Steady-State Analysis of Self-Excited Induction Generator

Using Real and Reactive Power Balances

S. M. Alghuwainem, Senior Member, IEEE
Department of Electrical Engineering
King Saud University
Riyadh, Saudi Arabia
Email: saadalgh@ksu.edu.sa

AbstractThis paper presents an efficient, and flexible analysis of the SEIG under various operating conditions is
method of steady-state performance analysis of the three- listed in [9]-[11].
phase self-excited induction generator. The method Steady-state analysis of the SEIG is usually based on the
equates the total real and reactive power of the equivalent analysis of its simplified equivalent circuit, shown in Fig. 1,
circuit to zero. Unlike the previous methods which equate where Rs , f X s , Rr , f X r , Rc , f X m , and X c / f represent
total node admittance or total loop impedance to zero, the the stator resistance, stator leakage reactance, rotor resistance,
proposed method produces two real equations directly, rotor leakage reactance, core loss resistance, magnetizing
without going through the lengthy, tedious, and error- reactance, and excitation capacitor reactance respectively, all
prone derivations of the real and imaginary parts of the are evaluated at per unit speed X and per unit frequency f.
total admittance or impedance. The method is easy to Rs f Xs
formulate and flexible to include any changes in the circuit
without repeating lengthy derivations to reanalyze the circuit.
The method is evaluated on a laboratory-size 380-V, 750-W Is Ir
induction generator, and the results obtained were found to be in fX r
agreement with the published results obtained by several authors XC Vt Eg f Xm
RL f XL Rc
using the previous methods of analysis. Some of the results f f Rr
obtained using the proposed method are also verified
f X
experimentally.

Index TermsInduction generator, self-excited induction Fig. 1. Per-phase, per unit equivalent circuit of a SEIG.
generator (SEIG), steady-state analysis, wind energy.
All circuit parameters in Fig. 1 are assumed to be independent
of voltage and current except the magnetizing reactance X m
I. INTRODUCTION
which depends nonlinearly on E g through the magnetization

S ELF-EXCITED induction generators are increasingly

being used in remote locations where renewable energy
sources such as mini-hydro, wind, or other
curve which can be approximated in several ways, such as an
exponential function [10], a piece-wise linear function [11], or
a higher order polynomial [12]-[14]. In this paper, it is
nonconventional energy sources are available. The self-
represented by a third-order polynomial of the form
excited induction generator (SEIG) is attractive for such
Eg
applications since it has several features, including low cost, k0  k1 X m  k2 X m2  k3 X m3 (1)
reduced maintenance, high reliability, rugged construction, f
self-protection against short-circuits, etc. However, since it is The coefficients k0 to k3 are obtained by curve-fitting of the
not connected to a utility grid, it has an inherently poor polynomial to the experimentally obtained magnetization
voltage and frequency regulations. In grid-connected mode, curve at base frequency.
the terminal voltage and frequency of the induction generator
are fixed by the grid, and hence its performance analysis is In the previous methods of analysis, the equivalent circuit in
straightforward. On the other hand, in the stand-alone mode Fig. 1 is solved to determine the air-gap voltage, terminal
the terminal voltage and frequency are not fixed but they vary voltage and frequency. The total admittance at air-gap node is
with speed and loading conditions, depending on the torque- equated to zero [12]-[14], or equivalently the total loop-
speed characteristics of the driving turbine [1]-[3[. impedance is equated to zero [15]-[18]. In both approaches,
Performance analysis of the SEIG is complicated further by the resulting equation is in complex form, and in order to
the presence of other nonlinearities such as magnetic solve it by a gradient method such as the Newton-Raphson
saturation of the load [4] or its transformer [5], or by modeling method, it has to be separated into two real equations. This
the load as a P-Q model [6], or by the method of connecting separation is a lengthy, tedious, time-consuming, and error-
the excitation capacitors as short or long shunt connections prone task. Moreover, the derived equations are valid only for
[7], [8]. A comprehensive overview of literature on the a particular machine and load model. A symbolic
CCECE 2014 Toronto, Canada.
978-1-4799-3101-9/14/$31.00 2014 IEEE
978-1-4799-3010-9/14/$31.00 2014 IEEE CCECE 2014 Toronto, Canada
programming technique to reduce the derivations of the Fig. 3, however the higher the speed and the lower the load,
equations is described in [19], and a number of optimization- the higher the frequency as expected. The effect of speed on
based techniques which minimize the required explicit terminal voltage and frequency for a constant excitation
algebraic expressions are developed in [20]. Despite all these capacitance is to increase terminal voltage in a way similar to
improvements, the admittance and impedance methods are increasing capacitance at constant speed, and to increase
still inflexible and need considerable amount of algebraic frequency linearly as shown in Fig. 4.
manipulations.
Speed= 1.0 p.u., noload
The method proposed in this paper is simple, flexible, and Speed = 0.8 p.u., noload
efficient. It is based on formulating two real equations, one for Speed= 1.0 p.u., RL = 1.0 p.u.

real power balance, and the other for reactive power balance. Speed= 0.8 p.u., RL = 1.0 p.u.

The equations can be easily modified to reflect any changes in 1.2

the load model or excitation capacitor connections without

Terminal Voltage (p.u.)

repeating lengthy derivations.
The proposed method is evaluated on a laboratory-size 750-
0.8
W, 380-V, induction generator operating under various
conditions. The results obtained are compared with
experimentally obtained results.
0.4

II. PROBLEM FORMULATION

0
Real power balance requires that power generated is equal to 0 10 20 30 40
the sum of power delivered to the load plus real power losses. Excitation Capacitance (PF)
Referring to Fig. 1, this relationship can be written as,
2
| Vt |2 | E g | f Rr Fig. 2. Effect of excitation capacitance on terminal voltage.
  | I s |2 Rs  | I r |2 0 (2)
RL Rc f v
Speed = 1 p.u., No-load
Similarly, reactive power balance requires that the reactive Speed = 1 p.u., RL = 1 p.u.
power generated by the excitation capacitor is equal to the sum Speed = 0.8 p.u., No-load
1
of reactive power delivered to the load plus reactive power
losses. Referring to Fig. 1, this relationship can be written as,
2
| Vt |2 | E g | f | Vt |2 (3)
  | I s | 2 f X s  | I r |2 f X r  0
Frequency (p.u.)

f XL f Xm XC 0.9

where,
E g  Vt Eg 1
Is , Ir , and X C ,
R f X
2 2 2
f R /( f  v)  X
2 2 2 2S (60)C 0.8
s s r r

where C is the excitation capacitance. Equations (1), (2), and

(3) can be solved for f , E g , and Vt , using a numerical
calculation software package such as MATLAB [21] and/or 0.7
10 20 30 40
MATHCAD [22].
Excitation Capacitance (PF)
Fig. 3. Effect of excitation capacitance on frequency.
III. PERFORMANCE ANALYSIS
Performance analysis of the generator whose parameters are No-Load, C = 25 PF.
RL = 1.0 p.u., C = 25 PF
Terminal Voltage (p.u.) & Frequency (p.u.)

given in Appendix A, is calculated under various operating 1.2 No-Load, C = 25 PF

using the proposed method. The effect of excitation RL = 1.0 p.u., C = 25 PF

capacitance on the terminal voltage and frequency at constant

speed is shown in Figs. 2 and 3, whereas the effect of speed
0.8
on the terminal voltage and frequency is shown in Fig. 4. The
effect of load power on the terminal voltage and frequency is
shown in Figs. 5 and 6. Performance without any constraints
except constant speed is the most basic mode of operation. In 0.4 Voltage
this case, for each constant speed there is a minimum value of Frequency

excitation capacitance to achieve self-excitation. For the same

speed, the minimum capacitance to achieve self-excitation
0
increases with load power. Increasing the excitation 0.5 0.6 0.7 0.8 0.9 1 1.1
capacitance further increases the terminal voltage as shown in Speed (p.u.)
Fig. 2. Increasing the excitation capacitance beyond the Fig. 4. Effect of speed on the terminal voltage and
minimum capacitance decreases the frequency as shown in frequency.
 the excitation capacitance beyond the minimum capacitance
1.2
decreases frequency as shown in Fig. 3, however the higher
the speed and the lower the load, the higher the frequency as
Speed = 1.0 p.u.
expected. The effect of speed on terminal voltage and
Terminal Voltage (p.u.)

1.1 C = 25 PF
frequency for a constant excitation capacitance is to increase
C = 30 PF
terminal voltage in a way similar to increasing capacitance at
constant speed, and to increase frequency linearly as shown in
1
Fig. 4. The effect of load power on the terminal voltage and
frequency is to decrease both until certain maximum load
power is reached at which the machine becomes unstable and
0.9
the terminal voltage and frequency collapse as shown in Figs.
5 and 6. For operation with constraints such that the terminal
voltage as well as the load power are constant requires
0.8
0.5 0.6 0.7 0.8 0.9 1
variation of excitation capacitance with speed as depicted in
Load Power (p.u.) Fig. 7 theoretically as well as experimentally. More excitation
capacitance is required at lower speeds whereas less
Fig. 5. Effect of load power on the terminal voltage. capacitance is required at the higher speeds. For operation
with the constraint that the terminal voltage as well as speed
are constant, the excitation capacitance must vary with load
Speed = 1.0 p.u. power as depicted in Fig. 8.
1.15
C = 25 PF
Terminal Voltage | Vt | = 1.0 p.u.
1.1 C = 30 PF
60 PL = 1.0 p.u.
PL = 0.5 p.u., calculated
Frequency (p.u.)

1.05
Excitation Capacitance (PF)
PL = 0.5 p.u., experimental

1
40

0.95

0.9
20

0.85
0.5 0.6 0.7 0.8 0.9 1
Load Power (p.u.)
0
Fig. 6. Effect of load power on frequency. 0.8 0.9 1 1.1
Speed (p.u.)

Fig. 7. Excitation capacitance versus speed for constant

IV. PERFORMANCE WITH CONSTANT TERMINAL VOLTAGE Terminal Voltage |Vt| = 1.0 p.u.
Performance analysis with constant terminal voltage is Speed = 0.9 p.u.
investigated by solving (1), (2), and (3) in addition to the 60 Speed = 1.1 p.u.
equation representing constant terminal voltage, namely
| Vt | 1.0 (4)
Excitation Capacitance (PF)

50
Fig. 7 shows the capacitance needed as the speed varies at
constant terminal voltage and constant power. Fig. 8 shows
the capacitance needed as the load power varies at constant 40
terminal voltage and constant speed.
30

V. DISCUSSION OF RESULTS
20
Performance without any constraints except constant speed
is the most basic mode of operation. In this case, for each
constant speed there is a minimum value of excitation 10
0.6 0.7 0.8 0.9 1
capacitance to achieve self-excitation. For the same speed, the
Load Power (p.u.)
minimum capacitance to achieve self-excitation increases with Fig. 8. Excitation capacitance versus load power for constant
load power. Increasing the excitation capacitance further terminal voltage.
increases the terminal voltage as shown in Fig. 2. Increasing
 VI. CONCLUSION [12] S. S. Murthy, O. P. Malik, and A. K. Tandor, Analysis
A novel method of steady-state performance analysis of of self-excited induction generators, in Proc. Inst. Electr.
three-phase self-excited induction generator is presented in Eng., vol. 129, no. 6, pp. 260265, 1982.
[13] S. S. Murthy, B. P. Singh, C. Nagamani, and K.V.V.
this paper. The method is more flexible in allowing different
Satyanarayana, Studies on the use of conventional
operation scenarios without repeating lengthy derivations.
induction motors as self-excited induction generators,
The results obtained using this method are in full agreement
IEEE Trans. Energy Convers., vol. 3, no. 4, pp. 842848,
with results obtained by previous methods. Some of the results Dec. 1988.
obtained by this method were verified experimentally. [14] S. P. Singh, B. Singh, and M. P. Jain, Performance
characteristics and optimal utilization of a cage machine
ACKNOWLEDGMENT as capacitor excited induction generator, IEEE Trans.
The author would like to thank Eng. Saleh Al-Senaidi and Energy Convers., vol. 5, no. 4, pp. 679685, Dec. 1990.
Eng. Umar Bawah, Department of Electrical Engineering, [15] I. Alolah and M. A. Alkanhal, Optimization-based
King Saud University for their help in setting up the steady state analysis of three phase self-excited induction
experiments for verification of some results. generator, IEEE Trans. Energy Convers., vol. 15, no. 1,
pp. 6165, Mar. 2000.
[16] L. Shridhar, B. Shing, and C. S. Jha, A step towards
REFERENCES improvements in the characteristics of self-excited
induction generator, IEEE Trans. Energy Convers., vol.
[1] S. Rajakaruna and R. Bonert, A technique for the steady 8, no. 1, pp. 4046, Mar. 1993.
state analysis of induction generator with variable [17] Singh, S. S. Murthy, and S. Gupta, Analysis and design
speed, IEEE Trans. Energy Convers., vol. 8, pp. 757 of STATCOM based voltage regulator for self-excited
761, Dec. 1993. induction generators, IEEE Trans. Energy Convers., vol.
[2] T. F. Chan, Self-excited induction generators driven by 19, no. 4, pp. 783790, Dec. 2004.
regulated and unregulated turbines, IEEE Trans. Energy [18] L. Quazene and G. McPherson, Analysis of the isolated
Convers., vol. 11, pp. 338343, June 1996. induction generator, IEEE Trans. Power App. Syst., vol.
[3] S. M. Alghuwainem, Steady-state analysis of an isolated PAS-102, no. 8, pp. 27932798, Aug. 1983.
self-excited induction generator driven by regulated and [19] T. F. Chan, Steady-state analysis of self-excited
unregulated turbine, IEEE Trans. Energy Convers., vol. induction generators, IEEE Trans. Energy Convers., vol.
14, pp. 718723, Sept. 1999. 9, no. 2, pp. 288296, Jun. 1994.
[4] S. M. Alghuwainem, Steady-state analysis of an [20] M. H. Haque, A novel method of evaluating
induction generator self-excited by a capacitor in parallel performance characteristics of a self-excited induction
with a saturable reactor, Electric Machines and Power generator, IEEE Trans. Energy Convers., vol. 24, no. 2,
Systems, vol. 26, pp. 617625, 1998. pp. 358365, June. 2009.
[5] S. M. Alghuwainem, Steady-state analysis of a self- [21] MATLAB, Mathworks, Natick, MA, U.S.A..
excited induction generator including transformer [22] MATHCAD, MathSoft, Inc, Needham, MA, USA
saturation, IEEE Trans. Energy Convers., vol. 14, no. 3,
pp. 667672, Sep. 1999.
[6] L. Shridhar, B. Singh, C. S. Jha, B. P. Singh, and S. S.
Murthy, Selection of capacitors for the self-regulated
short shunt self-excited generator, IEEE Trans. Energy
Convers., vol. 10, no. 1, pp. 1017, Mar. 1995.
[7] L. Wang and J. Y. Su, Effects of long-shunt and short-
shunt connections on voltage variations of a self-excited
induction generator, IEEE Trans. Energy Convers., vol.
12, no. 4, pp. 368374, Dec. 1997.
[8] M. H. Haque, Analysis of a self-excited induction
generator with PQ load model, IEEE Trans. Energy
Convers., vol. 25, no. 1, pp. 265267, March. 2010.
[9] R. C. Bansal, Three-phase self-excited induction
generators: An overview, IEEE Trans. Energy Convers.,
vol. 20, no. 2, pp. 292299, Jun. 2005.
[10] R. C. Bansal, T. S. Bhatti, and D. P. Kothari,
Bibliography on the application of induction generators
in nonconventional energy systems, IEEE Trans. Energy
Convers., vol. 18, no. 3, pp. 433439, Sep. 2003.
[11] G. K. Singh, Self-excited induction generator
researchA survey, Electr. Power Syst. Res., vol. 69,
pp. 107114, 2004.