WIND ENERGY SYSTEM VOLTAGE AND ENERGY ENHANCEMENT USING LOW COST DYNAMIC CAPACITOR COMPENSATIO SCHEME Adel M. Sharaf and Guosheng Wang Department of Electrical and Computer Engineering University of New Brunswick P. 0.Box 4400 / UNB, Fredericton, NB, Canada E-mail:
[email protected] Abstract - The paper presents a novel low cost dynamic capacitor compensation scheme using real time dynamic error tracking and harmonic ripple based PID controller to stabilize an isolated stand alone Wind Energy Conversion Scheme (WECS) with a induction generator. The novel Dynamic Capacitor Compensation scheme (DCC) serves as a voltage stabilization regulator and energy utilization enhancement compensator.
by using the low cost dynamic capacitor compensation scheme (DCC) and switching using a flexible dynamic errordriven PWM switched controller that mainly ensures voltage stability for standalone wind energy schemes. The effectiveness of this simple topology is validated using MatlabISimulink model of the unified wind energy conversion scheme utilizing an induction generator.
11.SAMPLE STUDYSYSTEM
Keyword - Dynamic Capacitor Compensation, Dynamic Tracking, Voltage and Energy Enhancements
I. INTRODUCTION Over the past two decades, wind has been the world's fastestgrowing energy source. Rising from 4,800 megawatts of generating capacity in 1995 to 31,100 megawatts in 2002, it increased a staggering six fold. Wind energy is popular because it is abundant, cheap, inexhaustible, widely distributed, climate-benign, and clean, so that no other energy source can match it. The cost of wind-generated electricity has dropped from 38 cents a kilowatt-hour in the early 1980s to about 4-5 cents a kilowatt-hour today on prime wind coastal sites [l-31. For standalone wind energy schemes, the terminal voltage and frequency are dependent on the rotor speed, shunt capacitance per phase and the load equivalent impedance, which are subject to both wind gusting and dynamic load excursiodchanging conditions [1-21. Such low-cost scheme is usually used in combined passive/motorized loads for driving water pumpslventilation and air circulatiodair conditioning loads, which are generally insensitive to small frequency variations [3]. The frequency excursions documented in literature vary within the range +- 5 1 0 % which is not very adverse to a remote localized /motorized load. The variations in frequency can be minimized using other dynamic control/dump resistive load strategies [ 1-31. The serious voltage instability and loss of excitation problem is usually a byproduct of load excursiodchanges and wind gusting conditions [3]. The value of selected self excitation capacitor bank is usually sized for the nominal wind velocity cut-in and cut-off turbine operating wind velocity range [l-21. The need for novel modulatedadjusted dynamic on-line capacitive compensation [ 1-31 is achieved
A standalone Wind Energy Conversion Scheme (WECS) using induction generator (IG) is studied in this paper under a sequence of Load Excursions and Wind variations. Figure 1 depicts the WECS with only the fixedlswitched linear test load and a final hybrid composite load is shown in Figure 2. Figure 3 shows the three GTO-switches forming the Dynamic Capacitor Compensation scheme (DCC), driven by the dynamic tri-loop tracking controller.
0-7803-8575-6/04/$20.00 02004 IEEE 804
Wind Wind n r b i n e
1 1 1
1.M 600
1.6
LVA
kV
7
Figure 1 Sampled 600 kVA Wind Energy Scheme Diagram
Figure 2 Hybrid Composite Load Model
c
Figure 3 Low Cost Dynamic Capacitor Compensation Scheme (DCC)
The values of capacitors were selected by a guided trial and error off-line simulation to ensure minimum generator voltage excursion for large wind or load variation. The unified system model parameters of the standalone WECS sample study system, comprising the induction generator, wind turbine, combined hybrid load and control parameters are given in the Appendix.
Figure 5 Dynamic Tri-loop Error Tracking and Generator Voltage Stabilization Control Scheme
111. SAMPLE MATLABNMULINK RESULTS Relay
Gain
PWM Output
a n d Hold
Figure 6 Pulse Width Modulator (PWM) Functional Model with a Pre-selected Switching Frequency
C.rnP"t#UO"
@-+a
Cl**
Dynamic Capacitor Compmsattontor Wind Energy System
Figure 4 MATLAB/SIMULINK-PSBlockset WECS Block -Functional Model of the Unified System Model
The sample WECS stand-alone scheme was subjected to severe combined Load switchingl Load variationl Load Excursion and Wind speed variationl Wind gusting. The novel dynamic tracking tri-loop control scheme is shown on Figure 5. Figure 6 depicts the Pulse Width Modulator (PWM) model used with a variable duty cycle ratio a, control at a selected constant switching frequency, fs/w=200 Hz.
Figures 7-8 show the system dynamic response including voltage (RMS), current (RMS), average generator power, and voltage-current-power 3-dimessional phase portraits for the wind energy system. These figures illustrate the system real time response under a combined load /wind excursion sequence--(+, - excursions) as follows: t=O. 1s Load excursion applied, +40%; t=0.3s Load excursion removed, +40%; t=OSs Load excursion applied, -40%; t=0.7s Load excursion removed, -40%; t=0.9s Wind Speed excursion applied, -30%; t=O.lls Wind Speed excursion removed, -30%; t=O. 13s Wind Speed excursion applied, +30%; t=0.15s Wind Speed excursion removed, +30%; The WECS dynamic performance is compared for the two cases, without and with the low cost dynamic capacitor compensator (DCC) as shown in figures 7-8. Figure 9 shows the control signals for the two controllers. Figure 10 shows -
the PWM signals s l and s2 for GTO 1&2. (Notes3 = s 2 , -
where ~2 is the complement of s2, and s3 is the PWM signal for GTO 3.)
805
....................
:. .. . . .. . .. .. ... ......................................................... .. . . .. . ... . . . . .. .. 1 3 .......: . .. ... ... .......:. . . . .. .. 2 ........ ......). .. ........+..,,.....+.... ...............,........,........ . . . ....... ................................................. 15
i
1.4
1
i
1 1
i
.
. . .. .. ... ... ... . ...................................... .. ... .. . . . . . . . . . . . 0 6 ........ ....................... .......................................... L
0.7 .......,i......i.... . I
I
I
-I----/---
. - .
0.7 ......i......i......................................... 0.6 ......L.....
.
O
6
: t.c., .
......i.........................
o.'s
014
i
o.'s
1'2
14
..
,
a bo 1.4
1.05
0
0.6 OS
Vg-nns
0.85
Figure 7 WECS perfonnance without Dynamic Capacitor Compensator
Figure 8 WECS performance Dynamic Capacitor Compensator
1 2 1
08
06 0 4
1
s2
0 2
0
-0 2 -0 4
0 2
1
I
I
0 4
06
08
I
1
12
1 4
0.2
I
0.1
0.6
0.8
16
1
1.2
1.4 t
t (SI
Figure 10
Figure 9
PWM Pulsing Sequences SI & s2 forGTOl& GT02 (note:
PID Controllers Control Sigmnals
806
1.6
6)
s3 = s2 )
IV. CONCLUSION The paper presents a novel low cost dynamic capacitor compensation scheme driven by a dynamic error tracking and ripple PID controller. The dynamic compensator is extremely effective in ensuring voltage stabilization and enhancing powedenergy utilization under severe load excursion and wind prime mover/ wind velocity excursion. The main functions can be assigned by proper selection of control scaling and signal weights.
V. REFERENCES Natarajan, K., Sharaf, A.M., Sivakumar, S. and Nagnarhan, S., “Modeling and Control Design for Wind Energy Conversion Scheme using Self-Excited Induction Generator”, IEEE Trans. On E.C., Vol. 2,No. 3,pp.506-512, Sept. 1987. Singh, S.P., Singh,B., and Jain, M.P., “Performance Characteristic and Optimum Utilization of a Cage Machine as a Capacitor excited Induction Genertor”, IEEE Trans. On E.C., Vol. 5, No.4, pp.679-685, Dec.1990. Hillowala, R.M., and Sharaf, A.M., “Modelling, simulation and analysis of variable speed constant fkquency wind energy conversion scheme using self excited induction generator”, South eastern symposium on Circuits and Systems, October 1990, South Carolina.
Constant machine parameters: Rs = 0.016, LIS = 0.06; R i = 0.015, Llr = 0.06; Lm =3.5; H=2, F=0, p=2. C. Combined Hybrid Load model (@ V=l.Opu) Linear Load (40%) PL = 0.4 PU , QL = 0.4 PU Nonlinear (Voltage-dependent type) Load (30%)
P = PO(-) Vg
(1)
p is the specific density of air (1.25 kg 1m2); A is the area swept by the blades; R is the radius of the rotor blades; CP is power conversion coefficient; ;1is the tip speed ratio; w, .is the wind turbine velocity in rpm; k is equivalent coefficient in per unit (0.745)
P z 2 - 3 (Nonlinearity Order) 3 phase squirrel cage Induction Motor load (30%) Power: SIM = 0.3 pu ,Pole pairs: 2 StatorRotor resistance and leakage inductance: R, = 0.0201 pu, LlS = 0.0349 p u a=2-3
D. Per Unit Base Values used SBase = 600 KVA, VBase = 1.6 K V ( L - L);
E. Dynamic Capacitor Compensator Scheme (DCC) C1 =lo0 uF, C2 =150 uF, C3 = 50 UF
F. PID Controllers Controller I K P --8, y - 1,
I-
Induction Generator Model (3 phase, 2 pairs of poles) Rating: V = 1.6 KV ( L - L ) , g Sg = 500 KVA , CseV = 150 uF Phase
v*.
-
“‘Pa
hr
I,,
L.,,
--
+ - 4.
Kr =0.8,
KD =0.1;
y = 0.5; --Control
r
scaling signal
weights To=20 ms Controller II
K p = 1,
KI = 0.05, KD = 0.01;
yv =1, 7., = O S ,
y -0.5;
hControl scaling and signal weights.
(D-Drh’b
+
P
R, = 0.0377 P U , Ll, = 0.0349 PU Magnetizing inductance: L , = 1.2082 p u
Simple Wind Turbine Model (Quasi-static model) 3 1 T - ---~ARc,v,~ = -pACpVw3= k L . - 21 20, OW
d-q model: + n.
Q = eo(-)vg
vgo , vgo . (2) where, Po = 0.3 pu ,Qo = 0.3pu, Vgo=l.Opu.
VI. APPENDIX
Where:
a
+ v;.
L.
I
G. P WM switching frequency fS/,=200 HZ
q. r1.
H. Average comparative generator power value
whereisthelowpassJilter,and T, = 30ms l+sT,
807
