unit prototype cost is based upon the current estimate of the installed equipment for the second prototype unit to be built by the Boeing Engineering and Construction Company.

Fixed charge rate The fixed charge rate (FCR) is a capital levelizing or annualizing factor which accounts for the return to investors, depreciation, allowance for retirement dispersion, income and other taxes, and other items such as insurance and working capital. It is a function of the design life of the unit, the general inflation rate, the debt/equity ratio of the utility and other financial parameters such as the weighted average cost of capital. A fixed charge rate of 18% has been assumed in computing the COE for large horizontal-axis wind turbines. This is a representative value for investor-owned utilities, assuming a general inflation rate of 6%, no allowance for tax preferences, an after-tax weighted average cost of capital of 8.0% (10% before tax) and a 30-year life.

Levelizing factor In order to correctly compute the total levelized revenue requirement or COE of a wind turbine (or any utility powerplant for that matter), expenses such as O&M costs which will tend to increase with time due to inflation (and thus result in a variable stream of annual costs) must be levelized before adding them to the levelized capital investment. Levelization of expenses can be accomplished by multiplying the first year's expense by a levelizing factor. The levelizing factor is a function of the

332 general inflation rate, the cost of capital and any real escalation (above inflation) to which the expense may be subject. Using the assumed values of economic parameters described above and a 0% real escalation rate on O&M costs, the corresponding levelization factor is 2.0. Annual operations and maintenance costs Sufficient operational data are not yet available to determine an appropriate O&M cost for the prototype units. Furthermore, because the prototype units are aimed at providing in-service testing and hardware qualifications, larger O&M costs will be experienced in these early prototypes than are expected from production units. The annual levelized O&M cost for the prototypes was assumed to equal 2% of the total capital investment. A total fixed charge rate of 20% (18% on capital plus 2% for O&M) was therefore applied to the total capital investment to compute the COEs in Fig.7. During the Mod-2 design, Boeing performed detailed O&M estimates for production wind turbines operating in a 25-unit cluster. These O&M estimates are approximately 1% of the capital costs thereby making the 2% estimates reasonable for the second unit COE estimates. Annual energy The annual energy o u t p u t for any horizontal-axis wind turbine can be computed for a specific wind speed duration curve by computing the power output at each wind speed and integrating it over the appropriate time duration for each wind speed. The power output as a function of wind speed for the Mod-0A, Mod-1 and Mod-2 wind turbines is shown in Fig.8. 1600(- 90~ AVAILABILITY 1400(- -

S

OD-Z

10 00( MOO-1

i~0 4 OOC

500

2000

~ -"-'i 10 20 30 WIND SPEED(AT 30 ft), mph

4O

Fig.8. Wind turbine power output.

I0

MOD-O.A

J

J

I

12 14 16 18 SITEMEANWINDSPEED.

Fig.9. Wind turbine annual output.

f 20

333

The annual energy produced by the Mod-0A, Mod-1 and Mod 2 are illustrated in Fig.9 as a function of the mean wind speed at 30 feet above the ground. These annual electrical energy production curves account for the aerodynamic, electrical, and mechanical losses up to the busbar (output side of wind turbine's step-up transformer}. They also include a 90% availability factor, i.e., the wind turbine is assumed to be available for service 90% of the time that the wind speed is in its operating range.

Estimated cost-of-electricity for prototype large wind turbines The cost of electricity (COE) for the second prototype units of Mod-0A, Mod-1, and Mod-2 is shown in Fig.7 versus the site mean wind speed. The installed equipment costs, annual energy estimates, estimated O&M fixed charge rate and levelizing factor discussed above were used in computing these COEs. As noted in Fig.7, the significant reduction in COE from Mod-0A to Mod-1 is mainly attributable to economy of scale. The COE reduction from Mod-1 to Mod-2 is mainly the result of improved technology, i.e., moving from a rather stiff and heavy design to a relatively soft and lighter weight design. The COEs of the prototype units displayed in Fig.7 are not low enough to be generally attractive to utilities. However, in quantity production the capital costs for Mod-2 are expected to decrease substantially. The dotted curve shows the projected Mod-2 costs for the 100th production unit. The current estimate of the installed equipment costs (19775) for the 100th production unit of Mod-2, along with a current estimate of weights, are given in Table 1. TABLE 1 Estimate of equipment cost and weights for Mod-2 Costs are per-unit costs assuming a 25-unit cluster.

Machine Rotor subassembly Drive train subassembly Nacelle subassembly Tower subassembly Transportation and installation Transportation Site preparation Erection and checkout Initial spares and maintenance equipment Production facility depreciation

Weight on foundation (lbs)

100th unit cost ($k)

(588204) 169567 103892 63279 251466

(1163) 329 379 184 271 (328) 29 162 137 (35) (35)

Subtotal FEE (10%)

1561 156

Total capital for installed equipment

1717

334

We anticipate that the Mod-2 will be cost-competitive in certain areas with attractive wind sites where current fossil fuel costs are high. However, before there is broad market penetration, further reductions in COE must be achieved. We believe that these reductions can be made. New system developments currentiy planned will incorporate new technology which will result in lighterweight, lower-cost systems.

Development assessment

Technology Major improvements have been made in wind-turbine technology over the past five years. As a result, wind-turbine system design is understood. The design tools are in an advanced state of development and are available to U.S. industry. Operational machines at utility sites have validated the basic system, electrical, structural, and mechanical designs. The compatibility of the single unit wind turbine with utility interfaces has been successfully demonstrated. Further operational experience is required to address long-term reliability. Additional blade development is required and is under way to reduce cost and weight. Metal, fiberglass, and wood blades are all currently attractive candidates. A 150-ft. fiberglass blade has been fabricated and tested by Kaman Aerospace and shows promise of providing a cost-effective design approach (Fig.10). Two 100-ft. blades of the 150-ft. design technology are being built for testing on the Mod-1 in 1980.

Fig.10. Karnan 150-ft. fiberglass blade.

335

Environmental issues No serious environmental issues have been identified which would impede the development of large wind turbines. Experience to date has shown that large wind turbines can be designed to be safe, quiet, and clean; however, television interference is a siting consideration. Economics and market potential Large megawatt machines have the greatest potential for application in utility networks because of economies of scale and are the only machines that will generate significant amounts of electricity. In most applications, wind turbines must produce electricity at 2 to 3 cents per kilowatt hour to have wide application as a utility fuel saver at current fuel prices. The Mod-2 machine approaches these costs, and moderate COE reductions and/or higher fuel costs will result in substantial market potential. Concluding remarks First-generation technology large wind turbines (Mod-0A and Mod-1) have been designed and are in operation at selected utility sites. Second-generation machines (Mod-2) are scheduled to begin operations on a utility site in 1980. These second-generation machines are estimated to generate electricity at less than 4 cents per kilowatt hour when manufactured at modest production rates. However, to make a significant energy impact, costs of 2 to 3 cents per kilowatt hour must be achieved. The U.S. program will continue to fund the development by industry of wind turbines which can meet the cost goals of 2 to 3 cents per kilowatt hour when operating at sites with modest winds (14 mph mean at 30 ft.). These lower costs will be achieved through the incorporation of new technology and innovative system design to reduce weight and increase energy capture. The challenge, however, is associated with acceptance by the utilities of wind turbines as part of their energy generating capability and the creation of a competitive industry to produce wind turbines efficiently. The principals -- government, industry, and the utilities -- are currently involved in meeting this challenge in the United States.

References 1 J. Glasgow and W. Robbins, Utility operation experience on the NASA/DOE 200 kW wind turbine, DOE/NASA/1004-79/1, NASA TM-79084, Paper presented at Energy Technology VI Conf., Washington, D.C., Feb. 26--28, 1979. 2 Mod-1 wind turbine generator analysis and design report -- Executive summary, General Electric Co., Philadelphia, Pennsylvania, under NASA Contract NAS3-20058, Report No. DOE/NASA/0058-79/3, NASA CR-159497, March 1979. 3 J. Lowe, Status and outlook of megawatt size wind turbines for utility applications, Paper Presented at Energy Technology VI Conf., Washington, D.C., Feb. 26--28, 1979. 4 J. Rainier and R. Donovan, Wind turbines for electric utilities: Development status and economics, DOE/NASA/1028-79, NASA TM-79170.