CONTROL AND PERFORMANCE IDENTIFICATION FOR SMALL VERTICAL AXIS WIND TURBINES

Size: px

Start display at page:

Download "CONTROL AND PERFORMANCE IDENTIFICATION FOR SMALL VERTICAL AXIS WIND TURBINES"

Egbert Davis
5 years ago
Views:

1 CONTROL AND PERFORMANCE IDENTIFICATION FOR SMALL VERTICAL AXIS WIND TURBINES W.-L. Chen, Z.-C. Li, Y.-S. Lin, and B.-X. Huang Department of Electrical Engineering, Green Technology Research Center Chang Gung University, 59 Wen-Hwa 1st Rd. Kwei-Shan, Tao-Yuan 0, Taiwan Keywords: inverter, vertical axis wind turbine (VAWT), wind energy conversion system (WECS). Abstract This paper presents the controller design and a simple method to identify the aerodynamic performance for an installed vertical axis wind turbine (VAWT). The hardware realization of the wind energy conversion system consists of a 1.5kW VAWT-driven permanent magnet synchronous generator (PMSG), a PWM rectifier and a grid-tied PWM inverter. The PWM inverter is mandatory to maintain the dc capacitor voltage mainly adapted for operating in parallel with the power grid. The PWM rectifier, serving as a speed governor, is employed to modulate the revolution speed of the VAWT according to the desired power-speed curve so as to extract maximum power from the wind. To guarantee the aerodynamic performance of the VAWT, a simple method, based on the long-term recorded energy efficiency, can determine the optimum power-speed curve for the VAWT. The effectiveness of the proposed strategy is assessed experimentally by examining the closed-loop system response to various wind speeds. 1 Introduction The horizontal axis wind turbine (HAWT) designed for the production of electricity has to be oriented into the wind, which not only declines the power efficiency but also limits the installation to the open topography such as sea coast and plain [1, ]. The VAWT can be driven by winds from any direction without the need to reposition the rotor appropriate for the urban area application such as building-integrated WECS. In VAWTs, the efficiency of the Darrieus wind turbine (WT) is nearly as good as that of the HAWT but it is difficult to start up by itself when the wind speed increases. A smallscale VAWT unifies the Darrieus and Savonius rotors was patented in Taiwan by Hi-VAWT Technology Corporation as illustrated in Figure 1. The compound structure overcomes the self-starting problem by the Savonius rotor and takes the advantage of the Darrieus rotor to extract the wind power when the turbine is running. However, since the inertia moment of the small-scale WT is small, the maximum power point tracking (MPPT) control becomes difficult when the turbine is operated at random wind speed. The fuzzy-logicbased MPPT algorithm [] can search the MPP without knowing the WT characteristics but presents high complexity in tuning the control parameters. Figure 1: Photo of a 1.5kW VAWT with PMSG. (Source: Hi-VAWT Technology Corporation, Taiwan). The steepest ascent method can estimate the direction of control command according to the ratio of the deviations in electrical power and control command but the control is rather sensitive to the random variation in wind speed. The tip speed ratio (TSR) control [5] is reliable except the shortcomings of the need for WT characteristic and the expensive anemometer. The optimal torque (OT) control [6, 7] is suitable to regulate the revolution speed of the small scale WT such that the trajectory of the revolution speed can track the given optimal curve closely. In this paper, the OT control is adopted. To further identify the efficiency of the WT, an energy-based method was proposed to refine the power curve which is originally distorted by the variation in air density and the deformation of blades after long-running operation. Modelling and control of VAWTs.1 Aerodynamic model The mechanical power output of a wind turbine can be written as 1 Pm ACPVw. (1) The wind speed V w in (1) can be replaced by the tip speed ratio and revolution speed r as 1 r Pm AC R P. () Note that the aerodynamic performance C p is a strong function of and therefore not constant at all wind speeds, unless the rotor has variable speed to maintain constant. Accordingly, the criterion for needing constant optimum tipspeed ratio = opt can be interpreted as the need to maintain C p =C p,max, and then the optimum mechanical power extracted

2 from the wind can be deduced as the function of the revolution speed as P m, opt b r. () Figure shows a typical VAWT power curves under various wind speeds. For higher b (b 1 ), the power converter draws electrical power excessively and slows down the revolution speed of the VAWT, which decreases the influence of airstream on the blades and therefore reduces the WT power efficiency. For controller with smaller b (b ) would speed up the revolution speed of the VAWT, causing the energy in the airstream to be dissipated in turbulent motion and vortex shedding. For the VAWT operating at the optimum b (b ), the blades are strongly affected by the airstream and are capable of extracting more power from the wind. As illustrated in Figure 4, the electrical power measured from the stator circuit of the PMSG is first converted to speed command r * based on Equation (). To meet the desired speed, the kinetic energy stored in the rotor of the VAWTdriven PMSG can be altered by conditioning the generator output power. In this paper, the electrical power control for the PMSG is realized by a PI-based current controller with the current magnitude β as control output. It is noted that the phase of the current should be in phase with the terminal voltage of the PMSG so as to improve the power factor for the PWM rectifier. ( g g + g g ) v i v i Ki K p s v -1 g tan v g d dt poles Figure 4: PMSG electrical power control block diagram.. Estimating the optimum b by energy efficiency Figure : VAWT power curves at various wind speeds (V w ).. Controller design of power converters To attain optimum aerodynamic performance for the WT, the revolution speed of the WT should be confined along the optimum power-speed curve. The revolution speed of WT can be conditioned by controlling the electrical power of the PMSG. Figure shows the proposed control scheme for the VAWT driven PMSG. The electrical power generated by the PMSG is first converted by a PWM rectifier according to the MPPT algorithm. The PWM rectifier has better power quality in the aspect of power factor and current harmonics as compared with the bridge rectifier. The ac current shape of the PWM rectifier can be refined by adopting high frequency switching and phase-locked loop techniques. The boost converter is employed to boost the DC voltage level so as to adapt the PWM inverter to the voltage level of the power grid. Figure : Block diagram of the proposed wind energy conversion system. The WT can attain optimum aerodynamic performance at various wind speeds when the power output and revolution speed satisfy Equation (). However, even a strict windtunnel test is performed, the optimum b varies with the air density and the deformation of blades after long-running operation. In this paper, a simple method to determine the optimum b based on the energy efficiency rather than expensive wind-tunnel test is proposed. The energy efficiency (EE) can be represented as the ratio of the accumulated practical wind power in free space and the output power of the WT under specific b within a continuous time interval: EE k n n 1 b r. (4) 1 AVw k 1 Because electrical power generated by the PMSG is always smaller than wind power, the EE would tend to stable after a short period of time. As depicted in Figure 5, the setting time of the EE is close to 1000 sec. which can be employed as empirical parameter when determining EE for various b factors. It is also observed from Figure 5 that the optimum b factor (b=0.0468) can be determined as compared with higher b (b=0.070) or lower b (b=0.04). This conclusion can also be verified by Figure 6 that the performance of the VAWT with higher and lower b would be inferior to that of the b factor at To further ensure that the optimum b determined by EE is available for random wind speed conditions, a procedure used to measure the distribution of the EE for various b factors within a continuous time interval was proposed as illustrated in Figure 7. The wind power and electrical power output of the VAWT are measured in each repetition loop until the time t reach 1000 sec.

3 EE b= b=0.04 b=0.070 The distributions of steady EE to wind speed corresponding to various b factors are shown in Fig. 8. To further confirm the aerodynamic performance for the WT, the moving average technique is employed to phase out the poor b (with lower EE) as illustrated in Fig. 9. In this work, the optimum b is which was selected for Equation () to carry out the MPPT for the WT. The major benefit of Equation (4) is that the aerodynamic performance of the WT under various b factors or MPPT algorithms can be compared without at the same atmospheric condition Time (sec) Figure 5: Time responses of energy efficiency (EE) under various b factors. Figure 8: Measured EE-V w characteristics under various b factors. Figure 6: Trajectories of the VAWT power curve under various b factors. Start Initialize b and t b r Accumulate 1 and AV w t > 1000sec.? yes Calculate EE Average V w no b=b+0.01 Stop? End no yes Figure 7: Procedure to measure the distribution of the optimum b under random wind speed condition. Figure 9: Performance comparisons among various b factors. Experimental setup Figure 10 shows an experimental setup of a WECS. The electrical power generated by a VAWT-driven PMSG is converted to the power grid through a fully digital-controlled laboratory prototype which consists of a three-phase PWM rectifier, a push-pull converter, and a grid-tied three-phase PWM inverter. The three-phase PWM rectifier is used to regulate the WT speed such that the wind can influence the moving blade in an efficient way without causing turbulent vortex. The push-pull converter boosts the voltage generated by the PWM rectifier so as to adapt the PWM inverter to the voltage level of the grid utility. The grid-tied PWM inverter is mandatory to maintain the dc-link voltage in order to deliver the wind power to the grid utility. The fully-digital controller based on the TMS0F85 DSP hosts the sampling, A/D conversion and the control purposes. It is noted that the grid voltages and currents, generator output voltages and currents and DC-link voltages, and wind speed were simultaneously sampled using a multi-channel data acquisition system in order to validate the performance of the control strategies.

Figure 10: Experimental setup for a VAWT-driven PMSG system.

order to cope with the random variation in wind speed.

It is clear that the experimental result ( r -P e ) coincides with the power curve ( r * -P e * ) based on the optimum b factor.

4 Figure 10: Experimental setup for a VAWT-driven PMSG system. 4 Experimental results The experimental result in Figure 11 shows that the electrical power output of the VAWT-driven PMSG can be effectively controlled in order to cope with the random variation in wind speed. Figure 1 shows the phase portrait of the power curve derived from the time responses of r and P e in Figure 11. It is clear that the experimental result ( r -P e ) coincides with the power curve ( r * -P e * ) based on the optimum b factor. Figure 1 shows that the proposed power converter is not only capable of paralleling with the power grid and the PMSG but maintaining high power factor for the WECS. Figure 1: Trajectories of the WT revolution speed control under optimum b factor (b=0.0468). Figure 11: Dynamic responses for the VAWT operated under random wind speed conditions. Figure 1: Steady-state responses for the VAWT operated in parallel with the power grid.

5 5 Conclusion The performance of a WT would decline due to incorrect b factor. Strictly speaking, the optimum b is not only the function of air density but is also affected by the deformation of blades after long-running operation. In other words, the optimum b must be refined regularly in order to achieve the optimum aerodynamic performance for the wind turbine. This paper focuses on the development of control algorithms to identify the performance for an installed small VAWT. To demonstrate the feasibility of the proposed method under random wind speed, a 1.5kW VAWT installed in the spacious environment was first performed. In addition, the power converters were adopted to condition the power output of the VAWT so as to achieve the control requirements. The experimental results show that the proposed strategy can effectively differentiate the aerodynamic performance between b factors under random wind speed conditions. Acknowledgements This work was supported in part by the National Science Council of the Republic of China under Contracts NSC E and NSC 99-0-E and in part by the Green Technology Research Center (GTRC) of Chang Gung University. Appendix System Parameters VAWT driven PMSG system (1.5kW, 4 poles) Cut-in wind speed.5m/s Rated wind speed 1m/s WT height WT radius.m 1.4m PWM rectifier Ac filter (inductor) L f 4.5mH AC filter (capacitor) Dc-link capacitance C f L C dc 10F 4700F 15kHz PWM inverter Ac filter (inductor) L sf 1.5mH AC filter (capacitor) C sf 10F Dc-link capacitance H C dc 50F 15kHz Push-pull converter Input capacitor C i 000F Output filter (inductor) Output filter (capacitor) L o C o1, C o 4.1mH 0F 1.5kHz References [1] M. Chinchilla, S. Arnaltes, and J. C. Burgos, Control of permanent-magnet generators applied to variable-speed wind-energy systems connected to the grid, IEEE Trans. on Energy Conversion, vol. 1, no. 1, pp , (006). [] N. Stannard and J. R. Bumby, Performance aspects of mains connected small-scale wind turbines, IET Gener. Transm. Distrib., vol. 1, no., pp , (007). [] Q. Wang and L. Chang, An intelligent maximum power extraction algorithm for inverter-based variable speed wind turbine systems, IEEE Trans. on Power Electronics, vol. 19, no. 5, pp , (006). [4] E. Koutroulis and K. Kalaitzakis, Design of a maximum power tracking system for wind-energy-conversion applications, IEEE Trans. on Industrial Electronics, vol. 5, no., pp , (006). [5] G. Poddar, A. Joseph, and A. K. Unnikrishnan, Sensorless variable-speed controller for existing fixedspeed wind power generator with unity-power-factor operation, IEEE Trans. on Industrial Electronics, vol. 50, no., pp , (006). [6] S. Morimoto, H. Nakayama, M. Sanada, and Y. Takeda, Sensorless output maximization control for variablespeed wind generation system using IPMSG, IEEE Trans. on Industrial Applications, vol. 41, no. 1, pp , (005). [7] C. T. Pan and Y. L. Juan, A novel sensorless MPPT controller for a high-efficiency microscale wind power generation system, IEEE Trans. on Energy Conversion, vol. 5, no. 1, pp , (010).

ADVANCED CONTROL TECHNIQUES IN VARIABLE SPEED STAND ALONE WIND TURBINE SYSTEM

ADVANCED CONTROL TECHNIQUES IN VARIABLE SPEED STAND ALONE WIND TURBINE SYSTEM V. Sharmila Deve and S. Karthiga Department of Electrical and Electronics Engineering Kumaraguru College of Technology, Coimbatore,