•
Improve efficiently soft-starter transients' immunityBy Delcho Penkov
& Alain Côte Schneider Electric
Summary
Abstract ........................................................................................................ 1
Nomenclature ................................................................................................. 1
Introduction .................................................................................................... 2
Overwiew of SCR functionning and electrical transient issues ......................... 4
Case study ..................................................................................................... 5
Modeling of the RVSS and case study power system in EMTP-ATP ............... 6
Analysis of the thyristor turn on current transient ............................................. 7
Overview of the voltage transient during switching .......................................... 9
Estimation of the risk of high current transient ............................................... 11
Development of protection sizing tool ........................................................... 12
Conclusion ................................................................................................... 14
Acknowledgements ...................................................................................... 15
Appendices .................................................................................................. 15
Vita ............................................................................................................... 16
COM-POWER-WP--EN Rev1 | 1
Improve efficiently soft-starter transients' immunity
Abstract
In this paper the authors present results of measurements and
mathematical analysis on MV Silicon Controlled Rectifiers (SCR, Thyristors)
for the purposes of the power electronics components protection during
turn on. Current transient was identified as responsible for damaging
soft-starters in field applications. This work is focused on the identification
of the major parameters playing role in the current transient.
Formulae for calculating the current rate-of-rise and risk identification
procedure are derived. A simple tool for practical SCR transient protection
design is described.
Index Terms — Soft-starter, current transient, EMTPATP
Nomenclature
φ Phase shift between current and voltage, in ms.
α Thyristor turn on delay with respect to voltage zero crossing, in ms.
COM-POWER-WP--EN Rev1 |
Improve efficiently soft-starter transients' immunity
Introduction
Silicon Controlled Rectifiers (SCR) or Reduced Voltage Soft-Starters (RVSS) are modern techniques used for smooth motor starting in MV power systems. Thus many papers on SCR design discuss motor or system protection issues like torque pulsations or harmonic reduction, however little is said about protection of SCR itself. Major electrical transient constraints on the semiconductors are overvoltages during switching off, and current rate of rise while turning on. For power systems with rated voltages of 5.5 kV and above the current transient may become important, and depending on the application, damage the semiconductors, some believe because long connecting cables are used. Since it is common to use one RVSS to start several motors sequentially, failure of the RVSS can lead to substantial production losses. Generally a 100 μH series reactor is likely to be sufficient. However, from an economical point of view and lack of space, it would be much better to determine a more rigorous way to size this reactor. Risk assessment is crucial, together with a clear explanation of the current rate of rise phenomenon. In this paper the authors investigate field measurements and mathematical models of the SCR in order to understand the mechanisms the current transient depends on. In a first step it came out that the transients depend on the immediate environment around the SCR including both upstream and downstream installation, and not only on the motor cable length. Further analysis helped us to build a mathematical constant parameters model in EMTP-ATP that reproduces accurately the system behaviour during thyristor turn on. This was important step since it allowed us to go further and by hand analysis establish a formula to calculate the current transient on turn on without performing a simulation. It allowed in-depth understanding of what the current transient was the most depending on. Consideration of the overall behaviour of the switching angle and motor acceleration during start made it possible to put forward recommendations to reduce the potential risk of SCR damage before installation of additional protection. Yet if this is not sufficient, calculation tool gives recommendation on the sizing of the protection reactance together with the expected current transient. The solution is generalised and user may account also for the installed protection capacitors. Unfortunately, due to a very high installation conditions dependency simple rule, just built on voltage/rated power/cable, is difficult to be formulated, given the necessity to make some preliminary calculations. Required input data concerns immediate upstream and downstream installation details, and how the soft-starter is set up. Practical tests confirmed the accuracy of the developed tool. Thus full SCR service continuity in optimized installation conditions will be ensured. The paper is organized as follows: Section II introduces the principles of the soft-starter functioning and the constraints it is exposed to. Section III is a brief overview of a field application where a soft-starter was damaged during motor start. Section IV presents the modelling of the soft-starter and the simplification assumptions that were validated by comparison with field measurements.Sections V and VI will describe the mathematical analyses made for deriving of the current rate-of-rise computation formula and estimation of critical transient voltage values. Section VII defines the critical moments (zones) during the motor start. Section VIII discusses the development of risk assessment tool capable of inductance sizing calculation.
COM-POWER-WP--EN Rev1 |
Improve efficiently soft-starter transients' immunity
The SCR driven motor start is based on the use
of parallel thyristors connected in reverse parallel
configuration to each other that are switched on
by a command signal. The command signal
is intentionally delayed from the voltage zero-
crossing so that a smaller current is provided
to the motor sufficient to start but lower than
the rated starting current.
The overall behavior is described in the Fig. 1.
(V,I)
t(s)Current zero time
Ithyristor
Vupstream
Figure 1 – Principle of controlled current in soft-starter.
tcurrent_zero = α - φ
The current through the thyristor stops naturally
on zero crossing. The moment of zero crossing
depends on the phase shift between the current
and voltage.
This phase shift varies during the motor start as
Overwiew of SCR functionning and electrical transient issues
the motor is equivalent to variable impedance and
decreases with the acceleration of the motor.
The command signal (represented by α ) is also
varied during the motor start.
Globally the behavior shown in Fig. 2 is observed:
(ms)
t(s)
t(s)
t(s)
leffinit = Veffinit.kd
Veffinit
init = f (Veffinit)
leff (% x Ir)
Veff (%)
Ramp time
Current limit
Rated current
Rated voltage
Full wavevoltage
Figure 2 – Basic evolution of main soft-starter depending variables.
Somewhere after passing in full-wave conduction
of the thyristors a parallel by-pass contactor is
closed and RVSS stops.
The user dependent settings are:
1. initial voltage, in % of rated voltage
2. ramp time (time to go to the current limit), in s
3. current limitation, in % of rated motor current.
Thyristor Functioning Principles
Electrical transient switching constraints
These are the surge voltage on switching off and
current rate-of-rise on turn on. The surge voltage
issue is solved by a parallel connected snubber,
sized according to the expected overvoltage and
energy. This is a standard issue, scheduled during
the development of the soft starter. Fig. 3 shows
an example on a 6kV power system.
The overvoltage on the thyristors goes up to
270 % of the peak rated single phase voltage.
Connecting 2 or 3 thyristors in series reduces
the overvoltage applied individually on them.
t(ms)55 59 63 67 71-12
-8
-4
0
4
8
12k(V)
Upstreamvoltage
Voltage acrossthe thyristor
: VP2F-VaP2F
: VaP2F
Figure 3 – Measurement of Voltage transient during a complete fundamental period.
(1)
COM-POWER-WP--EN Rev1 |
Improve efficiently soft-starter transients' immunity
The current rate-of-rise issue differs because
it depends on the installation conditions of the
soft-starter. Its solving is not generalized but case
dependent. An example of current turn on rate of
rise on 6 kV power system is shown in Fig. 4.
t(ms)55 59 63 67 71-1500
-1000
-500
0
500
1000
1500(A)
Turn on currenttransient
: IP1F
Figure 4 – Measurement of current through the thyristor during a complete fundamental frequency period.
As it can be seen there is a high current transient
on turn on.
If it overpasses the thyristor limit (100-200 μA/s
typically) the thyristor is damaged as well as the
softstarter itself. This is the subject of this paper.
Fig. 5 shows a zoom on the current during this
transient:
t(ms)63 63.05 63.10 63.15 63.20 63.25
(A)
-100
100
200
300
400
500
0
: IP1F
Figure 5 – Current transient on switching on of the thyristor.
In this paper the authors will focus on the main
parameters this transient depends on.
Case study
The problem of high current rate-of-rise emerged
on a 6 kV power system in an oil refinery based in
Spain (Fig. 6).
Grid
Power transformer
6 kV bus
Upstream cable
Soft-starter
Downstream cable
Motor
Figure 6 – Case study power system.
The power system is described hereafter:
1. Rated voltage of 6 kV
2. 30-40 m of 120 mm² cable upstream to
the softstarter, 2 conductors per phase
3. 320 m of 120 mm² cable downstream of
the softstarter, 2 conductors per phase
4. 3.1 MW Motor.
The motor has been started several times before
the soft-starter was damaged. Substantial
measurements on site, shown in Fig. 3 and 5
revealed high current rate-ofrise, over 170 A/μs
whereas the thyristors were only able to withstand
up to 150 A/μs repetitive rate-of-rise.
After inserting a reactance of 100 μH in series
with the softstarter the current rate-of-rise was
measured as 40 A/μs during the current limitation
period. The solution was very effective, however
bulky. We focused our analysis on the identification
of the origins of the problem.
COM-POWER-WP--EN Rev1 |
Improve efficiently soft-starter transients' immunity
EMTP-ATP is a software package dedicated to
modeling of power system transients.
It allows analysis of high frequency phenomena,
i.e. switching transients, and was chosen as most
adequate to the aim of our modeling.
Moreover EMTP-ATP allows in depth modeling
of the power system together with suitable RVSS
model, critical, since the analysis aimed to see
the impact of the surrounding power system
on the semiconductors.
Please note that the modeling was focused on
an accurate simulation of the transients during
switching on. At first, we modeled the complete
power system using frequency-dependent cable
models and RVSS with its command circuit,
capable to represent the power system during
the complete motor start. The comparison with
real test measurements showed the very good
accuracy of the models. However, due to the very
high number of setting parameters, such as
the RVSS settings and motor, cable specifications,
such approach was not applicable for large scale
analysis on the current rate-of-rise, since it is
difficult to establish, monitor and explore
the very high number of various study cases it
would require. Furthermore, indepth analysis
showed that the current rate-of rise varies during
motor start, which requires simulation of
the complete process. With a timestep of 100 ns
for an average of 15 s to be simulated,
the generated output files would have been
difficult to handle.
We proceeded to simplification of the used models:
1. The motor was implemented as series R-L
fixed impedance with phase-to-earth parasitic
capacitance. This assumption is valid since the
transients on the thyristors are much faster than the
speed variation of the motor shaft, responsible
for the variation of it’s equivalent impedance
(this is also stated in IEC 62271-100)
2. Power feed was assumed as source behind an
equivalent short-circuit impedance (comprising
the transformer), with phase-to-earth
equivalent capacitance, including protection
capacitors if any.
3. Cables were modeled in π, however in order to
avoid discharging of the immediate upstream
of the RVSS cable capacitance into the
downstream one, a different approach was
applied: the upstream and downstream cables
have been considered as an equivalent cable
whose capacitances were placed on its ends.
The softstarter is placed along this single cable,
according to the real data. Its precise position
does not have any impact on the current
rate of rise since the cable capacitances are
concentrated at the ends of this equivalent cable.
Thus the resulting power system model was
greatly simplified as shown in Fig. 7.
Cupstream
Upstreamcable
DownstreamcableZcc Motor
Cdownstream
Figure 7 – Simplified power system model.
Cupstream = Csystem + Cprotection capacitors + Cupstream cable
Cdownstream = Cmotor + Cdownstream cable
The results obtained with this simplified model
were still precise (<5 % error) with respect to
the measured values of current rate-of-rise,
oscillation frequency and attenuation.
Example curves are shown in Appendix C.
However this simplified model, built with only linear
elements had another important advantage:
it allowed clear analysis of the switching events.
Modeling of the RVSS and case study power system in EMTP-ATP
(2)
COM-POWER-WP--EN Rev1 |
Improve efficiently soft-starter transients' immunity
As it was stated before the simplification of
the power system model led to the possibility
of using simple mathematics in order to derive
the behavior of the current in the first instants
after turning on.According to the measurements
(see Fig. 5) the current transient may be
approximated to a second order step response.
Analysis of the thyristor turn on current transient
Thus in electrical terms this means that it may
be summarized as a second order circuit
comprised of an inductor, capacitor and resistor,
the latter providing the attenuation of the signal.
The estimation of their values will be given by
the mathematical analysis.
1. During switching on, due to the very high
frequency of the current transient (50-200 kHz)
the upstream power circuit, except the
connecting cable, is equivalent to a capacitor
connected to earth. In fact, if this was not
the case, the current rate-of-rise will be
quite limited by the short-circuit equivalent
inductance of the upstream circuit
2. During switching, again due to the very high
frequency, the motor is assumed as
a capacitor connected to earth, its equivalent
RL circuit having a very high value.
3. During switching on, the fundamental (50 Hz)
current increase is neglected with respect to
the first quarter of the oscillation of
the transient current
Hypothesis
The following parameters are defined:
C1, C2 equivalent capacitors of the power
system upstream and downstream of
the soft-starter, the value being the sum
of the upstream or motor equivalent
capacitance and that of the upstream or
downstream cable
R, L equivalent parameters of the connecting
cable comprising both the upstream and
downstream part, the resistance being
calculated at the presumed oscillating
frequency of the current
VC1, VC2 phase to earth voltages at the moment
of switching, upstream and downstream
to the soft-starter
The turn on of a thyristor is represented by
the equivalent circuit in Fig. 8.
C1
RL I
VC1VL VR
C2 VC2
Figure 8 – Equivalent circuit for current transient analysis.
Equivalent circuit of the system during switching on
COM-POWER-WP--EN Rev1 |
Improve efficiently soft-starter transients' immunity
The following equation is derived from Fig. 8:
From (3) the following expression of the current
in the first moments after thyristor turn on is
obtained:
Obviously from (4) the current transient depends
both on upstream and downstream circuit.
However, as often in practical cases the upstream
cable is quite short, it is often said that the current
transient is only dependent on the downstream
cable. This is valid for the equivalent cable
inductance; however the upstream capacitance,
together with the one downstream will determine
the equivalent capacitance of the circuit and thus
the frequency of the current transient.
The current rate-of-rise is calculated within 1 μs
interval. Since the initial value of the current is zero,
the current rate-of-rise 1 μs after turn on is equal
to the value of current at 1 μs which is given in (5).
Equation (5) allows the current rate-of-rise to be
computed, once the power system is simplified
to the equivalent circuit of Fig. 8. Should any
protection inductance be incorporated into
the circuit, its value is to be added to the cable
inductance.
It should be noted that the current rate-of-rise is
proportional to the power system rated voltage.
The higher the rated voltage, the higher will be
the current rate of rise. For example, in a 6.6 kV
system it will be approximately twice as high as for
a 3.3 kV system. There are two important variables
that do not depend on the power system itself but
on the switching conditions. Those are the values
of the upstream and downstream voltages on
the capacitors. Their estimation requires
consideration of the transients on the voltages
during switching off and on of the thyristors.
Mathematical expression of the current rate-of-rise during switching on
(3)
(4)
(5)
R.I + L. + .∫I.dt + VC1 t=0+ .∫I.dt + VC2 t=0
= 0dIdt
1 C2
1C1
I(t) = .e -t/T .sin(w.t)(VC1 t=0
+ VC2 t=0 )w. L
T = 2LR
LCw = 4.R 2
12L√
C = C1.C2
C1 + C2
dI(t)dt = I(t)
t=1µs=
- VC1 t=0 + VC2 t=0)(w.L .e-10 -6/ T.sin(w.10 -6)
COM-POWER-WP--EN Rev1 |
Improve efficiently soft-starter transients' immunity
A measurement of the voltage transient during
switching off and on is given in Fig. 9
t(ms)60 61 62 63 64-7000-5250
-3500
-1750
1750
0
3500
5250
7000
(V)
DownstreamVoltage
: VP2F
: VaP2F
Figure 9 – Measured Voltages on the soft-starter outgoing terminals during on-off switching.
The analysis of these measured voltages results in
the use of another equivalent circuit representing
the power system between on-off switching.
Overview of the voltage transient during switching
Frequency of the voltage transient
The calculation of the transient voltage frequency
requires establishing the equivalent circuit of the
power system at the moment of switching off.
Given the measured frequency of the transient
voltage, 3-4 kHz, the following simplifying
assumptions are made:
1. The upstream circuit, including cable is
represented by the short-circuit impedance of
the power system, whose value is quite low
compared to the one of the capacitors to earth.
This assumption is valid when the transformer
neutral earthing is made with small impedance,
which is generally the case.
2. The cable impedance is negligible compared to
that of the motor
3. The short-circuit impedance is negligible
compared to that of the motor, thus
the upstream circuit is considered to be
directly earthed.
According to the above assumptions,
the following equivalent circuit for the voltage
transient may be drawn:
Lm Rm
CdownstreamCsnubber
Figure 10 – Equivalent circuit for the voltage transient during switching off.
The frequency of the voltage transient is therefore:
The calculated frequency is slightly higher than
the real one because of the assumptions made for
the short-circuit impedance of the upstream circuit.
This will be accounted for in the next point.
Fdownstream_voltage
= 1√2.. L
eq.C
eq
Leq
= Lm + Lm = Lm 12
32
Ceq
= Csnubber
+ Cdownstream
(4)
COM-POWER-WP--EN Rev1 | 10
Improve efficiently soft-starter transients' immunity
Consideration of the transient voltage value
The transient voltage is the difference between
the upstream and downstream phase-to-earth
voltage. Depending on the turn on instant,
the voltage across the thyristor will be different.
Each side of the soft-starter should be considered
individually. The upstream voltage value will be
determined by the moment of switching on, i.e.
the delay given by the control circuit, (α).
This value is easily estimated. The downstream
voltage is defined both from the commutation
moment and the transient evolution.
The instantaneous fundamental frequency value
of the downstream voltage is half the one on
the other side, as shown in appendix A.
Precise estimation of the voltage requires having
the complete power system parameters, especially
for the attenuation constant calculation.
This data is unavailable for many reasons.
Thus it is necessary to formulate a simplified
method for the downstream voltage estimation.
The next figure presents how its value may be
approximated, of course in excess of the real one:
(V,pu)
-0.5
-1.0
-2.0
t(s)
Vapproximated
Vreal
Figure 11 – Approximation of the downstream voltage evolution after turn off of the thyristor.
In other terms, and with respect to the commutation
angle and the time interval between switching off
and on, the voltage is approximated as:
t - period of downstreamvoltage oscillation
Current zero time
Vdownstream(pu of V(t))
0.51
2
< t/2 > 2.t
Figure 12 – Approximation of the downstream voltage value as function of current zero time.
The value of the upstream voltage being always
1pu of itself, the voltage across the thyristor is
therefore:
(6)Vth_turn_on t=α
= (1+Vdownstream
(α _ φ)).Vupstream t=α
COM-POWER-WP--EN Rev1 | 11
Improve efficiently soft-starter transients' immunity
Estimation of the risk of high current transient
The above equations and considerations allow the computation of the current rate-of-rise at each thyristor turn on. Yet this requires selection of particular moments when this calculation is critical for the estimation of the need of adding a protection reactor. First let’s have a look on the evolution of the turn on delay (α) and the current zero delay (φ) with respect to the voltage zero crossing:
(ms)
T(s)
init = f (Veffinit)
init
Seconddanger zone
Firstdanger zone
Full wavevoltage
End oframp time
Closing ofbypass contactor
Rated
Figure 13 – Time variation of the control delay and current phase shift during the motor start.
The critical instants during motor start are those where the current zero time is the smallest, in this case the voltage across the breaker would be at its potentially highest value. As it can be seen, there can be designated two critical zones:1. First danger zone – it is the moment of from
start to ramp end, where the control delay (α) will make a first brake, according to the requested current limitation, and slow its decrease, the current phase shift (φ) will continue decreasing with the acceleration of the motor
2. Second danger zone – the period after ramp end until going into full wave conduction. Current limitation means to maintain the current to the desired value. With the speed increase the current is getting smaller and the control angle has to be decreased. This increases in turn the current, much like a proportional control. The control angle decrease may lead to experience very high di/dt. Earlier it happens after the ramp end and higher is the current rate of rise.
Between these two danger zones, the second one is potentially more critical.
This is because of the certainty that during this period the current zero time will decrease, to values lower than half of the oscillation period (Fig. 12) and the voltage across the thyristor must be taken as 3 times the upstream voltage, (6). If this happens during the current limitation phase the control angle will be still relatively high, i.e. the voltage will be close to its’ maximum on turn on. Generally, in order to consider the worst case, the current rate of rise is to be calculated immediately after the ramp end.The estimation of the voltage constraint on the thyristor requires a preliminary estimation of the control delay and the current phase shift. The control delay can be derived from the required current limitation. First it is necessary to estimate the voltage rms value on the motor at ramp end:
Where:Imax requested current limitation, pu of rated
currentkd motor starting current, in pu of the rated
current
The estimation of the control delay for a certain rms voltage at ramp end is made by dedicated algorithm, explained in Appendix B. The exact value of the phase shift can not be estimated with sufficient precision. This would require a close look on the motor equivalent parameters and speed evolution during start, which in turn is very load dependent. That is why it was preferred to use the stalled rotor phase shift. Parametric analysis on the phase shift impact showed that a higher power factor leads to a slightly higher di/dt (5-6 % over). Generally this should not play a role since selected reactor values will always be higher than the exactly needed, because choice is generally made among fixed by manufacturer values. Taking the immediate greater value will normally add additional security margin, far beyond 5-6 %. Of course, depending on the available reactor values, it may be sometimes important to increase the di/dt values by 5-6 % before going to selection.With the above assumptions, the ratio between the current rates of rise Zone2/Zone1 will be either 150 % or 200 %.
(7)Vrmsramp_end
= ImaxK
d
COM-POWER-WP--EN Rev1 | 1
Improve efficiently soft-starter transients' immunity
The above equations and considerations were
implemented in a tool capable of calculating
the current rate of rise for the two designated
danger zones.
Its’ application will ensure optimized installation
conditions for the SCR application as well as its
service continuity.
Development of protection sizing tool
Required Input Data
The case study data was entered for the
estimation of the risk of current rate-of-rise and it’s
evolution with the protection reactance. The next
figure shows the results of calculation, sorted as:
- Zone 1, when only danger zone 1 is
accounted, danger zone 2 will not take place
because of propitious installation/setting/load
conditions
- Zone 2, (usual case), when only the danger
zone 2 is accounted, being more constraining
than danger zone 1
400
350
300
250
200
50
100
150
00 10 20 30 40 50 60 70 80 90
Zone 1
Current limit set to: 380 % In
Current rate of rise as functionof installed protection inductance
Thyristor limit
Zone 2
Protection inductance (µjjjH)
Current rate of rise (A/µjjjs)
Figure 14 – Results of calculation of current rate-of-rise.
Designation UnitsRated voltage kVRated frequency HzUpstream capacitance nFUpstream cable material resistivity ΩmUpstream cable length mUpstream cable relative insulation dielectric constant -Number of cables per phase -Downstream cable material ΩmDownstream cable length mDownstream cable insulation dielectric constant -Number of cables per phase -Motor rated power kWMotor rated power factor -Motor Starting power factor -Motor Efficiency %Motor Starting current x InRVSS Snubber capacitance nFCurrent limitation setting % of InThyristor current rate-of-rise limit A/μs
Results
Input data for the protection sizing tool.
COM-POWER-WP--EN Rev1 | 13
Improve efficiently soft-starter transients' immunity
The obtained results show that insertion of 100 μH
inductance leads to 50 A/μs current rate-of-rise for
danger zone 1, whereas the measured value was
about 40 A/μs. Comparing the required inductance
sizing for Zone 1 and on Zone 2, it may be seen
that the required inductance is more than 5 times
higher. This higher value will cost more because
installation requirements will differ.
Sometimes, in order to get rid of Zone 2 constraint,
an earlier by-pass closing may be convenient.
But this solution is not sure to be efficient.
Also earlier by-pass closing is not really a solution;
one may ask why we need a soft-starter if it is to
not use it completely?
An elegant solution in order to reduce current
rate of rise would be to control α in a way that
the current zero time remains higher than twice
the voltage transient period.
This will keep the current rate of rise at the smallest
possible value. This will require signal processing
of the upstream or downstream voltage oscillation
immediately after the current interruption in
the thyristor. The drawback of this solution is that
it will not be efficient when the current limitation
setting is very close to the minimum acceptable
value, under which the motor will simply not start.
In fact, reducing the current rate of rise requires
increasing of the current zero time intervals that
decrease the current rms value.
Of course, in some existing installations over sizing
the inductance may be the simplest and the most
convenient solution.
Discussion of the results
COM-POWER-WP--EN Rev1 | 1
Improve efficiently soft-starter transients' immunity
Conclusion
In this paper we presented in-depth analysis of
the mechanisms and estimation of current rate-of-
rise in MV soft-starter semiconductors.
Based on field measurements and reasonable
simplification assumptions a mathematical
definition of the current transient immediately
after the thyristors turn on was established.
It showed that the current transient issues
increases with the application rated voltage.
Furthermore, soft-starter control analysis allowed
defining two zones of major importance for correct
sizing of protection solutions: one immediately after
ramp end (beginning of current limitation phase)
and another starting shortly after, until closing
of the by-pass contactor.
The second one requires greater value of
the protection inductance. Also there were
proposed partial solutions for decreasing
the current rate of rise, to close earlier the by-pass
contactor or to control the turn on delay (α).
The most effective solution is sizing of
the protection inductance by considering
the second danger zone. All the considerations
and mathematical formulae were implemented
in a tool which was also presented and validated
by comparison with field test measurements.
This advancement will ensure that soft-starter
installation conditions are optimized and safe for
full SCR service continuity.
COM-POWER-WP--EN Rev1 | 1
Improve efficiently soft-starter transients' immunity
Acknowledgements
Authors would like to thank Mr. C. Durand for his
work on the simulation models, Mr. P.A. Claudel
and Mr. R. Catalan-Herrero for the measurements
they performed on site and Mr. R. Henri for his
benefic contribution to the final version of this paper.
Appendices
Zm
VnVa
Vb
Vc
Figure A-1 – Equivalent circuit after current interruption in one phase.
Calculation of downstream voltage
Vn = Va _ Ia.Zm = Vb _ Ib.Zm
Ia = _Ib
Ia =Va _ Vb 2.Zm
Vn = = Va + Vb 2
Vc 2
_
(A-1)
Calculation of control delay as function of motor rms voltage
Calculation of the control delay requires set up of
the limits of control angle variation.
Since the motor is isolated from earth, only control
delay < 2/3 of the fundamental half period is
possible, so that there always at least two
thyristors conducting. With this assumption
the voltage variation on one phase of the motor,
for 6kV power system is given on the next figure:
-6000
-4000
-2000
0
2000
4000
6000
: X0001A
: MOTA-N_MOT
(V)
t(ms)0 10 20 30 40
Phase toneutralvoltage
50
Fig. A-2 Time variation of the phase to neutral motor voltage compared to the fundamental phase to earth voltage upstream of the soft-starter.
Mathematically the rms value calculation is
expressed as:
++++++
++−++
=
∫∫∫∫
∫∫∫
+
+
+
+
−
−
−
−
2
6
26
6
262
6
26
6
26
0
2
32
21
0
32
21
2t
tt
TT
tt
tt
rms
)wt(sin)t.wsin(.)wtsin()wt(sin
)wt(sin)t.wsin(.)wtsin()wt(sin
tV
α
α
ϕ
ϕ
α
α
ϕ
ϕ
α
α
ϕ
ϕ
π
π
Where:
t: period of the fundamental signal.
A dedicated calculation algorithm iterates on α
until the required value of the rms voltage is
achieved.
(A-2)
COM-POWER-WP--EN Rev1 | 1
Improve efficiently soft-starter transients' immunity
(A)
-1500
1500
-1000
1000
-500
5000
0.05
: simu_bpoil_v5. pl4
: RVSSA -RVSSA_A
0.06 0.07 0.08 0.09 0.10 t(s)
Measured
Simulated
Fig. A-3 Comparison of line currents, simulated and measured.
500(A)
400300
200100
0-100-200
61
: simu_bpoil_v5. pl4
: RVSSA -RVSSA_A
61.5 62 62.5 63 63.5 t(ms)
Measured
Simulated
Fig. A-4 Zoom to Fig. A-3, thyristor turn on.
1060104010201000
1140112011001080
98064 65 66 67 68 69 70
(A)
t(ms)
Measured
Simulated
: simu_bpoil_v5.pl4
: RVSSA-RVSS_A
Fig. A-5 Zoom to Fig. A-3, current crest values.
Comparison of measured and simulated waveforms
-10-7.5
-5-2.5
57.510
02.5
(V)
t(s)0.05 0.06 0.07 0.08 0.09
Measured
: simu_bpoil_v5.pl4
: RVSSA-RVSS_A
Simulated
Fig. A-6 Comparison of phase voltages, simulated and measured.
(V)
t(ms)
Measured
-10-7.5
-5-2.5
57.510
02.5
59 60 61 62 63 64 65 66
: simu_bpoil_v5.pl4
: RVSSA-RVSS_A
Simulated
Fig. A-7 Zoom to Fig. A-6, during current zero in the thyristor.
Vita
Delcho Penkov was born in Haskovo, Bulgaria.
Hegraduated from Technical University of Sofia in
2002(MSC). In 2006 he received his PhD degree
in ElectricalEngineering from the Institut National
Polytechnique deGrenoble (INPG).
He is currently working for Schneider Electric as
Power Systems Engineer. Member of IEEE.
Alain Côte received his Electrical Engineering
degree from the National Polytechnic Institute
of Grenoble in 1983. He is currently working
on electrical network analyses such as stability,
harmonic and over voltage studies.
He has been personally involved in several
instances of expertise on equipment failure or
malfunctioning in different fields of industrial plants,
particularly about transient phenomena.
Schneider Electric Industries SAS35, rue Joseph MonierCS 30323F- 92506 Rueil Malmaison Cedex
RCS Nanterre 954 503 439Capital social 896 313 776www.schneider-electric.com
05-2010AMTED110112EN
© 2
010
- S
chne
ider
Ele
ctric
- A
ll rig
hts
rese
rved
.
This document has been printed on ecological paper