Small on-site turbines have a long history. Now the arrival of small, domestic-scale wind turbines on the shelves of hardware stores has been welcomed by many but criticized by others as ‘vanity’ development, rather than a serious commitment to low-carbon generation. Sander Mertens and Eize de Vries explore the issues and focus on urban wind turbines operating in the highly complex flow of the built environment.The output of any wind turbine depends, to a large extent, on four main factors: prevailing wind speed at the rotor axis height or hub height, rotor swept area, overall system reliability, and total power conversion efficiency from wind to electricity. These factors, particularly the influence of swept area and hub height, might in theory preclude the use of small turbines, but reliable micro-electricity generating wind turbines up to about 20 kW have been in use since the 1920s or 1930s. Indeed, installations of famous makes like Jacobs have provided battery-powered electricity to rural farms and other remote applications without access to grid power, often for decades.
However, one technical difference between earlier turbines and today’s modern machines is that many variable-speed micro wind turbines can now be grid-connected with the aid of a frequency converter, whereas in the past, battery intermediate storage was the common choice. Modern power electronics vastly increase the chances of micro wind turbines becoming a substantial power source in a future energy infrastructure based upon the widespread application of renewables.
Small blades, global influence
While micro wind turbine deployment may well become a substantial contributor to a sustainable remote power supply in developing, as well as industrialized, countries, an additional application that is enjoying considerable interest is ‘microgeneration’ in the built environment – a space that is more familiar to solar PV.
Rural settings provide a comparatively easy environment for micro wind turbines – characterized by an open landscape with relatively undisturbed airflow and modest average wind speeds of 4–6 m/s. However, this potentially favourable operational picture is dramatically different from the complex environmental and operational wind conditions found in built-up areas.
In the built environment average wind speeds are generally low, around 2–4 m/s, because buildings tend to shelter installations and slow down wind. Nonetheless, recent research has shown that the built environment offers plenty of opportunities for exploiting wind power. However, this requires skill and clever combinations of carefully selected wind technology destined for specific building types, together with the right technical and environmental conditions (see textbox, page 137). Each set of urban conditions requires a wind technology capable of coping with site-specific circumstances, together with a built-in capacity to optimize and guarantee the owner-operators reliable long-term energy yield at an acceptable cost.
Power output in formula
Unchanging for all wind turbines – big or small – are a number of crucial factors that together determine the annual energy-generating potential in kWh/m2 of rotor swept area. Key factors that impact potential energy yield and their physical relationships are expressed in the image gallery.
P = wind turbine power performance fed into the grid (watts)
CP = aerodynamic efficiency of conversion of wind power into mechanical power, often called the power coefficient
Πme = conversion efficiency of mechanical power in the rotor axis into mechanical power in the generator axis. Encompasses all combined losses in the bearings, gearbox, and so on
Πel = conversion efficiency of mechanical power into electric power fed into the grid, encompassing all combined losses in the generator, frequency converter, transformer, switches etc
p = air density in kg/m3 (~1.25 kg/m3 depending on environmental conditions)
v = wind speed some three rotor diameters upwind from the rotor plane in m/s
A = rotor swept area in m2
Each of the elements of the performance formula has its own distinct contribution to total wind turbine power output and resulting yearly energy yield.
Factors affecting performance
One of the key performance differences in wind turbine designs is caused by the driving mechanism of the rotor. A wind turbine rotor can either be turned around by the aerodynamic drag of rotor blades or by the aerodynamic lift of rotor blades.
With drag-driven wind turbines, the rotor blades are pushed around by the wind and move in the same direction. Such turbines typically combine a low aerodynamic efficiency or small CP number (<11%) with a high materials input requirement and, consequently, are usually rather expensive to manufacture when set against their comparatively limited power output. A basic drag type machine is shown below.
The distinctively shaped Savonius rotor, named after the Finnish inventor, is one of the best known drag-type micro wind turbine models, as shown below (centre) in a common application.
By contrast, in turbines operating according to the lift principle, each rotor blade is turned by aerodynamic lift force, identical to the force that keeps an aircraft airborne. Rotor blades that operate based on this driving mechanism feature an airfoil shape, like the wings of an aircraft. Lift-driven rotor concepts combine a high CP of up to the theoretical maximum of 59.3%, called the Betz limit, with a much more favourable materials input. The photo below (right) shows an elderly example of this type of design.
The differences between both rotor concepts are striking: lift-driven wind turbines combine a high CP with a low material use, while a drag-driven wind turbine links a low CP to high material use. It is therefore difficult to design a micro drag-driven wind turbine that is economical in terms of costs of energy in €/kWh/20 years. Even so, some manufacturers of micro drag-driven wind turbines argue that the rotor, and therefore their complete turbine, is by comparison inexpensive to manufacture and that, as a result, this fully offsets the much smaller energy yield. However, a micro wind turbine rotor on average counts for only about 20% of total installation costs.
To complicate matters, a certain Savonius-type design concept that features a gap between the rotor blades combines a relatively high CP of 25% with a ‘drag-driven’ type wind turbine look. This particular rotor design is therefore not only driven by drag, but also by lift forces. Vortices around the rotor blades introduce an additional driving force. Furthermore, small state-of-the-art lift-driven wind turbines generally feature a lower CP of 30%–42% compared with a CP of 50% for their bigger equivalents. Enercon in 2003 claimed to have reached the Betz limit with its new generation rotor blades fitted on its turbines in the 100 kW to 6 MW+ range. This substantial CP difference is a consequence of the increase in drag and a decrease of the lift as the blade size gets smaller.
The functional difference between horizontal or vertical axis wind turbine rotor type classification is rather straightforward, as the names suggest. With regard to performance behaviour, wind tunnel measurements of CP for horizontal and vertical axis wind turbines do not show much difference as both types share the same driving force, lift driven or drag driven. Yet substantial differences in performance between both rotor types can occur in harsh and complex wind conditions, especially in the built environment.
Mechanical efficiency (Πme)
Mechanical losses in a wind turbine are mainly due to friction in components such as bearings and gearbox components. With many micro wind turbines, the rotor drives the generator directly in a so-called direct drive or gearless system. In other wind systems a gearbox or a belt drive is applied in the drive train to step-up the relatively slow rotor speed to a much faster generator speed. The required step-up ratio is a function of generator nominal speed divided by rotor nominal speed. Mechanical losses in a micro direct-drive system are potentially less compared to a geared or belt-driven drive train as they inherently contain fewer rotating components. Total mechanical efficiency of micro wind systems lies typically in the 96%–99% range.
Electrical efficiency (Πel)
Electrical efficiency encompasses all combined electric power losses in the generator, converter, switches, controls, and cables. The converter, as a key power electronics component in variable speed wind turbines, turns AC with variable frequency into fixed frequency AC grid-compliant power at 50 Hz or 60 Hz. For small state-of-the-art wind turbines in the 0.5–10 kW range, total electric efficiency is usually in the 60%–70% range. Permanent magnet-type synchronous generators offer superior partial-load efficiency compared to equivalent electric machines, but with external field excitation. As a rule, as generator power rating increases so does the electrical efficiency. For instance, for a squirrel-cage asynchronous generator in the 5 kW range, a maximum efficiency of 84%–85% is a state-of-the-art value, but, according to electric power experts, this increases to about 96%–97% for a similar product in the 2.5–3 MW class.
Influences on air density (p)
The Boyle-Gay-Lussac Law shows the impact of temperature and pressure on density, whereby density is proportional to pressure divided by temperature. The difference in air density at pedestrian level and at the top of a skyscraper is imited to only a few percent. Air density difference as a function of temperature is somewhat more significant, and is for instance equal to roughly 10% for a 30ºC temperature gradient. The influence of air density on wind turbine performance is therefore limited, and importantly one cannot influence air density except by choosing a different site for the turbine.
Average wind speed (v)
In the operational output range, wind power generated increases with wind speed cubed. In other words, at a wind speed of 5 m/s, the power output is proportional with 5 cubed (53) = 125, whereas at a wind speed of 6 m/s, the power output is proportional to 63 or 216. This shows that an increase of wind speed of only 20% gives rise to a power increase of 73%. It is obvious that the average wind speed at the hub height is the most important variable in terms of potential contribution to total energy yield. From the nominal speed range and up, wind turbine power output is usually curbed for technical and economic reasons and preferably kept at a constant value. This is achieved with the aid of technical measures like a mechanical system and/or in combinatio n with power electronics.
When a micro wind turbine operates in an open area, its output – and therefore overall installation economics – can be raised substantially by increasing the tower height. This is especially beneficial for sites where many obstacles like trees, and/or man-made structures intervene. A second benefit of higher towers is that rotors turn in a much more stable wind regime. This works out positively in terms of a much-reduced rotor wind shear, and fewer yaw actions. The result is a substantially reduced contribution to materials fatigue, a major factor in operational life expectancy.
Rotor swept area (A)
Rotor swept area is a function of the rotor diameter squared and is the second key wind turbine output variable. When the rotor diameter of a microturbine increases for instance by 20% from 5 metres to 6 metres, swept area increases by 44%. As a rule of thumb, the lower the average wind speed at a given hub height, the larger the rotor swept area required in relation to generator capacity to maintain
an ‘acceptable’ energy yield and closely linked favourable economics.
Inherently, most of the time wind turbines operate under partial load conditions. They also run a much smaller proportion of total hours under full-load conditions. A third condition is not running (stationary) – either due to a lack of wind, wind speeds higher than the operational range, or during maintenance and/or repair downtime. The number of full-load hours a given wind turbine produces on an annual basis is expressed as the capacity factor, Cf. Assuming a 5 kW wind turbine on a coastal location generates annually 10 MWh, if that same installation had run – theoretically – 24 hours a day and 365 days a year at full load, it would have generated 43.8 MWh. The capacity factor (Cf) is 10/43.8 = 0.23. However, if the same wind turbine is installed at a medium wind speed inland site, annual production can drop to 6 MWh or less, with obvious impacts on Cf. Average Cf for all wind turbines in the Netherlands during 2006 was 0.23. For small wind turbines a Cf of 0.23 is regarded as excellent performance. However, due to insufficient wind at low hub heights and/or legal restrictions on the maximum permitted installation height, generally a much lower Cf figure for microturbines is, unfortunately, more realistic. Finally, as a measure to achieve a favourable Cf, and therefore optimized economics, the ratio between generator capacity and rotor swept area has to match well. In fact, each wind turbine with a certain power rating should be offered with a range of rotor diameters for specific wind locations as an optimizing measure.
The majority of micro wind turbines operate with a predefined cut-in speed; certain makes and types also operate with a predefined cut-out wind speed. The difference between cut-in and cut-out wind speeds is the operational range. Micro wind turbines typically operate in relatively modest wind speeds. Under these conditions it is beneficial that the installation starts up at a relatively low cut-in wind speed of about 1.5–2 m/s. However, based on manufacturers’ product information, 3–4 m/s seems a more typical value.
When a given installation is ready to operate all the time within a predefined wind speed range, availability is defined as 100%. In reality availability is never 100%, as each installation tends to break down occasionally, and, in addition, there is always downtime due to service visits, repairs, grid failures, and such like.
Traditionally wind turbines applied in an open field are horizontal-axis designs fitted with an upwind rotor. The rotor thereby faces the wind when turning in front of the tower and these installations typically feature two, three or four aerodynamically shaped rotor blades when based on the lift principle. Many microturbine designs are, for cost reasons, fitted with fixed angle blades and operate on the stall principle. Others have pitch-variable rotor blades, a mechanically or electrically operated control and safety system aimed at limiting output beyond the nominal wind speed. Both upwind stall and pitch variable-type wind turbines often feature a mechanical system like a tail, to continuously redirect the rotor towards the prevailing wind direction. However, in many of the stall-type micro wind systems as a second main function the tail also acts as a safety device for power output limitation/control during high wind speed conditions. This function is achieved by gradually folding the tail towards the rotor plane with increasing wind speeds. As a result the tail still follows the prevailing wind direction, but the rotor is forced into inclined inflow conditions. This gradual repositioning often begins from about 5–6 m/s. The effective area of the inclined rotor perpendicular to the prevailing wind direction thereby changes gradually from a circle at low wind speed into an ellipse at medium and higher wind speeds. Power output, in turn, is curbed to a maximum range due to a combined effect of the reduced effective rotor swept area, and a collapse in aerodynamic rotor efficiency. However, due to the fact that this phenomenon typically begins to initiate even at low wind speeds, this type of installation rarely fully exploits its entire rotor swept area. This hidden loss factor is directly linked to this specific safety system.
With an alternative mechanical safety system, the rotor plane pivots backwards from a certain wind speed. The functional result is similar, with output limitation due to a reduced effective rotor area and a collapse in aerodynamic rotor efficiency.
Some state-of-the-art micro wind turbines, such as the 2.5 kW WES5 Tulipo, are also equipped with an active yaw system that redirects the rotor to the prevailing wind direction and comprises a wind direction sensor, yaw motor(s), and sometimes a yaw brake. Fitting wind turbines with an active yaw system and a yaw brake can be highly advantageous at complex wind sites with a high degree of turbulence. This is because these installations do not continuously follow the typically frequent changes in wind direction. This in turn substantially reduces risks of premature material fatigue-induced component and system failures, such as rotor failure in the blade foot and main shaft. However, such systems are inevitably more costly to manufacture compared with free-yawing tail-type systems, although they can offer superior yield levels and elongated service life.
The rapid development of advanced power electronics offers opportunities to eliminate both folding tail or rotor tilt back type safety systems for variable speed upwind stall turbines. Instead a system called Rotor Speed Control (RSC) is applied which electronically limits power output from a certain wind speed.
A functionally similar rotor speed control system is applied in the American Skystream 2.7, a 1.8 kW downwind-type microturbine manufactured by Southwest Windpower. In a downwind turbine the rotor turns behind the tower and directs itself continuously towards the prevailing wind direction without the need for a tail system.
Vertical-axis wind turbines operate independent of wind direction, and therefore do not require a system to redirect the rotor to the wind. However, any rotor with fixed blades that continuously faces the prevailing wind, even under high-wind and storm conditions, still requires a safety device.
Some manufacturers fit a fail-safe disk brake that can bring the installation to a full stop when a maximum wind speed is being exceeded. Depending on specific design choices, fail-safe disk brakes applied on microturbines can turn out to be a continuous and substantial hidden energy-loss factor.
Vertical-axis-type micro Darrieus rotors are generally not self-starting. In order to start up these wind turbines an additional power source is required. Usually these wind turbine types are fitted with a dual-mode electric machine that can act as a generator as well as an electric motor. Under conditions with a high degree of turbulence, typically 30% and up, it is likely that there might be three or more start-up actions in one hour. Therefore, depending on rotor size and other contributing variables, annual energy consumption for start-up operations can easily add up to hundreds of kWh.
In order to compare performance, safety and other key micro wind characteristics, several countries including the Netherlands, the US and the UK are individually working towards some form of certification. Certification has become a hot issue in the emerging micro wind turbine industry, but at the same time many small suppliers cannot afford the high costs normally associated with large wind turbine certification.
Nonetheless some form of micro wind turbine pre-certification is essential for building a viable professional industry in the coming years. In time it should preferably become a single wind industry standard valid for multiple countries and geographical regions. Such a certification scheme should include, among other things, different guidelines for wind turbines operating in the urban built environment and for those installations aimed at operating in the open field. Equally important is that output of all wind turbine makes and types is measured minimally at three different standard average wind speeds, such as 3 m/s, 5 m/s and 7 m/s. Three wind speeds in the lower speed range are recommended as that is typically the operational environment in which these micro installations work.
Reliable measured and validated testing data forms a key part of any certification procedure and for urban wind turbines in particular, proven performance should not only be based on wind tunnel testing, but be verified, too, at real live sites on representative building roofs.
Micro machines, big potential
Micro wind turbines offer considerable potential to become a major power source for the world’s future energy needs. The example of reliable American wind turbine designs developed nearly a century ago underlines this point. But for their modern successors aimed at both the built environment and operation in the open field, much work needs to be done to optimize overall performance, reliability and cost-effectiveness. Careful site selection is one of the primary performance variables as it determines average wind speed, the most important contributing factor for the potential annual energy output.
Sander Mertens PhD MSc BSc is Managing Director of the Voorburg (NL)-based engineering consultancy, and of project management company Ingreenious BV
Eize de Vries is Wind Technology Correspondent for Renewable Energy World Magazine
Wind power in the built environment
The site for a proposed urban wind turbine should be selected carefully, and the installation choice is crucial. Once the turbine has been selected, the prevailing wind speed at hub height is the only variable left in their physical relationship and is therefore a key-determining factor for an expected energy yield.
Building roofs towering well above surrounding buildings have, in general, good prospects for offering economic wind speeds and, as a rule of thumb, wind speed generally increases with height. But, close to the roof the situation is more complex.
Figure 1: Computer simulation of the airflow pattern around a rectangular building for wind coming from the left.
As Figure 1 shows, the wind hits the upwind building facade and flows around the structure in all directions. Part of this wind flow is bent downward and causes wind hindrance at pedestrian level. Part of the airflow also moves upward towards the roof. Following the facade, it makes a 90 degree angle with the roof at the roof edge and consequently leaves behind a wind shadow close to the roof.
The wind speed and direction are shown as arrows whose length and colour are a measure for the magnitude. Red denotes high wind speeds while blue denotes a low wind speed.
In order to avoid the area with low wind speeds, an urban wind turbine should be moved towards an area well above the roof surface. There are some rules of thumb for the calculation of the required height, but for a ‘standard’ office building example, it is approximately half the roof width. For a roof width of 10 metres the required height therefore is 5 metres when positioning the wind turbine above the roof centre. It can be shown that wind speed in this area above the roof surface is some 20% higher than the undisturbed wind speed upwind of the building. According to the power output formula, power increase in the operational area caused by this accelerated wind speed is 1.2 cubed or 1.7. In other words power output for any given system is 70% above the power output at roof height in the absence of the building.
But, instead of just striving for higher wind speed locations in the built environment, it is also possible to purpose-design specific buildings. This needs to be done in such a manner that specific shape and other measures increase wind speed at the planned location for a proposed wind turbine.
Frequent changes in wind speed and direction in the built environment may be observed, for instance when passing buildings on a windy day. These buildings form large vortices that move downwind away from the building (Figure 2).
Such large vortices passing wind measurement equipment show sudden changes in wind speed and direction. A wind turbine located in the built environment consequently has to deal with this complex airflow. Some manufacturers of urban wind turbines choose a vertical axis rotor concept in order to avoid the frequent yawing actions of a horizontal axis rotor in the direction of the prevailing wind, while others rely on the well-known horizontal axis concept. Moreover, if the wind frequently changes direction, a horizontal axis rotor nearly constantly yaws and, as a consequence, is never aligned fully with the prevailing wind direction. The resulting average misalignment with the wind direction can cause a significant drop in CP and therefore energy yield. The magnitude of this energy drop depends on the specific design features, but a 20% drop in efficiency is a realistic assumption. Additional power and yield loss occurs due to these yaw actions because the rotor is continuously accelerating and slowing down again.
Vertical axis rotor concepts inherently do not suffer from wind directional misalignment trouble, but turbine types like the H-Darrieus are not self-starting.
Drag-driven vertical axis wind turbines on the other hand are self-starting, but have the problem of a low efficiency combined with a high material use.