- Email: [email protected]
Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting
Engineering Science and Technology, an International Journal xxx (xxxx) xxx
Contents lists available at ScienceDirect
Engineering Science and Technology, an International Journal journal homepage: www.elsevier.com/locate/jestch
Full Length Article
Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting A. Aranizadeh a, A. Zaboli a, O. Asgari Gashteroodkhani b, B. Vahidi a,⇑ a b
Electrical Engineering Department, Amirkabir University of Technology, Tehran 1591634311, Iran Department of Electrical and Biomedical Engineering, University of Nevada, Reno, USA
a r t i c l e
i n f o
Article history: Received 6 September 2018 Revised 10 June 2019 Accepted 29 August 2019 Available online xxxx Keywords: Wind turbine Ultra-capacitor Harvested energy Microgrid Wind speed forecasting
a b s t r a c t Wind energy source has a complex control situation because of dependence of its torque and output power on wind speed and its fluctuations. Based on this, in order to improve its control condition and dynamic efficiency, when connecting to the microgrid, ultra-capacitor which has a fast charging and discharging speed is used. Furthermore, the maximum energy derived from wind turbine and ultracapacitor by the microgrid is of high importance which must be considered besides decreasing output power fluctuations. In this paper, for increasing the harvested energy, the Wind Speed Forecasting (WSF) model is used. So, the control method is applied by using WSF. In the proposed method, the gained energy is more than the lost energy. In fact, we increased harvested energy using a predictive control method. The considered predictive control is applied to the induction generator rotational speed variations. The considered wind turbine model in this paper produces an active power of 50 kW and is a variable speed induction generator (VSIG) with an apparent power of 50 kVA. All of the simulations are performed in MATLAB/SIMULINK software. Ó 2019 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Recently, European countries have been concerned about climate changes and have decided to reduce greenhouse gases. In this manner, they have adopted new policies to produce cleaner energies in order to reduce the produced CO2 until 2020 by 20% and until 2050 by 50%. In this regard, increase of renewable energies and hybrid generation would be the way to reach this aim [1]. Less pollution at consumption centers, easier and less hazardous as well as more efficient transmission, controllability at generation and consumption centers and more flexibility in transformation to other energy types at consumption centers are among the factors in making electricity more attractive than other types of energy [2]. Fossil fuels such as coal, oil and natural gas and nuclear energy are non-renewable energies and their available resources are also limited. Therefore, finding new energy sources are one of the important concerns of human in recent century [3,4]. Increasing use of energy consumption, high and ever increasing cost and non-renewable nature of fossil fuels as well as bad situation of global environment have caused much attraction to environment-
⇑ Corresponding author. E-mail address: [email protected] (B. Vahidi). Peer review under responsibility of Karabuk University.
friendly energy resources [5]. One of the most important resources of renewable energy production is wind source. The main problems with the energy from wind turbines are the low energy density, output power severe oscillations and uncertainty in the obtained energy from them. In order to energize all the loads, the producers must access all the required predictions about the power and energy from these resources. These predictions must have high accuracy so that no problem arises for power grid. To predict the instantaneous wind speed and increase the obtained energy from wind turbines, short term prediction must be used to control the dynamic condition of wind turbines [6]. In [7], power grid and market conditions are analyzed by predicting hourly energy from wind considering the increase of wind energy penetration level to the power grid. In [8], the authors determine the size of energy storage system in order to increase the wind energy penetration level in power network considering the grid frequency fluctuations. According to this reference, frequency deviation of power systems caused by grid-connected wind power fluctuations. In [9], calculation of the wind turbine maximum output energy in active distribution networks is elaborated and a multi-period optimal power flow analysis is proposed. In [10], the wind turbine output power fluctuations is decreased by utilizing a new method through continuous prediction of wind speed. In this method, the wind speed in every second and continuously is predicted and by applying these predicted fluctuations to
https://doi.org/10.1016/j.jestch.2019.08.006 2215-0986/Ó 2019 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
2
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
the grid controllers, the wind turbine output power is improved. The method proposed in [11] decreases the wind turbine output power fluctuations utilizing artificial neural network (ANN) and capacitor system controller. In this reference, an ANN controller is considered for controlling the DC-bus connected Energy Capacitor System (ECS). The authors in [12] lessen wind turbine output power fluctuations using prediction and supervisory control unit based on energy storage system. In [13], the transferred energy from wind turbine to the grid is increased using short term wind speed prediction. In [14], wind energy is increased through energy storage and demand response. In [15], scheduling of microgrid operation including wind turbines, photovoltaic systems, energy storages and responsive loads based on stochastic is analyzed. The authors in [16] elaborate the effect of short and long term prediction on the performance of energy storage system. In [17], the author presents maximizing profit of wind-battery supported power station based on wind power and energy price forecasting. In [1,18–20], precise models of wind turbine controllers design are presented in order to improve dynamic condition. Reference [21] dynamically simulates the wind turbine and hybrid energy systems. Considering the aforementioned explanations, in the second section of this paper, design and modelling of wind turbine, ultra-capacitor energy storage system and the procedure of connecting wind turbine and ultra-capacitor to the microgrid and also modelling of the wind speed forecasting system are elaborated. In the third section, some explanations about the control process of wind turbine blade’s angle control and also the specifications of the control system of connection of ultra-capacitor to the DC link are presented. In section four, increase of obtained energy using the predictive control is conducted. In section five, simulation results for the condition with and without predictive control are demonstrated. Finally, the paper is concluded in section six. 2. Design and modeling 2.1. Wind turbine specifications In general condition, the mathematical formulation of extraction power of wind energy using momentum theory is expressed by (1) and (2) [22]:
Pwind ¼
1 1 qAv 3 ¼ qpR2 v 3 2 2
ð1Þ
2.2. Ultra-capacitor specifications Different ultra-capacitor models have been presented up to now. Some of these models are merely for analyzing the ultracapacitor itself and obtaining thermal and electrical characteristics of it [23]. These models are only appropriate for analyzing the characteristics of ultra-capacitors and are not efficient for analyzing in electrical grids due to their complexity. An example of extended model of ultra-capacitor is shown in Fig. 1a. This ultracapacitor model is analyzed in [11] and its aim is to smooth the output power of wind turbine. The proposed model in [24] is presented in Fig. 1b. The same model is used in this paper. In this paper, a 2F (Farad) ultra-capacitor with an internal resistance of 1.5 mO is used. Each of the ultra-capacitor’s cells has the initial voltage of 50 V which is placed in 16 in series to reveal an initial voltage of 800 V. The capacity of the series is 0.125F. 2.3. System specifications of the grid System specifications of the grid with the connection of wind turbine and ultra-capacitor energy storage system are demonstrated in Fig. 2 [10,25]. As shown, wind turbine and ultracapacitor system are connected to a microgrid with a weak network. This microgrid is severely reacting against power fluctuations and transferred energy. Based on this, controlling power and output energy of wind turbine in this condition is of high importance. In Fig. 2, the combination of wind turbine, ultracapacitor energy storage and microgrid supply AC and DC loads. 2.4. The prediction system specifications The prediction method depends on the available data and data’s time period. In wind turbines dynamic control, data are wind speed and data’s time periods are the seconds. On this bases, the forecasting method of wind speed in this paper is the linear forecasting method [10,26]. This method can properly handle data of wind speed in a short time period. This prediction control method is demonstrated in Fig. 3. According to this figure, having the previous changed time (t0) and the changed predicted time (t0 + T), the variation rate is obtained as expressed by (4). Afterwards, having the rate from (4), the line equation obtained in (5) can be calculated.
V ðtÞ ¼ mððt 0 þ T Þ t0 Þ þ Vðt 0 Þ ) V ðt Þ ¼ mT þ Vðt0 Þ
where, q is air density in kg/m3, A is the swept area of the rotor in m2, R is the radius of swept area in meter and v is the wind speed in m/s. Therefore, the wind turbine obtainable power can be expressed by (2):
Pt ¼
1 qpR2 v 3 C p ðk; hÞ 2
ð2Þ
R b3 Cb2
In this equation, CP is the power coefficient of the wind turbine. This value is related to tip speed ratio (TSR) and wind turbine blade pitch angle. The ratio of tip speed ratio to wind speed is demonstrated by k which is dimensionless and is expressed by (3):
TSR ¼ k ¼
R xr V wind
R b1
R b2
C b3
ð4Þ
Cb1
Ucap
(a)
Ri
ð3Þ
where, Vwind is the wind speed, R is the blade radius and xr is the angular speed at the wind turbine blades. The wind speeds lower than the rated wind speed utilize the Maximum Power Point Tracking (MPPT) strategy through controlling rotor speed and making TSR constant in its optimum value. Furthermore, for the wind speeds more than rated wind speed, the strategy of limiting power is gained via controlling blades pitch angle.
Rp
Uc
C
Ucap
(b) Fig. 1. Ultra-capacitor modelling (a) The extended equivalent model of ultracapacitor based on [11] (b) The compact equivalent model of ultra-capacitor based on [24].
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
3
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
3.1. Wind turbine blade pitch angle control system specifications
DC Bus Wind Turbine
The control method of wind turbine that should be applied to wind turbine blades is demonstrated in Fig. 4. Blade pitch angle controller is activated when the generated power is more than generator rated power. Furthermore, blade pitch angle changes has some inertia which depends on the blade’s variation time constant value and the initial value of wind turbine changes. Furthermore, wind turbine blade’s pitch angle variation must be in a specified limit [28]. The pitch angle has limitation to change. In this paper, the maximum changes for pitch angle are 3°/s to 10°/s and delay time is 0.02 s. These limitations depend on size and characteristics of wind turbine. The PI control parameters are P = 3 and I = 30.
AC / DC
Compensator Copacitor
Energy Storage
DC / DC
DC Load
DC / AC
AC Load
DC / DC
Convertor
Other sources
3.2. Ultra-capacitor connection to DC link control system specifications
Micro grid
Wind Speed
Fig. 2. Microgrid (connection of wind turbine, energy storage system, weak source and loads).
Wind speed value to be predicted (future sample)
Wind speed samples
...
Modelling window
...
t0-mT
t 0-T t0 t0+T
time
Vt0+T
D1 ¼
Vt0+
V UC V dc
ð6Þ
Also, the duty ratio of switch S2 in boost mode is obtained through (7):
Vt0 t0
t0+
D2 ¼ 1 D1
t0+T
Fig. 3. Block diagram of Wind Speed Forecasting (WSF) model.
m¼
Fig. 5 shows a schematic of ultra-capacitor and converter control system connected to DC link. As shown, the ultra-capacitor consists of a capacitor bank which is connected to the DC link through a DC/DC converter [29]. The DC/DC converter has two separated IGBT switches with separately operating switching controllers. Switches S1 and S2 are switched in this controller in such a way to regulate Pg for power transmission between microgrid and wind turbine. In this structure, the DC/DC converter operates in two different modes of buck mode and boost mode which depend on the switching method of IGBT switches. When switch S1 is open, the DC/DC converter goes to the boost mode. In this mode, the ultra-capacitor acts as a source which transmits active power to the grid and by this operation, ultra-capacitor voltage (VUC) decreases. When switch S2 is open, the DC/DC converter conducts current in the buck mode. In this mode, ultra-capacitor performs as active power sink and by this operation, VUC increases. Based on this and according to the aforementioned explanation, the duty ratio of switch S1 in buck mode is estimated by (6):
Vðt 0 þ TÞ Vðt0 Þ T
ð5Þ
Having this line equation, the predicted data in any instance of time can be observed [10,27]. As shown, this prediction control method is appropriate for those inputs with constant rate changes and the predicted value follows a determined procedure. In this paper, the wind speed data must be forecasted and the forecasted values must be applied to the predicting control system. In fact V(t0 + T) is applied to predictive control instead of V(t0). Also amount of T is calculated according to wind turbine size and inertia. 3. Wind turbine and ultra-capacitor control system specifications In order to control wind the turbine system containing the ultra-capacitor to the grid shown in Fig. 2, two control methods i.e.,wind turbine blade pitch angle control system and ultracapacitor control system are analyzed.
ð7Þ
The on/off period of switches is obtained through the output of current controller and by comparing the reference current with the present current of the ultra-capacitor. The saturation block is used to regulate output signal of the controller in the range of 0–1. Then, this value is compared to the carrier signal value with 10 kHz frequency in this paper and is signed with the sign function. Furthermore, ultra-capacitor voltage level is controlled at every instance so that it is not out of the allowed predefined limits. The PID control parameters are P = 50, I = 1 and D = 0.3. 4. Increasing harvested energy using predictive control It should be noted that when the wind speed drastically changes, the generator rotational speed should change so as to Delay time
Pref (pu)
- +
PI Controller
Pitch angle WT system
Pmeasured (pu)
(a)
Fig. 4. Block diagram for pitch angle control.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
4
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
S1 DC Link
Less
+
-
+
S2
Iref
+ -
PID Controller
Imeasured
Saturation block
- + Min/Max voltage limit Sign function
Repeating Sequence
-
Ultra-capacitor Bank
(U+1)/2
Command
Charge/Discharge mod
Fig. 5. Control system of connectivity of ultra-capacitor to DC link.
keep power coefficient in its maximum value. However, due to the generator inertia, the change in the generator rotational speed is slow. In order to overcome the inertia, in the proposed method, the process of generator rotational speed control begins before the instance of the wind speed change. According to the Fig. 6a, in tvirtual, the generator rotational speed is changed prior to tactual where the actual wind speed is changed. In [tvirtualtactual], the output power decreases in correspondence to the change in tip speed ratio, and in turn, it leads to loss of energy in the proposed method in comparison with the conventional method. However, after tactual, since the generator rotational speed inertia is overcome, the output power reaches its maximum in less time, and it gains more energy compared to the conventional state. By noting to the Fig. 6a, it can be observed that in the proposed method, the gained energy after tactual is more than the lost energy in the [tvirtualtactual]. To elucidate more on why the gained energy is more than the lost energy in the proposed method, by considering the Fig. 6b, it can be seen that in the conventional method, when the wind speed
rapidly changes, tip speed ratio considerably changes that makes the operating point be near the rapid slope of the CP-k curve, and consequently the output power decreases to a great extent. However, in the proposed method, since the operating point is near the slow slope of the CP-k curve, the output power decreases in a lesser extent. Accordingly, total gained energy in the proposed method is more than that of conventional method. Fig. 7 shows a block diagram of the proposed method. As shown, using wind speed forecasting, in an instance before the change in actual wind speed, the generator rotational speed controller operates. It must be considered that time difference between application of the controller and the wind speed change depends on the size of wind turbine. The bigger the size of wind turbine, the bigger inertia of induction generator rotational speed variations is obtained, and thus the predictive controller should be applied earlier.
5. Simulation results Conventional Method
P
Proposed Method
P2 Gained Energy in Proposed Method
P1
Lost Energy in Proposed Method
tvirtual
CP
t
tactual (a)
∆CP-Proposed
Conventional Method
To analyze the simulation results, a 50 kW wind turbine with an induction generator is utilized. The specifications of wind turbine and induction generator are presented in Tables 1 and 2 [30], respectively. The ultra-capacitor storage system specifications are also presented in Table 3 [21]. The wind turbine model in this section is a variable speed doubly-fed induction generator (DFIG) and perturb and observe (P&O) mode for MPPT technique is used in this paper. This simulation is implemented in MATLAB/SIMULINK software. Fig. 8 shows the SIMULINK model of the proposed system in MATLAB. This figure consists of wind turbine model, ultracapacitor model, load model and other sources connected to each other with convertors. The Ultra-capacitor controller with switching and pitch angle controller is explained in Sections 3.2 and 3.1, respectively. Also, predicted wind speed is applied to generator speed.
Propo sed Proposed Method M t ∆CP-Conventional
Supercapacitor energy storage system
Vpresent
λ (b) Fig. 6. A comparison between the proposed and conventional methods (a) in harvested energy (b) in power coefficient.
WSF
Vpredicted
Rotational speed controller by prediction
Omegapredicted
Pitch angle Pitch angle controller
WT system Power
Micro Grid
Vpresent
Fig. 7. Block diagram of variable speed wind turbine in the proposed method.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx Table 1 Parameters of wind turbine [29]. Parameter
Value
Air density Rotor diameter Rated power Cut-in/cut-out speed Rated wind speed
1.225 kg/m3 15 m 50 kW 3/25 m/s 12 m/s
Table 2 Parameters of induction generator [29]. Parameter
Value
Rated power Stator voltage/frequency Stator resistance Rotor resistance Stator leakage inductance Rotor leakage inductance Mutual inductance
50 KVA 380 V/60 Hz 0.016 pu 0.015 pu 0.06 pu 0.06 pu 3.5 pu
Table 3 Parameters of ultra-capacitor [21]. Parameter
Value
Ultra-capacitor voltage NrS/NrP Ri Ultra-capacitor Upper voltage limit Lower voltage limit
800 V/each 50 V 16/1 1.5 mO 0.125F/each 2F 850 V 750 V
For analyzing this issue and applying it to the wind turbine model with ultra-capacitor storage system, wind speed is applied to the wind turbine according to Fig. 9a. As can be seen from Fig. 9a, the wind speed is 8.5 m/s at first and after 2.5 s, it reduces to 7 m/s. At this wind speed,
5
wind turbine output power becomes less than load power which is considered 23 MW and thus causes the discharge of ultra-capacitor. Afterward, the wind speed increases up to 9.5 m/s. At this wind speed, the wind turbine generated power becomes more than load power which leads to charge of ultracapacitor and thus causes the ultra-capacitor to reach its charging limitation.
5.1. Reach of ultra-capacitor to the charge limitation without using wind speed prediction In this part, wind speed prediction is not used and generator speed control is in such a way that as the wind speed changes the generator speed also alters for harvesting maximum power from wind. The generator speed in this part is shown in Fig. 9b. As shown, wind speed fluctuations have some inertia that considering small size of the studied wind turbine in this paper, this inertia is not very high. Anyway, in larger wind turbines this inertia is remarkable. As shown in Fig. 9b, as wind speed changes at 2.5 s, the generator rotational speed changes to receive the maximum harvesting power. The voltage of ultra-capacitor in this part is shown in Fig. 9c. Before 2.5 s, considering low wind speed, ultra-capacitor is in discharge mode. Increase in wind speed and as a consequence, increase in wind turbine produced power puts the ultra-capacitor in charge mode which is obviously observable in the voltage of ultra-capacitor. Continuation of charging makes the ultra-capacitor to reach its charging limitation at 4.55 s. When the ultra-capacitor reaches its charge limitation, ultra-capacitor storage system becomes separated from the grid and the wind turbine with the power grid supply the load demand. Also, with the separation of ultra-capacitor, the power grid generated power increases and it increases power generation cost. Fig. 9d demonstrates wind turbine output power, injected or received power from ultra-capacitor and also total power of wind turbine and ultra-capacitor for this case.
Fig. 8. SIMULINK model of the proposed system in MATLAB.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
6
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx 1.9
Rotational Speed (rad/s)
Wind Speed (m/s)
9.5 9 8.5 8 7.5 7 0
1
2
3
4
1.8 1.7 1.6 1.5 1.4 0
5
1
2
4
5
(b)
(a) 860
40 P
850
Total
30
Power (kW)
Capacitor Voltage (V)
3
time (s)
time (s)
840 830 820 810
=P
U.C.
+P
P
W.T.
W.T.
20 P
Load
10 P
U.C.
0
800 790 0
1
2
3
4
-10 0
5
1
2
3
4
5
time (s)
time (s)
(d)
(c)
2
860
Capacitor Voltage (V)
Rotational Speed (rad/s)
Fig. 9. Without using wind speed prediction (a) Wind speed variations applied to the wind turbine (b) Generator rotational speed variations (c) Ultra-capacitor voltage (d) Wind turbine output power, power from ultra-capacitor and also total power of wind turbine and ultra-capacitor as well as load power.
1.9 1.8 1.7 1.6 1.5 1.4 0
1
2
3
4
850 840 830 820 810 800 790 0
5
1
2
3
4
5
time (s)
(b) 40
850
30
P
Total
840
Power (kW)
Capacitor Voltage (V)
(a) 860
Without Prediction
830 With Prediction
820 810 3
3.5
4
4.5
=P
W.T.
+P
P
W.T.
U.C.
20 P
Load
10
P
U.C.
0
5
-10 0
1
2
3
time (s)
time (s)
(c)
(d)
4
5
Fig. 10. With using wind speed prediction (a) Generator rotational speed variations (b) Ultra-capacitor voltage (c) Comparing ultra-capacitor voltage with and without predictive control (d) Wind turbine output power, power from ultra-capacitor and also total power of wind turbine and ultra-capacitor as well as load power.
5.2. Reach of ultra-capacitor to the charge limitation using wind speed prediction Contrary to the previous section, generator speed control in this section is in such a way that before wind speed variation, generator speed varies for receiving maximum wind power. In fact, in this section, with predicting wind speed in future, generator speed controller changes in a way that it applies the occurred fluctuations of future in a sooner time. According to the aforementioned matters, generator speed is according to Fig. 10a. Generator speed variations occur at 2 s (0.5 s sooner than wind speed actual change). Ultra-capacitor voltage is also shown in Fig. 10b. In this part, at 2.5 s considering the wind speed, the ultra-capacitor is in discharge
mode and with the wind speed increase and therefore the increase of generated power from wind turbine, the ultra-capacitor gets into charge mode. According to Fig. 10b, the ultra-capacitor has reached to its charge limitation at 4.59 s. As shown, in this method, the ultracapacitor has reached to its charge limitation 0.04 s later. The comparison between Figs. 9c and 10b is presented in Fig. 10c. This figure obviously shows that by applying the predictive control, the ultra-capacitor has reached to its charge limits later. Fig. 10d demonstrates wind turbine output power, injected or received power from ultra-capacitor and also total power of wind turbine and ultra-capacitor for the mode of control with wind speed control.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
34
Power (kW)
32 30
P
without Predictive Control
P
with Predictive Control
Total Total
[5] [6]
28 [7]
26 24 22 3
[8]
3.5
4
4.5
5
time (s) Fig. 11. Comparison between the transmitted power from the grid to the load in both cases of without and with predictive control.
Fig. 11 demonstrates the generated power by wind turbine and ultra-capacitor in both conditions of with and without predictive control. According to this figure, it is obvious that sum of wind turbine and ultra-capacitor generated power with wind speed prediction control is more than the case with no wind speed prediction control. This has occurred with what has been thoroughly explained in Fig. 10c. The result is the increase of wind turbine and ultra-capacitor system generated power. According to Fig. 11, the injected power from the grid to the load in the state of predicted control has increased later. Therefore, the generated energy by wind turbine and ultra-capacitor for energizing the load in the case with predictive control is almost 296 J more than the case without predictive control. The amount of increased energy depends on the turbine size so that an increase in the turbine size increases the harvested energy.
[9]
[10]
[11]
[12]
[13]
[14] [15]
[16]
[17]
6. Conclusion
[18]
In this paper, a WSF method is applied to monitor the future wind speed data. Afterward, by applying this control method on wind turbine, the harvested energy from wind turbine and ultracapacitor energy storage is increased and the microgrid condition is improved. In this paper, all of modelling is carried out in MATLAB/SIMULINK software. By running simulations, it is observed that by utilizing WSF method for wind speed data and online monitoring of wind speed in the future and then applying predictive control to induction generator rotational speed, the obtained energy is increased by 296 J. Therefore, when this method is used, the produced energy from large-scale wind farm is increased and the network costs are reduced. Also, wind turbine with more inertia needs more prediction time. This paper is a proof of this method. In the future work, the optimal value of T can be found for different size of turbine.
[19]
[20]
[21] [22] [23]
[24]
[25]
[26]
References [27] [1] L.Y. Pao, K.E. Johnson, Control of Wind Turbines’, IEEE Control Syst. 31 (2) (2011) 44–62. [2] A. Forooghi Nematollahi, A. Dadkhah, O. Asgari Gashteroodkhani, B. Vahidi, Optimal sizing and siting of DGs for loss reduction using an iterative-analytical method, J. Renewable Sustainable Energy 8 (5) (2016) 055301. [3] T. Burton, D. Sharpe, N. Jenkins, et al., Wind Energy Handbook’, Wiley Press, 2001. [4] O. Asgari Gashteroodkhani, B. Vahidi, A. Zaboli, Time-time matrix z-score vector-based fault analysis method for series-compensated transmission lines,
[28] [29] [30]
7
Turkish Journal of Electrical Engineering & Computer Sciences 25 (2017) 2647– 2659. ‘International Energy Outlook, Energy Information Administration (EIA)’, http://www.eia.doe.gov/iea. Y. Izumi, A. Pratab, Y. Kinjyo, et al., A fuzzy-based output power smoothing of WECS using short-term ahead prediction of wind speed, in: IEEE International Conference on Fuzzy Systems (FUZZ), Taipei, 2011, pp. 2414–2419. M. Hedayati-Mehdiabadi, J. Zhang, K.W. Hedman, Wind power dispatch margin for flexible energy and reserve scheduling with increased wind generation, IEEE Trans. Sustainable Energy 6 (4) (2015) 1543–1552. Y. Liu, W. Du, L. Xiao, et al., A method for sizing energy storage system to increase wind penetration as limited by grid frequency deviations, IEEE Trans. Power Syst. 31 (1) (2016) 729–737. P. Siano, P. Chen, Z. Chen, et al., Evaluating maximum wind energy exploitation in active distribution networks, IET Gener. Transm. Distrib. 4 (5) (2010) 598– 608. A. Aranizadeh, B. Vahidi, A. Rahiminejad, Wind turbine power output smoothing in microgrid using ultra-capacitor with continuous wind speed forecasting and online supervisory control, J. Renewable Sustainable Energy 8 (3) (2016) 1–17. S.M. Muyeen, H.M. Hasanien, J. Tamura, Reduction of frequency fluctuation for wind farm connected power systems by an adaptive artificial neural network controlled energy capacitor system, IET Renew. Power Gener. 6 (4) (2012) 226–235. F. Islam, A. Al-Durra, S.M. Muyeen, Smoothing of wind farm output by prediction and supervisory-control-unit-based FESS, IEEE Trans. Sustainable Energy 4 (4) (2013) 925–933. J. Kiviluoma, P. Meibom, A. Tuohy, et al., Short-Term Energy Balancing with Increasing Levels of Wind Energy, IEEE Trans. Sustainable Energy 3 (4) (2012) 769–776. P.F. Ribeiro, B.K. Johnson, M.L. Crow, et al., Energy storage systems for advanced power applications, Proc. IEEE 89 (12) (2001) 1744–1756. S. Talari, M. Yazdaninejad, M.R. Haghifam, Stochastic-based scheduling of the microgrid operation including wind turbines, photovoltaic cells, energy storages and responsive loads, IET Gener. Transm. Distrib. 9 (12) (2015) 1498–1509. A. Kusko, J. DeDad, Short-term, long-term, energy storage methods for standby electric power systems, Fortieth, in: IAS Annual Meeting. Conference Record of the 2005 Industry Applications Conference, 2005, pp. 2672–2678. A. Kulka, Pitch and Torque Control of Variable Speed Wind Turbine (M.Sc. Thesis), Chalmers University of Technology, Sweden, 2004. R. Datta, V.T. Ranganathan, Variable-speed wind power generation using doubly fed wound rotor induction machine-a comparison with alternative schemes, IEEE Trans. Energy Convers. 17 (3) (2002) 414–421. K.H. Kim, T.L. Van, D.C. Lee, et al., Maximum output power tracking control in variable-speed wind turbine systems considering rotor inertial power, IEEE Trans. Ind. Electron. 60 (8) (2013) 3207–3217. T. Senjyu, R. Sakamoto, N. Urasaki, et al., Output power levelling of wind turbine generator for all operating regions by pitch angle control, IEEE Trans. Energy Convers. 21 (2) (2006) 467–475. M.J. Khan, M.T. Iqbal, Dynamic modeling and simulation of a small wind–fuel cell hybrid energy system, Renewable Energy 30 (3) (2005) 421–439. J.F. Manwell, J.G. McGowan, A.L. Rogers, Wind Energy Explained: Theory, Design and Application, Wiley Press, 2002. W. Lajnef, J.M. Vinassa, S. Azzopardi, et al., First step in the reliability assessment of ultracapacitors used as power source in hybrid electric vehicles, Microelectron. Reliab. 44 (9–11) (2004) 1769–1773. J.V. Mierlo, P.V. Bossche, G. Maggetto, Models of energy sources for EV and HEV: fuel cells, batteries, ultracapacitors, flywheels and engine-generators, J. Power Sources 128 (1) (2004) 76–89. X. Control Zhang, Strategy of Cascaded H-Bridge Multilevel Inverter with PV System as Separate DC Source (M.Sc Thesis), KTH Royal Institute of Technology, 2011. G.H. Riahy, M. Abedi, Short term wind speed forecasting for wind turbine applications using linear prediction method, Renewable Energy 33 (1) (2008) 35–41. S. Amin-Nejad, T. Asgari Gashteroodkhani, K. Basharkhah, A comparison of MVDR and LCMV beamformers’ floating point implementations on FPGAs, Wireless Pers. Commun. 98 (2) (2018) 1913–1929. T. Ackermann, Wind Power in Power Systems, Wiley Press, 2005. L. Qu, W. Qiao, Constant Power Control of DFIG Wind Turbines with Supercapacitor Energy Storage, IEEE Trans. Ind. Appl. 47 (1) (2011) 359–367. C. Wang, Modelling and Control of a Hybrid Wind/Photovoltaic/Fuel Cell Distributed Generation Systems (Ph.D. dissertation), Montana State University, Bozeman, Montana, USA, 2006.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
Contents lists available at ScienceDirect
Engineering Science and Technology, an International Journal journal homepage: www.elsevier.com/locate/jestch
Full Length Article
Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting A. Aranizadeh a, A. Zaboli a, O. Asgari Gashteroodkhani b, B. Vahidi a,⇑ a b
Electrical Engineering Department, Amirkabir University of Technology, Tehran 1591634311, Iran Department of Electrical and Biomedical Engineering, University of Nevada, Reno, USA
a r t i c l e
i n f o
Article history: Received 6 September 2018 Revised 10 June 2019 Accepted 29 August 2019 Available online xxxx Keywords: Wind turbine Ultra-capacitor Harvested energy Microgrid Wind speed forecasting
a b s t r a c t Wind energy source has a complex control situation because of dependence of its torque and output power on wind speed and its fluctuations. Based on this, in order to improve its control condition and dynamic efficiency, when connecting to the microgrid, ultra-capacitor which has a fast charging and discharging speed is used. Furthermore, the maximum energy derived from wind turbine and ultracapacitor by the microgrid is of high importance which must be considered besides decreasing output power fluctuations. In this paper, for increasing the harvested energy, the Wind Speed Forecasting (WSF) model is used. So, the control method is applied by using WSF. In the proposed method, the gained energy is more than the lost energy. In fact, we increased harvested energy using a predictive control method. The considered predictive control is applied to the induction generator rotational speed variations. The considered wind turbine model in this paper produces an active power of 50 kW and is a variable speed induction generator (VSIG) with an apparent power of 50 kVA. All of the simulations are performed in MATLAB/SIMULINK software. Ó 2019 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Recently, European countries have been concerned about climate changes and have decided to reduce greenhouse gases. In this manner, they have adopted new policies to produce cleaner energies in order to reduce the produced CO2 until 2020 by 20% and until 2050 by 50%. In this regard, increase of renewable energies and hybrid generation would be the way to reach this aim [1]. Less pollution at consumption centers, easier and less hazardous as well as more efficient transmission, controllability at generation and consumption centers and more flexibility in transformation to other energy types at consumption centers are among the factors in making electricity more attractive than other types of energy [2]. Fossil fuels such as coal, oil and natural gas and nuclear energy are non-renewable energies and their available resources are also limited. Therefore, finding new energy sources are one of the important concerns of human in recent century [3,4]. Increasing use of energy consumption, high and ever increasing cost and non-renewable nature of fossil fuels as well as bad situation of global environment have caused much attraction to environment-
⇑ Corresponding author. E-mail address: [email protected] (B. Vahidi). Peer review under responsibility of Karabuk University.
friendly energy resources [5]. One of the most important resources of renewable energy production is wind source. The main problems with the energy from wind turbines are the low energy density, output power severe oscillations and uncertainty in the obtained energy from them. In order to energize all the loads, the producers must access all the required predictions about the power and energy from these resources. These predictions must have high accuracy so that no problem arises for power grid. To predict the instantaneous wind speed and increase the obtained energy from wind turbines, short term prediction must be used to control the dynamic condition of wind turbines [6]. In [7], power grid and market conditions are analyzed by predicting hourly energy from wind considering the increase of wind energy penetration level to the power grid. In [8], the authors determine the size of energy storage system in order to increase the wind energy penetration level in power network considering the grid frequency fluctuations. According to this reference, frequency deviation of power systems caused by grid-connected wind power fluctuations. In [9], calculation of the wind turbine maximum output energy in active distribution networks is elaborated and a multi-period optimal power flow analysis is proposed. In [10], the wind turbine output power fluctuations is decreased by utilizing a new method through continuous prediction of wind speed. In this method, the wind speed in every second and continuously is predicted and by applying these predicted fluctuations to
https://doi.org/10.1016/j.jestch.2019.08.006 2215-0986/Ó 2019 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
2
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
the grid controllers, the wind turbine output power is improved. The method proposed in [11] decreases the wind turbine output power fluctuations utilizing artificial neural network (ANN) and capacitor system controller. In this reference, an ANN controller is considered for controlling the DC-bus connected Energy Capacitor System (ECS). The authors in [12] lessen wind turbine output power fluctuations using prediction and supervisory control unit based on energy storage system. In [13], the transferred energy from wind turbine to the grid is increased using short term wind speed prediction. In [14], wind energy is increased through energy storage and demand response. In [15], scheduling of microgrid operation including wind turbines, photovoltaic systems, energy storages and responsive loads based on stochastic is analyzed. The authors in [16] elaborate the effect of short and long term prediction on the performance of energy storage system. In [17], the author presents maximizing profit of wind-battery supported power station based on wind power and energy price forecasting. In [1,18–20], precise models of wind turbine controllers design are presented in order to improve dynamic condition. Reference [21] dynamically simulates the wind turbine and hybrid energy systems. Considering the aforementioned explanations, in the second section of this paper, design and modelling of wind turbine, ultra-capacitor energy storage system and the procedure of connecting wind turbine and ultra-capacitor to the microgrid and also modelling of the wind speed forecasting system are elaborated. In the third section, some explanations about the control process of wind turbine blade’s angle control and also the specifications of the control system of connection of ultra-capacitor to the DC link are presented. In section four, increase of obtained energy using the predictive control is conducted. In section five, simulation results for the condition with and without predictive control are demonstrated. Finally, the paper is concluded in section six. 2. Design and modeling 2.1. Wind turbine specifications In general condition, the mathematical formulation of extraction power of wind energy using momentum theory is expressed by (1) and (2) [22]:
Pwind ¼
1 1 qAv 3 ¼ qpR2 v 3 2 2
ð1Þ
2.2. Ultra-capacitor specifications Different ultra-capacitor models have been presented up to now. Some of these models are merely for analyzing the ultracapacitor itself and obtaining thermal and electrical characteristics of it [23]. These models are only appropriate for analyzing the characteristics of ultra-capacitors and are not efficient for analyzing in electrical grids due to their complexity. An example of extended model of ultra-capacitor is shown in Fig. 1a. This ultracapacitor model is analyzed in [11] and its aim is to smooth the output power of wind turbine. The proposed model in [24] is presented in Fig. 1b. The same model is used in this paper. In this paper, a 2F (Farad) ultra-capacitor with an internal resistance of 1.5 mO is used. Each of the ultra-capacitor’s cells has the initial voltage of 50 V which is placed in 16 in series to reveal an initial voltage of 800 V. The capacity of the series is 0.125F. 2.3. System specifications of the grid System specifications of the grid with the connection of wind turbine and ultra-capacitor energy storage system are demonstrated in Fig. 2 [10,25]. As shown, wind turbine and ultracapacitor system are connected to a microgrid with a weak network. This microgrid is severely reacting against power fluctuations and transferred energy. Based on this, controlling power and output energy of wind turbine in this condition is of high importance. In Fig. 2, the combination of wind turbine, ultracapacitor energy storage and microgrid supply AC and DC loads. 2.4. The prediction system specifications The prediction method depends on the available data and data’s time period. In wind turbines dynamic control, data are wind speed and data’s time periods are the seconds. On this bases, the forecasting method of wind speed in this paper is the linear forecasting method [10,26]. This method can properly handle data of wind speed in a short time period. This prediction control method is demonstrated in Fig. 3. According to this figure, having the previous changed time (t0) and the changed predicted time (t0 + T), the variation rate is obtained as expressed by (4). Afterwards, having the rate from (4), the line equation obtained in (5) can be calculated.
V ðtÞ ¼ mððt 0 þ T Þ t0 Þ þ Vðt 0 Þ ) V ðt Þ ¼ mT þ Vðt0 Þ
where, q is air density in kg/m3, A is the swept area of the rotor in m2, R is the radius of swept area in meter and v is the wind speed in m/s. Therefore, the wind turbine obtainable power can be expressed by (2):
Pt ¼
1 qpR2 v 3 C p ðk; hÞ 2
ð2Þ
R b3 Cb2
In this equation, CP is the power coefficient of the wind turbine. This value is related to tip speed ratio (TSR) and wind turbine blade pitch angle. The ratio of tip speed ratio to wind speed is demonstrated by k which is dimensionless and is expressed by (3):
TSR ¼ k ¼
R xr V wind
R b1
R b2
C b3
ð4Þ
Cb1
Ucap
(a)
Ri
ð3Þ
where, Vwind is the wind speed, R is the blade radius and xr is the angular speed at the wind turbine blades. The wind speeds lower than the rated wind speed utilize the Maximum Power Point Tracking (MPPT) strategy through controlling rotor speed and making TSR constant in its optimum value. Furthermore, for the wind speeds more than rated wind speed, the strategy of limiting power is gained via controlling blades pitch angle.
Rp
Uc
C
Ucap
(b) Fig. 1. Ultra-capacitor modelling (a) The extended equivalent model of ultracapacitor based on [11] (b) The compact equivalent model of ultra-capacitor based on [24].
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
3
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
3.1. Wind turbine blade pitch angle control system specifications
DC Bus Wind Turbine
The control method of wind turbine that should be applied to wind turbine blades is demonstrated in Fig. 4. Blade pitch angle controller is activated when the generated power is more than generator rated power. Furthermore, blade pitch angle changes has some inertia which depends on the blade’s variation time constant value and the initial value of wind turbine changes. Furthermore, wind turbine blade’s pitch angle variation must be in a specified limit [28]. The pitch angle has limitation to change. In this paper, the maximum changes for pitch angle are 3°/s to 10°/s and delay time is 0.02 s. These limitations depend on size and characteristics of wind turbine. The PI control parameters are P = 3 and I = 30.
AC / DC
Compensator Copacitor
Energy Storage
DC / DC
DC Load
DC / AC
AC Load
DC / DC
Convertor
Other sources
3.2. Ultra-capacitor connection to DC link control system specifications
Micro grid
Wind Speed
Fig. 2. Microgrid (connection of wind turbine, energy storage system, weak source and loads).
Wind speed value to be predicted (future sample)
Wind speed samples
...
Modelling window
...
t0-mT
t 0-T t0 t0+T
time
Vt0+T
D1 ¼
Vt0+
V UC V dc
ð6Þ
Also, the duty ratio of switch S2 in boost mode is obtained through (7):
Vt0 t0
t0+
D2 ¼ 1 D1
t0+T
Fig. 3. Block diagram of Wind Speed Forecasting (WSF) model.
m¼
Fig. 5 shows a schematic of ultra-capacitor and converter control system connected to DC link. As shown, the ultra-capacitor consists of a capacitor bank which is connected to the DC link through a DC/DC converter [29]. The DC/DC converter has two separated IGBT switches with separately operating switching controllers. Switches S1 and S2 are switched in this controller in such a way to regulate Pg for power transmission between microgrid and wind turbine. In this structure, the DC/DC converter operates in two different modes of buck mode and boost mode which depend on the switching method of IGBT switches. When switch S1 is open, the DC/DC converter goes to the boost mode. In this mode, the ultra-capacitor acts as a source which transmits active power to the grid and by this operation, ultra-capacitor voltage (VUC) decreases. When switch S2 is open, the DC/DC converter conducts current in the buck mode. In this mode, ultra-capacitor performs as active power sink and by this operation, VUC increases. Based on this and according to the aforementioned explanation, the duty ratio of switch S1 in buck mode is estimated by (6):
Vðt 0 þ TÞ Vðt0 Þ T
ð5Þ
Having this line equation, the predicted data in any instance of time can be observed [10,27]. As shown, this prediction control method is appropriate for those inputs with constant rate changes and the predicted value follows a determined procedure. In this paper, the wind speed data must be forecasted and the forecasted values must be applied to the predicting control system. In fact V(t0 + T) is applied to predictive control instead of V(t0). Also amount of T is calculated according to wind turbine size and inertia. 3. Wind turbine and ultra-capacitor control system specifications In order to control wind the turbine system containing the ultra-capacitor to the grid shown in Fig. 2, two control methods i.e.,wind turbine blade pitch angle control system and ultracapacitor control system are analyzed.
ð7Þ
The on/off period of switches is obtained through the output of current controller and by comparing the reference current with the present current of the ultra-capacitor. The saturation block is used to regulate output signal of the controller in the range of 0–1. Then, this value is compared to the carrier signal value with 10 kHz frequency in this paper and is signed with the sign function. Furthermore, ultra-capacitor voltage level is controlled at every instance so that it is not out of the allowed predefined limits. The PID control parameters are P = 50, I = 1 and D = 0.3. 4. Increasing harvested energy using predictive control It should be noted that when the wind speed drastically changes, the generator rotational speed should change so as to Delay time
Pref (pu)
- +
PI Controller
Pitch angle WT system
Pmeasured (pu)
(a)
Fig. 4. Block diagram for pitch angle control.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
4
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
S1 DC Link
Less
+
-
+
S2
Iref
+ -
PID Controller
Imeasured
Saturation block
- + Min/Max voltage limit Sign function
Repeating Sequence
-
Ultra-capacitor Bank
(U+1)/2
Command
Charge/Discharge mod
Fig. 5. Control system of connectivity of ultra-capacitor to DC link.
keep power coefficient in its maximum value. However, due to the generator inertia, the change in the generator rotational speed is slow. In order to overcome the inertia, in the proposed method, the process of generator rotational speed control begins before the instance of the wind speed change. According to the Fig. 6a, in tvirtual, the generator rotational speed is changed prior to tactual where the actual wind speed is changed. In [tvirtualtactual], the output power decreases in correspondence to the change in tip speed ratio, and in turn, it leads to loss of energy in the proposed method in comparison with the conventional method. However, after tactual, since the generator rotational speed inertia is overcome, the output power reaches its maximum in less time, and it gains more energy compared to the conventional state. By noting to the Fig. 6a, it can be observed that in the proposed method, the gained energy after tactual is more than the lost energy in the [tvirtualtactual]. To elucidate more on why the gained energy is more than the lost energy in the proposed method, by considering the Fig. 6b, it can be seen that in the conventional method, when the wind speed
rapidly changes, tip speed ratio considerably changes that makes the operating point be near the rapid slope of the CP-k curve, and consequently the output power decreases to a great extent. However, in the proposed method, since the operating point is near the slow slope of the CP-k curve, the output power decreases in a lesser extent. Accordingly, total gained energy in the proposed method is more than that of conventional method. Fig. 7 shows a block diagram of the proposed method. As shown, using wind speed forecasting, in an instance before the change in actual wind speed, the generator rotational speed controller operates. It must be considered that time difference between application of the controller and the wind speed change depends on the size of wind turbine. The bigger the size of wind turbine, the bigger inertia of induction generator rotational speed variations is obtained, and thus the predictive controller should be applied earlier.
5. Simulation results Conventional Method
P
Proposed Method
P2 Gained Energy in Proposed Method
P1
Lost Energy in Proposed Method
tvirtual
CP
t
tactual (a)
∆CP-Proposed
Conventional Method
To analyze the simulation results, a 50 kW wind turbine with an induction generator is utilized. The specifications of wind turbine and induction generator are presented in Tables 1 and 2 [30], respectively. The ultra-capacitor storage system specifications are also presented in Table 3 [21]. The wind turbine model in this section is a variable speed doubly-fed induction generator (DFIG) and perturb and observe (P&O) mode for MPPT technique is used in this paper. This simulation is implemented in MATLAB/SIMULINK software. Fig. 8 shows the SIMULINK model of the proposed system in MATLAB. This figure consists of wind turbine model, ultracapacitor model, load model and other sources connected to each other with convertors. The Ultra-capacitor controller with switching and pitch angle controller is explained in Sections 3.2 and 3.1, respectively. Also, predicted wind speed is applied to generator speed.
Propo sed Proposed Method M t ∆CP-Conventional
Supercapacitor energy storage system
Vpresent
λ (b) Fig. 6. A comparison between the proposed and conventional methods (a) in harvested energy (b) in power coefficient.
WSF
Vpredicted
Rotational speed controller by prediction
Omegapredicted
Pitch angle Pitch angle controller
WT system Power
Micro Grid
Vpresent
Fig. 7. Block diagram of variable speed wind turbine in the proposed method.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx Table 1 Parameters of wind turbine [29]. Parameter
Value
Air density Rotor diameter Rated power Cut-in/cut-out speed Rated wind speed
1.225 kg/m3 15 m 50 kW 3/25 m/s 12 m/s
Table 2 Parameters of induction generator [29]. Parameter
Value
Rated power Stator voltage/frequency Stator resistance Rotor resistance Stator leakage inductance Rotor leakage inductance Mutual inductance
50 KVA 380 V/60 Hz 0.016 pu 0.015 pu 0.06 pu 0.06 pu 3.5 pu
Table 3 Parameters of ultra-capacitor [21]. Parameter
Value
Ultra-capacitor voltage NrS/NrP Ri Ultra-capacitor Upper voltage limit Lower voltage limit
800 V/each 50 V 16/1 1.5 mO 0.125F/each 2F 850 V 750 V
For analyzing this issue and applying it to the wind turbine model with ultra-capacitor storage system, wind speed is applied to the wind turbine according to Fig. 9a. As can be seen from Fig. 9a, the wind speed is 8.5 m/s at first and after 2.5 s, it reduces to 7 m/s. At this wind speed,
5
wind turbine output power becomes less than load power which is considered 23 MW and thus causes the discharge of ultra-capacitor. Afterward, the wind speed increases up to 9.5 m/s. At this wind speed, the wind turbine generated power becomes more than load power which leads to charge of ultracapacitor and thus causes the ultra-capacitor to reach its charging limitation.
5.1. Reach of ultra-capacitor to the charge limitation without using wind speed prediction In this part, wind speed prediction is not used and generator speed control is in such a way that as the wind speed changes the generator speed also alters for harvesting maximum power from wind. The generator speed in this part is shown in Fig. 9b. As shown, wind speed fluctuations have some inertia that considering small size of the studied wind turbine in this paper, this inertia is not very high. Anyway, in larger wind turbines this inertia is remarkable. As shown in Fig. 9b, as wind speed changes at 2.5 s, the generator rotational speed changes to receive the maximum harvesting power. The voltage of ultra-capacitor in this part is shown in Fig. 9c. Before 2.5 s, considering low wind speed, ultra-capacitor is in discharge mode. Increase in wind speed and as a consequence, increase in wind turbine produced power puts the ultra-capacitor in charge mode which is obviously observable in the voltage of ultra-capacitor. Continuation of charging makes the ultra-capacitor to reach its charging limitation at 4.55 s. When the ultra-capacitor reaches its charge limitation, ultra-capacitor storage system becomes separated from the grid and the wind turbine with the power grid supply the load demand. Also, with the separation of ultra-capacitor, the power grid generated power increases and it increases power generation cost. Fig. 9d demonstrates wind turbine output power, injected or received power from ultra-capacitor and also total power of wind turbine and ultra-capacitor for this case.
Fig. 8. SIMULINK model of the proposed system in MATLAB.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
6
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx 1.9
Rotational Speed (rad/s)
Wind Speed (m/s)
9.5 9 8.5 8 7.5 7 0
1
2
3
4
1.8 1.7 1.6 1.5 1.4 0
5
1
2
4
5
(b)
(a) 860
40 P
850
Total
30
Power (kW)
Capacitor Voltage (V)
3
time (s)
time (s)
840 830 820 810
=P
U.C.
+P
P
W.T.
W.T.
20 P
Load
10 P
U.C.
0
800 790 0
1
2
3
4
-10 0
5
1
2
3
4
5
time (s)
time (s)
(d)
(c)
2
860
Capacitor Voltage (V)
Rotational Speed (rad/s)
Fig. 9. Without using wind speed prediction (a) Wind speed variations applied to the wind turbine (b) Generator rotational speed variations (c) Ultra-capacitor voltage (d) Wind turbine output power, power from ultra-capacitor and also total power of wind turbine and ultra-capacitor as well as load power.
1.9 1.8 1.7 1.6 1.5 1.4 0
1
2
3
4
850 840 830 820 810 800 790 0
5
1
2
3
4
5
time (s)
(b) 40
850
30
P
Total
840
Power (kW)
Capacitor Voltage (V)
(a) 860
Without Prediction
830 With Prediction
820 810 3
3.5
4
4.5
=P
W.T.
+P
P
W.T.
U.C.
20 P
Load
10
P
U.C.
0
5
-10 0
1
2
3
time (s)
time (s)
(c)
(d)
4
5
Fig. 10. With using wind speed prediction (a) Generator rotational speed variations (b) Ultra-capacitor voltage (c) Comparing ultra-capacitor voltage with and without predictive control (d) Wind turbine output power, power from ultra-capacitor and also total power of wind turbine and ultra-capacitor as well as load power.
5.2. Reach of ultra-capacitor to the charge limitation using wind speed prediction Contrary to the previous section, generator speed control in this section is in such a way that before wind speed variation, generator speed varies for receiving maximum wind power. In fact, in this section, with predicting wind speed in future, generator speed controller changes in a way that it applies the occurred fluctuations of future in a sooner time. According to the aforementioned matters, generator speed is according to Fig. 10a. Generator speed variations occur at 2 s (0.5 s sooner than wind speed actual change). Ultra-capacitor voltage is also shown in Fig. 10b. In this part, at 2.5 s considering the wind speed, the ultra-capacitor is in discharge
mode and with the wind speed increase and therefore the increase of generated power from wind turbine, the ultra-capacitor gets into charge mode. According to Fig. 10b, the ultra-capacitor has reached to its charge limitation at 4.59 s. As shown, in this method, the ultracapacitor has reached to its charge limitation 0.04 s later. The comparison between Figs. 9c and 10b is presented in Fig. 10c. This figure obviously shows that by applying the predictive control, the ultra-capacitor has reached to its charge limits later. Fig. 10d demonstrates wind turbine output power, injected or received power from ultra-capacitor and also total power of wind turbine and ultra-capacitor for the mode of control with wind speed control.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006
A. Aranizadeh et al. / Engineering Science and Technology, an International Journal xxx (xxxx) xxx
34
Power (kW)
32 30
P
without Predictive Control
P
with Predictive Control
Total Total
[5] [6]
28 [7]
26 24 22 3
[8]
3.5
4
4.5
5
time (s) Fig. 11. Comparison between the transmitted power from the grid to the load in both cases of without and with predictive control.
Fig. 11 demonstrates the generated power by wind turbine and ultra-capacitor in both conditions of with and without predictive control. According to this figure, it is obvious that sum of wind turbine and ultra-capacitor generated power with wind speed prediction control is more than the case with no wind speed prediction control. This has occurred with what has been thoroughly explained in Fig. 10c. The result is the increase of wind turbine and ultra-capacitor system generated power. According to Fig. 11, the injected power from the grid to the load in the state of predicted control has increased later. Therefore, the generated energy by wind turbine and ultra-capacitor for energizing the load in the case with predictive control is almost 296 J more than the case without predictive control. The amount of increased energy depends on the turbine size so that an increase in the turbine size increases the harvested energy.
[9]
[10]
[11]
[12]
[13]
[14] [15]
[16]
[17]
6. Conclusion
[18]
In this paper, a WSF method is applied to monitor the future wind speed data. Afterward, by applying this control method on wind turbine, the harvested energy from wind turbine and ultracapacitor energy storage is increased and the microgrid condition is improved. In this paper, all of modelling is carried out in MATLAB/SIMULINK software. By running simulations, it is observed that by utilizing WSF method for wind speed data and online monitoring of wind speed in the future and then applying predictive control to induction generator rotational speed, the obtained energy is increased by 296 J. Therefore, when this method is used, the produced energy from large-scale wind farm is increased and the network costs are reduced. Also, wind turbine with more inertia needs more prediction time. This paper is a proof of this method. In the future work, the optimal value of T can be found for different size of turbine.
[19]
[20]
[21] [22] [23]
[24]
[25]
[26]
References [27] [1] L.Y. Pao, K.E. Johnson, Control of Wind Turbines’, IEEE Control Syst. 31 (2) (2011) 44–62. [2] A. Forooghi Nematollahi, A. Dadkhah, O. Asgari Gashteroodkhani, B. Vahidi, Optimal sizing and siting of DGs for loss reduction using an iterative-analytical method, J. Renewable Sustainable Energy 8 (5) (2016) 055301. [3] T. Burton, D. Sharpe, N. Jenkins, et al., Wind Energy Handbook’, Wiley Press, 2001. [4] O. Asgari Gashteroodkhani, B. Vahidi, A. Zaboli, Time-time matrix z-score vector-based fault analysis method for series-compensated transmission lines,
[28] [29] [30]
7
Turkish Journal of Electrical Engineering & Computer Sciences 25 (2017) 2647– 2659. ‘International Energy Outlook, Energy Information Administration (EIA)’, http://www.eia.doe.gov/iea. Y. Izumi, A. Pratab, Y. Kinjyo, et al., A fuzzy-based output power smoothing of WECS using short-term ahead prediction of wind speed, in: IEEE International Conference on Fuzzy Systems (FUZZ), Taipei, 2011, pp. 2414–2419. M. Hedayati-Mehdiabadi, J. Zhang, K.W. Hedman, Wind power dispatch margin for flexible energy and reserve scheduling with increased wind generation, IEEE Trans. Sustainable Energy 6 (4) (2015) 1543–1552. Y. Liu, W. Du, L. Xiao, et al., A method for sizing energy storage system to increase wind penetration as limited by grid frequency deviations, IEEE Trans. Power Syst. 31 (1) (2016) 729–737. P. Siano, P. Chen, Z. Chen, et al., Evaluating maximum wind energy exploitation in active distribution networks, IET Gener. Transm. Distrib. 4 (5) (2010) 598– 608. A. Aranizadeh, B. Vahidi, A. Rahiminejad, Wind turbine power output smoothing in microgrid using ultra-capacitor with continuous wind speed forecasting and online supervisory control, J. Renewable Sustainable Energy 8 (3) (2016) 1–17. S.M. Muyeen, H.M. Hasanien, J. Tamura, Reduction of frequency fluctuation for wind farm connected power systems by an adaptive artificial neural network controlled energy capacitor system, IET Renew. Power Gener. 6 (4) (2012) 226–235. F. Islam, A. Al-Durra, S.M. Muyeen, Smoothing of wind farm output by prediction and supervisory-control-unit-based FESS, IEEE Trans. Sustainable Energy 4 (4) (2013) 925–933. J. Kiviluoma, P. Meibom, A. Tuohy, et al., Short-Term Energy Balancing with Increasing Levels of Wind Energy, IEEE Trans. Sustainable Energy 3 (4) (2012) 769–776. P.F. Ribeiro, B.K. Johnson, M.L. Crow, et al., Energy storage systems for advanced power applications, Proc. IEEE 89 (12) (2001) 1744–1756. S. Talari, M. Yazdaninejad, M.R. Haghifam, Stochastic-based scheduling of the microgrid operation including wind turbines, photovoltaic cells, energy storages and responsive loads, IET Gener. Transm. Distrib. 9 (12) (2015) 1498–1509. A. Kusko, J. DeDad, Short-term, long-term, energy storage methods for standby electric power systems, Fortieth, in: IAS Annual Meeting. Conference Record of the 2005 Industry Applications Conference, 2005, pp. 2672–2678. A. Kulka, Pitch and Torque Control of Variable Speed Wind Turbine (M.Sc. Thesis), Chalmers University of Technology, Sweden, 2004. R. Datta, V.T. Ranganathan, Variable-speed wind power generation using doubly fed wound rotor induction machine-a comparison with alternative schemes, IEEE Trans. Energy Convers. 17 (3) (2002) 414–421. K.H. Kim, T.L. Van, D.C. Lee, et al., Maximum output power tracking control in variable-speed wind turbine systems considering rotor inertial power, IEEE Trans. Ind. Electron. 60 (8) (2013) 3207–3217. T. Senjyu, R. Sakamoto, N. Urasaki, et al., Output power levelling of wind turbine generator for all operating regions by pitch angle control, IEEE Trans. Energy Convers. 21 (2) (2006) 467–475. M.J. Khan, M.T. Iqbal, Dynamic modeling and simulation of a small wind–fuel cell hybrid energy system, Renewable Energy 30 (3) (2005) 421–439. J.F. Manwell, J.G. McGowan, A.L. Rogers, Wind Energy Explained: Theory, Design and Application, Wiley Press, 2002. W. Lajnef, J.M. Vinassa, S. Azzopardi, et al., First step in the reliability assessment of ultracapacitors used as power source in hybrid electric vehicles, Microelectron. Reliab. 44 (9–11) (2004) 1769–1773. J.V. Mierlo, P.V. Bossche, G. Maggetto, Models of energy sources for EV and HEV: fuel cells, batteries, ultracapacitors, flywheels and engine-generators, J. Power Sources 128 (1) (2004) 76–89. X. Control Zhang, Strategy of Cascaded H-Bridge Multilevel Inverter with PV System as Separate DC Source (M.Sc Thesis), KTH Royal Institute of Technology, 2011. G.H. Riahy, M. Abedi, Short term wind speed forecasting for wind turbine applications using linear prediction method, Renewable Energy 33 (1) (2008) 35–41. S. Amin-Nejad, T. Asgari Gashteroodkhani, K. Basharkhah, A comparison of MVDR and LCMV beamformers’ floating point implementations on FPGAs, Wireless Pers. Commun. 98 (2) (2018) 1913–1929. T. Ackermann, Wind Power in Power Systems, Wiley Press, 2005. L. Qu, W. Qiao, Constant Power Control of DFIG Wind Turbines with Supercapacitor Energy Storage, IEEE Trans. Ind. Appl. 47 (1) (2011) 359–367. C. Wang, Modelling and Control of a Hybrid Wind/Photovoltaic/Fuel Cell Distributed Generation Systems (Ph.D. dissertation), Montana State University, Bozeman, Montana, USA, 2006.
Please cite this article as: A. Aranizadeh, A. Zaboli, O. Asgari Gashteroodkhani et al., Wind turbine and ultra-capacitor harvested energy increasing in microgrid using wind speed forecasting, Engineering Science and Technology, an International Journal, https://doi.org/10.1016/j.jestch.2019.08.006