Professional Documents
Culture Documents
Asynchronous Machine
ElmAsm, TypAsmo
DIgSILENT GmbH
Heinrich-Hertz-Str. 9
72810 - Gomaringen
Germany
http://www.digsilent.de
info@digsilent.de
r1211
Copyright ©2012, DIgSILENT GmbH. Copyright of this document belongs to DIgSILENT GmbH.
No part of this document may be reproduced, copied, or transmitted in any form, by any means
electronic or mechanical, without the prior written permission of DIgSILENT GmbH.
Contents
1 General Description 4
2.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.6 Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.7.3 Reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
B Parameter Definitions 24
C References 25
List of Figures 26
List of Tables 27
1 General Description
The general asynchronous machine model of DIgSILENT PowerFactory is the so-called Type
2 Asynchronous Machine model (TypAsmo) that has been available since version 12.0. Prior
to this, a different asynchronous machine model (TypAsm) was used, but this model has since
been deprecated and is no longer recommended for use.
The model is represented as the classical asynchronous machine model with a frequency (or
slip) dependent rotor impedance (see the equivalent circuit in Figure 1.1).
Stator voltages and currents in these equivalent circuit diagrams are represented as instanta-
neous phasors in a steady reference frame. Rotor voltages and currents are represented in a
reference frame that rotates with mechanical frequency. Hence, all quantities in these equiva-
lent circuits are represented in their natural reference frame. The machine model is supposed to
be unearthed, which explains why no equations for the zero sequence components are given.
The rotor impedance is referred to the stator side, and thus the rotating transformer in Figure
1.1) does not show any winding ratios.
The stator winding resistance Rs , the stator leakage reactance Xs , the magnetizing reactance
Xm and the rotor impedance Zrot characterise the model.
As mentioned above, Zrot can be frequency dependent, which allows the modelling of asyn-
chronous machines over a wide speed (or slip) range. PowerFactory offers several ways to
model a frequency dependent rotor impedance Zrot . The rotor impedance models vary in de-
tail and complexity and should be carefully selected depending on the types of studies to be
performed and the availability of input data.
The single cage model is the simplest of the rotor impedance models. The is characterised by
a single R-L branch with a slip dependent rotor resistance (see Figure 1.2. The advantage of
this model is its simplicity, particularly when manufacturer data is incomplete or unavailable.
This model is adequate for describing wound rotor motors, however it is unsuitable for motor
starting studies with squirrel cage motors. This is because the single cage model cannot de-
scribe the torque-speed characteristics of a squirrel cage motor over the full range of slip values
(as a result of the current displacement effect). When applying the single cage model, either the
starting characteristics or the full-load characteristics are modelled accurately, but not both [1].
Squirrel cage rotors exhibit current displacement on starting and at low speeds (i.e. at high slip)
due to the skin effect. The current displacement effect can be modelled by adding an additional
R-L branch in parallel to the single cage rotor (see Figure 1.3). The additional R-L branch is
modelled to represent the squirrel cage rotor on starting, where the rotor leakage reactance
predominates. As the speed increases, the influence of the additional R-L branch decreases.
An additional R-L element (RrA0 and XrA0 in Figure 1.5) represents a user-defined slip-independent
rotor leakage impedance, which provides additional modelling flexibility. By default, the rotor
leakage impedance is set to zero.
The double cage rotor is modelled in a similar way to the squirrel cage rotor with two parallel R-L
branches representing the outer (A) and inner (B) cages (see Figure 1.4). There key difference
is that in the double cage model, a rotor leakage reactance (Xrm) is included in lieu of the
slip-independent impedance (RrA0 and XrA0).
A deep bar double cage rotor, where the outer cage exhibits current displacement on starting,
can be modelled by applying a double cage model with current displacement. Altogether, the
frequency dependence of the rotor impedance can be approximated by up to three parallel R-L
branches (see Figure 1.5).
Motor data can be entered into an asynchronous machine type (TypAsmo) either by directly
specifying the resistances and reactances of the equivalent circuit diagrams (”Electrical Param-
eter” option) or by specifying characteristic points on the torque-slip and current-slip curves of
the machine (”Slip-Torque/Current Characteristic” option).
On the second page of the load flow tab, there are curves showing the torque-speed or the
current-speed characteristics of the machine (see Figure 1.7). These characteristics are always
calculated from the steady state equations of the equivalent circuit. Hence, they truly represent
the machines characteristics.
If you have motor electrical parameter data on hand (e.g. Rs , Xs , Xm , RrA , etc), then by se-
lecting the ”Electrical Parameter” input mode, you can enter this data directly into asynchronous
motor type (in the Load Flow tab).
Sometimes motor manufacturers supply measured test values of the equivalent rotor impedance
according to Figure 1.1 at different motor speeds (or slip-values), typically at no-load (i.e. close
to synchronous speed) and at locked rotor / stand still.
For such cases, PowerFactory offers a pocket calculator to convert these test measurements
into equivalent circuit parameters for two parallel R-L branches (i.e. for rotors with current dis-
placement). The pocket calculator can be accessed when the input mode is ”Electrical Parame-
ters” and the ”Consider Current Displacement (Squirrel Cage Rotor)” options are selected (see
Figure 1.8.
The general formula that relates the equivalent rotor impedance to equivalent circuit parameters
of a rotor circuit approximated by two ladder circuits (as in Figure 1.3) is:
RA1 RA2 (RA1 + RA2 ) + s2 RA1 XA2
2 2
+ RA2 XA1
Rr (s) = 2 2 (1)
(RA1 + RA2 ) + s2 (XA1 + XA2 )
2 2
RA1 XA2 + RA2 XA1 + s2 (XA1 + XA2 ) XA1 XA2
Xr (s) = 2 2 (2)
(RA1 + RA2 ) + s2 (XA1 + XA2 )
The slip-independent rotor leakage impedance (RrA0 and XrA0) is assumed to be zero in this
case.
2 2
RA1 XA2 RA2 XA1
Xr (0) = 2 (4)
(RA1 + RA2 )
2 2
RA1 RA2 (RA1 + RA2 ) + RA1 XA2 + RA2 XA1
Rr (s) = 2 2 (5)
(RA1 + RA2 ) + (XA1 + XA2 )
2 2
RA1 XA2 + RA2 XA1 + (XA1 + XA2 ) XA1 XA2
Xr (s) = 2 2 (6)
(RA1 + RA2 ) + (XA1 + XA2 )
This set of non-linear equations can be solved by an iterative procedure. In the pocket calculator,
these equations are solved using a Newton-Raphson algorithm.
RA1 Rr (0)
RA2 = (7)
(RA1 − Rr (0))
XA1 Xx
XA2 = (8)
(XA1 − Xx )
The auxiliary variable Xx can directly be calculated from the given values for Zrot and is defined
by:
2
Xr (1)2 − (Rr (1) − Rr (0))
Xx = (9)
Xr (0) − Xr (1)
The Newton-Rahpson algorithm in the pocket calculator uses the following initial values:
1
RA1 = RA2 (10)
5
5
XA1 = XA2 (11)
2
If the input mode is set to slip-torque/current characteristic, the parameters of the equivalent cir-
cuit diagram are estimated (using an in-built parameter estimation algorithm) from the following
set of motor performance parameters:
The rated mechanical power, the rated power factor, the efficiency at nominal operation and the
nominal speed of the machine specify the nominal operating point.
Pressing the Calculate button starts the parameter estimation algorithm (which converts the
performance parameters to equivalent circuit electrical parameters). The algorithm aims to
solve for the following parameters:
The following parameters are input parameters and therefore assumed to be known:
The parameter estimation algorithm solves a non-linear least squares problem using the Newton
Raphson method. For single cage models, the problem formulation is as per the four equations
below:
For squirrel cage and double cage models, the following two equations are added to the problem
formulation:
If the parameter estimation algorithm fails, one of two corresponding error messages will appear:
For representing asynchronous machines in load flow analysis, the user has the choice between
two representations:
The ”slip iteration” representation is the more accurate representation and is based on the
equivalent circuit diagrams according to Figure 1.1 to Figure 1.5. Here, the model equations are
evaluated in steady state. The user defines only the (electrical) active power of the machine.
During the load flow iteration, the corresponding slip is calculated from the steady state model
equations and the reactive power (Q) is resulting.
The ”P-Q” representation corresponds to the classical way of representing asynchronous ma-
chines in load flow programs. By assuming that the machine operates at a certain power factor,
independent of the bus bar voltage, the machine can be approximated by a standard P-Q load
model.
The ”slip iteration” is of course the more precise method of representing asynchronous ma-
chines in load flow programs. Since this model is consistent with dynamic models it should
always be used when the load flow is used for initializing a transient analysis. However, it re-
quires the full machine characteristics why it is sometimes more suitable to use the simple P-Q
approach, especially in load flow planning studies, when no transients have be calculated or
when no concrete data are available.
For short circuit analysis, a voltage source behind the subtransient impedance (rs + jx00 ) gen-
erally represents asynchronous machines (see Figure 1.10).
The value of the subtransient impedance is either directly taken from the speed current charac-
teristic (”Consider Transient Parameter”) or it can be entered separately. This is sometimes the
more accurate approach because under short circuit conditions saturation effects of the leakage
reactance that are not represented in the standard model can occur.
The relationship between the locked rotor current ratio and the subtransient impedance is the
following:
Ia Zn
= 00 (19)
In Z
The actual value of the subtransient voltage depends on the short circuit method applied. Also,
the model according to Figure 1.10 is only able to represent the subtransient behaviour cor-
rectly. For calculating DC time constants, transient or permanent short circuit currents, the rules
defined in the individual short circuit standards are applied.
The asynchronous machine model for harmonics analysis can directly be derived from the
equivalent circuits according to Figure 1.1 to Figure 1.5.
The value of this impedance is either calculated from the equivalent circuits according to Figure
1.1 to Figure 1.5. For higher frequencies, the asynchronous machine impedance corresponds to
the subtransient value. Only for frequencies around fundamental frequency, the actual slip de-
pendence is important. This accurate representation is especially required for subsynchronous
resonance studies or self-excitation studies of asynchronous machines.
It is possible to neglect the effect of slip dependence by disabling the flag ”consider transient
parameters”.
2.1 General
The dynamic models for RMS (stability) and EMT-simulations can be derived from the equivalent
circuits according to Figure 1.1 to Figure 1.5.
Possible state variables of a general asynchronous machine model are either current or flux
variables.
As long as no saturation is considered, the actual choice of state variables doesn’t have any
influence to the results, only the numerical behaviour of the solution algorithm will depend on it.
The PowerFactory model uses stator currents and rotor flux as state variables because this
choice leads to the best decomposition of time frames and has therefore the best numerical
properties.
The voltage equations of an asynchronous machine model with a number of n R-L rotor-loops
are the following:
1 dψ S ωref
uS = rS iS + +j ψ (20)
ωn dt ωn S
1 dΨR ωref − ωR
0 = RR iR + +j ΨR (21)
ωn dt ωn
The equations are expressed in a rotating reference frame common to the stator and the rotor
equations. The dimension of the rotor-flux vector and the rotor-current vector is equal to the
number of rotor-loops.
For formulating the asynchronous machine equations with stator current and rotor flux as state
variables, the flux linkage equations must be solved for the non-state variables, which are stator
flux and rotor currents:
iR = −kRS iS + X−1
RR ΨR (25)
kRS = X−1
RR xRS (28)
ψ 00 = kTSR ΨR (30)
The flux-linkage and the resistance matrices of the single cage model according to Figure 1.1
and Figure 1.2 can be expressed as follows:
xSS = xS + xm (31)
xSR = xm (32)
xRS = xm (33)
rR = RrA (35)
The flux-linkage and the resistance matrices of the squirrel cage rotor model according to Figure
1.1 and Figure 1.3 are the following:
xSS = xS + xm (36)
xTSR = xm
xm (37)
x
xRS = m (38)
xm
x + xrA0 + xm xrA0 + xm
XRR = rA1 (39)
xrA0 + xm xrA2 + xrA0 + xm
RrA0 + RrA1 RrA0
RR = (40)
RrA0 RrA0 + RrA2
The flux-linkage and resistance matrices of the double cage model with three R-L-rotor loops
according to Figure 1.1 and Figure 1.5 are the following:
xSS = xS + xm (41)
xTSR = xm
xm xm (42)
xm
xRS = xm (43)
xm
xrA1 + xrA0 + xrm + xm xrA0 + xrm + xm xrm + xm
XRR = xrA0 xrm + xm xrA2 + xrA0 xrm + xm xrm + xm (44)
xrm + xm xrm + xm xrB + xrm + xm
RrA0 + RrA1 RrA0 0
RR = RrA0 RrA0 + RrA1 0 (45)
0 0 RrB
In the EMT simulation, PowerFactory uses a steady state reference frame for expressing the
stator equations. The stator-voltage equation in a steady state reference frame is:
00
!
x00 diS ωref 00 1 dψ
uS = rS iS + + j ψ + ejωref t (46)
ωn dt ωn ωn dt
In the EMT-model, the reference frame, in which the rotor equations are expressed, rotates with
nominal frequency, hence:
ωref = ωn (47)
This equation corresponds exactly to the equivalent circuit according to Figure 1.10, with the
following definition for the subtransient voltage:
00
!
ωn 1 dψ
u00 = j ψ 00 + ejωn t (49)
ωn ωn dt
For stability analysis, the asynchronous machine model has to be reduced. In accordance with
the steady state model of the electrical network that is applied in stability analysis, the stator
equations of the asynchronous machine model are reduced to steady state equations. The
following voltage equation is resulting:
ωref 00 ωref 00
uS = rS + j x iS + j ψ (50)
ωn ωn
This is a steady state representation of the equivalent circuit according to Figure 1.10. The
subtransient voltage is here defined as:
ωref 00
u00 = j ψ (51)
ωn
In the stability model, the stator equations are expressed in a reference frame that rotates with
the global system reference that is usually fixed to the rotor of the reference generator (or an
external network or a voltage source, depending on the load flow reference).
Because stator transients are neglected, the choice of the reference frame has actually an
influence to the stator voltage equations. For avoiding any dependence on the actual choice of
the reference machine, the influence of the reference frequency is not considered in subtransient
reactance of the PowerFactory stability model.
dωR
Jtot = Me − Mm (53)
dt
Jtot (1 − sn ) ωpzn ωn dn
!
dn
= Tag,tot = me − mm (55)
Pmn pz dt dt
The total inertia is calculated by adding the inertia defined in the type and the mechanical load
defined in the machine, as follows:
2
Jtot = Jtype + Jme · gratio (56)
From this, it follows that the total acceleration time constant is:
2
Tag,tot = Tag,type + Tag,me · gratio (57)
If no separate mdm model is defined, the asynchronous machine uses the speed-torque char-
acteristic of the built-in mdm-model:
mm = Ip nex (58)
2.6 Initialization
All state variables of the model are initialized from a preceding load flow calculation so that a
simulation starts from a steady state condition.
If the default orientation of the asynchronous machine is set to ”motor”, the mechanical load
torque xmdm is initialized. In case of ”generator” orientation, the turbine power pt is used for
establishing the active power balance of the model.
In case of a running machine, the proportional factor Ip of the built-in mdm or analogous factors
of separately modelled motor-driven machines are calculated during the initialization process. In
case of a disconnected machine, e.g. if a motor start-up is simulated, the user-defined variable
of the input dialogue is used instead.
Different starting methods are supported for the Motor Starting command, which require differ-
ent input data. The starting methods are the following:
• Directly Online
• Star-Delta
• Variable Rotor Resistance
• Reactor
• Auto Transformer
The method Directly Online behaves the same as when the option Use Motor Starting Meth-
od is unselected, which means that the motor is turned on and connected directly to the network
by the Motor Starting command.
In the star-delta method, the motor is first connected in star configuration in order to reduce the
voltage across its windings.
While connected in star configuration, the equations for voltage and current are expressed as:
1
um = u : bus1 · √ [p.u.] (59)
3
1
im = i : bus1 · √ [p.u.] (60)
3
After the time specified in Switch to ’D’ after or in Switch to ’D’ at Speed >=, the configuration
is changed to Delta, and so the equations for voltage and current are modified as follows:
With this starting method, a user-specified table determines the additional rotor resistance for
different times or speed values.
2.7.3 Reactor
With this method, at the starting time, the by-pass switch is closed and the motor connected to
the network. See Figure 2.1.
In addition, while the by-pass switch is open, the additional stator reactance, xrea , is connected
in series with the motor reactance, xstr . The total reactance is given by:
The value of the additional stator reactance is calculated in p.u. with reference to the motor
rated power. Given the Reactance value, xrea entered in %, and the Rated Apparent Power,
ratedStr in kV A:
xrea sgn
xrea = · [p.u.] (64)
100 ratedStr
Where sgn is the Rated Apparent Power specified in the motor’s type TypAsmo in kV A.
After the time specified in Bypass after or after the motor speed has reached the value specified
in Bypass at speed >=, depending on the option Triggered by.., the by-pass switch is closed,
thus ignoring the additional reactance.
With this starting method, the motor is connected, the star-contactor is closed and the by-pass
switch is open. See Figure 2.2.
Initially, both the star-contactor is closed and the by-pass switch is open. The equations of the
asynchronous machine for voltage and current depend on the selected T ap specified in %, and
the auto transformer reactance, xt , is added to the stator reactance. All these are expressed
as:
T ap
um = u : bus1 · [p.u.] (65)
100
T ap
im = i : bus1 · [p.u.] (66)
100
2
T ap
x0str = xstr + xt · [p.u.] (67)
100
After the specified time in Release Star Contactor after or when the speed is greater than the
value in Relase Star Contactor at speed >=, the star contactor is opened. Since the by-pass
switch is still open, the auto transformer reactance, xt , is still added to the stator reactance, xstr :
After the specified time in Bypass after or when the speed is greater than the value in Bypass
at speed >=, the by-pass switch is closed, thus ignoring the additional reactance of the auto
transformer.
The value of the additional reactance is calculated in p.u. with reference to the motor rated
power. Given the Reactance value, xrea entered in %, and the Rated Apparent Power, ratedStr
in kV A:
xrea sgn
xt = · [p.u.] (71)
100 ratedStr
Where sgn is the Rated Apparent Power specified in the motor’s type TypAsmo in kV A.
Vr2
Sr ; Vr ; Zb = (72)
Sr
B Parameter Definitions
C References
[1] J. Pedra. Estimation of typical squirrel-cage induction motor parameters for dynamic perfor-
mance simulation. IEEE Proceedings on Generation, Transmission and Distribution, 153(2),
2006.
List of Figures
List of Tables