Calculation - Method - ULF Unbalanced Load Flow ETAP PDF
Calculation - Method - ULF Unbalanced Load Flow ETAP PDF
Calculation - Method - ULF Unbalanced Load Flow ETAP PDF
The current injection method formulates and solves iteratively the following load flow equation:
abc abc
where DV is a 3-phase bus voltage vector in an incremental form, and DI is a 3-phase bus current
rm mr
abc
injection mismatch vector between specified value and calculated value; and Y is the corresponding
abc
Jacobian matrix. The off-diagonal elements of Y are identical to the corresponding elements of the node
abc
admittance matrix; and the diagonal elements of Y are dependent on both the corresponding elements of the
node admittance matrix and the load model adopted for each phase at a given bus.
The current injection method has relatively lower requirements of the bus initial voltage values compared to
the Newton-Raphson method and the fast-decoupled method. Instead of using bus real power and reactive
power mismatch as convergence criteria, the current injection method checks bus voltage magnitude tolerance
between two consecutive iterations to control the solution precision. The typical value for bus voltage
magnitude precision is set to 0.0001 pu.
Negative Impedance
Negative resistance and reactance should be avoided. As an example, the traditional method of modeling three-
winding transformers by a Y equivalent model, using one impedance and two two-winding transformers,
sometimes results in a negative impedance value for one of the impedance branches. In this case, the negative
impedance should be combined with other series circuit elements so that the result is a positive impedance
value. Load flow calculations may not converge if a large value of negative impedance is used. ETAP can now
model three-winding transformers directly with no need for users to do any conversions.
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017
Calculation Method - Unbalanced Load Flow Page 2 of 8
Solution convergence speed and computing time are functions of the initial voltages for load-type buses. The
closer the initial voltages are to their final profile, the faster the solution converges. The solution may not
converge if the initial voltages are too far from the final profile, thus it is recommended that the Update Bus
Voltage option be used to obtain a set of sound initial bus voltages.
Under open phase fault condition, the extremely low voltage profile and heavy constant KVA loading might
make the calculation diverge due to the system maximum loadability limits. The INI entry of
ImpedanceMotorVoltagePercentage sets the voltage threshold for induction motors to switch from constant
KVA loads to constant impedance loads when the terminal bus voltages are under the threshold voltages for
positive-sequence. The threshold voltage is calculated as the motor rated voltage times the percentage of the
INI entry. The percentage is default to be 65. For example, if the motor rated voltage is 4 kV, then the default
threshold voltage is 2.6 kV. When the positive-sequence voltage of the motor terminal bus is below 2.6 kV, the
motor constant KVA load for the positive sequence will be switched to a constant impedance load which has
the same KVA rating at voltage 2.6 kV. Please refer to Modeling of Loads for modeling of induction motors.
If there is a Double Open Phase fault in the system, a check will be performed on each bus in the system to
determine if there is any upstream source that can supply rotating torque for connected motors. This is
accomplished on the initial load flow run by using only swing and voltage control sources, temporarily
disconnecting any load except for static loads, then using a fictitious induction motor to determine if sufficient
rotating torque is present for connected motors. Theoretically, the rotating torque would be zero for any bus
subjected to a double open phase fault condition, but due to numerical convergence tolerance in the initial load
flow, a margin is added for the rotating torque. This margin is specified by the ETAPS.INI entry
RotationTorqueMargin and is defined as the percent of rotating torque when compared to 100% balanced
conditions, with a default value of 5%. Any bus will be treated as experiencing a double open phase condition,
if the initial load flow provides a rotating torque less than RotationTorqueMargin. Motors will be
automatically switched to their locked rotor impedance (static loading) if they are connected to a double open
phase bus. However, these motors can be considered as still operating/running based on their loading. The
ETAPS.INI entry of "DoubleOpenLockRotorThreshold" sets the loading threshold for induction motors and
synchronous motors to be considered as still operating/running under double open phase fault conditions, with
a default value of 30%. Percent loading is determined from the motors loading category sections or the ratio of
the operating value and the corresponding rating. Any motor connected to a double open phase bus, with
percentage loading below "DoubleOpenLockRotorThreshold" will still be considered as operating/running. For
conventional lump loads connected to a double open phase bus, the constant KVA loading portion will always
be automatically switched to a locked rotor impedance value regardless of loading.
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017
Calculation Method - Unbalanced Load Flow Page 3 of 8
Note: These internal impedances are ignored and the utility is represented in phase domain if Unbalanced is
selected in the rating page. The grounding connection is always considered to be solid grounded in this
scenario.
Swing Mode
The specified constraints for a swing power source are the magnitude and the phase angle of the positive
sequence voltage at the swing source terminal. The use of positive sequence representation for voltage
magnitude regulation makes it possible to properly represent the automatic voltage regulation (AVR)
mechanism of a power source, where in most cases, the average of voltage magnitudes of all three phases is the
voltage magnitude that is regulated. Under unbalanced conditions, the negative and zero sequence voltages
may be non-zero.
Modeling of Loads
Constant Power Load
Constant power loads include induction motors, synchronous motors, conventional and unbalanced lumped
loads with % motor load, UPS, and chargers. The power output remains constant for all changes in input
voltage. Below are the respective I-V and P-V curves for a constant power load:
Note: The constant power loads of synchronous motors are treated as the Mvar control mode sources with
negative power generations. These types of loads have the same structures as the synchronous generators of
Mvar control mode, and only the total of 3-phase power outputs / inputs for such a load remain constant for
unbalanced situations. Because of the difficulty of multiple solutions, ETAP does not allow for the constant
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017
Calculation Method - Unbalanced Load Flow Page 4 of 8
power unbalanced lumped loads (with % motor load) connected in Y with open neutral.
The constant power loads of 3-phase induction machines, conventional and unbalanced lumped loads with %
motor loading, are treated as combinations of the Mvar control mode sources with negative power generations
for positive sequence and constant impedance loads for negative-sequence and zero-sequence.
Note: When induction motors run under locked rotor conditions, they should act as static loads. Under different
voltages, the Power Factor might be different from that under the rated voltage.
Generic Load
Generic loads include lumped loads operating in either exponential, polynomial, or comprehensive modes.
A generic load model expresses the characteristics of the load at any time as algebraic functions of the bus
voltage magnitude and frequency at that instant.
Exponential Model:
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017
Calculation Method - Unbalanced Load Flow Page 5 of 8
where and are active and reactive components of the load when the bus voltage magnitude is and
is the frequency deviation . The subscript identifies the values of the respective variables at the
initial operating condition. Typically ranges from 0 to 3.0, and ranges from 2.0 to zero.
The parameters of this model are the exponents and . With these exponents equal to 0, 1, or 2, the model
represents constant power, constant current, or constant impedance characteristics, respectively.
Polynomial Model:
The polynomial model is composed of constant impedance, constant current, and constant power components.
The parameters of the model are the coefficients , , , , , and , which define the proportion of
each component.
Comprehensive Model:
Where:
The expression for the reactive component of the load has a similar structure. The reactive power compensation
associated with the load is represented separately.
Modeling of Transformers
ETAP uses a combined model of common transformers and regulating transformers. This model can
accommodate any phase shift, grounding impedance, and different positive/negative and zero sequence
impedance.
The (3) 1-phase transformer, open delta transformer, are modeled phase by phase based on the single phase
transformer parameters. Single phase center-tap transformer is modeled by assuming the solid connection to
the reference point of the power grid at the center-tap point of the transformer.
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017
Calculation Method - Unbalanced Load Flow Page 6 of 8
The No Load Loss is modeled as shunt impedances in T equivalent circuits for each pair of the coupling
windings.
For two winding transformer, the positive- and negative- sequence circuits will use the same shunt impedances
which are calculated from the positive-sequence no load loss testing data. The zero-sequence will use the shunt
impedances which are calculated from the zero-sequence no load loss testing data when there is no buried delta
winding.
For three winding transformer, the PS, PT and ST coupling windings will be modeled as T equivalent circuits
correspondingly with the same shunt impedances.
Buried delta winding will be modeled as an independent winding for zero-sequence circuit. The winding
parameters will be derived from the inputs of the Buried Delta Winding page. A two winding transformer with
a buried delta winding can be modeled equivalently as a three winding transformer with the tertiary winding
connected as delta and unconnected. A three winding transformer with a buried delta winding can be modeled
equivalently as a four winding transformer with the fourth winding connected as delta and unconnected.
Modeling of HVDC
An HVDC Link is not considered in the unbalanced load flow study. In the event any HVDC Link is detected,
an error will be posted.
Modeling of SVC
The voltage support capability of the SVC deteriorates with decreasing system voltage. The 3-phase SVC is
connected in Delta.
Modeling of UPS
In a Load Flow Study, the UPS is represented as a constant load at its input side and a swing source at its
output side (where the output side is energized).
When the UPS is selected as a load based on its loading category in its loading page, the system that is
connected to the UPS output side will be de-energized if there is not any other swing sources in the system and
the UPS is modeled as a pure constant load.
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017
Calculation Method - Unbalanced Load Flow Page 7 of 8
When the UPS is selected as a load based on its connected load, the UPS output bus will be modeled as a swing
bus with the UPS rated output voltage as its regulating voltage for its output bus. Then the calculated UPS
output bus loading will be treated/shared as the UPS output side loading.
If multiple UPS selected as connected loads share the same output bus, the calculated UPS output bus loading
will be shared among all the UPS by their maximum rated currents. The UPS output side loading will be
reflected onto the UPS input bus by considering its efficiency and the operating input power factor selection in
its loading page. For example, if the UPS share of its output bus loading is P + j*Q, then the UPS loading will
be reflected onto its input bus based on the operating input PF selection as a load:
1. P/EFF + j*P/EFF*sqrt(1-PF*PF)/PF where EFF is the UPS efficiency and PF is the rated or user-
defined power factor.
2. P/EFF + j*Q when connected load power factor is selected.
Modeling of VFD
In a Load Flow Study, the VFD is modeled the same way as the UPS except for:
Load editor Calculation of load for loading categories and voltage drop
Input for studies Calculation of load specification for load flow and initial load for motor starting and
transient stability studies
Studies results Calculation of load displayed in one-line diagram from load flow, motor starting, and
transient stability studies
The following two tables describe how these factors are used in these cases:
Factors Used for Motor Load Calculation
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017
Calculation Method - Unbalanced Load Flow Page 8 of 8
App. Factor *
Load Quantity x x x x x x x
Bus Diversity Factor * * * * *
Global Diversity Factor * * * * *
* Indicates the factor is used in calculation if you have specified it in the related load editor or study case.
Notes:
Motor load includes induction motor and generator, synchronous motor, MOV, and motor load portion
of lumped load.
Static load includes static load, capacitor, and static load portion of lumped load.
mk:@MSITStore:C:\ETAP%201260\Etaps.chm::/Calculation_Method_ULF.htm 5/8/2017