Professional Documents
Culture Documents
3, July 1994
1600
Augmentin a conventional (mechanicall switched) phaseanale reauyator with a static converter oYsmall MVA rating,
enables f i e phase angle regulator to perform the functions of an
SPS in the context of FACTS.
A stract:
TEST SYSTEMS
2.1
INTRODUCTION
D. Maratukulam(MIEEE)
Electrical Systems Division
Electric Power Research Institute
Palo Alto, CA 94303
2.2
o--ys-=to
xr
1601
CB2
cB3
CB4
System-Ill
System-IV
IBII
T1
Figure5:
(a)
(b)
(c)
(d)
Type-B SPS
Type-C SPS
Type-D SPS
Type-D2 SPS
1602
in [l]. For ease of reference, a brief description of-each SPS
scheme is given in this section.
Type-B SPS, Fig. 5(a), controls the magnitude of the injected
voltage in discrete steps. The phase angle of the injected voltage
is either +90" or -90" with respect to the corresponding system
phase voltage. Harmonic content of the injected voltage is
negligible, and Type-B SPS does not require harmonic filters.
Type-C SPS, Fig. 5(b), also operates based on the discrete step
control principle. The phase angle of the injected voltage can be
either +60 or -120" with respect to the corresponding system
phase volta e In contrast to Type-B SPS, the electronic
converter of !ype-C SPS is at low voltage level and it does not
require high voltage platform. Harmonic content of Type-C SPS
voltage, similar to that of Type-B SPS, is negligible.
Static converter of Type-D SPS, Fi . 5(c) is composed of two
GTO based voltage source inverters fVSl's), and a dc link. TypeD SPS provides continuous control over the magnitude and phaseangle of the injected voltage. The voltage phase angle can be
adjusted between 0" to 360". Output voltage of Type-D SPS
contains harmonic components. The harmoniccontent is minimized
by elaborate pulse-width modulation (PWM) switching approaches
and/or high pulse converter arrangements [l]. Practically Type-D
SPS requires harmonicfilters.
Type-DP SPS, Fig. 5(d), is a mechanical phase angle regulator
which is augmented with a static converter for rapid phase angle
regulation. The static converter is the same as that of Type-D
SPS. Type-DP SPS is intended to enhance power s stems
behaviour during steady-state and small-signal dynamics. {teadystate power flow regulation is carried out by mechanical tap
changing. Injected voltage Vp, Fig. 5(d), is controlled by the
electronic converter for rapid phase angle control to mitigate small
signal oscillations.
3.
3.1
STUDY RESULTS
s2+b
--sz +aS
S2
+b - s z +cS+d
60 Hz
High Pass
(1+sT,)(1+sT2)
k4
(1+ sT,xl+
(b)
ST,)
Compensator
Notch
sZ+b
---
mm
SI
(1 +ST,)(I
+ST,)
Figure 7
1603
For studies reported in this paper, no control is exercised on phase
angle SE of Type-D and Type-D2 SPS's. SE can be controlled
such that the net reactive power exchange between the static
converter system and the network is at a pre-specifiedvalue, e.g.
zero. When an SPS is used for functions other than mitigation of
torsional oscillations, its control path should be equipped with
proper torsional filter(s), similar to power system stabilizers
(PSS's), to avoid excitation of torsional modes.
0
3.1.1 Case-1
0.5
15
25
25
Time($
0.15
(b)
0.1
0.1
0.1
3.1.2 Case-7
To investi ate the impact of an SPS Location on its torsional
dampin esfect, eigenvalue studies were conducted on System-lI
(Fig. 1 Oy. Nodes A, 8, C, and D in Fig. 10 show different locations
of the SPS, which provides damping for the torsional modes of T-G
set #I.
T-G sets #2, #3, and #4 are represented by ideal voltage
sources behind fixed impedances.
Fi ure 1 1 shows the damping effect of a Type-D SPS on the first
S8R mode (7.6 Hz) of System-ll. Table 1 shows the maximum
magnitudes of the injected voltages by different SPS's for damping
the first mode of T-G set # I , when X,varies from 0.1 per unit to 0.6
per unit. Figure 1 1 and Table 1 show that:
0
(2)
05
15
Time(s)
0.15
0.1
e
a
-0.1
-0.15
0.5
15
25
20
20
Figure 8
A:
40
40
60
CompmsationLevel (96)
80
100
80
100
1604
0.4
60
..
(a)
(b)
-0.6
0.1
03
03
0.4
0.3
0.6
No SPS In service
Type-D SPS (at Node A) in service
(a)
(b)
0.6
0.4
4-
......;.........................................................................................
'-I
do
0
-0.6
0.1
(b)
0.2
03
0.4
0.3
I
0.6
-,
03
03
0.4
05
1.2
1.4
16
1.8
(a)
(b)
.....
..........................................................
..........
.......
0.6
(b)
I
03
02
0.4
05
0.6
No SPS in service
~ y p e - DSPS (at node D) in service
Figure 11: Real part (a) of the eigenvalue of the first torsional
mode of System-ll (Fig. I O )
3.2
Enhancement of Transient Stability
The reason for rotor swings as a result of large-signal disturbances
is the mismatch between input mechanical power and output
electric power of the generator after the disturbance inception.
Depending upon a system's parameters and its operating point, if
the mismatch is more thari a nlaximum value, the generator will lose
stability. The conceptual approach to mitigate rotor swings and
enhance transient stability is to reduce the power mismatch by
modulatin the generator output power after disturbance inception.
Since an gPS can rapidly modulate voltage phase-angle and as a
result, power transfer through the associated line, it is a natural
candidate for enhancing transient stability. An SPS can enhance
transient stabili based on either bang-bang or continuous control
strategies [iI,&
This section demonstrates technical feasibility of Type-D SPS for
enhancing transient stability of System-Ill, Fig. 3. Depending upon
the pre-disturbance operating conditions, type of fault, fault
location, and line switching process, the required phase shift to
enhance transient stability can be as high as 60'. Type-B and
Type-C SPS's provide the required voltage phase shift by
adjusting on1 the magnitude of the injected voltage. Thus, the
magnitude or the injected voltage and consequently, the MVA
ratings of Type-B and Type-C SPS's will be unacceptabty high to
enhance transient stability under extreme conditions. However, the
principles of operation of Type-D SPS also permits phase-angle
control of the injected voltage to achieve the required phase shift
based on the optimum magnitude of the injected voltage. Our
preliminary studies show that the MVA rating of Type-B SPS
......................
............. .......
. . . . .
...................
0.1
............ ....
0.4
0.8
(b)
I
0.6
(a)
(b)
No SPS in service
Type-D SPS (at node C) in service
(a)
(b)
0.6
-0.4
-0.6
0.4
Time(s)
...............................................................................................
2
i
(a'
0.1
0.2
_1
L 0O
-0.4
-0.6
No SPS in service
Type-D SPS (at Node B) in service
(a)
0.A
0.2
0.4
0.6
0.8
.
1
'...................................................
1.2
1.4
1.6
1.8
Time(s)
Figure 13: Generator angle deviation of System-Ill as a result of
a 3-15-3 cycles fault
should be-about 35% higher than that of Type-D SPS to maintain
the stability of System-Ill for close-in faults. The MVA rating of
T pe C is always higher than that of Type-B to maintain stability
o!Sy&em-lll after various large disturbances.
3.2.1
Case.-S
1605
3.2.2
Case 4
AREA-2
AREA- 1
When an SPS is used to enhance transient stability after a largesignal disturbance, rapid and significant changes of the phaseangle and/or magnitude of the injected. voltage result in air-gap
transient torques, which in turn can impose severe transient
torsional stresses on T-G shaft sections.
Figure 14 shows transient mechanical stresses imposed on the
shaft segment LPB-G of the T-G set of System-Ill, as a result of a
3-25 cycles fault at point C. Figure 14 indicates that when the SPS
attempts to enhance transient stability it can adversely affect
transient shaft stresses. This effect is particularly pronounced for
the first torsional frequency which is usual1 the dominant oscillatory
mode. Figure 15 shows similar adverse egect after a 3-20-3 cycles
disturbance.
One approach to reduce the above mentioned SPS adverse effect
is to ramp (up and down) the amplitude and phase-angle of the
injectedvoltage, instead of permitting the SPS to introduce sudden
changes. The SPS controls can be specifically designed to
alleviate transient shaft stresses as a result of lar e-signal
disturbances. This concept is further explored in Section 85.
.-
ad
0.3
I
1.2
I
1.5
Case-8
-2 3
-9
%!O
I2-3
0.3
I
1.2
'
-0.4
1
1.5
-0.2
0.2
Real (l/s)
Figure 17: Eigenvalue loci for System-li when power transfer
from Area-1 to Area-2 is increased from 180 MW to
600 MW
(a) SPS disabled
(b) SPS in service (at point A)
0.3
I
1.2
1
1.5
3.3
Case-5
0.2I
-0.2
0.2
Real (l/d
Figure 18: Eigenvalue loci for System-ll when power transfer
from Area-1 to Area-2 is increased from 180 MW to
600 MW
(a) SPS disabled
(b) SPS in service (at point B. Max Vsps4.18 Pu)
(c) SPS in service (at point B, Max Vsps=O.26 Pu)
1606
unit, the SPS damping effect is less than the case when it was
located at point A (Case-5).
3.3.3
Case-?
Figure 19 shows the eigenvalue loci for S stem4 when the SPS is
located at point C, and power transfer {om Area-1 to Area-2 is
increased from 180 MW to 600 MW. Figure 19 shows that in
contrast to Case-5 and Case-6, the SPS fails to provide adequate
damping for the inter-area oscillatory mode under heavy tie line
loading conditions. The main reason is that due to the series
capacitors (Fig. 16), line L4 is transferrin most of the inter-area
power exchange and power modulation ofline L5 is not adequate
to damp inter-area power oscillations
Figure 20 shows the system transient response to a 3-15 cycles
three-phase fault on line L5. Under stead state condition, both tie
lines transfer total of 430 MW from &ea-1 to Area-2. The
disturbance does not result in transient instability of the system;
however, it excites the inter-area oscillatory mode. Similar to the
ei envalue analysis results of Fig. 19, Fig. 20 also indicates that the
S8S (located at point C) does not have any noticeable impact on
the inter-areamodeat the given operating point.
Loop Flow Control
3.4
Case-8
3.4.1
desired path. Power flow analysis shows that a voltage phaseshift of 17.2" can prevent the loop flow. This is achieved by
injecting 0.31 per unit voltage at the angle 98.6", by means of a
Type-D SPS. Since speed of response is not a concern for
steady-state loop flow control, a conventional phase-angle
regulator can also perform the function. A conventional phaseangle re ulator can provide voltage phase shift in the range of -30"
to +30 ?or 0" to +60" or 0" to -SO0), in steps of 2". Although the
Type-D SPS of System-IV (Fig. 4) is primarily used for loop flow
control, it is equipped with additional controls for rapid power flow
adjustment in corridor "B"according to a predetermined strategy
during specific system transients. This mode of operations of the
SPS in this paper is referred to as "rapid loop flow control".
Rapid loop flow control can be used to maintain transient stability of
System-IV, when dc link (1) is subjected to a permanent fault.
Transient stability studies show that under a full-load operating
condition, a sudden drop in power transfer of dc link (l), from 3200
MW to 1600 MW, results in the loss of stability of T-G set number
one. Sudden power drop of dc link (1 can be a result of losing one
of the poles of the bipole system. kigure 21 shows rotor angle
deviations of T-G set number one when the power transfer of dc
link (1) is permanently reduced from 3200 MW to 1600 MW. Figure
21(a) corresponds to the case when the SPS only prevents loop
flow and does not react to the dc link disturbance. Figure 21(a)
indicates that the disturbance results in instability of T-G set
number one.
Figure 21(b) corresponds to the case when prior to the dc link
disturbance, the SPS is not in service, and the system experiences
loop flow through corridor " B . Figure 21(b) indicates that the
disturbance does not result in the instability of T-G set # l .
Figure 21(c) shows the system response to the same disturbance
when prior to the disturbance, the SPS is in service and prevents
70
0.81
imer-areamode
35
0-
-3s
-70
'
0.2
-0.2
Time(s)
10
0.2
0
Real (l/s)
-70
1 Time(s) k
1 Time(s) k
10
0-
-50
10
12
I
14
16
18
20
(8)
10
1607
steady-state loop flow through corridor "8". The SPS is controlled
to react to the disturbance and rapidly reduce the magnitude of the
injectedvoltage to zero, and remains inactive after the disturbance
detection. This requires communication among the SPS, the HVdc
converter station, and the T-G set #l. When compared to Fig.
21(a), Fig. 21(c) clearly illustrates that rapid response of the SPS
can prevent instability of T-G set #I. The reason is that the SPS
is rapidly disabted after the disturbance and permits loop flow,
thus, higher level of power transfer to the load centre is possible
which in turn assists transient stability.
Figure 21(d) also shows the system response to the same dc link
disturbance. Prior to the disturbance, the SPS is in service and
prevents loop flow through corridor "B". After the disturbance
inception, the SPS relinquishes its steady-state loop flow control
function, and dynamically adjusts power flow through corridor " B ,
to mitigate swings of T-G set # l . Rotor speed deviation is used as
the feedback signal and communicated to the SPS. Figure 21(d)
clearly shows that the SPS can effectively enhance the transient
response of T-G set # I . The SPS starts the steady-state loop
flow control again after the swing amplitudes subside below a
specified value.
It is interesting to note that the system transient responses
illustrated in Figs. 21(a) and 21(b) will be the same, if Type-D SPS
is replaced by a mechanicallyswitched phase-an le regulator. The
important conclusion is that loop flow controy by means of a
conventional phase-angle regulator has an adverse effect on
transient stability of T-G set # l . The instability problem can be
resolved by reducing steady-state power transfer limits of the
parallel lines of corridor A. This indicates that total power generation
of T-G set #1 should be reduced to assure stability, when loop
flow is prevented. This may not be an economically viable
strategy. However, if instead of a conventional phase-angle
regulator an SPS is used to prevent loop flow, not only does it not
result in instability (Fig. 21(c)), but it can assist in enhancing
system transient behaviour (Fig. 21(d)).
Alleviation of Transient Torques
3.5
Case-9
Figure 22 shows the torsional response of the T-G set to the 3-203 cycles disturbance (Section 3.2.2) with and without SPS's in
service. The maximum voltage that each SPS can inject in the
system is 0.58 per unit. SPS's are inactive during steady-state
conditions and react to the electrical disturbance only to mitigate
transient torsional oscillations. Figure 22 indicates that:
Both Type-B and Type-D SPS's can reduce peak transient
stresses of shaft sections.
For the same magnitude of injected voltage, Type-D SPS is
more effective for reducing shaft stresses than Type43 SPS.
The reason is that Type-D SPS also regulates the phase angle
of the injected voltage to maximize the required damping effect.
The impact of both SPS configurations on the first few peak
torques is relatively small.
The studies show that mitigation of transient torques by a SPS
does not have any noticeable adverse effect on the generator
swing mode of oscillations.
4.
CONCLUSIONS
2 3
7
&
. - -3
4-6
0.3
0.3
0.3
Timeb) o!9
1.2
I
1.5
6 3
bo
c-"-3
4
I
I Time(s) I
0.6
0.9
1.2
1.5
o16
1.2
I
1.5
Timeb) o!9
(E\
5.
REFERENCES
1608
Flows", IEEE Trans., Vol. PWRS-6, No. 3, pp. 1024-1034,
1991.
L. G ugi, "Unified Power-Flow Control Concept for Flexible
AC fransmission Systems", IEE Proceedings, Pt. C, Vol.
139,NO. 4,pp. 323-331,July, 1992.
[ll] R. Baker, G. Guth, W. Egli, P. Eglin, "ControlAlgorithm for a
Static Phase Shifting Transformer to Enhance Transient
Stabilitv of b r a e Power Svstems". IEEE Trans.. Vol. PAS-
['01