The capacity or size of wind turbines and what that means relative to how much electricity they will generate often confuses journalists, renewable energy advocates, policymakers, and even professionals in the wind industry.
In my 1995 book, Wind Energy Comes of Age, I referred to the size of wind turbines by their rotor diameter, including the wind turbine’s generator size in kilowatts only as needed to differentiate various models of the same size turbines.
Why I did so is explained in a long passage titled Turbine Rating. Here is an excerpt from the book explaining how wind turbines are “rated” in terms of generating capacity. Note especially the explanation of specific capacity.
In this discussion, the measure of size has been rotor diameter, not generator capacity. There is good reason for this. Wind turbine ratings, in kilowatts, give only a crude indication of how much electricity a wind turbine can produce. Worse yet, ratings in kilowatts can deceive the unwary. Rotor diameter is always a much more reliable indicator.
Power ratings grew out of the role electric utilities played in the industry’s formative years. Much of the early work was done by or for the utility industry. As a result, when utility engineers talked among themselves about wind turbines, they described turbines in the terms they knew best: generator size. They blithely applied the same terminology to wind turbines as they did to reactors at Three Mile Island or diesel generators at Nome, Alaska. Wind turbines, though, are different.
In conventional power plants, the operator controls the fuel to produce the power desired. Utilities try to run large base-load plants as close to full output as possible, leaving a little room as “spinning reserve.” Even when they bring small plants on line, they prefer to run them as close to their full capacity as possible for best performance. It makes sense, then, to call a nuclear reactor 1,000 MW if the utility operates it at nearly 1,000 MW most of the time.
Wind plants are dependent on the wind–the operator has no control over the fuel. All the plant operator can do is ensure that the turbines are in service when the wind is present. The wind may not be blowing, and even when it does the wind may often be insufficient to drive the turbine at its rated power. Wind turbines, unlike conventional plants, seldom operate for long periods at their “rated” power.
Utility engineers had to derive a different yardstick for comparing wind turbines to conventional power plants. Early wind turbine designers created a hybrid rating system that sufficed: the power output at some arbitrary wind speed. This method would work reasonably well if all agreed on the speed at which wind turbines would be “rated.” But there is no international consensus on what this speed should be. Rated wind speeds vary from 10 m/s to 16 m/s (22 mph to 36 mph).
This results partly from tailoring wind turbines to different wind regimes, and partly from different approaches to maximizing total generation. This rating approach also results from the early concept that turbines would reach their “rated” capacity, then limit output to the rated amount for wind speeds up to cut-out when they would turn themselves off. See Figure 9-7. Sample Power Curves. This is represented as a straight line on the power curve, a chart of the power produced at various wind speeds. Few wind turbines operate this neatly in the real world.
This is most apparent in wind turbines using aerodynamic stall to regulate power in high winds. Typically, power in these turbines peaks at higher than rated wind speeds, then declines until the cut-out wind speed (often 25 m/s for Danish wind turbines) is reached. The “rated” power only approximates the power these turbines will produce at wind speeds above “rated”. And among pitch-regulated turbines, the power will fluctuate above and below the rated value as the blade pitch mechanism adjusts to changing wind conditions. Power curves are only approximations of what the wind turbine will produce at any given instant, depending on whether the wind speed is increasing or decreasing.
This rating designation ultimately leads to confusion. The Carter model 25, for example, uses a rotor 10 meters in diameter to drive a 25 kW generator, loading the rotor to 0.33 kW/m2 of swept area. The Fayette 95 IIS uses a rotor about the same size to drive a 95 kW generator for a specific capacity rating of 0.95 kW/m2, nearly three times that of the Carter turbine. An unsuspecting investor could easily be led to believe that the Fayette turbine is three times more productive than its competitor, the Carter 25, because of its higher rating. The peak outputs from the two turbines do differ markedly, because of differing rotor design and generator size. Generation also differs as well. But the Fayette turbine, when it is operating, generates much less than three times the energy of the Carter turbine, in spite of what would be indicated by its higher rating. See Figure 9-8. Fayette 95-IIS.
Specific ratings today range from 0.3 kW/m2 to 0.5 kW/m2. See Table 9-7. Some turbines, such as the Carter 300 and WindMaster 300 kW, have specific ratings of up to 0.7 kW/m2, but these are outside the norm. Other manufacturers of the same size turbines would rate them at 200 kW or 0.4 kW/m2. Historically, specific ratings were even lower.
In the early 1970s, NASA designed its 100 kW Mod-O prototype with a rotor 38 meters in diameter, giving it a specific loading of only 0.1 kW/m2. Its successor, the Mod-OA, used a generator twice the size of the Mod-O, pushing the specific rating up to 0.2 kW/m2.
Several years later, the private sector began introducing its own turbines. The early Enertech E44 (13-meter), for example, was rated at 20 kW for about 0.14 kW/m2. This was similar to Bergey Windpower Co.’s small 1 kW (2.8 m) model at 0.16 kW/m2. Later models of both companies pushed ratings higher. Bergey rated its 7 meter (23-foot) diameter Excel at 10 kW or 0.26 kW/m2. Enertech introduced its 44/40 as a 40 kW commercial version of its earlier machine for wind plants with a specific rating of 0.28 kW/m2. Subsequently, Enertech increased the rating further to 60 kW for a specific capacity of 0.43 kW/m2.
During the 1970s and early 1980s, there was a clear trend among U.S. manufacturers toward higher rotor loading in successive models. Turbines introduced from 1983 to 1985, for example, had higher loadings than those introduced during the 1970s. During the years of peak development in California, manufacturers were tempted to raise the kilowatt rating. A higher rating lowered a turbine’s relative cost, in dollars per kilowatt, in the eyes of unsophisticated buyers, and was used to competitive advantage by some manufacturers. Since the tax credits expired, there has been less incentive to artificially boost a turbine’s kilowatt rating.
Some engineers argue that ratings are important, as they to tell project designers how to size the wind plant’s transformers, collectors, and substation. Few use a turbine’s rating as such a guide. To size the power collection system properly, the peak capacity given on the generator’s nameplate must be used. And generators are rated in a different manner than wind turbines are. Those who develop and finance wind projects today are solely interested in cost and annual generation, as well they should be. Wind turbine manufacturers continue to “rate” their turbines more out of a necessity to identify their product as much as for any other reason.
Tailoring a wind turbine’s performance to a specific wind regime does require a skillful matching of rotor and generator performance. If the rated speed is too low, at a very energetic site too much energy will be lost in high winds. Conversely, if the rating is too high, performance at a less energetic site will suffer. Picking a rated speed is as much an art as a science, because gearboxes and generators are not manufactured in continuous increments. There are only discrete sizes available. The rated capacity and the size of the gearbox and generator are partly determined by the manufacturers of the generator and gearbox manufacturer. Small variations from the optimal rating are not critical for most temperate latitude sites where wind turbines are currently deployed.
The influence of rating on overall performance is more important where wind turbines will operate in wind regimes with characteristics far differenct from those of Northern Europe or the continental United States, such as the Caribbean. The optimal rated power at trade wind sites, as on the island of Curaçao off the coast of Venezuela, should be somewhat lower than that of machines now in use elsewhere because there are few periods with high winds. Generators over-sized for the occasional high winds of temperate climates would operate nearly all the time at a fraction of their capacity on Curaçao, resulting in lower efficiency.[1]
To avoid the rating dilemma, most European manufacturers refer to their products by rotor diameter. Thus, Vestas’ V39 is a wind turbine 39 meters in diameter. American manufacturers eventually adopted a hybrid designation: diameter followed by generator size. For example Enertech’s 44/40 or U.S. Windpower’s 56-100 designate wind turbines respectively 44 and 56 feet in diameter, powering 40 kW and 100 kW generators.
Eric Miller of U.S. Windpower notes that the rating of conventional power plants is also subject to interpretation. The ratings of many cogeneration (combined heat and power) plants, for instance, vary seasonally with changes in air density because they use aeroderivative jet engines. U.S. Windpower has abandoned the hybrid designation for its newest model, once known as the 33-300 because of its 33-meter (110-foot) rotor and 300 kW generator. The new turbine is dubbed simply the 33M-VS, which refers to its rotor diameter in meters and its variable speed operation. The rating in kilowatts of the new turbine, says Miller, becomes a function of wind regime. The turbines will have a higher rating (360 kW) for windy sites along the Columbia River Gorge, than for those in Minnesota (340 kW) or in California’s Solano county (300 kW). Like their competitors, U.S. Windpower will tailor the turbine’s power curve to the site as required.
Danish manufacturers take a slightly different approach. They will supply different rotors for different wind regimes. For example, Vestas supplies a 29-meter (95-foot) rotor with its 225 kW model to the Midwest (0.34 kW/m2) while it ships the same turbine with a rotor 27 meters (90 feet) in diameter (0.39 kW/m2) for a windy ridge in the Tehachapi Pass.
This illustrates the weakness of the rated power designation; the rotors differ in size, the power curves differ, and the rated speed differs. But both turbines reach the same rated power. The turbine with the 29-meter rotor reaches its rated capacity at a slower wind speed (15 m/s) than the 27-meter rotor (14 m/s). The larger rotor shifts the power curve slightly to the left, so the turbine will generate more power at low wind speeds. The turbine will endure greater loads at higher winds than the smaller rotor, but winds at these speeds occur less frequently at Midwestern sites than at Tehachapi, justifying the trade-off between greater energy generation and greater wear and tear.
Engineers use the power curve to estimate the turbine’s annual energy output or AEO. By matching the power curve to the wind speed distribution for a specific site, they can more accurately project production than by using the turbine’s swept area and guessing at its overall efficiency.
For example, consider the power curve for the 500 kW, variable-pitch turbine shown previously in Figure 9-7. This turbine uses a rotor 39 meters (130-feet) in diameter (1,195 m2 of swept area) to drive a 500 kW generator that reaches its rated capacity at 16 m/s (36 mph). See Figure 9-9. At a 7 m/s (16 mph) site with a Rayleigh distribution, winds at this speed occur about 75 hours per year. Winds at the rated speed contribute some 37,000 kilowatt-hours per year to the turbine’s total generation. All winds above rated produce about 90,000 kilowatt-hour per year, or only 7% of total generation at this site. Most of the energy generated by this wind turbine at this site is produced at wind speeds lower than rated, once again illustrating that rated capacity says little about how much energy the turbine will generate.
At a 7 m/s site with a Rayleigh distribution, the energy density is 3,515 kWh/m2/y (See Table 9-2). The 500 kW wind turbine in this example captures 1,080 kWh/m2/y of swept area or 31% of the energy in the wind.
The performance projection for this 500 kW machine is typical for medium-sized turbines. Small wind turbines are considerably less productive at such windy sites. See Figure 9-10. Typical Specific Yields.
Wind Energy Comes of Age, by Paul Gipe, John Wiley & Sons, New York, 1995, 536 pages, ISBN 0-471-10924-X, pp. 155-161.