- Email: [email protected]
A new synchronous generator based wind energy conversion system feeding an isolated load through variable frequency transformer
Renewable Energy 86 (2016) 106e116
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
A new synchronous generator based wind energy conversion system feeding an isolated load through variable frequency transformer Farhad Ilahi Bakhsh*, Dheeraj Kumar Khatod Alternate Hydro Energy Centre, Indian Institute of Technology Roorkee, Roorkee 247667, India
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 February 2015 Received in revised form 20 June 2015 Accepted 30 July 2015 Available online xxx
This paper aims to explore the possibility of synchronous generator (SG) based wind energy generation system feeding an isolated load using a latest power transmission technology i.e. variable frequency transformer (VFT). The proposed configuration does not employ any power electronics based interface as in conventional SG based stand-alone wind energy conversion systems (SWECS). For analysis, the simulation models of proposed configuration as well as conventional configuration have been developed under MATLAB-Simulink environment. A series of studies on power fed from the SG to the different loads at various SG input speeds has been carried out with the developed models. Further to analyze the effectiveness of the proposed method; the efficiency, total harmonic distortion (THD) of output voltage and THD of output current of the proposed method have been compared with those of the conventional method. From obtained results, it is observed that the proposed method is simple and does not produce harmonics. Moreover to validate the proposed scheme, an experimental analysis has been carried out. Further, the cost analysis of both systems has also been carried out. From the cost analysis, it is observed that the proposed system is cheaper than the conventional system. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Isolated loads Synchronous generator (SG) Power fed Variable frequency transformer (VFT) Standalone wind energy conversion system (SWECS)
1. Introduction Due to concern over the environment, the advent of renewable energy (RE) based power generation (wind, hydro, etc.), has increased their utilization in different forms. A customer with limited RE based generation (e.g. wind turbine), needs a connection with distribution system of bulk power grid for reliable power supply. In remote locations, isolated RE based power generation systems can be utilized to serve the local loads, eliminating the grid connection, transmission losses and transmission costs [1]. Nowadays, the wind based power generation has been emerged as one among the most promising RE based generation system. Due to this small installations of stand-alone wind energy conversion systems (SWECS) are quite promising in remote area electrification programmes. To utilize the wind energy, fixed speed and variable speed SWECS are used. In fixed-speed SWECS, the wind turbines are mostly equipped with an induction generator because of its simplicity, ruggedness and less maintenance [2,3]. Here, the isolated induction generator is directly connected to the loads. In order
* Corresponding author. E-mail address: [email protected] (F.I. Bakhsh). http://dx.doi.org/10.1016/j.renene.2015.07.093 0960-1481/© 2015 Elsevier Ltd. All rights reserved.
to supply constant voltage and frequency to the load, the speed of the turbine is controlled by the gearbox. The use of a gearbox causes many problems such as it requires continuous maintenance, enlarges the weight and size of the SWECS, generates noise, increases the power losses and hence reduces the efficiency of SWECS [4]. Variable-speed wind turbines possess many advantages over fixedspeed, such as larger energy capture, operation at maximum power point, better efficiency and quality of power. Hence, variable-speed wind turbines are the dominating type of turbines as far as present trend is concerned and used in SWECS for feeding an isolated load. In this operating mode, the wind turbines are equipped with either synchronous or doubly fed induction generator. The application of doubly fed induction generator for feeding isolated load is discussed in Refs. [5,6], whereas the application of synchronous generator (wound field or permanent magnet) feeding the isolated load is discussed in Refs. [7,8]. Both of these methods require suitable power electronics conversion system for feeding power to the load. These power electronics converters are costly, require sophisticated control system, cause harmonic distortion and thereby deteriorate the power quality. Moreover, they require suitable compensation in order to meet the standards for harmonic pollution which further increases the cost and complexity of the system [9e11].
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
The main challenge faced in a synchronous generator (SG) based SWECS is the power quality. This challenge can be well answered by using a current power transmission technology termed as variable frequency transformer (VFT) [12,13]. The application of VFT for grid interconnection of wind power generation is discussed in Refs. [14e18]. But the application of VFT for SWECS has not been discussed so far. This paper deals with the analysis of a new configuration of SG based wind energy generation system for feeding power to an isolated load using VFT. The VFT is realized using a wound rotor induction machine (WRIM) whose rotor is mechanically coupled to the dc drive motor (DDM). The SG supplies power to the load at different levels of SG input speed. The requirements of costly power electronics converters are omitted. Hence, the proposed method is simple and does not produce harmonics. This paper has been organized in eight sections. The Section 2 of the paper describes the conventional SG based SWECS. Section 3 discusses the modeling and analysis of the proposed SG based SWECS. Sections 4 and 5 present the MATLAB-Simulink model; and results of the conventional method and the proposed method, respectively. Section 6 presents the experimental analysis of the proposed method. Section 7 shows the cost analysis of the proposed method and the conventional method. Finally Section 8 draws the conclusion of the paper.
107
Fig. 2. Proposed system for SG based SWECS.
feeding SG power to load. Here, the wind turbine is connected to the rotor of the SG without a gear box. The stator winding of the SG is connected to the rotor winding of the WRIM. The stator winding of the WRIM is connected to the isolated load as shown in Fig. 2. A DDM is coupled (mechanically) to the rotor of the WRIM in order to apply torque to the rotor of the WRIM.
2. Conventional method of SG based SWECS In a conventional method of SG based SWECS, the wind turbine is connected to the rotor of the SG with or without gear box. The output power of the SG is fed to the load through a power electronics based interface. The power electronics based interface mainly consists of an ac-to-dc rectifier followed by a dc-to-ac inverter as shown in Fig. 1. The SG output ac power is first rectified into a dc power using uncontrolled ac-to-dc rectifier. The capacitor across the output of ac-to-dc rectifier is connected in order to filter out the ripples in the dc power. Then this dc power is converted again into the ac power using self-commutated inverter, which are mainly pulse width modulated (PWM) inverter employing insulated gate bipolar transistors (IGBTs). In this type of inverter, in addition to the control of the active power, the reactive power is also controllable. This interface produces inter-harmonics (harmonic distortion) and thereby deteriorates the quality of power supplied. A filter is required at the inverter side in order to meet the standards for harmonic pollution which further increases the cost and complexity of the system. 3. Proposed method In the proposed method of SG based SWECS, VFT is used for
3.1. Modeling of proposed method In modeling, the VFT is represented as a doubly-fed WRIM whose rotor is mechanically coupled to the DDM [19,20]. Here, DDM supplies mechanical power ‘Pd’ to the rotor of the WRIM as shown in Fig. 3. The three phase windings are available on both rotor and stator sides of the WRIM. The isolated load is connected to the stator windings of the WRIM having voltage ‘VL’ and load angle ‘qL’. The rotor windings of the WRIM is connected to the stator windings of the SG, energized by voltage ‘Vs’ with phase angle ‘qs’. The rotor of the SG is mechanically coupled with the dc motor, where dc motor is working as prime mover of the SG in order to simulate the wind turbine. To transfer power from SG to load, the Pd is applied at the rotor of the WRIM. The power transfer through VFT is controlled by the magnitude of the applied Pd. In the power flow process, only real power transfer has being considered. 3.2. Analysis of proposed method Considering VFT as an ideal and lossless machine and neglecting its magnetizing current and leakage reactance, the power balance equation for it will be written as:
Fig. 1. Conventional method of SG based SWECS.
108
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
It is well known that frequency of rotor is slip times-frequency of stator. Thus, putting fs ¼ sfL in (6), we get
f rm ¼ sf L f L ¼ ðs 1Þf L
(7)
and
Nrm ¼ f rm 120 Np urm ¼
2p 2p Nrm ¼ f rm 120 Np 60 60
(8)
(9)
where,
Fig. 3. Model representation of the proposed system.
PL ¼ Ps þ Pd
(1)
s ¼ slip of the VFT, frm ¼ VFT rotor mechanical speed in Hz, Np ¼ number of poles in the VFT, Nrm ¼ VFT rotor mechanical speed in rpm, and urm ¼ VFT rotor mechanical speed in rad/s. Combining the above relationships gives the power exchanged with the drive system as
where, PL ¼ electrical power fed to the load, Ps ¼ electrical power available at output of the SG, and Pd ¼ mechanical power supplied by the DDM.
Pd ¼ PL Ps ¼ VL IL Vs Is VL fs IL N1 ¼ V L IL N2 N1 fL N2
Since the VFT behaves like a transformer, the ampere-turns must balance between stator and rotor windings:
N1 I L ¼ N2 I s
(2)
¼ VL IL ½ðsVL Þ IL ¼ VL IL ð1 sÞ or;
Pd ¼ PL ð1 sÞ
where,
(10) It shows that the electrical power flowing into the load being dependent on the mechanical power of the DDM and slip of WRIM.
N1 ¼ number of turns available on stator winding of VFT, N2 ¼ number of turns available on rotor winding of VFT, IL ¼ current fed to load, and Is ¼ current supplied by the SG.
a) If PL is constant, then
Since, both the rotor and stator windings of VFT link the same magnetic flux, therefore
VL ¼ N1 f L Ja ;
(3)
Vs ¼ N2 f s Ja ;
(4)
(5)
sff s and f s fNs
s ¼ kNs where,
where, Vs ¼ voltage magnitude of SG on rotor side of VFT, fs ¼ frequency of voltage applied across rotor winding of VFT (Hz), VL ¼ voltage magnitude of load on stator side of VFT and fL ¼ frequency of voltage available across stator winding of VFT (Hz), and Ja ¼ air-gap flux. The nature of the VFT is such that in steady state condition, the rotor mechanical speed in frequency (electrical) is equal to the difference in the frequency (electrical) on the stator and rotor windings,
f rm ¼ f s f L
Since load frequency is maintained constant, therefore
Hence,
and
Vs =N2 ¼ ðVL =N1 Þ ðf s =f L Þ
Pd fð1 sÞ and s ¼ f s =f L
(6)
k ¼ constant, Ns ¼ SG speed. Thus,
Pd fð1 kNs Þ
(11)
For constant load, the mechanical power of the DDM decreases with increase in the SG speed. b) If the slip of WRIM is maintained constant, then
Pd fPL
(12)
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
At a constant slip, the electrical power fed to the load is being only proportional to mechanical power of the DDM. 4. MATLAB-Simulink analysis of conventional method 4.1. MATLAB-Simulink model For simulation in MATLAB-Simulink, SG is simulated with the synchronous machine pu standard having rating 3 phase, 8.1 kVA, 1500 rpm, 400 V, 50 Hz and the load is simulated using three-phase series R load as shown in Fig. 4. The SG is connected to rectifier side of the power electronics based interface and the load is connected to inverter side of the power electronics based interface. Here, rectifier is simulated with the universal diode bridge and the inverter is simulated with universal IGBT bridge using PWM scheme. A constant block named as dc motor is used to apply rated speed at input of the SG. The voltage and frequency of different loads are kept constant i.e. 400 V (LeL) and 50 Hz. Then the simulated model (Fig. 4) is used to analyze the performance of the conventional method.
109
Table 1 Power Fed to Different Loads at 1500 rpm. S. No.
Ps (W)
PL (W)
% h ¼ PL =Ps 100
1 2 3 4
1095 2148 3246 4395
1000 2000 3000 4000
91.32 93.11 92.42 91.01
connected to the stator windings of WRIM and SG (3 phase, 8.1 kVA, 1500 rpm, 400 V, 50 Hz) is connected to rotor windings of WRIM. A block named as dc motor is used to apply various speed at the input of the SG. The DDM is simulated with separately excited dc motor. The output mechanical power of the DDM applied to the rotor of WRIM is of mechanical torque ‘Tm’ as shown in Fig. 5. The voltage and frequency of the load is kept constant and the input speed of the SG is varied. The three-phase capacitor branch is connected across the stator windings of the WRIM in order to supply reactive power required at stator side of the WRIM. Then the simulated model (Fig. 5) is used to analyze the performance of the proposed method.
4.2. MATLAB-Simulink results 5.2. MATLAB-Simulink results Table 1 shows the power fed to different loads by conventional SG based WECS at rated value of input SG speed. It is evident from this table that the efficiency of the conventional system increases with the increase in value of load (at 1 kW load the efficiency is 91.31%), reaches to maximum value (93.80%) at 2 kW load and then starts decreasing with further increase in load. 5. MATLAB-Simulink analysis of proposed method 5.1. MATLAB-Simulink model For simulation in MATLAB-Simulink, VFT is simulated with the asynchronous machine SI units which is basically a doubly-fed, 3 phase, 7.5 kW, 400 V, 50 Hz, 1440 rpm WRIM and the isolated load is simulated with three-phase R (resistive) load having rated voltage 400 V at rated frequency of 50 Hz. The three-phase R load is
5.2.1. For constant load In this case 1 kW, 2 kW, 3 kW and 4 kW loads are supplied at different values of SG input speeds. Since the load frequency is dependent upon slip of the VFT, the mechanical power from DDM is varied which controls the speed of WRIM and hence the slip of the VFT. Thus by controlling the slip, the load frequency is maintained constant at 50 Hz. 5.2.1.1. For 1 kW load. To realize the performance of proposed method, the SG is operated at different speeds for feeding a 1 kW load. The DDM power is varied by DDM controller which varies the torque applied at the rotor of the WRIM, thus controlling the slip of the machine in order to supply power to the load at 50 Hz. Various variables obtained at different SG input speed conditions are given in Table 2.
Fig. 4. MATLAB-Simulink model of the conventional SG based SWECS.
110
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Fig. 5. MATLAB-Simulink model for the proposed SG based SWECS.
Table 2 Power Fed to 1 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300
394.0 367.2 340.5 313.8 287.1 260.5 233.9 207.4 180.9 154.4 128.0 101.6 75.18
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00
1089 1013 937.1 861.8 786.7 711.8 637.2 562.8 488.6 414.6 340.8 267.2 194.1
0.025 72.01 143.44 214.23 284.29 353.48 421.41 487.98 552.50 614.02 671.25 720.75 756.12
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20
400.0 399.3 398.5 397.6 396.7 395.6 394.4 392.9 391.1 388.7 385.5 380.9 373.7
1000 996.4 992.5 988.2 983.5 978.2 972.1 964.8 955.8 944.4 929.1 907.0 872.9
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
91.83 91.83 91.85 91.84 91.83 91.83 91.83 91.82 91.81 91.81 91.8 91.81 91.86
Table 2 shows that when SG is operating at rated speed i.e. 1500 rpm, the voltage available across the stator terminals of the SG is 394 V which is near to rated voltage (400 V) and its frequency is 50 Hz (rated frequency). The output power of the SG is 1089 W. A little amount of DDM power (0.025 W) is required to keep the rotor of WRIM stationary (or to achieve unity slip) so that the load frequency becomes 50 Hz. At this value of slip, the load voltage and power both are having their rated values. On reducing the value of SG speed from 1500 rpm to 1400 rpm, the output voltage, frequency and output power of SG also reduces from 394 V to 367.2 V, 50 Hz to around 46.67 Hz and 1089 W to 1013 W, respectively. But the DDM mechanical power increases to 72.01 W for maintaining the slip around 0.93 so that the load frequency becomes 50 Hz. At this value of slip, the load voltage is reduced to 399.3 V and accordingly load power also reduces to 996.4 W. Since the load is constant, on further reducing the SG input speed, the SG (voltage, frequency and power) reduces but the mechanical power of DDM increases in order to supply the constant load. The torque applied by the DDM is such that the rotor of the WRIM rotates at a speed which is difference of the WRIM synchronous speed and the SG speed. At this speed of WRIM, the obtained slip value will be such that the voltage available at the stator terminals of VFT will have a
frequency of 50 Hz. But the magnitude of load voltage and power is reduced from the rated value with the decrease in SG speed. Moreover, the THD of load voltage and current remains constant i.e. 0.01% which is negligible. The variation of SG output power, DDM power, load power and % efficiency with the SG input speed of Table 2 are plotted in Fig. 6. When the SG input speed becomes 300 rpm the load voltage becomes 373.7 V which is not within permissible limit of voltage variation i.e. ±5%. Thus it is required to increase the load voltage in order to achieve power from the SG at low level of input speed. This can be done by various methods i.e. using auto-transformer, ac voltage regulator or by excitation control of SG. In this work, field excitation of the SG is controlled in order to achieve the rated load voltage. The various parameters achieved using excitation control is presented in Table 3. From Fig. 6, it is clear that with the decrease in SG input speed, the SG output power decreases but the DDM power increases. Though there is decrement in load power but the efficiency is almost constant for all operating conditions due to the constant load. Table 3 shows the variation of load profile with the SG input speed at different field voltage (Vf), where Vf is expressed in pu. In this case when SG is operating at rated speed the Vf is 1 pu, thus the
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
111
Fig. 6. The variation of SG power, DDM power, load power and % efficiency with the SG input speed at 1 kW load.
Fig. 7. The variation of SG power, DDM power, load power and % efficiency with excitation control for 1 kW load.
SG (output voltage, frequency and output power), DDM power, load (voltage, frequency and power) and THD of load voltage and current is same as in earlier case (without excitation control). On reducing the value of SG input speed from 1500 rpm to 1400 rpm, the SG output voltage reduces from 394 V to 368 V, frequency reduces from 50 Hz to 46.67 Hz and output power also reduces form 1089 to 1016 W. The reduction in SG output voltage and power is lesser than the earlier case because of 0.2% increase in excitation voltage. But the DDM mechanical power increases from 72.01 W to 72.3 W from previous case for maintaining the slip around 0.93 in order to achieve 50 Hz load frequency. Moreover, due to excitation control, the load voltage and power remains constant at their rated values. On further reducing the SG input speed, the value of Vf increases, the SG (voltage, frequency and power) reduces, the mechanical power of DDM increases but the load voltage and power is maintained constant. Moreover, the THD of load voltage and current remains constant as in earlier case. Thus, by controlling the excitation of the SG, the load voltage and power can be maintained constant (which generally reduces with the reduction in SG input speed) even at 100 rpm SG input speed. Hence, the SG power which is normally wasted due to low level of wind speed can be utilized effectively using the proposed method. Similarly the results are obtained for 2 kW, 3 kW and 4 kW loads. The variation of SG output power, DDM power, load power and % efficiency with the SG input speed of Table 3 are plotted in Fig. 7. Fig. 7 shows the variation of SG power, DDM power, load power and efficiency in % with excitation control. From Fig. 7, it is clear
with the increase in SG input speed, the SG output power increases but the DDM power decreases such that the load power is maintained constant at all values of SG input speed. 5.2.1.2. For 2 kW load. The load is increased from 1 kW to 2 kW and it is kept constant for all operating conditions. The power flow from SG to the load at different values of SG input speed without and with excitation control are obtained in Tables 4 and 5, respectively. 5.2.1.3. For 3 kW load. The SG is operated at different values of input speed and the load is kept constant i.e. 3 kW. Then the power fed to the load at different values of SG input speed without and with excitation control are presented in Tables 6 and 7, respectively. 5.2.1.4. For 4 kW load. The power fed from SG to 4 kW load at different levels of SG input speed without and with excitation control are evaluated in Tables 8 and 9, respectively. From Tables 2e9, it is clear that VFT can be used as an interface between SG based wind energy generation system and isolated loads, and power can be supplied to load at different levels of input SG speed without or with excitation control. With excitation control, the load power remains almost constant and thus the proposed method is able to effectively utilize the small power available at the output of the SG and hence avoids the wastage of power due to the low level of input wind speed. Further, the efficiency of the proposed SWECS method with or without excitation control for a
Table 3 Power Fed to 1 kW Load with Excitation Control. Vf in pu Ns (rpm) Vs (V)
fs (Hz) Ps (W)
1.0000 1.0020 1.0040 1.0060 1.0084 1.0112 1.0145 1.0182 1.0230 1.0290 1.0375 1.0500 1.0705 1.1110 1.2326
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
394.0 368.0 341.8 315.7 289.5 263.4 237.3 211.2 185.0 158.9 132.8 106.6 80.48 54.33 28.20
Pd (W)
1089 0.025 1017 72.3 944.7 144.6 872.2 216.8 800.0 289.07 727.7 361.28 655.6 433.6 583.5 505.94 511.3 578.22 439.0 650.31 366.8 722.57 294.6 795.4 222.4 866.83 150.1 938.65 77.88 1010.6
Nrm (rpm) S ¼ ½ð1500 Nrm Þ=1500
VL (V) PL (W) % THD of VL % THD of IL % h ¼ ½PL =ðPs þ Pd Þ 100
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00
91.83 91.80 91.80 91.83 91.82 91.83 91.81 91.79 91.78 91.80 91.80 91.74 91.81 91.85 91.87
112
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Table 4 Power Fed To 2 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800 700 600
396.2 368.9 341.8 314.6 287.6 260.5 233.6 206.6 179.7 152.9
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00
2136 1988 1836 1684 1533 1382 1232 1082 933.4 785.1
1.057 140.53 279.18 415.63 549.57 680.68 807.39 928.76 1043.8 1147.0
0 100 200 300 400 500 600 700 800 900
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40
400.0 399.1 397.8 396.4 394.8 392.9 390.6 387.9 384.6 380.2
2000 1991 1979 1964 1948 1929 1907 1881 1849 1807
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
93.59 93.54 93.56 93.54 93.54 93.52 93.51 93.55 93.51 93.53
Table 5 Power Fed to 2 kW Load with Excitation Control. Vf in pu
Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1.0000 1.0021 1.0054 1.0090 1.0132 1.0182 1.0239 1.0310 1.0400 1.0520 1.0685 1.0928 1.1332 1.2139 1.4563
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
396.2 369.7 343.6 317.5 291.4 265.3 239.1 213.0 186.9 160.8 134.7 108.5 82.4 56.27 30.14
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
2136 1996 1856 1714 1573 1433 1292 1151 1010 868.8 727.9 586.7 445.6 304.5 163.6
1.057 141.06 282.11 423.17 564.23 705.29 846.34 987.4 1128.5 1269.5 1410.6 1551.6 1692.7 1833.7 1974.8
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00
93.59 93.59 93.54 93.58 93.58 93.53 93.53 93.53 93.53 93.53 93.52 93.53 93.53 93.53 93.53
Table 6 Power Fed To 3 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800 700
396.5 369.1 341.6 314.2 286.8 259.4 232.1 204.9 177.7
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33
3224 2994 2759 2525 2291 2059 1827 1597 1368
3.905 210.17 416.57 618.26 814.72 1003.7 1184.4 1353.9 1508.8
0 100 200 300 400 500 600 700 800
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47
400.0 398.6 396.8 394.7 392.4 389.6 386.4 382.4 377.6
3000 2979 2952 2921 2886 2846 2799 2743 2673
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
92.94 92.97 92.96 92.93 92.93 92.92 92.95 92.95 92.92
particular load under all operating conditions are almost same. Moreover, the THD of output voltage and current of the proposed method is almost negligible.
5.2.2. For constant slip In this case, the SG input speed is kept constant but the load is varied from 1 kW to 4 kW. Since the SG input speed is constant, the
Table 7 Power Fed to 3 kW Load with Excitation Control. Vf in pu
Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1.0000 1.0036 1.0082 1.0134 1.0195 1.0266 1.0352 1.0458 1.0594 1.0773 1.1024 1.1619 1.2019 1.3266 1.7025
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
396.5 370.4 344.4 318.4 292.4 266.3 240.3 214.3 188.3 162.2 136.2 110.2 84.19 58.19 32.22
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
3224 3016 2804 2593 2381 2170 1958 1747 1535 1324 1113 900.9 689.3 477.9 266.4
3.905 211.64 423.28 634.92 846.55 1058.2 1269.8 1481.5 1693.1 1904.8 2116.4 2328.0 2539.7 2751.3 2962.9
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00
92.94 92.95 92.96 92.94 92.95 92.93 92.94 92.92 92.93 92.92 92.90 92.91 92.91 92.90 92.90
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
113
Table 8 Power Fed To 4 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800
394.7 367.0 339.3 311.7 284.2 256.7 229.2 201.8
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67
4392 4064 3737 3411 3087 2764 2443 2124
4.12 282.32 557.95 825.61 1083.2 1329.4 1559.5 1769.6
0 100 200 300 400 500 600 700
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53
400 397.9 395.5 392.8 389.7 386.1 381.8 376.5
4000 3958 3911 3858 3797 3727 3644 3545
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00
90.989 91.065 91.061 91.063 91.05 91.049 91.043 91.048
Table 9 Power Fed to 4 kW Load with Excitation Control. Vf in pu
Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1.0000 1.00538 1.01140 1.01832 1.02639 1.03599 1.04770 1.06230 1.08100 1.10580 1.14070 1.19280 1.28050 1.45725 1.99450
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
394.7 369.0 343.2 317.4 291.7 265.9 240.1 214.4 188.6 162.9 137.1 111.4 85.62 59.90 34.22
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
4392 4108 3823 3537 3252 2967 2682 2397 2112 1826 1541 1256 970.5 685.2 399.8
4.12 285.36 570.72 856.08 1141.4 1426.8 1712.2 1997.5 2282.9 2568.3 2853.6 3134.0 3423.1 3708.3 3993.6
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
90.989 91.046 91.039 91.052 91.045 91.037 91.03 91.022 91.015 91.028 91.021 91.013 91.042 91.043 91.046
mechanical speed of WRIM and thus the slip of the VFT for achieving 50 Hz load frequency is also maintained constant. The results obtained without and with excitation control are given in Tables 10 and 11 where SG input speed is kept constant at 1200 rpm and load is changing. Tables 10 and 11 show that when the SG input speed is kept fixed at 1200 rpm, the SG output power and the DDM mechanical power without and with excitation control increases with the increase in load. The increase in DDM power results in constant speed of WRIM rotor (300 rpm) i.e. slip becomes constant (0.8). Similarly the results can be obtained at other SG input speeds. Various plots obtained from the Tables 10 and 11 are shown in Figs. 8 and 9. Fig. 8 shows the variation of load power with the DDM power at 1200 rpm SG input speed. It is evident from Fig. 8 that the power fed to load with and without excitation control are directly proportional to the DDM power and the load power with excitation control is slightly greater than without excitation control. Fig. 9 shows the efficiency variation of proposed SWECS with the power fed to load at SG input speed of 1200 rpm. It is clear from Fig. 9 that the efficiency of proposed method with excitation control is slightly more than without excitation control and both increases with the increase in load from 91.8%, reaches to the maximum value and then starts decreasing with the further increase in load. The efficiency, THD of output voltage and THD of output current
of proposed system without excitation control have been compared with the conventional system at rated speed with different loads and presented in Table 12. From Table 12, it is clear that in conventional SWECS, the THDs of inverter output voltage and current without filter are very high and violate permissible limits. Hence, filter is required in order to bring these THD's within permissible limits. After filter both THD's comes within permissible limits. Whereas, in proposed SWECS, the THDs of VFT output voltage and current are very less than the conventional system and are within permissible limits. Thus, here no filter is required. Further, the efficiency of both conventional and proposed methods increases with the increase in load from 1 kW, gets a maximum value at 2 kW and then starts decreasing with the further increase in load. Moreover, the efficiency of proposed method is slightly greater than the conventional system which is around 0.5%. Hence from Table 12, it is concluded that the proposed method is having slightly greater efficiency with almost negligible THD. 6. Experimental analysis 6.1. Experimental setup Fig. 10 shows the block diagram of practical setup. The dc motor is mechanically coupled to the SG in order to simulate the wind
Table 10 Power Fed To Load at 1200 rpm SG Input Speed (Frequency ¼ 40 Hz) without Excitation Control. Load (kW)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
fL (Hz)
PL (W)
% h ¼ ½PL =ðPs þ Pd Þ 100
1 2 3 4
861.8 1684 2525 3411
214.23 415.63 618.26 825.61
300 300 300 300
0.8 0.8 0.8 0.8
397.6 396.4 394.7 392.8
50 50 50 50
988.2 1964 2921 3858
91.84 93.54 92.93 91.06
114
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Table 11 Power Fed To Load at 1200 rpm SG Input Speed (Frequency ¼ 40 Hz) with Excitation Control. Load (kW)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
fL (Hz)
PL (W)
% h ¼ ½PL =ðPs þ Pd Þ 100
1 2 3 4
872.2 1714 2593 3537
216.8 423.17 634.92 856.08
300 300 300 300
0.8 0.8 0.8 0.8
400.0 400.0 400.0 400.0
50 50 50 50
1000 2000 3000 4000
91.83 93.58 92.94 91.05
phase, 10.669 kVA). The auto-transformer #1 is used for matching the stator voltage of WRIM to the load voltage. The DDM is energized by SG via DDM controller. The controller of DDM converts the 3 phase ac supply to controlled dc supply and feeds the dc supply to DDM. The pictorial view of the experimental setup is shown in Fig. 11. Here, WRIM is mechanically coupled to a dc motor #1 (shunt type), where dc motor #1 is working as DDM. 6.2. Experimental results
Fig. 8. Variation of load power with and without excitation control w.r.to the DDM power at 1200 rpm.
In experimental analysis, the SG input speed is varied while the load is kept constant i.e. 1.2 kW, 400 V rated voltage and 50 Hz rated frequency. The experimental results observed are given in Table 13. The experimental results obtained in Table 13 shows that with decrease in SG speed the SG output voltage, SG frequency and the SG output power decreases but the DDM power increases and thus the efficiency of the SWECS remains almost constant for all operating conditions. Hence the experimental results validate the simulation results. In Table 13 the load is 1.2 kW but it is showing the output power more than load power. This is because the output power is sum of auto-transformer #1 power and load power. The variations of SG power, DDM power, output power and % efficiency with the SG input speed of Table 13 are plotted in Fig. 12. From Fig. 12 it is again clear that for a constant load, with the decrease in SG input speed the SG power decreases while the DDM power increases but the output power and the efficiency remains almost constant. 7. Cost analysis
Fig. 9. Efficiency of proposed SWECS with and without excitation control w.r.to the load power at 1200 rpm.
turbine. The SG supplies power to auto-transformer #2 (3 phase, 415 V, 22.793 kVA), which steps down the SG voltage to rated value of rotor voltage and transfer the power to rotor of the WRIM (3 phase, 2.2 kW, 400 V, 50 Hz, 1440 rpm). The DDM supplies mechanical power to rotor of the WRIM in order to control the power flow from rotor to stator side of WRIM, such that the transferred power at stator side is having rated frequency. The stator of WRIM supplies power to isolated loads through auto-transformer #1 (3
The cost analysis of conventional and proposed systems have been performed for 3 kW rating and shown in Table 14. The data for the cost analysis has been taken from Ref. [21]. From Table 14, it is clear that the proposed system is cheaper than the conventional system by 40 Euro. 8. Conclusion A new method for feeding an isolated load from synchronous generator (SG) based wind power generation has presented in this paper. The proposed method uses variable frequency transformer (VFT) for standalone wind energy conversion system (SWECS). For this, digital simulation models of the conventional method and the proposed method have been developed under MATLAB-Simulink environment. Few studies on power flow from SG to the isolated
Table 12 Comparison of Proposed Method With Conventional Method At 1500 rpm. Load (kW)
1 2 3 4
Conventional method of SWECS
Proposed method of SWECS
% THD of VL without filter
% THD of IL without filter
% THD of VL with filter
% THD of IL with filter
%h
% THD of VL
% THD of IL
%h
67.93 68.15 67.96 68.06
17.48 16.37 17.76 16.68
3.96 2.75 2.91 2.72
3.96 2.75 2.71 2.69
91.32 93.11 92.42 91.01
0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01
91.83 93.63 93.05 91.07
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
115
Fig. 10. Block diagram of the experimental setup.
Fig. 11. The experimental setup.
Table 13 Experimental Results for Power Fed to 1.2 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
fL (Hz)
Output power, Po (W)
% h ¼ ½PL =ðPs þ Pd Þ 100
972 890 845 799 753 703
228.4 199.3 183.3 167.0 150.6 133.0
32.4 29.7 28.2 26.7 25.1 23.5
1199.6 1107.6 1070.4 1002.8 977.2 925.2
410.38 468.00 515.95 591.56 630.72 691.95
528.1 610.7 655.7 701.9 747.5 797.8
0.65 0.59 0.56 0.53 0.5 0.47
400.15 400.14 400.14 400.12 400.53 400.54
50.039 50.074 50.044 50.016 50.014 50.014
1224 1224 1224 1236 1252 1260
76.03 77.68 77.16 77.52 77.86 77.91
Fig. 12. The variation of SG power, DDM power, output power and % efficiency with the SG input speed at 1.2 kW load.
loads by conventional method at rated SG speed i.e. 1500 rpm have been carried out using MATLAB-Simulink model. A number of studies on power flow from SG to the isolated loads have been carried out by MATLAB-Simulink model of proposed method without and with excitation control under different values of SG input speed. A number of quantities such as SG output power, dc drive motor (DDM) mechanical power, load power and efficiency have been obtained. From the simulation results, it is evident that with decreases in SG input speed, the SG output power decreases but the DDM power increases in order to supply power to the load at constant frequency and voltage. Even though with the increase in DDM power, the load power reduces slightly but the efficiency is maintained constant. Then the power fed to variable load by the SG based SWECS at fixed speed i.e. 1200 rpm without and with excitation control has been analyzed using simulation model of proposed method. From here it is concluded that the power fed to load increases linearly with the DDM power. The load power as well as efficiency with excitation control are slightly higher than without excitation control. Both increases with the increase in load, reaches to the maximum value and then starts decreasing with the further
116
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Table 14 Comparison of cost (Euro). Equipments
Iron (kg) Copper (kg) Total (kg) SG active material cost (Euro) SG construction cost (Euro) WRIM active material cost (Euro) DDM active material cost (Euro) (WRIM þ DDM) construction cost (Euro) Converter cost (Euro) Total cost (Euro)
Direct-drive SG Conventional system
Proposed system SG
WRIM
DDM
Total
32.5 12.6 45.1 287 160 e e e 120 567
32.5 12.6 52.08 287 160 30 10 40 e 527
4.03 1.21
1.34 0.40
37.87 14.21
increase in load. Moreover, the THD of output voltage, THD of output current and efficiency of the proposed method have been compared with those of the conventional method. From the comparison results, it is obtained that the proposed system does not produce harmonics and having slightly higher efficiency than the conventional system. Moreover, in order to validate the proposed method a practical analysis has been performed. The response characteristics of efficiency and different powers under various SG input speed are achieved. The achieved practical results validate the simulation results. Hence, from simulation and practical results it can be concluded that the proposed method can be used as an SWECS for SG based wind energy generation system. Further, cost analysis has also been carried out for the proposed system and it has been compared with that of conventional system. From the cost analysis, it is observed that the proposed system is cheaper than conventional system.
[9]
[10]
[11]
[12]
[13]
[14]
[15]
References [1] L.A. de S. Ribeiro, O.R. Saavedra, S.L. Lima, J.G. de Matos, G. Bonan, Making isolated renewable energy systems more reliable, Renew. Energy 45 (2012) 221e231. [2] F.I. Bakhsh, M.M. Shees, M.S.J. Asghar, Performance of wound rotor induction generators with the combination of input voltage and slip power control, Russ. Electr. Eng. 85 (6) (June 2014) 403e417. [3] S. Sharma, B. Singh, Asynchronous generator with battery storage for standalone wind energy conversion system, IEEE Trans. Ind. Appl. 50 (4) (July/ August 2014) 2760e2767. [4] A.J.G. Westlake, J.R. Bumby, E. Spooner, Damping the power-angle oscillations of a permanent-magnet synchronous generator with particular reference to wind turbine applications, IEE Proc.-Electr. Power Appl. 143 (3) (May 1996) 269e280. [5] M.M. DJesus, D.S. Martin, S. Arnaltes, E.D. Castronuovo, Optimal reactive power allocation in an offshore wind farms with LCC-HVdc link connection, Renew. Energy 40 (2012) 157e166. [6] H. Ghoddami, M.B. Delghavi, A. Yazdani, An integrated wind-photovoltaicbattery system with reduced power-electronic interface and fast control for grid-tied and off-grid applications, Renew. Energy 45 (2012) 128e137. rdenas, R. Pen ~ a, P. Wheeler, J. Clare, C. Juri, Control of a matrix converter [7] R. Ca for the operation of autonomous systems, Renew. Energy 43 (2012) 343e353. [8] P. Moutis, S.A. Papathanassiou, N.D. Hatziargyriou, Improved load-frequency
[16]
[17]
[18]
[19]
[20]
[21]
control contribution of variable speed variable pitch wind generators, Renew. Energy 48 (2012) 514e523. C.N. Bhende, S. Mishra, S.G. Malla, Permanent magnet synchronous generatorbased standalone wind energy supply system, IEEE Trans. Sustain. Energy 2 (4) (October 2011) 361e373. C. Mi, M. Filippa, J. Shen, N. Natarajan, Modeling and control of a variablespeed constant-frequency synchronous generator with brushless exciter, IEEE Trans. Ind. Appl. 40 (2) (March/April 2004) 565e573. N. Mendis, K. Muttaqi, An integrated control approach for standalone operation of a hybridised wind turbine generating system with maximum power extraction capability, Electr. Power Energy Syst. 49 (2013) 339e348. E. Larsen, R. Piwko, D. McLaren, D. McNabb, M. Granger, M. Dusseault, LP. Rollin, J. Primeau, Variable Frequency Transformer e a New Alternative for Asynchronous Power Transfer, Canada Power, Toronto, Ontario, Canada, September 28-30, 2004. P. Doyon, D. McLaren, M. White, Y. Li, P. Truman, E. Larsen, C. Wegner, E. Pratico, R. Piwko, Development of a 100 MW Variable Frequency Transformer, Canada Power, Toronto, Ontario, Canada, September 28-30, 2004. F.I. Bakhsh, D.K. Khatod, Application of variable frequency transformer (VFT) for grid interconnection of PMSG based wind energy generation system, Sustain. Energy Technol. Assess. 8 (2014) 172e180. L. Wang, S.-R. Jan, C.-N. Li, H.-W. Li, Y.-H. Huang, Y.-T. Chen, Analysis of an integrated offshore wind farm and seashore wave farm fed to a power grid through a variable frequency transformer, in: IEEE Power and Energy Society General Meeting, San Diego, CA, 24-29 July, 2011. L. Wang, L.-Y. Chen, Reduction of power fluctuations of a large-scale gridconnected offshore wind farm using a variable frequency transformer, IEEE Trans. Sustain. Energy 2 (3) (July 2011) 226e234. T. Geetha, V. Jayashankar, Variable frequency transformers for increased wind penetration, in: Joint International Conference on Power System Technology and IEEE Power India Conference (POWERCON 2008), New Delhi, 12-15 Oct., 2008. F.I. Bakhsh, D.K. Khatod, A novel method for grid integration of synchronous generator based wind energy generation system, in: IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES), IIT Bombay, India, 16th to 19th Dec., 2014. A. Merkhouf, P. Doyon, S. Upadhyay, Variable frequency transformerdConcept and electromagnetic design evaluation, IEEE Trans. Energy Convers. 23 (4) (December 2008) 989e996. F.I. Bakhsh, D.K. Khatod, Digital simulation of VFT applications between power system networks, in: Fourth International Joint Conference on Advances in Engineering and Technology, NCR, India, 13-14 Dec., 2013, pp. 60e74. H. Polinder, F.F.A. van der Pijl, G.-J. de Vilder, P. Tavner, Comparison of directdrive and geared generator concepts for wind turbines, IEEE Trans. Energy Convers 21 (3) (Sept. 2006) 725e733.
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
A new synchronous generator based wind energy conversion system feeding an isolated load through variable frequency transformer Farhad Ilahi Bakhsh*, Dheeraj Kumar Khatod Alternate Hydro Energy Centre, Indian Institute of Technology Roorkee, Roorkee 247667, India
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 February 2015 Received in revised form 20 June 2015 Accepted 30 July 2015 Available online xxx
This paper aims to explore the possibility of synchronous generator (SG) based wind energy generation system feeding an isolated load using a latest power transmission technology i.e. variable frequency transformer (VFT). The proposed configuration does not employ any power electronics based interface as in conventional SG based stand-alone wind energy conversion systems (SWECS). For analysis, the simulation models of proposed configuration as well as conventional configuration have been developed under MATLAB-Simulink environment. A series of studies on power fed from the SG to the different loads at various SG input speeds has been carried out with the developed models. Further to analyze the effectiveness of the proposed method; the efficiency, total harmonic distortion (THD) of output voltage and THD of output current of the proposed method have been compared with those of the conventional method. From obtained results, it is observed that the proposed method is simple and does not produce harmonics. Moreover to validate the proposed scheme, an experimental analysis has been carried out. Further, the cost analysis of both systems has also been carried out. From the cost analysis, it is observed that the proposed system is cheaper than the conventional system. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Isolated loads Synchronous generator (SG) Power fed Variable frequency transformer (VFT) Standalone wind energy conversion system (SWECS)
1. Introduction Due to concern over the environment, the advent of renewable energy (RE) based power generation (wind, hydro, etc.), has increased their utilization in different forms. A customer with limited RE based generation (e.g. wind turbine), needs a connection with distribution system of bulk power grid for reliable power supply. In remote locations, isolated RE based power generation systems can be utilized to serve the local loads, eliminating the grid connection, transmission losses and transmission costs [1]. Nowadays, the wind based power generation has been emerged as one among the most promising RE based generation system. Due to this small installations of stand-alone wind energy conversion systems (SWECS) are quite promising in remote area electrification programmes. To utilize the wind energy, fixed speed and variable speed SWECS are used. In fixed-speed SWECS, the wind turbines are mostly equipped with an induction generator because of its simplicity, ruggedness and less maintenance [2,3]. Here, the isolated induction generator is directly connected to the loads. In order
* Corresponding author. E-mail address: [email protected] (F.I. Bakhsh). http://dx.doi.org/10.1016/j.renene.2015.07.093 0960-1481/© 2015 Elsevier Ltd. All rights reserved.
to supply constant voltage and frequency to the load, the speed of the turbine is controlled by the gearbox. The use of a gearbox causes many problems such as it requires continuous maintenance, enlarges the weight and size of the SWECS, generates noise, increases the power losses and hence reduces the efficiency of SWECS [4]. Variable-speed wind turbines possess many advantages over fixedspeed, such as larger energy capture, operation at maximum power point, better efficiency and quality of power. Hence, variable-speed wind turbines are the dominating type of turbines as far as present trend is concerned and used in SWECS for feeding an isolated load. In this operating mode, the wind turbines are equipped with either synchronous or doubly fed induction generator. The application of doubly fed induction generator for feeding isolated load is discussed in Refs. [5,6], whereas the application of synchronous generator (wound field or permanent magnet) feeding the isolated load is discussed in Refs. [7,8]. Both of these methods require suitable power electronics conversion system for feeding power to the load. These power electronics converters are costly, require sophisticated control system, cause harmonic distortion and thereby deteriorate the power quality. Moreover, they require suitable compensation in order to meet the standards for harmonic pollution which further increases the cost and complexity of the system [9e11].
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
The main challenge faced in a synchronous generator (SG) based SWECS is the power quality. This challenge can be well answered by using a current power transmission technology termed as variable frequency transformer (VFT) [12,13]. The application of VFT for grid interconnection of wind power generation is discussed in Refs. [14e18]. But the application of VFT for SWECS has not been discussed so far. This paper deals with the analysis of a new configuration of SG based wind energy generation system for feeding power to an isolated load using VFT. The VFT is realized using a wound rotor induction machine (WRIM) whose rotor is mechanically coupled to the dc drive motor (DDM). The SG supplies power to the load at different levels of SG input speed. The requirements of costly power electronics converters are omitted. Hence, the proposed method is simple and does not produce harmonics. This paper has been organized in eight sections. The Section 2 of the paper describes the conventional SG based SWECS. Section 3 discusses the modeling and analysis of the proposed SG based SWECS. Sections 4 and 5 present the MATLAB-Simulink model; and results of the conventional method and the proposed method, respectively. Section 6 presents the experimental analysis of the proposed method. Section 7 shows the cost analysis of the proposed method and the conventional method. Finally Section 8 draws the conclusion of the paper.
107
Fig. 2. Proposed system for SG based SWECS.
feeding SG power to load. Here, the wind turbine is connected to the rotor of the SG without a gear box. The stator winding of the SG is connected to the rotor winding of the WRIM. The stator winding of the WRIM is connected to the isolated load as shown in Fig. 2. A DDM is coupled (mechanically) to the rotor of the WRIM in order to apply torque to the rotor of the WRIM.
2. Conventional method of SG based SWECS In a conventional method of SG based SWECS, the wind turbine is connected to the rotor of the SG with or without gear box. The output power of the SG is fed to the load through a power electronics based interface. The power electronics based interface mainly consists of an ac-to-dc rectifier followed by a dc-to-ac inverter as shown in Fig. 1. The SG output ac power is first rectified into a dc power using uncontrolled ac-to-dc rectifier. The capacitor across the output of ac-to-dc rectifier is connected in order to filter out the ripples in the dc power. Then this dc power is converted again into the ac power using self-commutated inverter, which are mainly pulse width modulated (PWM) inverter employing insulated gate bipolar transistors (IGBTs). In this type of inverter, in addition to the control of the active power, the reactive power is also controllable. This interface produces inter-harmonics (harmonic distortion) and thereby deteriorates the quality of power supplied. A filter is required at the inverter side in order to meet the standards for harmonic pollution which further increases the cost and complexity of the system. 3. Proposed method In the proposed method of SG based SWECS, VFT is used for
3.1. Modeling of proposed method In modeling, the VFT is represented as a doubly-fed WRIM whose rotor is mechanically coupled to the DDM [19,20]. Here, DDM supplies mechanical power ‘Pd’ to the rotor of the WRIM as shown in Fig. 3. The three phase windings are available on both rotor and stator sides of the WRIM. The isolated load is connected to the stator windings of the WRIM having voltage ‘VL’ and load angle ‘qL’. The rotor windings of the WRIM is connected to the stator windings of the SG, energized by voltage ‘Vs’ with phase angle ‘qs’. The rotor of the SG is mechanically coupled with the dc motor, where dc motor is working as prime mover of the SG in order to simulate the wind turbine. To transfer power from SG to load, the Pd is applied at the rotor of the WRIM. The power transfer through VFT is controlled by the magnitude of the applied Pd. In the power flow process, only real power transfer has being considered. 3.2. Analysis of proposed method Considering VFT as an ideal and lossless machine and neglecting its magnetizing current and leakage reactance, the power balance equation for it will be written as:
Fig. 1. Conventional method of SG based SWECS.
108
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
It is well known that frequency of rotor is slip times-frequency of stator. Thus, putting fs ¼ sfL in (6), we get
f rm ¼ sf L f L ¼ ðs 1Þf L
(7)
and
Nrm ¼ f rm 120 Np urm ¼
2p 2p Nrm ¼ f rm 120 Np 60 60
(8)
(9)
where,
Fig. 3. Model representation of the proposed system.
PL ¼ Ps þ Pd
(1)
s ¼ slip of the VFT, frm ¼ VFT rotor mechanical speed in Hz, Np ¼ number of poles in the VFT, Nrm ¼ VFT rotor mechanical speed in rpm, and urm ¼ VFT rotor mechanical speed in rad/s. Combining the above relationships gives the power exchanged with the drive system as
where, PL ¼ electrical power fed to the load, Ps ¼ electrical power available at output of the SG, and Pd ¼ mechanical power supplied by the DDM.
Pd ¼ PL Ps ¼ VL IL Vs Is VL fs IL N1 ¼ V L IL N2 N1 fL N2
Since the VFT behaves like a transformer, the ampere-turns must balance between stator and rotor windings:
N1 I L ¼ N2 I s
(2)
¼ VL IL ½ðsVL Þ IL ¼ VL IL ð1 sÞ or;
Pd ¼ PL ð1 sÞ
where,
(10) It shows that the electrical power flowing into the load being dependent on the mechanical power of the DDM and slip of WRIM.
N1 ¼ number of turns available on stator winding of VFT, N2 ¼ number of turns available on rotor winding of VFT, IL ¼ current fed to load, and Is ¼ current supplied by the SG.
a) If PL is constant, then
Since, both the rotor and stator windings of VFT link the same magnetic flux, therefore
VL ¼ N1 f L Ja ;
(3)
Vs ¼ N2 f s Ja ;
(4)
(5)
sff s and f s fNs
s ¼ kNs where,
where, Vs ¼ voltage magnitude of SG on rotor side of VFT, fs ¼ frequency of voltage applied across rotor winding of VFT (Hz), VL ¼ voltage magnitude of load on stator side of VFT and fL ¼ frequency of voltage available across stator winding of VFT (Hz), and Ja ¼ air-gap flux. The nature of the VFT is such that in steady state condition, the rotor mechanical speed in frequency (electrical) is equal to the difference in the frequency (electrical) on the stator and rotor windings,
f rm ¼ f s f L
Since load frequency is maintained constant, therefore
Hence,
and
Vs =N2 ¼ ðVL =N1 Þ ðf s =f L Þ
Pd fð1 sÞ and s ¼ f s =f L
(6)
k ¼ constant, Ns ¼ SG speed. Thus,
Pd fð1 kNs Þ
(11)
For constant load, the mechanical power of the DDM decreases with increase in the SG speed. b) If the slip of WRIM is maintained constant, then
Pd fPL
(12)
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
At a constant slip, the electrical power fed to the load is being only proportional to mechanical power of the DDM. 4. MATLAB-Simulink analysis of conventional method 4.1. MATLAB-Simulink model For simulation in MATLAB-Simulink, SG is simulated with the synchronous machine pu standard having rating 3 phase, 8.1 kVA, 1500 rpm, 400 V, 50 Hz and the load is simulated using three-phase series R load as shown in Fig. 4. The SG is connected to rectifier side of the power electronics based interface and the load is connected to inverter side of the power electronics based interface. Here, rectifier is simulated with the universal diode bridge and the inverter is simulated with universal IGBT bridge using PWM scheme. A constant block named as dc motor is used to apply rated speed at input of the SG. The voltage and frequency of different loads are kept constant i.e. 400 V (LeL) and 50 Hz. Then the simulated model (Fig. 4) is used to analyze the performance of the conventional method.
109
Table 1 Power Fed to Different Loads at 1500 rpm. S. No.
Ps (W)
PL (W)
% h ¼ PL =Ps 100
1 2 3 4
1095 2148 3246 4395
1000 2000 3000 4000
91.32 93.11 92.42 91.01
connected to the stator windings of WRIM and SG (3 phase, 8.1 kVA, 1500 rpm, 400 V, 50 Hz) is connected to rotor windings of WRIM. A block named as dc motor is used to apply various speed at the input of the SG. The DDM is simulated with separately excited dc motor. The output mechanical power of the DDM applied to the rotor of WRIM is of mechanical torque ‘Tm’ as shown in Fig. 5. The voltage and frequency of the load is kept constant and the input speed of the SG is varied. The three-phase capacitor branch is connected across the stator windings of the WRIM in order to supply reactive power required at stator side of the WRIM. Then the simulated model (Fig. 5) is used to analyze the performance of the proposed method.
4.2. MATLAB-Simulink results 5.2. MATLAB-Simulink results Table 1 shows the power fed to different loads by conventional SG based WECS at rated value of input SG speed. It is evident from this table that the efficiency of the conventional system increases with the increase in value of load (at 1 kW load the efficiency is 91.31%), reaches to maximum value (93.80%) at 2 kW load and then starts decreasing with further increase in load. 5. MATLAB-Simulink analysis of proposed method 5.1. MATLAB-Simulink model For simulation in MATLAB-Simulink, VFT is simulated with the asynchronous machine SI units which is basically a doubly-fed, 3 phase, 7.5 kW, 400 V, 50 Hz, 1440 rpm WRIM and the isolated load is simulated with three-phase R (resistive) load having rated voltage 400 V at rated frequency of 50 Hz. The three-phase R load is
5.2.1. For constant load In this case 1 kW, 2 kW, 3 kW and 4 kW loads are supplied at different values of SG input speeds. Since the load frequency is dependent upon slip of the VFT, the mechanical power from DDM is varied which controls the speed of WRIM and hence the slip of the VFT. Thus by controlling the slip, the load frequency is maintained constant at 50 Hz. 5.2.1.1. For 1 kW load. To realize the performance of proposed method, the SG is operated at different speeds for feeding a 1 kW load. The DDM power is varied by DDM controller which varies the torque applied at the rotor of the WRIM, thus controlling the slip of the machine in order to supply power to the load at 50 Hz. Various variables obtained at different SG input speed conditions are given in Table 2.
Fig. 4. MATLAB-Simulink model of the conventional SG based SWECS.
110
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Fig. 5. MATLAB-Simulink model for the proposed SG based SWECS.
Table 2 Power Fed to 1 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300
394.0 367.2 340.5 313.8 287.1 260.5 233.9 207.4 180.9 154.4 128.0 101.6 75.18
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00
1089 1013 937.1 861.8 786.7 711.8 637.2 562.8 488.6 414.6 340.8 267.2 194.1
0.025 72.01 143.44 214.23 284.29 353.48 421.41 487.98 552.50 614.02 671.25 720.75 756.12
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20
400.0 399.3 398.5 397.6 396.7 395.6 394.4 392.9 391.1 388.7 385.5 380.9 373.7
1000 996.4 992.5 988.2 983.5 978.2 972.1 964.8 955.8 944.4 929.1 907.0 872.9
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
91.83 91.83 91.85 91.84 91.83 91.83 91.83 91.82 91.81 91.81 91.8 91.81 91.86
Table 2 shows that when SG is operating at rated speed i.e. 1500 rpm, the voltage available across the stator terminals of the SG is 394 V which is near to rated voltage (400 V) and its frequency is 50 Hz (rated frequency). The output power of the SG is 1089 W. A little amount of DDM power (0.025 W) is required to keep the rotor of WRIM stationary (or to achieve unity slip) so that the load frequency becomes 50 Hz. At this value of slip, the load voltage and power both are having their rated values. On reducing the value of SG speed from 1500 rpm to 1400 rpm, the output voltage, frequency and output power of SG also reduces from 394 V to 367.2 V, 50 Hz to around 46.67 Hz and 1089 W to 1013 W, respectively. But the DDM mechanical power increases to 72.01 W for maintaining the slip around 0.93 so that the load frequency becomes 50 Hz. At this value of slip, the load voltage is reduced to 399.3 V and accordingly load power also reduces to 996.4 W. Since the load is constant, on further reducing the SG input speed, the SG (voltage, frequency and power) reduces but the mechanical power of DDM increases in order to supply the constant load. The torque applied by the DDM is such that the rotor of the WRIM rotates at a speed which is difference of the WRIM synchronous speed and the SG speed. At this speed of WRIM, the obtained slip value will be such that the voltage available at the stator terminals of VFT will have a
frequency of 50 Hz. But the magnitude of load voltage and power is reduced from the rated value with the decrease in SG speed. Moreover, the THD of load voltage and current remains constant i.e. 0.01% which is negligible. The variation of SG output power, DDM power, load power and % efficiency with the SG input speed of Table 2 are plotted in Fig. 6. When the SG input speed becomes 300 rpm the load voltage becomes 373.7 V which is not within permissible limit of voltage variation i.e. ±5%. Thus it is required to increase the load voltage in order to achieve power from the SG at low level of input speed. This can be done by various methods i.e. using auto-transformer, ac voltage regulator or by excitation control of SG. In this work, field excitation of the SG is controlled in order to achieve the rated load voltage. The various parameters achieved using excitation control is presented in Table 3. From Fig. 6, it is clear that with the decrease in SG input speed, the SG output power decreases but the DDM power increases. Though there is decrement in load power but the efficiency is almost constant for all operating conditions due to the constant load. Table 3 shows the variation of load profile with the SG input speed at different field voltage (Vf), where Vf is expressed in pu. In this case when SG is operating at rated speed the Vf is 1 pu, thus the
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
111
Fig. 6. The variation of SG power, DDM power, load power and % efficiency with the SG input speed at 1 kW load.
Fig. 7. The variation of SG power, DDM power, load power and % efficiency with excitation control for 1 kW load.
SG (output voltage, frequency and output power), DDM power, load (voltage, frequency and power) and THD of load voltage and current is same as in earlier case (without excitation control). On reducing the value of SG input speed from 1500 rpm to 1400 rpm, the SG output voltage reduces from 394 V to 368 V, frequency reduces from 50 Hz to 46.67 Hz and output power also reduces form 1089 to 1016 W. The reduction in SG output voltage and power is lesser than the earlier case because of 0.2% increase in excitation voltage. But the DDM mechanical power increases from 72.01 W to 72.3 W from previous case for maintaining the slip around 0.93 in order to achieve 50 Hz load frequency. Moreover, due to excitation control, the load voltage and power remains constant at their rated values. On further reducing the SG input speed, the value of Vf increases, the SG (voltage, frequency and power) reduces, the mechanical power of DDM increases but the load voltage and power is maintained constant. Moreover, the THD of load voltage and current remains constant as in earlier case. Thus, by controlling the excitation of the SG, the load voltage and power can be maintained constant (which generally reduces with the reduction in SG input speed) even at 100 rpm SG input speed. Hence, the SG power which is normally wasted due to low level of wind speed can be utilized effectively using the proposed method. Similarly the results are obtained for 2 kW, 3 kW and 4 kW loads. The variation of SG output power, DDM power, load power and % efficiency with the SG input speed of Table 3 are plotted in Fig. 7. Fig. 7 shows the variation of SG power, DDM power, load power and efficiency in % with excitation control. From Fig. 7, it is clear
with the increase in SG input speed, the SG output power increases but the DDM power decreases such that the load power is maintained constant at all values of SG input speed. 5.2.1.2. For 2 kW load. The load is increased from 1 kW to 2 kW and it is kept constant for all operating conditions. The power flow from SG to the load at different values of SG input speed without and with excitation control are obtained in Tables 4 and 5, respectively. 5.2.1.3. For 3 kW load. The SG is operated at different values of input speed and the load is kept constant i.e. 3 kW. Then the power fed to the load at different values of SG input speed without and with excitation control are presented in Tables 6 and 7, respectively. 5.2.1.4. For 4 kW load. The power fed from SG to 4 kW load at different levels of SG input speed without and with excitation control are evaluated in Tables 8 and 9, respectively. From Tables 2e9, it is clear that VFT can be used as an interface between SG based wind energy generation system and isolated loads, and power can be supplied to load at different levels of input SG speed without or with excitation control. With excitation control, the load power remains almost constant and thus the proposed method is able to effectively utilize the small power available at the output of the SG and hence avoids the wastage of power due to the low level of input wind speed. Further, the efficiency of the proposed SWECS method with or without excitation control for a
Table 3 Power Fed to 1 kW Load with Excitation Control. Vf in pu Ns (rpm) Vs (V)
fs (Hz) Ps (W)
1.0000 1.0020 1.0040 1.0060 1.0084 1.0112 1.0145 1.0182 1.0230 1.0290 1.0375 1.0500 1.0705 1.1110 1.2326
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
394.0 368.0 341.8 315.7 289.5 263.4 237.3 211.2 185.0 158.9 132.8 106.6 80.48 54.33 28.20
Pd (W)
1089 0.025 1017 72.3 944.7 144.6 872.2 216.8 800.0 289.07 727.7 361.28 655.6 433.6 583.5 505.94 511.3 578.22 439.0 650.31 366.8 722.57 294.6 795.4 222.4 866.83 150.1 938.65 77.88 1010.6
Nrm (rpm) S ¼ ½ð1500 Nrm Þ=1500
VL (V) PL (W) % THD of VL % THD of IL % h ¼ ½PL =ðPs þ Pd Þ 100
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00
91.83 91.80 91.80 91.83 91.82 91.83 91.81 91.79 91.78 91.80 91.80 91.74 91.81 91.85 91.87
112
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Table 4 Power Fed To 2 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800 700 600
396.2 368.9 341.8 314.6 287.6 260.5 233.6 206.6 179.7 152.9
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00
2136 1988 1836 1684 1533 1382 1232 1082 933.4 785.1
1.057 140.53 279.18 415.63 549.57 680.68 807.39 928.76 1043.8 1147.0
0 100 200 300 400 500 600 700 800 900
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40
400.0 399.1 397.8 396.4 394.8 392.9 390.6 387.9 384.6 380.2
2000 1991 1979 1964 1948 1929 1907 1881 1849 1807
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
93.59 93.54 93.56 93.54 93.54 93.52 93.51 93.55 93.51 93.53
Table 5 Power Fed to 2 kW Load with Excitation Control. Vf in pu
Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1.0000 1.0021 1.0054 1.0090 1.0132 1.0182 1.0239 1.0310 1.0400 1.0520 1.0685 1.0928 1.1332 1.2139 1.4563
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
396.2 369.7 343.6 317.5 291.4 265.3 239.1 213.0 186.9 160.8 134.7 108.5 82.4 56.27 30.14
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
2136 1996 1856 1714 1573 1433 1292 1151 1010 868.8 727.9 586.7 445.6 304.5 163.6
1.057 141.06 282.11 423.17 564.23 705.29 846.34 987.4 1128.5 1269.5 1410.6 1551.6 1692.7 1833.7 1974.8
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00
93.59 93.59 93.54 93.58 93.58 93.53 93.53 93.53 93.53 93.53 93.52 93.53 93.53 93.53 93.53
Table 6 Power Fed To 3 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800 700
396.5 369.1 341.6 314.2 286.8 259.4 232.1 204.9 177.7
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33
3224 2994 2759 2525 2291 2059 1827 1597 1368
3.905 210.17 416.57 618.26 814.72 1003.7 1184.4 1353.9 1508.8
0 100 200 300 400 500 600 700 800
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47
400.0 398.6 396.8 394.7 392.4 389.6 386.4 382.4 377.6
3000 2979 2952 2921 2886 2846 2799 2743 2673
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
92.94 92.97 92.96 92.93 92.93 92.92 92.95 92.95 92.92
particular load under all operating conditions are almost same. Moreover, the THD of output voltage and current of the proposed method is almost negligible.
5.2.2. For constant slip In this case, the SG input speed is kept constant but the load is varied from 1 kW to 4 kW. Since the SG input speed is constant, the
Table 7 Power Fed to 3 kW Load with Excitation Control. Vf in pu
Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1.0000 1.0036 1.0082 1.0134 1.0195 1.0266 1.0352 1.0458 1.0594 1.0773 1.1024 1.1619 1.2019 1.3266 1.7025
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
396.5 370.4 344.4 318.4 292.4 266.3 240.3 214.3 188.3 162.2 136.2 110.2 84.19 58.19 32.22
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
3224 3016 2804 2593 2381 2170 1958 1747 1535 1324 1113 900.9 689.3 477.9 266.4
3.905 211.64 423.28 634.92 846.55 1058.2 1269.8 1481.5 1693.1 1904.8 2116.4 2328.0 2539.7 2751.3 2962.9
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000 3000
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00
92.94 92.95 92.96 92.94 92.95 92.93 92.94 92.92 92.93 92.92 92.90 92.91 92.91 92.90 92.90
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
113
Table 8 Power Fed To 4 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1500 1400 1300 1200 1100 1000 900 800
394.7 367.0 339.3 311.7 284.2 256.7 229.2 201.8
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67
4392 4064 3737 3411 3087 2764 2443 2124
4.12 282.32 557.95 825.61 1083.2 1329.4 1559.5 1769.6
0 100 200 300 400 500 600 700
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53
400 397.9 395.5 392.8 389.7 386.1 381.8 376.5
4000 3958 3911 3858 3797 3727 3644 3545
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00
90.989 91.065 91.061 91.063 91.05 91.049 91.043 91.048
Table 9 Power Fed to 4 kW Load with Excitation Control. Vf in pu
Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
PL (W)
% THD of VL
% THD of IL
% h ¼ ½PL =ðPs þ Pd Þ 100
1.0000 1.00538 1.01140 1.01832 1.02639 1.03599 1.04770 1.06230 1.08100 1.10580 1.14070 1.19280 1.28050 1.45725 1.99450
1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
394.7 369.0 343.2 317.4 291.7 265.9 240.1 214.4 188.6 162.9 137.1 111.4 85.62 59.90 34.22
50.00 46.67 43.33 40.00 36.67 33.33 30.00 26.67 23.33 20.00 16.67 13.33 10.00 6.67 3.33
4392 4108 3823 3537 3252 2967 2682 2397 2112 1826 1541 1256 970.5 685.2 399.8
4.12 285.36 570.72 856.08 1141.4 1426.8 1712.2 1997.5 2282.9 2568.3 2853.6 3134.0 3423.1 3708.3 3993.6
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1.00 0.93 0.87 0.80 0.73 0.67 0.60 0.53 0.47 0.40 0.33 0.27 0.20 0.13 0.07
400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
90.989 91.046 91.039 91.052 91.045 91.037 91.03 91.022 91.015 91.028 91.021 91.013 91.042 91.043 91.046
mechanical speed of WRIM and thus the slip of the VFT for achieving 50 Hz load frequency is also maintained constant. The results obtained without and with excitation control are given in Tables 10 and 11 where SG input speed is kept constant at 1200 rpm and load is changing. Tables 10 and 11 show that when the SG input speed is kept fixed at 1200 rpm, the SG output power and the DDM mechanical power without and with excitation control increases with the increase in load. The increase in DDM power results in constant speed of WRIM rotor (300 rpm) i.e. slip becomes constant (0.8). Similarly the results can be obtained at other SG input speeds. Various plots obtained from the Tables 10 and 11 are shown in Figs. 8 and 9. Fig. 8 shows the variation of load power with the DDM power at 1200 rpm SG input speed. It is evident from Fig. 8 that the power fed to load with and without excitation control are directly proportional to the DDM power and the load power with excitation control is slightly greater than without excitation control. Fig. 9 shows the efficiency variation of proposed SWECS with the power fed to load at SG input speed of 1200 rpm. It is clear from Fig. 9 that the efficiency of proposed method with excitation control is slightly more than without excitation control and both increases with the increase in load from 91.8%, reaches to the maximum value and then starts decreasing with the further increase in load. The efficiency, THD of output voltage and THD of output current
of proposed system without excitation control have been compared with the conventional system at rated speed with different loads and presented in Table 12. From Table 12, it is clear that in conventional SWECS, the THDs of inverter output voltage and current without filter are very high and violate permissible limits. Hence, filter is required in order to bring these THD's within permissible limits. After filter both THD's comes within permissible limits. Whereas, in proposed SWECS, the THDs of VFT output voltage and current are very less than the conventional system and are within permissible limits. Thus, here no filter is required. Further, the efficiency of both conventional and proposed methods increases with the increase in load from 1 kW, gets a maximum value at 2 kW and then starts decreasing with the further increase in load. Moreover, the efficiency of proposed method is slightly greater than the conventional system which is around 0.5%. Hence from Table 12, it is concluded that the proposed method is having slightly greater efficiency with almost negligible THD. 6. Experimental analysis 6.1. Experimental setup Fig. 10 shows the block diagram of practical setup. The dc motor is mechanically coupled to the SG in order to simulate the wind
Table 10 Power Fed To Load at 1200 rpm SG Input Speed (Frequency ¼ 40 Hz) without Excitation Control. Load (kW)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
fL (Hz)
PL (W)
% h ¼ ½PL =ðPs þ Pd Þ 100
1 2 3 4
861.8 1684 2525 3411
214.23 415.63 618.26 825.61
300 300 300 300
0.8 0.8 0.8 0.8
397.6 396.4 394.7 392.8
50 50 50 50
988.2 1964 2921 3858
91.84 93.54 92.93 91.06
114
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Table 11 Power Fed To Load at 1200 rpm SG Input Speed (Frequency ¼ 40 Hz) with Excitation Control. Load (kW)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
fL (Hz)
PL (W)
% h ¼ ½PL =ðPs þ Pd Þ 100
1 2 3 4
872.2 1714 2593 3537
216.8 423.17 634.92 856.08
300 300 300 300
0.8 0.8 0.8 0.8
400.0 400.0 400.0 400.0
50 50 50 50
1000 2000 3000 4000
91.83 93.58 92.94 91.05
phase, 10.669 kVA). The auto-transformer #1 is used for matching the stator voltage of WRIM to the load voltage. The DDM is energized by SG via DDM controller. The controller of DDM converts the 3 phase ac supply to controlled dc supply and feeds the dc supply to DDM. The pictorial view of the experimental setup is shown in Fig. 11. Here, WRIM is mechanically coupled to a dc motor #1 (shunt type), where dc motor #1 is working as DDM. 6.2. Experimental results
Fig. 8. Variation of load power with and without excitation control w.r.to the DDM power at 1200 rpm.
In experimental analysis, the SG input speed is varied while the load is kept constant i.e. 1.2 kW, 400 V rated voltage and 50 Hz rated frequency. The experimental results observed are given in Table 13. The experimental results obtained in Table 13 shows that with decrease in SG speed the SG output voltage, SG frequency and the SG output power decreases but the DDM power increases and thus the efficiency of the SWECS remains almost constant for all operating conditions. Hence the experimental results validate the simulation results. In Table 13 the load is 1.2 kW but it is showing the output power more than load power. This is because the output power is sum of auto-transformer #1 power and load power. The variations of SG power, DDM power, output power and % efficiency with the SG input speed of Table 13 are plotted in Fig. 12. From Fig. 12 it is again clear that for a constant load, with the decrease in SG input speed the SG power decreases while the DDM power increases but the output power and the efficiency remains almost constant. 7. Cost analysis
Fig. 9. Efficiency of proposed SWECS with and without excitation control w.r.to the load power at 1200 rpm.
turbine. The SG supplies power to auto-transformer #2 (3 phase, 415 V, 22.793 kVA), which steps down the SG voltage to rated value of rotor voltage and transfer the power to rotor of the WRIM (3 phase, 2.2 kW, 400 V, 50 Hz, 1440 rpm). The DDM supplies mechanical power to rotor of the WRIM in order to control the power flow from rotor to stator side of WRIM, such that the transferred power at stator side is having rated frequency. The stator of WRIM supplies power to isolated loads through auto-transformer #1 (3
The cost analysis of conventional and proposed systems have been performed for 3 kW rating and shown in Table 14. The data for the cost analysis has been taken from Ref. [21]. From Table 14, it is clear that the proposed system is cheaper than the conventional system by 40 Euro. 8. Conclusion A new method for feeding an isolated load from synchronous generator (SG) based wind power generation has presented in this paper. The proposed method uses variable frequency transformer (VFT) for standalone wind energy conversion system (SWECS). For this, digital simulation models of the conventional method and the proposed method have been developed under MATLAB-Simulink environment. Few studies on power flow from SG to the isolated
Table 12 Comparison of Proposed Method With Conventional Method At 1500 rpm. Load (kW)
1 2 3 4
Conventional method of SWECS
Proposed method of SWECS
% THD of VL without filter
% THD of IL without filter
% THD of VL with filter
% THD of IL with filter
%h
% THD of VL
% THD of IL
%h
67.93 68.15 67.96 68.06
17.48 16.37 17.76 16.68
3.96 2.75 2.91 2.72
3.96 2.75 2.71 2.69
91.32 93.11 92.42 91.01
0.01 0.01 0.01 0.01
0.01 0.01 0.01 0.01
91.83 93.63 93.05 91.07
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
115
Fig. 10. Block diagram of the experimental setup.
Fig. 11. The experimental setup.
Table 13 Experimental Results for Power Fed to 1.2 kW Load. Ns (rpm)
Vs (V)
fs (Hz)
Ps (W)
Pd (W)
Nrm (rpm)
S ¼ ½ð1500 Nrm Þ=1500
VL (V)
fL (Hz)
Output power, Po (W)
% h ¼ ½PL =ðPs þ Pd Þ 100
972 890 845 799 753 703
228.4 199.3 183.3 167.0 150.6 133.0
32.4 29.7 28.2 26.7 25.1 23.5
1199.6 1107.6 1070.4 1002.8 977.2 925.2
410.38 468.00 515.95 591.56 630.72 691.95
528.1 610.7 655.7 701.9 747.5 797.8
0.65 0.59 0.56 0.53 0.5 0.47
400.15 400.14 400.14 400.12 400.53 400.54
50.039 50.074 50.044 50.016 50.014 50.014
1224 1224 1224 1236 1252 1260
76.03 77.68 77.16 77.52 77.86 77.91
Fig. 12. The variation of SG power, DDM power, output power and % efficiency with the SG input speed at 1.2 kW load.
loads by conventional method at rated SG speed i.e. 1500 rpm have been carried out using MATLAB-Simulink model. A number of studies on power flow from SG to the isolated loads have been carried out by MATLAB-Simulink model of proposed method without and with excitation control under different values of SG input speed. A number of quantities such as SG output power, dc drive motor (DDM) mechanical power, load power and efficiency have been obtained. From the simulation results, it is evident that with decreases in SG input speed, the SG output power decreases but the DDM power increases in order to supply power to the load at constant frequency and voltage. Even though with the increase in DDM power, the load power reduces slightly but the efficiency is maintained constant. Then the power fed to variable load by the SG based SWECS at fixed speed i.e. 1200 rpm without and with excitation control has been analyzed using simulation model of proposed method. From here it is concluded that the power fed to load increases linearly with the DDM power. The load power as well as efficiency with excitation control are slightly higher than without excitation control. Both increases with the increase in load, reaches to the maximum value and then starts decreasing with the further
116
F.I. Bakhsh, D.K. Khatod / Renewable Energy 86 (2016) 106e116
Table 14 Comparison of cost (Euro). Equipments
Iron (kg) Copper (kg) Total (kg) SG active material cost (Euro) SG construction cost (Euro) WRIM active material cost (Euro) DDM active material cost (Euro) (WRIM þ DDM) construction cost (Euro) Converter cost (Euro) Total cost (Euro)
Direct-drive SG Conventional system
Proposed system SG
WRIM
DDM
Total
32.5 12.6 45.1 287 160 e e e 120 567
32.5 12.6 52.08 287 160 30 10 40 e 527
4.03 1.21
1.34 0.40
37.87 14.21
increase in load. Moreover, the THD of output voltage, THD of output current and efficiency of the proposed method have been compared with those of the conventional method. From the comparison results, it is obtained that the proposed system does not produce harmonics and having slightly higher efficiency than the conventional system. Moreover, in order to validate the proposed method a practical analysis has been performed. The response characteristics of efficiency and different powers under various SG input speed are achieved. The achieved practical results validate the simulation results. Hence, from simulation and practical results it can be concluded that the proposed method can be used as an SWECS for SG based wind energy generation system. Further, cost analysis has also been carried out for the proposed system and it has been compared with that of conventional system. From the cost analysis, it is observed that the proposed system is cheaper than conventional system.
[9]
[10]
[11]
[12]
[13]
[14]
[15]
References [1] L.A. de S. Ribeiro, O.R. Saavedra, S.L. Lima, J.G. de Matos, G. Bonan, Making isolated renewable energy systems more reliable, Renew. Energy 45 (2012) 221e231. [2] F.I. Bakhsh, M.M. Shees, M.S.J. Asghar, Performance of wound rotor induction generators with the combination of input voltage and slip power control, Russ. Electr. Eng. 85 (6) (June 2014) 403e417. [3] S. Sharma, B. Singh, Asynchronous generator with battery storage for standalone wind energy conversion system, IEEE Trans. Ind. Appl. 50 (4) (July/ August 2014) 2760e2767. [4] A.J.G. Westlake, J.R. Bumby, E. Spooner, Damping the power-angle oscillations of a permanent-magnet synchronous generator with particular reference to wind turbine applications, IEE Proc.-Electr. Power Appl. 143 (3) (May 1996) 269e280. [5] M.M. DJesus, D.S. Martin, S. Arnaltes, E.D. Castronuovo, Optimal reactive power allocation in an offshore wind farms with LCC-HVdc link connection, Renew. Energy 40 (2012) 157e166. [6] H. Ghoddami, M.B. Delghavi, A. Yazdani, An integrated wind-photovoltaicbattery system with reduced power-electronic interface and fast control for grid-tied and off-grid applications, Renew. Energy 45 (2012) 128e137. rdenas, R. Pen ~ a, P. Wheeler, J. Clare, C. Juri, Control of a matrix converter [7] R. Ca for the operation of autonomous systems, Renew. Energy 43 (2012) 343e353. [8] P. Moutis, S.A. Papathanassiou, N.D. Hatziargyriou, Improved load-frequency
[16]
[17]
[18]
[19]
[20]
[21]
control contribution of variable speed variable pitch wind generators, Renew. Energy 48 (2012) 514e523. C.N. Bhende, S. Mishra, S.G. Malla, Permanent magnet synchronous generatorbased standalone wind energy supply system, IEEE Trans. Sustain. Energy 2 (4) (October 2011) 361e373. C. Mi, M. Filippa, J. Shen, N. Natarajan, Modeling and control of a variablespeed constant-frequency synchronous generator with brushless exciter, IEEE Trans. Ind. Appl. 40 (2) (March/April 2004) 565e573. N. Mendis, K. Muttaqi, An integrated control approach for standalone operation of a hybridised wind turbine generating system with maximum power extraction capability, Electr. Power Energy Syst. 49 (2013) 339e348. E. Larsen, R. Piwko, D. McLaren, D. McNabb, M. Granger, M. Dusseault, LP. Rollin, J. Primeau, Variable Frequency Transformer e a New Alternative for Asynchronous Power Transfer, Canada Power, Toronto, Ontario, Canada, September 28-30, 2004. P. Doyon, D. McLaren, M. White, Y. Li, P. Truman, E. Larsen, C. Wegner, E. Pratico, R. Piwko, Development of a 100 MW Variable Frequency Transformer, Canada Power, Toronto, Ontario, Canada, September 28-30, 2004. F.I. Bakhsh, D.K. Khatod, Application of variable frequency transformer (VFT) for grid interconnection of PMSG based wind energy generation system, Sustain. Energy Technol. Assess. 8 (2014) 172e180. L. Wang, S.-R. Jan, C.-N. Li, H.-W. Li, Y.-H. Huang, Y.-T. Chen, Analysis of an integrated offshore wind farm and seashore wave farm fed to a power grid through a variable frequency transformer, in: IEEE Power and Energy Society General Meeting, San Diego, CA, 24-29 July, 2011. L. Wang, L.-Y. Chen, Reduction of power fluctuations of a large-scale gridconnected offshore wind farm using a variable frequency transformer, IEEE Trans. Sustain. Energy 2 (3) (July 2011) 226e234. T. Geetha, V. Jayashankar, Variable frequency transformers for increased wind penetration, in: Joint International Conference on Power System Technology and IEEE Power India Conference (POWERCON 2008), New Delhi, 12-15 Oct., 2008. F.I. Bakhsh, D.K. Khatod, A novel method for grid integration of synchronous generator based wind energy generation system, in: IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES), IIT Bombay, India, 16th to 19th Dec., 2014. A. Merkhouf, P. Doyon, S. Upadhyay, Variable frequency transformerdConcept and electromagnetic design evaluation, IEEE Trans. Energy Convers. 23 (4) (December 2008) 989e996. F.I. Bakhsh, D.K. Khatod, Digital simulation of VFT applications between power system networks, in: Fourth International Joint Conference on Advances in Engineering and Technology, NCR, India, 13-14 Dec., 2013, pp. 60e74. H. Polinder, F.F.A. van der Pijl, G.-J. de Vilder, P. Tavner, Comparison of directdrive and geared generator concepts for wind turbines, IEEE Trans. Energy Convers 21 (3) (Sept. 2006) 725e733.