Improving the transient performance of DFIG wind turbine using pitch angle controller low pass filter timing and network side connected damper circuitry
2020; Institution of Engineering and Technology; Volume: 14; Issue: 7 Linguagem: Inglês
10.1049/iet-rpg.2019.1124
ISSN1752-1424
Autores Tópico(s)Multilevel Inverters and Converters
ResumoIET Renewable Power GenerationVolume 14, Issue 7 p. 1219-1227 Research ArticleFree Access Improving the transient performance of DFIG wind turbine using pitch angle controller low pass filter timing and network side connected damper circuitry Kenneth E. Okedu, Corresponding Author Kenneth E. Okedu kenokedu@yahoo.com Department of Electrical and Electronic Engineering, Kitami Institute of Technology, Hokkaido, JapanSearch for more papers by this author Kenneth E. Okedu, Corresponding Author Kenneth E. Okedu kenokedu@yahoo.com Department of Electrical and Electronic Engineering, Kitami Institute of Technology, Hokkaido, JapanSearch for more papers by this author First published: 16 April 2020 https://doi.org/10.1049/iet-rpg.2019.1124Citations: 6AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract In wind turbine systems, pitch angle controllers are employed to reduce the aerodynamic power gained during high wind speeds. On the other hand, passive filters are usually used to mitigate disturbances in grid-connected voltage source converters (VSCs). To avoid the risk of instability in the grid-connected VSC, as a result of resonance in the capacitive and inductive components, it is necessary to consider damping. In this study, the low pass filter timing constant of the pitch angle controller of the doubly-fed induction generator (DFIG) wind turbine was varied considering different time constants. The best timing response of the low pass filter was further used to analyse different filter topologies, in addition to a new control strategy that uses two trap passive filters with shunt resistor–capacitor as an active damper, in augmenting the DFIG wind turbine during grid fault. Simulation studies in PSCAD/EMTDC were carried out to compare the performance of the proposed scheme with some other conventional filter solutions for the DFIG wind generator, during a severe bolted three-phase to ground fault. The simulation results demonstrate the improved performance and faster recovery of the wind generator variables after the fault considering the proposed filter scheme. Nomenclature List of symbols shaft torque electromagnetic torque mechanical torque rotor inertia turbine inertia stator voltage rotor voltage stator current rotor current stator inductance magnetising inductance rotor inductance damping inductor rotor leakage inductance sum of inductance stator resistance rotor resistance damping resistor initial resistance stator flux linkage rotor flux linkage stator angular frequency rotor angular frequency rotor leakage factor stationary frames doubly-fed induction generator (DFIG) rotor and stator quantities G DFIG grid-side converter circuit quantity m component total power reference power stator power grid power rotor speed turbine speed synchronous speed V voltage I current Ѱ flux vector quantity L inductance R resistance C capacitance damping capacitor slip speed dc capacitor sum of capacitance dc-link voltage dc power q-axis rotor current d-axis rotor current dc capacitor current shaft twist angle density of air R radius of the blade wind speed nominal power power coefficient pitch angle reference pitch angle ratio of blade tip speed to actual wind speed optimum value of time constant proportional gain integrator gain error of the pitch angle mechanical power S slip characteristics of filter at lower frequency characteristics frequency characteristics of filter higher frequency Q filter overall quality factor filter input impedance filter reverse trans-admittance filter attenuation or forward trans-admittance filter output admittance 1 Introduction One of the major challenges of the doubly-fed induction generator (DFIG) is its ability to ride through fault during periods of grid disturbances. The German electric transmission operators [1] were the first organisation to proposed fault ride-through (FRT) of wind turbines connected to the grid. Many countries thereafter, specified their own FRT with different specifications on the lowest voltage level and reactive power provision during grid disturbances [2, 3]. The DFIG variable speed wind turbine (VSWT) has the ability to maintain constant output voltage during the dynamics of the grid, while at the same time providing active and reactive power control. Hence, they are expected to remain grid-connected during transient conditions [4, 5]. Several methods have been proposed in the literature on the enhancement of the DFIG FRT capability ranging from the traditional crowbar [6], DC-chopper [7, 8], series dynamic braking resistor at the stator [9] and rotor circuit [10] of the DFIG wind generator. The use of the crowbar involves the blocking of the rotor side converter, hence leads to the consumption of more reactive power with energy dissipation in the rotor circuitry due to the addition of more resistance. The DC-chopper presents a simple and good control of the DC-link voltage of the DFIG, however, it is quite slow in the recovery of the DFIG variables such as the crowbar system after grid disturbances. The application of the series resistor in the DFIG system leads to the dissipation of heat and thus increases the losses of the generator system, even though an improved performance may be achieved. The use of the dynamic voltage restorer [11, 12], fault current limiters [7], as means of protecting the DFIG and dissipation of unbalanced power, is quite expensive, even though they can effectively reduce the electromagnetic transients and improve the DC-link voltage during fault conditions [13, 14]. In [15], a parallel capacitor was proposed as an additional circuitry to the conventional DFIG system to protect the DC-link voltage during transient. However, the charging and discharging of the capacitor system is quite a challenge. The use of an additional parallel grid side rectifier and series grid side converter (GSC) was reported in [16], while in [17], the conventional DFIG converters were interleaved for both the rotor and the GSCs. Though these schemes enhanced the DFIG wind turbine, however, they increase the cost of the system, with more circuit design intricacies by altering the traditional DFIG structure. Another category of DFIG FRT enhancement scheme is the utilisation of external expensive reactive power compensation systems such as the static synchronous compensator [18], the super-capacitor based energy storage [19], and the superconducting magnetic energy storage system [20]. The application of these solutions helps in the transient performance of the DFIG wind generator by the provision of additional reactive power to the entire system. However, these external devices increase the installation cost of the system. One of the traditional VSWT control strategies is to employ the rotor speed as a function of the wind speed. In the literature, there are several works on blade pitch angle control mechanisms. The control strategy of using speed variations of the rotor of the wind generator and its derivatives added to the wind speed factor was reported in [21]. In [22], the permanent magnetic synchronous generator was equipped with a blade pitch controller with the main aim of replacing the rotor speed in the wind generator controller input. The use of fuzzy input in the mechanical power variation and scenarios of sudden changes in wind speed fluctuations as compared to the use of traditional proportional–integral controllers that was implemented in the work. Furthermore, the fuzzy controller was applied with a dead zone to reduce fluctuations in the voltage and frequency variables in Island mode in [23]. However, the main drawback of the use of these fuzzy controllers is the difficulty in regulating the fuzzy system. This shortcoming was taken care of in [24], where a neuro-fuzzy controller was applied in controlling the pitch angle of a wind generator. In [25], a fuzzy-neural control strategy was used to adjust and control the input between the chord line of the pitch blades and the wind direction. This control strategy has the main benefit of making the fuzzy-neural adaptive system to acquire new learning techniques and adjust themselves to the newly developed data. By doing this, the performance of the rotor speed and power output of the wind generator would be more reliable. To reduce cost and size, it is usual to consider high-order passive filters to eliminate high-frequency harmonics in grid-connected voltage source converters (VSCs). The harmonic content at the point of common coupling (PCC) in a grid-connected VSC is usually affected by the connection standards [26-28]. Filter design basically depends on the strategy of the filter to be adopted and requires the multi-objective optimisation of the conversion system in relation to cost, size, and efficiency [29, 30]. On the general assumption that the impedance of the grid is basically inductive, the inductor–capacitor filter can be referred to as an inductor–capacitor–inductor (LCL) filter. The LCL filter [31] is generally an interfaced (T equivalent circuit) high-order filter model that is capable of describing the stability of VSCs in a grid-connected system [32, 33]. Owing to the resonance peaks attributed to high-order passive filters, damping element is usually used to achieve stability of the grid system. As a result, at the PCC, the high risk of the amplification of harmonics could be avoided by the implementation of the shunt passive damped filters [34] or some active damping measures [35]. Considering the cost and optimisation of the filters, based on low inductive and high capacitive passive components, damping of the system is more challenging [36-38]. The use of high capacitance in the filter circuitry leads to high ripples at the grid-connected VSCs. Again, the employment of passive damping strategies increases the overall filter reactive power ratings and damping losses [39]. In such a scenario, for high harmonic current in the shunt capacitor, there is limitation on how important the passive damping losses could be because their implementation is not practical [40]. It is, therefore, necessary to carry out their evaluation to minimise damping losses. Several passive filter strategies used to dampen the resonance of filters have been reported in the literature [41-43]. However, their detail implementation considering the grid-connected DFIG wind turbine under transient condition has not been widely reported. In [44], the mitigation of harmonics of a DFIG grid-connected system was presented using a single-tuned notched LCL filter. An active filter was proposed to reduce the DFIG stator and rotor current harmonics in [45], while in [46], the DFIG system was integrated with an active filter at the GSC for harmonics control and slip power transfer. This study presents the enhancement of the DFIG wind turbine transient performance considering the dual control strategies of the pitch angle controller and a newly proposed topology using a two-trap filter with shunt resistor–capacitor (RC) damper at the network side of the wind generator. Firstly, the low pass filter time constant was varied to see the effects it has on the wind generator variables. When the time constant of the low pass filter of the pitch controller was too small or too high, more oscillations and delay in recovery and settling time of the wind generator variables were observed. However, a moderate selected time constant for the low pass filter gave an improved performance of the wind generator. The obtained optimal time constant performance of the pitch angle low pass filter was further used in analysing the wind generator performance with several damping filter circuitries. Simulations were run in power system computer-aided design and electromagnetic transient including DC (PSCAD/EMTDC) environment [47], considering a severe three-line to ground fault. The results obtained were compared for better performance of the DFIG variables with other conventional filter solutions for DFIG FRT. The other considered filter schemes for the DFIG behaviour during transient are LCL filter with series resistor (case 1), LCL filter with shunt RC damper (case 2), LCL filter with shunt resistor–inductor–capacitor (RLC) damper (case 3), LCL filter with series RLC damper (case 4), trap filter with shunt RC damper (case 5), and the proposed DFIG two-trap filter with shunt RC damper (case 6). The proposed DFIG VSC two-trap filter topology was found to give better performance of the DFIG grid-connected system during transient condition considering the optimal low pass filter timing constant of the pitch angle controller. Thus, protecting the DFIG fragile power converters and also achieving the stipulated grid requirement for wind turbines FRT based on the set grid codes. 2 DFIG wind turbine and its modelling The schematic representation of power flow breakdown in a classical DFIG wind turbine configuration that uses bidirectional power converters is shown in Fig. 1a [48]. The wind turbine modelling and parameters could be obtained in [49]. One of the salient features of the DFIG is the economical utilisation of the power electronic VSCs only in the rotor circuitry as compared to the synchronous-based wind generators that operate using fully-rated converters. From Fig. 1b, optimum active power is obtained during normal operation based on the wind power conditions by the setting of the d-axis reference current in relation to the maximum power point tracking characteristics. The pitch angle control of the DFIG could also help in the active power regulation by providing power margins which allow frequency control and grid support [50]. The rotor d-axis current could be given preference in the control system to generate active power in the GSC, which indirectly controls the DC-link voltage by maintaining it at 1.0 pu. The rotor and GSCs can contribute to the reactive power dissipation by regulating the grid voltage as shown in Fig. 1c. Fig. 1Open in figure viewerPowerPoint DFIG wind turbine system (a) DFIG structure and system control, (b) DFIG rotor side converter control circuit, (c) DFIG GSC control circuit The two main sections of the wind turbine system are the mechanical power exploitation and its conversion into electrical power by the wind generator [51]. The two-mass model is normally employed in the transient stability analysis of the modelling and connection of both the mechanical and electrical sections of the wind turbine system [52], according to the following equations: (1) (2) (3) From (1) to (3), is the shaft torque, and are the electromagnetic and mechanical torques of the wind turbine, respectively. The rotor and turbine constant inertia is given by and , while the rotor and turbine angular frequencies are given by and , respectively. The shaft twist angle is . The expression for the mechanical torque of the wind turbine is given by (4) (5) (6) (7) From (4) to (7), is the air density in , R is the radius of the blade in m, is the wind speed in , and are the nominal power and power coefficient of the wind turbine, is the pitch angle, and is the ratio of the blade tip speed to the actual wind speed. 3 Wind generator pitch angle with low pass filter circuit From (7), there is a unique speed of rotation called maximum power point tracking (MPPT)-based on certain speed. Considering the derivative of (5), the power coefficient expressed as a function of is given as (8) From the above equations, it could be ascertain that for different pitch angles of , the blade tip speed ratio to the wind speed can be obtained. The power curve with power limitation in the pitch angle sector [53] is shown in Fig. 2a. Fig. 2b shows the general concept of the pitch angle controller, whereby, once the speed of the rotor exceeds its upper limit, there will be an increase in the angle of the blade to reduce the pressure of the aerodynamics. An actuator can be considered as an integrator in an enclosed system or a typical delayed first-order system with a time constant . The dynamic behaviour of the pitch actuator angle is expressed as (9) From (9), is the pitch angle and is the reference value obtained via different schemes. Usually, the response of the pitch angle relies on the time constant of the actuator with a range of 0.2–0.25 s [54]. Also, from Fig. 2b, it is necessary to include the rate limiter, to reflect the actual output of the response of the controller. Consequently, the following equations are used to determine the real power: (10) (11) From (10) and (11), and are the proportional and integrator gains. The typical pitch angle controller shown in Fig. 2b is modified in Fig. 2c with the low pass filter timing circuitry used in this study. The modified-pitch angle controller with low pass filter has the ability to influence the responses of the entire wind generator variables when the time constant is been varied because it would boast the response time of the actuator based on the stipulated timing range of the pitch angle controller. Fig. 2Open in figure viewerPowerPoint Wind generator pitch angle scheme (a) DFIG MPPT curve, (b) Traditional pitch angle controller for DFIG, (c) Modified pitch angle controller with low pass filter for DFIG 4 Dynamics of traditional filter solutions for DFIG wind turbine system and control A brief description of some of the passive filter solutions for the DFIG wind turbine is given as follows: A. LCL filter with series resistor (case 1): In this topology, a damping resistor is connected in series with the filter capacitor as shown in Fig. 3a. B. LCL filter with shunt RC damper (case 2): An improved damping topology of case 1 could be to connect a shunt RC to the filter as shown in Fig. 3b for the DFIG model system. One of the major benefits of this strategy is that the power loss in the resistor can be mitigated considering a proper choice of the split capacitor ratio. The details of this damping filter solution are given in [55]. C. LCL filter with shunt RLC damper (case 3): An improved frequency attenuation with considerable low damping loss can be achieved using the damped second-order filter (RLC) circuit connected in parallel with the filter capacitor as shown in Fig. 3c. In this topology, the fundamental current in the resistor circuitry is bypassed by the extra damping inductance . D. LCL filter with series RLC damper (case 4): The use of a selective resonant circuit as shown in Fig. 4a, with the filter containing an additional parallel RLC circuit in series with the reference filter capacitor, gives a similar benefit as in case 3. In Fig. 3d, during low and high frequencies, the damping inductor and capacitor bypasses the damping resistor of the system. In this topology, the maximum resonance attenuation that could be obtained is more limited when compared to other previous topologies. Also, this strategy gives room for placing the damping element in either series or parallel connection with the VSC. E. Trap filter with shunt RC damper (case 5): In this filter topology, the size of the filter is reduced as compared to the LCL filters, due to its close to zero impedance around the switching harmonics. However, the trap filter is purely inductive above the tuning frequency, thus, the frequency attenuation can be increased by an additional capacitor that is connected in parallel with the trap filter as shown in Fig. 4a. The major setback of this filter is the creation of multiple resonances that lead to the amplification of resonance between the grid-connected VSC and the filter. F. Proposed DFIG two-trap filter with shunt RC damper (case 6): The proposed DFIG two-trap filter with a shunt RC damper (case 6) is shown in Fig. 4b. Basically, the LCL filter or a trap filter is used to reduce the ripple content as a result of the switching operation of devices. The passive damping strategy used to constrain the resonance condition of the filters is based on traditional shunt passive damped filters relating to power systems [56, 57] or based on shunt or series damping passive strategies in conventional low power dc–dc converters [34, 58]. In light of the above, the mathematical dynamics of the filter topologies that could be employed in the grid-connected DFIG wind turbine system is described as follows. In the conventional LCL filter topology, a damping resistor is connected in series with the filter capacitor. The filter capacitor could also be connected in parallel with the damping resistor, but this will result in high ripple current in the resistor [59]. The filter attenuation admittance helps in the attenuation performance of the filter. For the LCL filter with damping resistor, it would be [30] (12) where is the characteristics of the filter at lower frequencies than the characteristic frequency , is the characteristics of the filter above , Q is the filter overall quality factor. is derived based on (13) where is the filter input admittance, is the filter reverse trans-admittance, is the filter attenuation or forward trans-admittance, and is the filter output admittance. The expressions for , , , and Q in (12) are [60] (14) A new topology of further reducing the size of the LCL filter is possible by employing two single-tuned filters. The DFIG model in Fig. 4b shows the shunt RC damper used for the two-trap filter topology to improve the performance of the wind turbine during transient based on the effective operation of the circuit. G. Parameter selection of the passive components: The selection of the parameters of the filter is based on [61], which ensures that the harmonics in the connected grid current are lower than the specified values. The performance of the filter is based on the admittance transfer function represented by the grid current to the converter voltage. Fig. 3Open in figure viewerPowerPoint DFIG model with LCL filter solutions (a) DFIG model with LCL and series R, (b) DFIG model with LCL shunt RC damper, (c) DFIG model with LCL shunt RLC damper, (d) DFIG model with LCL series RLC damper Fig. 4Open in figure viewerPowerPoint DFIG model with trap filter solutions (a) DFIG model with trap filter shunt RC damper, (b) DFIG model with proposed two-trap filter shunt dc damper Theoretically, a trap filter produces infinite attenuation at the switching frequency and frequencies above the switching frequency, the filter attenuation rapidly decreases. Based on the above features and analysis, the two-trap shunt RC damper will give a good damping performance of the variables of a grid-connected VSC system under network disturbance. The parameters of the DFIG wind generator and its excitation circuitry are given in Tables 1 and 2, respectively. The ratings and parameters for the various DFIG filter topologies including the proposed two-trap filter scheme used in this study are given in Table 3. Table 1. Parameters of the DFIG wind turbine DFIG Parameters Symbols Ratings rated power P 2 MVA rated voltage V 690 V stator resistance r1, pu 0.01 stator leakage reactance x1, pu 0.1 magnetising reactance Xmu, pu 3.5 rotor resistance r21, pu 0.035 rotor leakage reactance x21, pu 0.030 inertia constant H, s 1.5 Table 2. Parameters of the DFIG excitation circuit Parameters Ratings DC-link voltage 1.5 kV DC-link capacitor 50 mF device for power converter insulated gate bipolar transistor pulse width modulation carrier frequency 2 kHz upper circuit voltage limit 120% Table 3. Ratings and parameters of the various filter topologies Filter type Line inductor, mH Filter capacitor, μF Damp resistor, Ω Damp inductor, mH Damp capacitor, μF Trap inductors, mH case 1 1.5 4.7 5.5 — — — case 2 1.6 2.35 41.4 — 2.15 — case 3 1.5 2.35 20 3.3 2.15 — case 4 1.5 4.7 6.2 0.62 5.6 — case 5 0.7 2.15 20 — 2.35 0.113 case 6 0.1 2.15/0.21 5.4 — 2.35 0.113/0.3 5 Evaluation of the system performance 5.1 Effect of the pitch angle controller low pass filter timing The evaluation of the system performance for this study is given below. First and foremost, the effect of the low pass filter timing of the pitch angle controller was analysed considering different time constants. The time constants considered are 2.5, 5, 7.5, 10 and 12.5 s, respectively. The DFIG VSWT was evaluated using a severe three-phase to ground-bolted fault. The DFIG is operating at the rated power under 15 m/s wind speed in all cases considered. Simulations were carried out using PSCAD/EMTDC environment. A 100 ms fault is considered to occur at 5.1 s. The circuit breakers on the faulted line are opened and reclosed at 0.2 and 1.0 s, respectively. Some of the simulation results are presented in Fig. 5. Fig. 5a shows the rated speed of 15 m/s for the DFIG wind turbine used for the study. The rated speed of the wind generator was used for the study to effectively judge its performance under severe transient conditions. Fig. 5Open in figure viewerPowerPoint Effect of the pitch angle controller low pass filter timing constant (1) (a) DFIG wind turbine rated speed, (b) Response of the DFIG wind turbine pitch angle control, (c) Response of the DFIG wind turbine power coefficient Figs. 5b and c show the pitch angle controller and the power coefficient of the wind turbine responses, respectively. From both figures, a low pitch angle controller time constant of 2.5 s that is used for the traditional DFIG scheme gave a high peak with many oscillations and high delayed settling or recovering time. Also, it could be observed that a higher time constant of 12.5 s gave much lower oscillation, but the recovery time of the DFIG variables is still slow. However, a time constant of 5 s for the pitch angle controller, gave a fewer oscillation and the same time a faster recovery and settling time for the DFIG variables. From Fig. 6a, a low pass filter time constant for the pitch angle controller gave a similar response for the other DFIG variables of high transients and much oscillations like those in Fig. 5. Again, a higher time constant of 12.5 s shows similar behaviour as earlier discussed with slow recovery or settling time. For Fig. 6b, although the low time constant of 2.5 s did not result in high peak or transient response of the wind generator rotational speed, however, the recovery or settling time is still low as compared to low pass filter time constant of 5 s that has a considerable peak and faster settling time response of the DFIG variable. Fig. 6Open in figure viewerPowerPoint Effect of the pitch angle controller low pass filter timing constant (2) (a) DFIG wind turbine torque, (b) DFIG wind generator rotational speed Figs. 7a–c show the DC-link voltage, reactive power of the GSC and the terminal voltage of the wind generator during transient, where a moderate time constant of 5 s gave a better response for the DFIG wind generator variables. In Fig. 7c, although the low pass filter time response for 7.5 s improved the terminal voltage response relatively, however, the settling time is delayed compared to the response of 5 s time constant. Fig. 7Open in figure viewerPowerPoint Effect of the pitch angle controller low pass filter timing constant (3) (a) DFIG DC-link voltage, (b) DFIG GSC reactive power, (c) DFIG terminal voltage 5.2 Analyses of the various filter configurations for DFIG Based on the best pitch angle controller time constant response of 5 s on the DFIG variables during the transient, a further analysis was carried out to improve the performance of the wind generator during the transient condition. The DFIG VSWT was evaluated for the six cases considered, using the various filter topologies earlier discussed, with a severe three-phase to ground bolted fault. The DFIG was still operating at the rated power under 15 m/s wind speed in all cases considered. Simulations were also carried out using PSCAD/EMTDC. A 100 ms fault is considered to occur at 0.1 s. The circuit breakers on the faulted line are opened and reclosed at 0.2 and 1.0 s, respectively, as in the previous condition. Some of the DFIG variables simulation results are presented in Fig. 8. Fig. 8Open in figure viewerPowerPoint Responses of the various filter control strategies (ai) DFIG terminal voltage, (aii) Zoom of (ai), (b) DFIG rotational speed, (c) DFIG reactive power of GSC Figs. 8ai and aii show the responses of the DFIG terminal voltage during the grid fault. It is observed that the performance of the two-trap shunt RC damper gave a better performance due to the fact that the damping circuit is designed to keep the same resonance attenuation to improve its dynamic performance. The two-trap topology also satisfies the grid requirement, considering the stipulated recovery time of 0.9 pu for the wind turbine voltage after transient condition. In Fig. 8b, more oscillations of the wind turbin
Referência(s)