CHAPTER 4 PITCH CONTROL OF WIND TURBINE GENERATORS

Transcription

1 62 CHAPTER 4 PITCH CONTROL OF WIND TURBINE GENERATORS 4.1 INTRODUCTION The use of wind power has in the last decade increased in the central parts of Europe and at the west coast of the U.S. The rest of the world is following these and today this is the fastest growing energy source on a world basis. India has one of the fifth largest wind power installed capacity in the world. This chapter describes the modeling of the various components in a pitch controlled wind energy system. It also describes the design of the pitch controller and discusses the response of the pitchcontrolled system to wind velocity variations. The pitch control system is found to have a large output power variation and a large settling time. The pitch function gives full control over the mechanical power and if the most common method is used for the variable speed wind turbines. At wind speeds below the rated power of the generator, the pitch angle is at its maximum though it can be lower to help the turbine accelerate faster. Above the rated wind speed, the pitch angle is controlled to keep the generator power at rated power by reducing the angle of blades. The pitch control system is one of the most widely used control techniques to regulate the output power of a wind turbine generator. The method relies on the variation in the power captured by the turbine as the

2 63 pitch angle of the blades is changed. Hydraulic actuators are used to vary the pitch angle. The wind turbine generator describes the design of the pitch controller and discusses the performance of the system in the presence of disturbances. The control system consists of two controllers are inverter controller that keeps the load voltage constant and the pitch controller acting on the blades angle by using simulation. The variable speed induction generator using Volts/Hertz control enables efficient wind energy capture and is shown in Figure 4.1. The output of the generatorside converter can be varied to control the speed of the induction generator (Schians 2007). The gridside converter can be controlled to inject the desired power and reactive power into the grid. Thus, the wind system is capable of providing reactive power support, if required. The disadvantage of this configuration is the large cost associated with the power converter, since the converter has to be rated for the maximum power output of the generator. Figure 4.1 Volts / Hertz controlled induction generator

3 64 For fixed speed turbines, the active stall can be used to limit the power, not by reducing the pitch angle as in pitch controlled, but by increasing the pitch angle to a point where the blade stalls and in that way reduces the force on the turbine. By regulating, the angle to be on the limit of stalling, fast torque changes from the wind will be neutralized (Schinas et al 2007; Nayar and Bundell 1987). 4.2 PRINCIPLE OF CONTROL Aerodynamic Power Control for Wind Turbines When a generator reaches rated power, the turbines must limit the mechanical power delivered to the generator. This is valid because the generator reaches the rated power at for instance 15 m/s while the maximum speed is typically 25 m/s for a wind turbine. Control is done by three different methods called stall, pitch and a combination called active stall. There are no moving parts in the stallcontrolled blades and the challenge is in the construction of the blades to avoid vibration and make them stall gradually. The pitch angle is controlled to keep the generator power at rated power by reducing the angle of the blades. By regulating, the angle to be on the of stalling, fast torque changes from the wind will be reutilized (Nayar and Bundell 1987). The power captured by the turbine is given by Pm = Pw Cp

4 65 Figure 4.2 Airstreams around a turbine The mass flow rate is a constant for upstream (0), the rotor (1) and down stream (2) mass flow rate mass as shown in Figure 4.2. m= pav 0 0 = pav 1 1 = pav 2 2 (4.1) The turbine model represents the power captured by the turbine. The power in the wind (P w ) in an area A is given by, P w 1 3 = ρ AVw (4.2) 2 where ρ is the density of air and V w is the wind speed. However, the turbine captures only a fraction of this power. The power captured by the turbine (P m ) previously expressed as P m = P w C p, where C p is a fraction called the power coefficient. The power coefficient represents the fraction of the power in the wind captured by the turbine and has a theoretical maximum of The power coefficient can be expressed by a typical empirical formula as, Cp 1 2 ( γ 0.022β 5.6) 2 e 0.17 γ = (4.3)

5 66 The variation of C P with the pitch angle β for various values of the tip speed ratio γ. Thus, by varying the pitch angle, the power coefficient can be changed and the power captured by the turbine can be controlled. If the C p γ performance characteristic is available, the turbine parameters can also be modeled from data fields containing the group of curves derived from measurement or from calculation or analytical function. The groups of power coefficient and tip speed ratio curves obtained by measurement or by computation can also be approximated in closed form by nonlinear functions as shown in Figure 4.3. Figure 4.3 Power Coefficient vs tip speed ratio for various values of pitch angle Normal operation, blade pitch adjustments with rotational speeds are approximately expected at β n = 5 to 10 / S = 0.09 to 0.17 rad/sec (Siegfried 2006) Wind Turbine Controller The name of the controller for the wind turbine is the multi processor controller, which is an abbreviation of multi processor controller. The controller monitors and controls all functions in the turbine, in order to ensure that the performance of the turbine is optimal at any wind speed.

6 67 The control will stop the turbine if supervision detects an error. At an operator panel, data of current operation is displayed. The multi processor controller is divided in to two parts: 1. Ground controller and 2. Topcontroller. Ground processor takes care of cut in and cutout of the generator and the capacitors and of current and voltage measurement. Top processor takes care of the tasks in the nacelle, example: Speed, pitch and power control yawing and internal temperature control ( Data Collection The multi processor controller collects continuously data about the performance of the turbine example: Rotor and generator speed, Wind speed, Hydraulic pressure, Temperature, Power and energy production and Pitch. If some irregularities or errors arise, the data is stored in a LOG and/or an ALARM LOG, making it possible to analyze errors in the turbine System of Parameters The software to the MPcontroller system is basically constructed so that all variables can be initialized. Examples of parameters are power

7 68 reference, various alarm limits, calibration values for anemometer etc. To every type of turbine and variants, a set of parameters is present, which is selected by startup of the turbine Turbine Control with OptiTip When the turbine is stopped (PAUSE, STOP or EMERGENCY STOP), the blades will be present in a pitch angle of 90 (out of the wind), as shown in Figure 4.5. When the turbine is in RUNmode, it is able to produce electrical energy, but the momentary wind conditions are determining for how much. This is controlled by the OptiTip control system of the multi processor controller. The momentary wind conditions can be divided into four categories, shown by the power curve in Figure 4.4, Figure 4.4 Power Curve

8 69 1. Low wind, the generator is not connected to grid. 2. Medium wind, the generator is connected, but does not produce nominal power. 3. Higher wind, the generator is connected and produces nominal power. 4. Stop wind, the generator is disconnected and the turbine is stopped. When the wind is very low and the rotor does not rotate with a very low speed, the pitch angle will be approximately This will give maximum start moment to rotor, which gives a quicker start, when the wind increases. The MPcontroller will then pitch the blades to 0 (into the wind). The rotating speed for the rotor and the generator will increase towards the nominal, which the multi processor controllers will try to keep (speed control). The same situation occurs with connected generator. When the wind decreases and the produced power become negative, the generator will be disconnected from the grid and the MP controller will control the speed. If the wind decreases even more then the rotating speed will decrease below nominal value and the rotor will run freely. At medium wind speed, the rotating speed is regulating to the nominal value and if the pitch angle can be kept at 5 0 (which says that there would be enough energy in the wind), the generator is connected to the grid. With the generator connected and but enough energy in the wind to produce nominal power, the pitch angle is regulated as a function of the wind speed.

9 70 This function is calculated with precision, simulated and evaluated by measurements. It is called OptiTip. This is implemented in turbines to optimize the aerodynamics of the blades, which will optimize the energy production as shown in Figure 4.5. Figure 4.5 Pitch setting at different operating states When the generator is producing power, it causes a torque, in opposition to the mechanical torque of the rotor. The MPcontroller based fully automatic systems controls the produced powers, so the rotor speed is kept constant within a narrow band, which is called slip, which is the procentual relation between actual and

10 71 synchronous rotating speed and optimal adjustment of the angle of the blade in relation to the prevailing wind, thus generating maximum power. Such processes are currently used by machines in 30 KW to MW range that employ power electronics to convert the frequency to that required with their design in the limits, such models permit Power Factor Correction An asynchronous generator contains no permanent magnet and is not separately excited. That means, that an asynchronous generator has to get exciting current from the grid and the magnetic field is only established, when generator is connected to the grid. When the generator excites current from the grid, the generator consumes reactive power from the grid. The current in the cable to the generator will consist of two parts, 1. Active current corresponding to the active power production (KW) and 2. Reactive current corresponding to the reactive power consumption (KVAr). Because of the exciting current, there is a phase shift between the current and the voltage. The current delayed compared to the voltage by an angle (θ). A term of the phase shift is the power factor (cos θ). A part of the generator s exciting current / reactive power is delivered from the capacitors, which are called power factor correction. The capacitors are connected to the grid a little later than the generators are disconnected, before the generator is disconnected from the grid.

11 72 The advantage of the power factor correction is that the loss in the grid decreases, because the grid current decreases. At noload the grid current is about zero Amp, because the generator s noload current (only reactive current) is delivered from the capacitors. 4.3 PITCH CONTROL SYSTEM DESIGN Pitch control means that the blades can pivot upon their own longitudinal axis. The pitch control used for speed control, optimization of power production and to start and step the turbine. The control system structure used to generate the pitch angle reference is given in Figure 4.6 (Murdoch et al 1982). The pitch controller consists of two paths a nonlinear feed forward path, which generates β 0 and a linear feedback path, which generates β. Figure 4.6 Pitch angle reference generator The feed forward path uses the information about the desired power output, wind velocity and the turbine speed to determine the pitch angle required. Equation (4.4) gives the pitch angle as a function of the measured variables. 1 2Pref e β0 = γ Pw 0.17γ (4.4)

12 73 However, the feed forward term assumes that all the components are ideal and does not account for the losses in the system. The feedback path compensates for the losses by decreasing the pitch angle, if the output power is less than the desired power, to increase the power captured. The PI controller (Proportional integration) for the system is designed using the ZeiglerNicholas rules for tuning PID (Proportional integration and differentiation) controllers Zeigler Nicholas Rules The method used here is the first method applicable for plants without integrators or dominant complex conjugate poles, as the plant response for the wind system is found to correspond to these requirements of the ZeiglerNicholas rules. For a plant response to a unit step input as given in Figure 4.7, the intersection of the tangent line at the inflection point with the start and end values of the curve gives the delay time (L) and the time constant (T). Figure 4.7 Obtaining the time constant and delay time for a response to a unit step

13 74 For a controller of the type K P (1+1/T i s) The values of K P and T i are given as K P =0.9(T/L); T i =L/ PI Controller Design In order to use the ZeiglerNicholas rules for designing the PI controller, the step response of the plant is required. The block diagram showing the plant and controller used for generating the step response is shown in Figure 4.8. The actuator is modeled as an integrator in a feedback loop, as shown in Figure 4.9. The rate limiter limits the rate of change of the pitch angle, as most pitch actuators cannot change the pitch angle more than a particular degrees / sec. The value used for the rate limiter in the simulations is 5 0 / sec. Figure 4.8 Plant and controller definition

14 75 Figure 4.9 Hydraulic actuator model used in the simulation The unitstep response of the plant, obtained by applying a onedegree step input is shown in Figure 4.10 (Pena et al 1996). The measured values of the time delay L and the time constant T are, L = 0.03s and T = 0.155s. Thus, the values of the proportional and integral constants are, K P = 4.65 and K i = Figure 4.10 Plant response to a unit step input

15 76 The SIMULINK block diagram of the pitch controlled wind energy system is shown in Appendix A ANALYSIS OF FIELD TEST RESULTS Wind Shear Correction No correction needs to be applied to the wind speedreading for anemometer heights different from hub height Correction of Power for Air Density Variations Before data analysis using the method of bins is carried out, corrections to the data sets for air density variations must be applied. In this procedure, the wind speed range of operation of the wind turbine is divided into a series of intervals (bins). The aim of the corrections is to bring the power curve and the calculated mean power as close as possible to the values which would be obtained if the measurements were all carried out at a standard air density at sea level of Kg / m 3 ( mbar, dry air, degree Celsius or degrees Kelvin). For a stallcontrolled wind turbine, each 10 min average net power values shall be corrected by applying the following formula: [ ] P = P ρ (4.5) s T T where, P s = Power corrected to standard conditions, P T = Uncorrected average power, ρ T = Test air density and

16 77 P T is calculated from degrees Kelvin. [ ][ ] ρ T = T B (4.6) where, B T t = Barometric pressure, (mbar) = t and = Air temperature in degrees Celsius. For a pitchregulated WTG, the correction is the same as for a stall controlled WTG as long as the measured power levels are below 70 % of rated power. For measured power levels above 70 % of rated power, the correction applies instead to wind speed (and not to power) according to the following expression. V s t T [ 1.225] 1/3 = V ρ (4.7) where V t is the measured, uncorrected wind speed (m / s) and V s is the wind speed corrected to standard conditions. n i P = 1 n P (4.8) i i ij j= 1 where, V ij = j th 10 minute average of wind speed in the i th bin, P ij = j th 10 minute average of net power in the j th i th bin and n i = Number of data sets in the i th bin. The ensemble averages (V i, P i ) are then plotted and a curve fitted through the plotted points. This curve is the WT power curve. The minimum conditions of shorter preaveraging time does not reduce the total time, as the

17 78 number of data sets per bin multiplied by the pre averaging time shall be constant and must be met before the curve is established. The power curve is a linearly scaled Cartesian coordinate system graph of WT net power corrected for air density variations (ordinate) versus wind speed (abscissa) as shown in Figure Both scales start at zero. The ordinate scale should extend to at least 110 % of the WT maximum power. The abscissa should extend to a wind speed of at least 20 meters / second. The power curve is to be displayed graphically as indicated in Figure 4.11 and in form of a table as shown in Table 4.1. Care should be taken not to apply air density corrections to fractions of the power, which are not dependent on air density, such as gearbox and generator losses. Figure 4.11 Output power profile of the pitch controlled wind turbine generator

18 79 Table 4.1 The results of the method of bins analysis Bin no 1 Bin interval Number of datasets m/s n 1 speed V 2 Wind (m/s) Bin averages Net speed P 1 (kw) 6.22 Rotor speed (rpm) Nominally fixed speed If, for instance, the relation between net power P and rotor power P R is of the form P = αp R β (4.9) with α and β constant, then the correction of power for air density variations should be ( P ) P = ρ α β (4.10) s T R

19 80 This will be the case with gridconnected, constant speed WTG, where the electric losses normally will be proportional to the power produced, while the mechanical losses in the drive train will be rotor speed dependent and hence constant for a constant speed WTG. 4.5 RESULTS OF THE SIMULATION The power output of the pitchcontrolled wind turbine generator to the wind velocity variation of Figure 4.12, is shown in Figure The desired output power is 12 KW. The output power variation is large (12 %) and the settling time of the system is about 15 sec. The pitch angle reference generated by the pitch controller and the actual pitch angle of the blades (output of the actuator) are plotted in Figure The response of the pitch actuator is slow (the time constant is 1 second) and this limits the performance of the system. The plot of the rotor speed is shown in Figure The rotor speed is always above 1.0 p.u, indicating that the generator is operating at supersynchronous speeds and has pitch regulated rotor blades for high performance and for highly effective start and break. Figure 4.12 Wind speed profile used for simulation

20 81 Figure 4.13 Output power profile of the pitch controlled wind turbine generator Figure 4.14 Plots of the pitch angle reference and the output of the pitch actuator Figure 4.15 Rotor speed of the induction generator

21 THE MERITS OF PITCH REGULATED WIND ENERGY GENERATORS The two basic types of wind energy generators that have been installed in India are Pitchregulated and stallregulated. Pitchregulated Wind Energy Generator (WEG) is manifold. In Stallregulated WEG s as the blades are firmly connected to the rotor the flow over the profile of the blade is lost and less torque is produced on the rotor shaft thereby limiting the power output. From the Figure 4.16 can be seen that a Pitch Regulated WEG is preferable to a StallRegulated WEG. For a PitchRegulated WEG the generation even after reaching maximum remains constant when speed increases whereas in the StallRegulated WEG it starts to drop down. Figure 4.16 Comparison of Power Curves of PitchRegulated and Stall Regulated WEG

22 83 The Pitch Regulated WEG has the following additional merits. Stopping of Rotor: The Pitch system makes it possible to rotate the blades to a position, which stops and starts the rotor at any wind speed. Active Control: The Pitch system is useful for active regulation by which nominal power can always be reached. However, in stall regulated, power curve can be changed only by effecting a permanent change in blade angle. Speed: The Speed can be controlled prior to connecting the WEG to the grid. That is, the pitch control system facilitates rotation at proper speed for a longer period of time before connecting to the grid. Lower wind speeds: Even at lesser winds in PitchRegulated WEG s maximum torque is obtained and the rotor starts whereas in Stall Regulated WEG s higher wind speed is required, as the design is such. This in turn may lead to decrease in the lifetime of the electrical components. Power: The power can be limited at locations with a weak grid. The WEG can easily be regulated by the Pitch control system and there is no need to change the WEG design to produce less nominal power. Pitch Control System: The pitch control system is not significantly affected by the change in air density and change in temperature, about placement of the WEG s at certain heights above sea level. Maximum power production will be always be attained in a pitch controlled WEG while in a stall

23 84 controlled WEG it is poor when there is change in temperature or and when the blades are dirty. Calculation of Blade and Rotor Loads: In the case of Pitch Regulated WEG s the blades always experience laminar flow across the profile and turbulent, chaotic flow has no influence. This means that WEG loads in a Pitch Controlled WEG are determined in a more correct manner. Greater Flexibility: The PitchRegulated WEG can readily be transformed to other turbine by changing the Pitch control strategy and programming a different blade angle. Better test facilities: As the blade angle can be turned to, a welldefined position is easier; it is simulation of specific production or testing condition. Power optimization below nominal Power: The power is optimized by rotating the blade to an angle, which results in a better and optimized power production. This is done by the pitch control system, which could compare the power measured as a certain wind speed with a master curve and change the pitch angle accordingly. Blade geometry: It is possible to optimize blade geometry without worrying about power regulation and the same can only be done by a Pitch system and not by Stall. Greater Capacity Factor: It is easy to get a greater capacity factor for a Pitch Regulated WEG, since only a new pitch control strategy (Software) and a smaller generator at same speed is needed.

24 85 Reduced loading on the tower and foundation: The bending load on the tower and foundation from the rotor is relatively small as horizontal forces becomes smaller due to reduced drags across the blade profile at high wind speeds. Less Noise at large wind speed: At large wind speeds, the blade angle are greater and this in turn results in less noise from the blade tip. Blade as one Structural Element: Each blade is one structural element without hydraulic system and/or mechanical systems to rotate the blade tip as an aerodynamic brake. For PitchRegulated WEG s, the total blade is an aerodynamic brake. As such, even at low wind velocity, the Pitch regulated blade has a large surface area against the wind and as the windvelocity increases this surface area decreases gradually resulting in optimum production at all times. 4.7 SCOPE FOR THE FUTURE For adapting the usage of the pitch angle mechanism of the stepper motor as shown in Figure 4.17 in place of the hydraulic actuator, following mechanism is involved given below: 1. Stepper motor 10 Kg torque, 60 RPM, 24 V and 2 amps, 10 degree. 2. Single axis travel mechanism. 3. Piston inside the shaft. 4. Converter for linear movement to circular movement. Based on the RPM sense, the stepper motor will rotate clockwise or

25 86 anticlockwise of travel mechanism. The travel mechanism will move the piston to front and back (that is some linear movement). The linear movement of the piston gets converted to circular movement of the blade axis by reciprocating links and it is an easy mechanism. This converts linear moment to circular moment, there by making the blade rotate circularly. Figure 4.17 Pitch angle mechanism of the stepper motor The cost is very low comparing to hydraulic actuator mechanism. This mechanism can be used up to 30 KW wind turbine. 4.8 SUMMARY The pitch control system is easy to implement, but the response is slow and output power variation is large. The aerodynamic power control for the variable speed wind turbine can limit the power into the turbine by reducing the angle of the blades. The pitch control has a large response time.

26 87 Therefore, fast torque changes caused by wind gusts will affect the generator. The speed range of the generator is depending on the converter voltage and the load situation. A higher voltage available for the converter in the rotor circuit, will give a higher range of speed. A speed was chosen to make the generator choose the mechanical speed, which gives the best power output related to the wind speed. A rate limiter was chosen to avoid fast changes in the mechanical speed. The scaling of the converter could be investigated and compared with the winding ratio of the machine, to find the optimal winding ratio compared with bus voltage in the converter. The response time is mainly dictated by the hydraulic actuator, which limits the performance of the system. The controller should be used in conjunction with the pitch control system to avoid large losses during steady state operation. The pitch angle mechanism of the steeper motor is also discussed.