Power in Non Sinusoidal Situations
Power in Non Sinusoidal Situations
Power in Non Sinusoidal Situations
SEMINAR REPORT
ON
POWER IN NONSINUSOIDAL
SITUATIONS
DEFINITIONS AND PHYSICAL
MEANING
Submitted in partial fulfillment of the requirements for the degree of
MASTER OF TECHNOLOGY
(Power Systems)
By
GOKUL VENUGOPAL IYER
(P13PS009)
: Supervisor:
Prof. Pranav B. Darji
November 2014.
CERTIFICATE
This is to certify that the seminar report entitled POWER IN NONSINUSOIDAL
SITUATION: DEFINITIONS AND PHYSICAL MEANING submitted by GOKUL
VENUGOPAL IYER, P13PS009 is a record of bonafide work carried out by him in partial
fulfillment of the requirement for the award of the degree of MASTER OF TECHNOLOGY
IN ELECTRICAL ENGINEERING (Power Systems).
Date: 22/11/2013
Place: SURAT
PG In charge
Head of Department
SVNIT
ACKNOWLEDGMENT
First of all, I am grateful to The Almighty God for establishing me to complete this
seminar
I wish to express my sincere thanks to my project supervisor Prof. Pranav B. Darji,
Associate Professor, Department of Electrical Engineering, for his valuable suggestions and
time.
I place on record, my sincere gratitude to Mr. H.R. Jariwala Sir, PG In charge, Power
Systems, for his constant encouragement.
I take this opportunity to record my sincere thanks to all the faculty members of the
Department of Electrical Engineering, for their help and encouragement.
I also place on record, my sense of gratitude to one and all who, directly or indirectly,
have lent their helping hands in this venture.
ii
ABSTRACT
The development of efficient, high-power, semi-conductor switching devices and their
application to the control of large electrical machines and industrial processes has imposed
severe problems on the electrical utilities, particularly with respect to the measurement of
energy flow and the optimum use of transmission networks. In power systems it is not only
necessary to determine the active power but also to know the amount of reactive power in the
system. Knowledge of reactive power allows proper selection and design of compensation
devices for its control thereby reducing losses in the system.
It is well known that meters that measure energy (kWh) and active power (kW) provide
accurate measurements also under non sinusoidal or unbalanced conditions; nevertheless,
meters dedicated to apparent power (kVA) and non-active power (kVAr) measurements are
prone to significant errors when the current and voltage waveforms are distorted. The main
reason for such uncertainties stems from the inadequate power definitions that dictate the
conceptual design of such instrumentation.
The Budeanu and Fryzes model gives some insight in the definitions and components of the
reactive power in non sinusoidal situations but fall short in explaining the physical meaning
of it. This article is meant to help in understanding the physical meaning of the instantaneous
power components and its characteristic values (amplitude, frequency of oscillation and
phase) to provide a basic model for apparent power.
iii
TABLE OF CONTENTS
ACKNOWLEDGMENT............................................................................................................ii
ABSTRACT ............................................................................................................................. iii
TABLE OF CONTENTS .......................................................................................................... iv
LIST OF FIGURES ................................................................................................................... v
ACRONYMS AND SYMBOLS .............................................................................................. vi
Chapter 1 Introduction ............................................................................................................... 1
1.1
1.2
2.2
Chapter 3 System with sinusoidal voltage and non sinusoidal current ...................................... 9
3.1 Reactive power definition in sinusoidal voltage and non sinusoidal current situations .. 9
3.2 Thyristorized circuits with sinusoidal voltage ............................................................... 12
Chapter 4 Systems with non sinusoidal voltage ...................................................................... 14
Chapter 5 Conclusion ............................................................................................................... 19
Chapter 6 REFERENCES ........................................................................................................ 21
iv
LIST OF FIGURES
Figures
Figure 2-1 dc Machine driven with a hypothetical shaft power Pm sin 2t .............................. 5
Figure 2-2 dc Machine coupled to a Flywheel........................................................................... 5
Figure 2-3 dc Machine coupled to a torsion bar ........................................................................ 6
Figure 2-4 dc Machine driving a Flywheel and Torsion bar ..................................................... 6
Figure 3-1 Generation of Reactive power in a circuit with a Triac controlled Resistor .......... 12
PV
Photovoltaic
2.
AC
Alternating current
3.
DC
Direct current
4.
Current
5.
Voltage
6.
MPPT
7.
VSI
8.
FB
Full Bridge
9.
HB
Half Bridge
10.
Diodes
11.
Power switch
12.
PWM
13.
Capacitance
14.
HERIC
15.
THD
16.
PF
17.
Inductance
18.
CG-PV
Parasitic capacitance
19.
Lf
Filter inductance
20.
Cf
Filter capacitance
vi
Chapter 1
Introduction
In 1888 the first reference was made to the fact that the oscillations of power between the
alternating source and the load is due to the difference in the phase of the voltage and the
current. The explanation to this was given by Stanley and Shalenberger which with minor
modifications is still found in standards.
The definitions of apparent power S, active power P and reactive power Q in sinusoidal
situations is widely accepted by the electrical engineering community without reservations.
The instrumentation used in the industry for the measurement of power is based on the same.
Load flow studies and energy management program depend on models where P jQ is the
complex variable on which the economical optimization is based.
Today the increased use of static power converters and adjustable speed drives has led to
creation of conditions prone to distortion of voltage waveform. While the effects of non
sinusoidal waveforms on the power delivery were clearly indicated in early researches, the
engineering community has not reached a consensus yet for a universally accepted definition
for the power in electrical system under non sinusoidal situations. Today two major models
dominate the approach to the definitions and components of reactive power. They are,
i.
ii.
P Pn Vn I n cos n
n
(1)
Where Vn and I n are the rms values of the voltage and current harmonics of the order n, and
n is the phase difference between them. Therefore, it makes sense in defining the reactive
power by,
Q Qn Vn I n sin n
n
(2)
which is proposed Budeanu. However this equation does not satisfy the power triangle
equation S 2 P 2 Q 2 . It was then found that,
2
S V I Vn I n cos n Vn I n sin n
n
n
n
n
2
n
2
n
(3)
(4)
S 2 P2 Q2 D2
(5)
The distortion power mainly consists of cross-products of voltage and current harmonics of
different orders and will be reduced to zero if the harmonics are reduced to zero, i.e. at
sinusoidal conditions.
ia
and
P
u
V2
ir i ia
(6)
(7)
The reason for this division is that the current ia is the current of a purely resistive load that,
for the same voltage, would develop the same power as the load measured on. That is, if i r
can be compensated, the source will see a purely resistive load and the power factor will be
equal to unity. It can easily be shown that ia and ir are orthogonal and then the rms-values can
be determined by
I 2 I a2 I r2
(8)
In fact,(6) gives the only possible amplitude of ia if it should be orthogonal to the residual
term ir and have the same wave shape as u. The apparent power can then be obtained as the
product of the rms current and the rms voltage:
S 2 V 2 I 2 V 2 ( I a2 I r2 ) P2 Q2
(9)
Fryze uses Pb instead of Q in his reactive power definition. In other literature reactive power
according his definition is often denoted by QF and named "fictitious power".
2
Many articles in journals of speciality and conferences have tried to produce a practical
model acceptable to both the utility and the end user of electricity. This article is meant to
help understand the physical meaning of the instantaneous power component and relate the
characteristic values of these components (amplitude, phase and frequency) to basic model of
apparent power. This study has two parts: First reviews the energetical factors in systems
with sinusoidal voltage. In the second part, the method of analysis for sinusoidal system is
extended to non sinusoidal voltage situations.
Chapter 2
sinusoidal
systems,
voltage
and
current
are
defined
as
v 2V sin t and
i 2I sin(t ) respectively. The resolution of current into active (in phase) and reactive
where
p vi p p pq
(10)
(11)
pq Q sin 2t ; Q VI sin
(12)
P is the active power or also called the effective or real power. The component p p has the
average value of P and it pulsates in unidirectional manner between 0 to 2P.
The component pq is an oscillation of power at double supply frequency. The component
transfers energy back and forth between the sources and the linear loads. The amplitude of
oscillation of pq is the reactive power Q .
It is necessary to use the terms received and delivered to designate the flow direction of
reactive power. The time average of the energy transferred by pq is zero. The oscillations in
pq are caused by the property of the capacitors and inductors to store the energy in form of
pm p q J
where J =moment of inertia
= mechanical angular velocity
d
dt
(13)
Since the system is assumed lossless pm = 0, (once the rotor is brought to synchronous speed
the steam to turbine can be cut) and (13) can be written as,
Q sin 2t J
d
dt
(14)
The angular velocity will oscillate around a synchronous value with an excursion of . The
maximum kinetic energy delivered or received is obtained from (14)
J 2 / 2 Q /
(15)
This equation emphasizes the energetical meaning of the reactive power as a value
proportional with maximum energy transferred back and forth between prime mover and
load.
If a perfectly balanced three-phase purely reactive load is supplied by a p-poles alternator the
mechanical torque developed by the prime mover has the equation
T Q
P
sin 2t sin 2(t 2 / 3) sin 2(t 2 / 3) 0
2
Here Q is the reactive power flowing in one phase. This result shows that under balanced
conditions the alternators do not deliver/receive kinetic energy under steady state operation.
The oscillations in pq takes place among the three reactive loads via the inductive coupling
of the armature and field windings.
Figure 2-1 to Figure 2-4 gives set of examples that will reinforce the physical meaning of the
instantaneous rate of energy transfer pq . Figure 2-1 shows a hypothetical dc machine is
coupled with a prime mover which delivers mechanical power pm' Pm sin 2t . Assuming the
dc machine to be lossless,
v k ; m sin t; m
2V
k
(16)
(17)
T ki
(18)
pm'
P sin 2t 2 Pm cos t
m
k k m sin t
k m
(19)
2 2VPm
sin t cos t Pm sin 2t
k m
Thus Q Pm . This result tells us that we are dealing with a pendulum where the energy is
changed back and forth between the prime mover delivering pm' and the kinetic energy
stored in the masses turning with the alternators shaft. The amplitude of this instantaneous
power is Q .
To illustrate further the physical meaning of this concept the power source pm' is replaced by
with a flywheel, Figure 2-2, or a torsion bar, Figure 2-3, or a combination, Figure 2-4.
In general case, Figure 2-4, the torque is
T ki J
d
K dt
dt
(20)
(21)
This is electrically equivalent to the parallel combination of inductor (Le) and capacitor (Ce),
J
k2
hence Ce 2 and Le
. The expression for pq obtained from (16) and (21) shows that
k
K
the amplitude of the oscillation of power is
K
J
Q V 2 2 2 WJ WK
k
k
(22)
J 2m
V
J CeV 2
2
k
2
WJ
where
K m2
V
2
WK
K
Le I
2
k
2
are the maximum energies stored in the flywheel and spring respectively in the process of
mechanical pendulation.
The definition of Q as the amplitude of an instantaneous rate of energy transfer with average
energy transfer nil can be extended to any type of energy conversion. Consider for example a
time-varying resistor with a conductance
g t g cos t
(23)
The current supplied by the source is i gv 2Vg cos t and the instantaneous power
pq V 2 g sin 2t
with the amplitude Q = V2g as a measure of the reactive power. The energy supplied by the
source is stored in the form of thermal energy in a heat sink when g 0 0 t / 2 and
returned from the heat sink to the source g 0 / 2 t and the resistor generates
electric energy. It is difficult to find such a resistor in nature; nevertheless this example paves
the road for the next paragraphs where more realistic time-varying devices are considered.
Chapter 3
where
pD1h QD1h Fc (1, h) ; pB1h QB1h Fs (1, h)
(24)
(25)
(26)
The instantaneous power components pD1h and pB1h are non sinusoidal oscillations which are
composed of two elementary sinusoidal oscillations of equal amplitudes at the frequencies
(27)
Since QD1h , QB1h can take both positive and negative values, it is possible to designate four
quadrant flow direction to the elementary reactive power Q1h .It is worth noting that making
h=1, pB1h oscillation becomes of the form (12) encountered in sinusoidal current case.
To investigate the nature of pD1h and pB1h we will address the case of time varying resistor
having conductance
g G1 G3 (1 2 cos 2t )
(28)
This particular load yields a current with a fundamental and a third harmonic.
i vg 2 I1 sin t 2 I 3 sin 3t
(29)
I1 VG1; I 3 VG3
(30)
where
p p P(1 cos 2t )
pq V 2G3 Fc (1,3)
(31)
Comparing (31) with (24) results in Q QD13 V 2G3 0 , i.e. delivered. The non sinusoidal
oscillation of pq are caused by the time varying conductance G3 (1 2 cos 2t ) which can be
viewed as a reversible electro-thermal energy converter similar to resistor by (23) in parallel
with a constant resistor of conductance G1.
The practical implication of the time varying resistor for filtering application is evident when
we analyze a simple but realistic non linear resistor with the i-v characteristics
i Av3 2 2 AV 3 sin 3 t
2 I1' sin t 2 I 3' sin 3t
(32)
where
I1' 3AV 3 / 2 ;
I3' I1' / 3
(33)
Comparing (29) with (32) results that a non linear resistor can be viewed as a time varying
element. Moreover, by adjusting G3 AV 3 / 2 , the third harmonic generated by the non linear
resistor can be cancelled by the third harmonic produced by the time varying resistor (28).
In a similar way the reactive power of type pB1h is produced when a non linear inductor is
energized. For example assuming the flux current characteristics ai1/3 and keeping in
mind that
vdt ( 2V / ) cos t
i 2 I1 cos t 2 I3 cos3t
10
(34)
3
I1 (V / a )3 ;
2
I 3 I1 / 3
(35)
The fundamental reactive power Q1 and the elementary harmonic power QB13 are amplitudes
of totally different frequencies of oscillations and cannot be added algebraically. The
elementary reactive power QB13 caused by the nonlinear inductor can be cancelled in the
following way:
A time varying resistor is connected in parallel with the non linear inductor. The conductance
of this resistor is given as
(36)
where
I q''1 VG3'' / 2;
I q'' 3 VG3" / 2
with
a 3rd
harmonic current
conductance
If QB" 13 V 2G3" QB13 VI3 then the 3rd harmonic current I3 in (34) will be cancelled by the
current I 3" . This observation reinforces the usefulness of a unified theory on the nature of
reactive power.
11
Figure 3-1 Generation of Reactive power in a circuit with a Triac controlled Resistor
(a) Circuit
(b) Voltage, Current and Fundamental Current
(c) Equivalent Conductance time variation
(d) Equivalent Circuit based on eq. (37)
(e) Equivalent circuit based on eq. (38)
A triac in series with a resistor R, Figure 3-1, is triggered at t . In this circuit the
fundamental current is lagging with respect to the voltage and a fundamental reactive power
Q1 as well as harmonic reactive powers are created. This circuit is examined in the same
way as in earlier examples. This circuit has an equivalent time varying conductance
g 0 for 0 t ; t
1
G for t ; t 2
R
g G
an cos 2nt bn sin 2nt
n 1,2,3
an
sin 2n
;
n
bn
12
1 cos 2n
n
(37)
sin 2
i 2VG
2
where
Bh
1 cos 2
h 3,5,7
an an 1
;
2
Dh
bn bn 1
;
2
(38)
h 2n 1
where
p p P(1 cos 2t );
P V 2G (sin 2 ) / 2 /
pQ1 Q1 sin 2t ;
Q1 V 2G 1 cos 2 / 2
QB1h V 2GBh
QD1h V 2GDh
and the function Fc (1, h) and Fs (1, h) are same as in (25) and (26).
For this purely resistive circuit there are no components able to store and return energy.
Nevertheless the equivalent circuits derived from(37), Figure. 3-1(d), or from (38), Figure 32e, show that besides a constant resistor, there are an infinite number of fictitious timevarying resistors all connected in parallel. The active power is dissipated by the constant
resistor. Each fictitious time-varying resistor has a time-average dissipated thermal energy
zero. The time-varying resistors however, are the sources of power oscillations of types pB1h
and pD1h . The fundamental reactive power is caused by a fictitious equivalent inductor,
Figure 3-1 (e), obtained in an identical manner as in(36).
13
Chapter 4
P Vh I h cos h ;h h h
(39)
(40)
PH ph Vh I h cos h
h 1
(41)
h 1
(42)
where
ia 2 ( I h cos h ) sin h
ir 2 ( I h sin h ) cos h ; h ht h
(43)
p pa pqR
(44)
where
pa P Ph cos 2 h
h 1
(V
m , n 1
m n
14
I cos m ) Fc (m, n)
m n
(45)
(46)
(47)
(48)
h 1
m , n 1
m n
n nt n
m mt m ;
and Fs (m, n), Fc (m, n) are generalized forms of functions (25) and (26).
The current can be resolved in two different components in other way than(42). The first
component i p is the in- phase component, having exactly the same waveform as the distorted
voltage. Assuming a scale factor K
i p 2 K Vh sin h
and from
2
1
K
P vi p dt 2 Vh sin h dt KV 2
T 0
T 0 h
results K P / V 2 , where V
2
h
i p 2( P / V 2 )Vh sin h
(49)
2
h
P /V.
The second component, called the quadrature component, is iq i i p and can be subdivided
in two components with harmonics 900 out of phase,
iq ir iqD
(50)
IR
(I
h 1
15
sin h ) 2
(51)
I QD
I
h 1
cos h ( P / V 2 )Vh
(52)
(53)
(54)
where
h 1
m , n 1
m n
(55)
h 1
Comparing (55) and (45) it is seen that both the instantaneous powers p p and pa have the
same average power and oscillation frequencies. The amplitudes of oscillations are however
not same. This difference is observed due to the fact that in Budeanus definition pqD is
included in pa . Subtracting (45) from (53) gives
p p pa pqD
(56)
Realizing that p p in (53) has an identical definition as the one used in (10), where the in
phase current happens to be sinusoidal and therefore has the same waveform as the voltage,
results that pa in (44) is not the intrinsic power.
The rms current computed from (49),(51) and (52) is
2
I 2 I p2 I R2 IQD
(57)
(58)
2
2
QR2 V 2 I R2 QBh
QBmn
(59)
where
m n
2
2
2
QD2 V 2 IQD
QDh
QDmn
h
QBh Vh I h sin h
16
(60)
m n
(61)
QBmn Vm I n sin n
(62)
(63)
(64)
Each of the elementary reactive power defined in(61), (62), (63), (64) are recognized as
amplitudes of the oscillations in the instantaneous power ( pq ) . From(53), (55) and (46)
results
pq QBh sin 2 h
h 1
m , n 1
m n
Bmn
(65)
m , n 1
mn
The total compensation of pq will lead to a unity power factor. The oscillations with the
amplitude QBh and QBmn are typical for linear or nonlinear inductors or capacitors and can be
compensated, in theory, with the help of simple shunt reactances (linear capacitors or
inductors). The terms with the oscillations of amplitude QDh and QDmn ,are typical for
nonlinear resistors or lossy nonlinear reactances and can be compensated only with the help
of time-varying impedances designed as harmonic cancellation devices or active filters.
The current division given in (42) leads to a widely accepted definition for apparent power
division
S 2 P2 QB2 D2
(66)
where
Q QBh
h 1
2
B
(67)
m , n 1
m n
2 2
m n
(68)
Inspecting the instantaneous power p qR given in (36), one will conclude that the components
of term D cannot be recognized in the amplitudes of the oscillations, therefore no physical
meaning can be attributed to D. Only the terms QBh can be recognized in the amplitudes of the
oscillations. Moreover QB may be nil while QBh 0 and minimization of QB does not mean
a reduction of the oscillations of p qR . Thus the equation (68) is misleading giving the
17
impression that QB can be partially or totally cancelled while in reality the oscillations of
power will take place.
18
Chapter 5
Conclusion
1. In non sinusoidal situations the apparent power can be defined from
S 2 P1 PH QF2
(69)
QF2 QB2 D 2
(70)
(71)
where
or
The definition (70) lacks physical meaning and the measured value of QB provides
useless information for the purpose of harmonic cancellation or power improvement.
2. No direction of flow can be attributed to the reactive power QF . The term QF can be
divided in a multitude of elementary reactive powers for which complex (four
quadrant) representation becomes possible. This is evident when (65) is rewritten in
the form where
pq QDh jQBh sin(2 h h )
h 1
m , n 1
mn
19
20
Chapter 6 REFERENCES
1. Mr. Roberto Gonzalez, Jesus Lopez, Pablo Sanchis, Luis Marroya,
Transformersless
4. Martina Calais, Johana Myrzik, Ted Spooner, Vassilios G. Agelidis, Inverters for
SinglePhase Grid Connected Photovoltaic Systems- An Overview, IEEE, 2002.
5. Fritz Schimpf, Lars E. Norum, Grid Connected Converters for Photovoltaic, State of the
Art,
Ideas for Improvement of Transformerless Inverters, Nordic Workshop on Power and
Industrial Electronics, June 9-11,2008.
6. Eduardo Roman, Ricardo Alonso, Pedro Ibanez, Intelligent PV Module for Grid-Connected
PV systems, IEEE, 2006
21