Three-Phase Two-Winding Transformer Modelling
1. Introduction
A three-phase two-winding transformer steps voltage between two levels using electromagnetic induction
across a shared magnetic core. In PSCAD/EMTDC, it is modelled as a lumped-parameter T-equivalent circuit
combining leakage impedance, core losses, and non-linear saturation. The modelling workflow has four steps:
(1) collect nameplate data, (2) derive per-unit parameters, (3) define the saturation curve, and (4) simulate and
validate inrush behaviour.
2. Input Data
Table 1: Example transformer input datasheet — nameplate ratings, vector group, test losses, magnetising data, and saturation
parameters used as PSCAD model inputs
Table 2: Per-unit parameter calculation worksheet — Xpu, loss resistances, Zbase, flux linkage M, 80% M threshold, and per-
phase residual flux values (yellow = PSCAD inputs)
Table 3: Maximum inrush current result from Multiple Run sweep — 7.180 kA peak under worst-case residual flux and closing
angle
The following data must be obtained from the transformer nameplate and factory test report before any
simulation can be set up:
Parameter Unit Example Value (GSUT)
Power Rating MVA 759 MVA
HV / LV Voltage kV 232 / 22.5 kV
Frequency Hz 50 Hz
Vector Group — YNd1
% Impedance % at rated MVA 19.5% @ 759 MVA
Full Load Loss kW 1600 kW
No Load Loss kW 230 kW
Magnetising Current % 3.52%
Knee Point Voltage % of rated V ~120%
Air Core Reactance H 0.0985 H
Inrush Decay Constant s 3.353 s
NGR Ω 17 Ω
3. Parameter Derivation
3.1 Base Impedance and Leakage Reactance
All parameters must be converted to per-unit (p.u.) on the transformer's own MVA and voltage base. The base
impedance referred to the HV side is:
Zbase = V²HV(kV) / MVAbase [Ω]
The leakage reactance in ohms is Xl = (Z%/100) × Zbase. The resistive components are split from the total
impedance using the measured loss data. The flux linkage M is derived from the rated voltage and frequency
using Faraday's law:
M = V / (4.44 × fr) [p.u.]
where V is the peak phase voltage of the reference winding and fr is the rated frequency in Hz. The 80%
threshold of M marks the lower boundary of the piece-wise linear saturation curve.
3.2 Derived Parameters — Worked Example
Parameter Formula Result (GSUT)
Zbase V²HV / MVA 70.914 Ω
R (Full Load) 1600 / (759×1000) 0.002108 p.u.
R (Eddy Loss) 230 / (759×1000) 0.000303 p.u.
X (Leakage) sqrt(Z² − R²) 0.195 p.u.
Xpu (effective) X with base correction 0.4364 p.u.
M (flux linkage) V / (4.44 × fr) 0.6034 p.u.
80% of M 0.8 × M 0.4827 p.u.
Max Inrush Current Multiple Run result 7.180 kA
4. Core Saturation Model
The non-linear B-H curve of the transformer core is represented in PSCAD using a piece-wise linear saturation
model. Three parameters define the curve:
• Knee Point Voltage — the voltage (% of rated) at which the core begins to saturate. Typically 110–
130% of rated voltage.
• Air Core Reactance — the reactance of the winding when the core is fully saturated (permeability →
1). Entered in Henries or p.u.
• Magnetising Current — the no-load excitation current as % of rated, which sets the linear region slope
of the B-H curve.
The 80% of M value (lower inflection point) and the knee voltage (upper inflection point) together define the
transition region between linear operation and full saturation.
5. Inrush Current Modelling
When a transformer is energised from the HV side, the core flux starts from its residual value and is driven
sinusoidally by the applied voltage. If residual flux and applied flux are in the same direction, the peak flux can
reach twice the rated value — driving the core deep into saturation and producing a large asymmetric inrush
current. The inrush is characterised by a high initial peak (5–10× rated), a DC offset, second-harmonic content,
and an exponential decay with time constant τ = L/R.
In PSCAD, the Multiple Run component sweeps the breaker closing angle across N steps (typically 5 steps: 0°–
150°) for each residual flux condition. The worst-case residual flux assumptions for a three-phase transformer
are:
• Phase A: +80% of peak rated flux
• Phase B: 0% to −40%
• Phase C: −80% to −40%
The peak detection subsystem uses |X| (absolute value) blocks on each phase current, a Max function across
all three phases, and a Measurement-Enable block that records the peak at the end of each run. The highest
value across all runs is the worst-case inrush current.
6. PSCAD Simulation Setup
The standard PSCAD circuit for inrush study consists of: an ideal three-phase voltage source at HV level (R =
0), a timed circuit breaker closing at t = 0 (swept by Multiple Run), the two-winding transformer with all derived
parameters, and a high-impedance open-circuit termination on the LV side (no-load energisation — worst case
for inrush). Three slider controls (Var_1, Var_2, Var_3) set the per-phase residual flux initial conditions.
Figure 1: PSCAD simulation circuit — Two-winding transformer energisation with timed breaker, three-phase voltage and inrush
current measurements, and Multiple Run worst-case peak detection subsystem (worked example: 759 MVA GSUT)
The voltage waveforms confirm balanced three-phase sinusoidal supply at rated voltage throughout the
simulation. The inrush current waveforms show the characteristic asymmetric shape: a large unidirectional peak
at t = 0, a slowly decaying envelope driven by τ = 3.353 s, and eventual convergence to the small steady-state
magnetising current. The maximum inrush current of 7.180 kA is approximately 3.7× the transformer rated HV
current (1.89 kA), which is consistent with the expected range for large GSUTs.