See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266481090

Comparison of flexible systems of reactive power compensation

Article

CITATIONS READS

7 225

2 authors, including:

Janis Zakis
Riga Technical University
112 PUBLICATIONS   915 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

New Converter Topologies and Control Methods for Electronic Power Distribution Networks View project

Electromagnetic Compatibility Issues in Wireless Power Transfer Systems View project

All content following this page was uploaded by Janis Zakis on 20 February 2015.

The user has requested enhancement of the downloaded file.

 5th International Symposium
„Topical Problems in the Field of Electrical and Power Engineering”,
Doctoral School of Energy and Geotechnology
Kuressaare, Estonia, January 14 – 19, 2008

Comparison of flexible systems of reactive power compensation

Janis Zakis, Ivars Rankis
Riga Technical University
janis_zakis@one.lv, rankis@eef.rtu.lv

Abstract As we can see in fig.1, there is capacitor and reactor

in each phase. Capacity of capacitor is accepted
The following paper presents two possible systems
constant and it corresponds to maximal
of flexible compensation of reactive power. One of
compensating lagging power of load, i.e.
them is system of constant capacitors and thyristor
regulated reactors and the second one is modulated QC = QLD max . If consumed reactive power of load is
rectifier. There are compared technical and decreasing in procession, then reactors are regulated
electrical parameters and their changes of both in order to decrease total reactive power. Total
systems. The calculation expressions of parameters reactive power of reactors is
of both compensation systems are given. QL max = k R × QC , (1)
Keywords where kR defines necessary dimensions and costs.
Common minimal reactive power of compensation
Capacitors, thyristor, reactors, transistor, system is
regulation, modulation, regulation angle.
Qk min = QC (1 − k R ) . (2)
Introduction Depending on ratio minimal power and nominal QC,
Compensation of reactive power is still actual in parameters of reactors will change. Approximate
many enterprises, where load is changing very fast weight of reactor is [3]
with time. It means that there is a necessity for G L ≈ 3I L2 L , (3)
flexible reactive power compensation system that
could regulate balance of reactive power fast and where in delta connection of reactors
smooth. We can get such regulation using combined Ul U2
QL max = k R QC = 3U l 3 = 3 l . (4)
LC compensation system [1], where reactor current ωL ωL
is regulated with a thyristor regulator, but capacity
From here we can get inductance for one reactor
of capacitors is constant. Besides this there is one
more possible solution where LC elements are used U l2 3 U2
L= =3 l (5)
together with modulated rectifier [2]. k R QC ω k R Qω
It is possible to realize both variations and they Effective current of reactor
provide flexible compensation of reactive power and
high-speed operation. The point and task of this
U l U l k R QC ω k R QC
paper is to compare technical parameters of such I Lef = = = . (6)
systems. ωL ωU l2 3 3U l
Finally the total weight of all reactors of the system
1 Thyristor regulated parameters of LC
k R2 QC2 × 3U l2 QC k R
system. GL ≈ 3 = . (7)
9U l2 k R QC ω ω
There is a smooth regulation system of reactive
power with thyristor regulated reactors and Accepting different kR values from 0,5...1, we can
capacitors with constant capacity [1] depicted in get necessary weight depending on maximal
fig.1. compensating reactive power QC. (fig.2.).

Table 1. The weight of reactors (kg) at different

powers of compensators
kR
Qc, kvar 0,5 0,7 0,9 1
20 32 45 57 64
40 64 89 115 127
60 96 134 172 191
Fig.1. Smooth regulation system of reactive power 80 127 178 229 255
with thyristor regulated reactors 100 159 223 287 318

99
 GL=f(Qc)

350
As we can see in fig.2, the weight of compensation
300 reactors grows in almost direct ratio to reactive
250 kR=0,5 power.
GL, kg

200 kR=0,7 The second element in compensation system is

150 kR=0,9

100 kR=1 capacitor the reactive power of which is QC. From

50
available data of capacitors there are system
0 parameters for different values of reactive power
0 20 40 60 80 100
Qc, kva r
given in table 2.

Fig.2. Weight of reactors depending on power of

compensator at different kR

Table 2. Parameters of compensation capacitors

Current, GC ,
Type Capacitors, kvar Steps A dimensions kg
Reactive power compensation devices with step 2,5 kvar
AKVS-22,5 2,5+5+5+5+5=22,5 1.2.2.2.2. 35 500x500x200 17,2
AKVS-37,5 2,5+5+10+20=37,5 1.2.4.8. 56 600x600x300 28,6
AKVS-57,5 2,5+5+10+(2x20)=57,5 1.2.4.8.8. 86 800x600x300 44
AKVS-77,5 2,5+5+10+(3x20)=77,5 1.2.4.8.8. 115 1000x600x300 59,2
Reactive power compensation devices with step 3,33 kvar
AKVS-22,5 3,33+6,67+12,5 1.2.4. 34 500x500x200 17,2
AKVS-35 3,33+6,67+12,5+12,5 1.2.4.4. 52 600x600x300 26,8
AKVS-72,5 3,33+6,67+12,5+(2x25) 1.2.4.8.8. 107 1200x800x300 55,4
AKVS-97,5 3,33+6,67+12,5+(3x25) 1.2.4.8.8. 143 1200x800x300 74,4

shoulder accordingly to reference between reference

2 Modulated compensators of reactive
power
With modulation, which is realized in diod-transistor
rectifier with C load in DC side (fig.3.), it is possible
to maintain quasi sinusoidal current, that leads or
lags behind network voltage for angle 90°.
Fig.3. Scheme of modulated compensator of reactive
power
Also in this scheme there is LC compensation unit,
but regulation of reactor current is realized switching
reverse transistors of full bridge rectifier each
phase current and the real one.
The algorithm of regulation is:
• If real current of the phase, which measured
with sensor (fig.3. in phase A there is
sensor SA) is bigger than reference signal
iRA of phase Au (in scheme it is acquired
proportional on 90 to leading voltage UCB,
t.i., compensator works with leading
current), then transistor VT1 is switched on
and capacitor voltage works against iA
increment;
• If real phase current iA is smaller than iRA,
then VT4 is switched on. As a result
capacitor C voltage contributes growth of
iA.
Accordingly current iA fluctuates near sinusoidal
curve of the referent instantaneous value. For
determination in system we can initiate an element,
which maintains fluctuations in constant level ∆I1. If
the upper transistor is switched on then differential
equation is
di , (8)
U Am sin ωt − 0,5U C = L
dt
but when lower transistor is switched on,
di
U Am sin ωt + 0,5U C = L . (9)
dt
If ∆I1 in both regimes is constant then equal growth
and drop decrease times of the current in switching

100
period U Am sin ωt = 0 . Then there will be the
highest switching frequency of transistors As we can see in fig. 4 capacitor voltage UC of
compensator depends on maximal control frequency
UC . (10)
f max = fmax of transistor that depends on modulated current
4 L∆I 1 difference ∆I1 of transistor.
This equation also can be used as calculation of At the previous assumptions inductance of reactor is
input reactor L using reference fmax. That is limited
2U l2 (18)
by power loses of transistor dependence on L=
switching frequency. Accordingly Q(U lk f f *10− 3 − 2ω )
UC . (11) And from this expression in fig.5 inductance L
L≥
4 f max ∆I 1 values depends on reactive power Q.
Voltage value of capacitor in steady state regime is L=f(Q)

approximately equal with input line voltage 0,02500

amplitude of the rectifier bridge. Capacitor voltage
at input current offset of phase on 90° is 0,02000

0,01500 kf=2

L, H
kf=3

U C ≈ 2U l + 3I 1mωL , (12) 0,01000 kf=4

where I1m is amplitude of the reactor current which 0,00500

is connected with compensating reactive power. 0,00000

20000 40000 60000 80000
As a result Q, v ar

Q 2
U C ≈ 2U l + ωL , (13)
Ul
Fig.5. Reactor inductance at different reactive
or, noticing L expression, powers of reactors and frequencies
2U l2 4 f max ∆I 1 . (14)
UC = As we can see necessary inductance L depends on
4U l f max ∆I 1 − 2Qω
reactive power and frequency. The smaller reactive
As we can see, to get normal operation
power the bigger is inductance L. Similarly L is
2Qω , (15) growing with decreasing of frequency.
f max .∆I 1 >
4U l The common weight of reactors in compensator can
be expressed as [ 3 ]
It means, if Q grows, ∆I1 must be increased, because Q2
increasing of fmax is not possible because of the GL ≈ L 2 ; (19)
Ul
power loses of transistor commutation.
Accepting, that ∆I1 is connected with Q modifying,
whereby 2Q (20)
GL =
Q , (16) (U l k f f * 10 −3 − 2ω
∆I 1 ≈
4 × 10 3
According to this expression calculated weight of
Then frequency must be
reactors depending on Q at different maximal
2ω10 3 frequencies are depicted in fig. 6.
f max > = f *. (17)
Ul G=f(Q)
*
At Ul=380V, ω=314 1/s this parameter f =1173 Hz. 300,0

In Fig. 4 and 5 from the gained expressions is 250,0

depicted L=f(Q), UC=f(f) at different fmax and ∆I1 200,0

kf=2
values. It is assumed that, that ∆I1 value is connected
G, kg

150,0 kf=3
kf=4
with Q value: the bigger is Q, the bigger can be ∆I1. 100,0

In fig. 4.is depicted UC=f (fmax) at different 50,0

kf=fmax/f*=2, 3, 4. 0,0
0 10000 20000 30000 40000 50000 60000 70000 80000 90000

Q, v ar

Uc=f(f)

1200,00 Fig.6. The common weight of reactors depending on

1000,00
reactive power and maximal frequency
800,00
Uc, V

600,00

400,00
As we can see, reactor weight grows almost
200,00
proportionally to reactive power of compensator;
0,00 however at higher frequencies weight is smaller. If
comparing reactor weight at f*=2 (fmax=2,35 kHz,
2346 3519 4692

f, Hz

which is optimal) with reactor weight in scheme

Fig.4. Capacitor voltage at different transistor with thyristor regulated reactors in fig.2. Here at
modulating frequencies fmax Q=80 kvar the weight is 255 kg which is equal with

101
 reactor weight in 1st scheme at kR=1. But in the
scheme with regulated reactors kR can be smaller
than 1 and then reactor weight in 1st scheme will be
smaller than in the modulated rectifier scheme.
Modulation type compensator capacitor is uni-polar
and it works with small voltage ripple which
accordingly to theory of three-phase full-bridge
rectifier [4] is with relative ratio versus UC
amplitude 0,057 and frequency 300 Hz (if network
frequency is 50 Hz). In addition capacity of
capacitor must provide this ripple with amplitude of
phase current during time interval T/2 (full
increment time interval
I 1m
C≥ . (21)
12 × f × 0,057 × 2 × U C
Such reactive power of capacitor is rather small
0,057 2 U C2
QC = C × 12 × π × f . (22)
2 parameters – control systems, costs of
It is much smaller than reactor necessary reactive
power therefore we can accept, that main reactive
elements in compensator are reactors. For example,
References
at Q=20000 var (I1m=43.2 A) and Uc=1070 V
necessary C=590 µF. Reactive power of this
capacitor is 2068 var. It means that weight of
capacitor; dimensions and costs are smaller than
reactors in modulated scheme.
Conclusions
1. Compensator where thyristor regulated LC
system parameters are used, power of
capacitors, weight and capacity are bigger
than in modulated compensator of reactive
power.
2. Modulated compensator capacitor voltage
UC depends on maximal control frequency
fmax of transistors, which must be connected
102
View publication stats
with transistor modulated reactor current
difference ∆I1. The bigger is fmax, the
smaller is compensator capacitor voltage
UC.
3. The weight of reactors of compensation
system with regulated reactors depends on
minimal compensated reactive power. The
weight of modulated reactor compensator
depends on compensator reactive power
and maximal frequency.
4. Comparing both compensation systems it is
visible that reactor weight, if only in
regulated reactor system capacitor power
must be fully compensated, but in
modulated system maximal frequency is
not forced, are almost the same at
appropriate reactive power Q. It means that
both systems according to reactor
parameters are equal evaluated and as
uppermost must bee considered other
semiconductor switches and other.
1. Zakis J., Plataiskalns O. , Rankis I.
Comparison of Thyristor Regulated Three-
phase Reactors’ Schemes. Elektron. ir
elektrotehn. Kaunas, Nr.3 (59), 2005. –
p.84-87
2. Muhammad H. Rashid Power Electronics
Handbook. Academic press, Canada 2001,
895 p.
3. Ранькис И.Я. Оптимизация параметров
тиристорных систем импульсного
регулирования тягового электропривода.
– Рига: Зинатне, 1985. – 183 стр.
4. RaĦėis I. Energoelektronika. Rīga 2004,
159 lpp.