Case Study / Power Systems Simulation

Grid-Connected PV Microgrid & 4MVA EV Infrastructure Simulation

A full electromagnetic-transient model of a 25 kV distribution feeder hosting solar generation, a heavy industrial motor, dynamic residential demand and a ten-bay EV fast charging depot — built in Simulink to quantify stability margins under worst-case switching.

  • MATLAB
  • Simulink
  • Simscape Electrical
  • Grid Modernization
  • Power Systems
Feeder rating
20 MVA
EV infrastructure
4 MVA
Peak transient
8.5 MW
Steady state
2.05 MW

02 / Architecture

System architecture

A schematic block view of the network, alongside the actual top-level Simulink model it was built from.

Utility Grid25 kV, 60 HzTransformer20 MVA · 25kV/630VMeasurement ScopeV / I / P / Q logging630 V AC busPV Subsystem3 MVA injectionIndustrial Load2 MVA motorResidential LoadMATLAB FunctionEV Charging10 × 400 kWPower AnalyserP, Q, THD, PFTransient Stability Metrics
Fig. 1 — Schematic single-line view: 25 kV source → 20 MVA transformer → 630 V bus feeding the four subsystems into the power analyser.
Top-level Simulink model of the grid-connected PV microgrid with EV charging stations
Fig. 2 — Actual Simulink model: three-phase source → V-I measurement → 20 MVA Yg–Yg transformer → 630 V bus with the PV, industrial, residential and EV subsystems.
PV Subsystem Simulink subsystem diagram
01

PV Subsystem

MATLAB Function block converts V, pf and S setpoints into three controlled phase currents injected at 25 kV, with positive-sequence P/Q measurement.

  • 3 MVA
  • 25 kV
  • Unity pf
Industrial Load Simulink subsystem diagram
02

Industrial Load

Asynchronous machine fed from the 630 V bus, instrumented for electromagnetic torque, rotor speed and input active/reactive power.

  • 2.054 MW
  • 600 V
  • Tm = 1 pu
Residential Load Simulink subsystem diagram
03

Residential Load

MATLAB Function derives per-phase R and L from S, V and pf, driving variable resistor and inductor branches for a dynamic demand profile.

  • 3 MVA
  • 600 V
  • Unity pf
10 EV Charging Stations Simulink subsystem diagram
04

10 EV Charging Stations

Ten breaker-switched RLC charging branches enabled in stages by a comparator bank driven by the occupancy input u.

  • 4 MVA
  • 10 stations
  • Staged breakers

03 / Parameters

Model parameters

Block-level ratings used across the Simscape Electrical network.

ComponentRatingVoltageConfigurationNotes
Three-Phase SourceUtility grid25 kV60 Hz, YgPoint of common coupling upstream of the transformer
Three-Phase Transformer20 MVA25 kV / 630 VYg–Yg, two windingsSteps the feeder down to the 630 V distribution bus
Asynchronous Motor2 MVA600 VSquirrel cage, Tm = 1 pu8.5 MW starting transient settling to 2.05 MW steady state
Residential Load3 MVA600 VVariable R–L, unity pfMATLAB Function computes Rph and Lph from S, V, pf
EV Charger Cluster4 MVA600 V10 stations, breaker switched3 MW → 4.2 MW step as stations come online
Solverpowergui—Discrete, Ts = 5e-05 s10 s simulation window

04 / Results

Waveform & results gallery

Each tab pairs an annotated reconstruction of the logged signals with the raw Simulink scope captures from the 10 s run.

Induction motor — active power & electromagnetic torque

  • Active power P (MW)
  • Torque Te (kN·m ÷ 5)
-2.00.83.56.39.00.000.501.001.502.008.5 MW transient peak2.05 MW steady stateTime (s)P (MW) / Te (kN·m ÷ 5)
Direct-on-line energisation drives an 8.5 MW inrush that decays within ~350 ms to a 2.05 MW steady-state draw; torque pulsation damps out over the same window.

Raw Simulink scope captures

Scope — induction motor input active power

Scope capture of induction machine input active power versus time
Starting transient peaks at ≈8.5 MW before collapsing to a flat 2.05 MW steady-state draw for the remainder of the 10 s run.

Scope — electromagnetic torque Te (pu)

Scope capture of electromagnetic torque in per unit versus time
Torque oscillates between ±5 pu during run-up, damps through a final 4 pu swing and settles at ≈1 pu at synchronous speed.

05 / Findings

Key takeaways

What the study established about hosting large EV infrastructure on an existing feeder.

  • Grid stability analysis

    With the 20 MVA transformer in place, feeder voltage stayed balanced through motor starting and every charger switching event, confirming the 4 MVA depot can be hosted without reinforcement.

  • MATLAB Function dynamic load modelling

    PV injection and residential demand are generated by MATLAB Function blocks that convert S, V and pf setpoints into currents or variable R–L branches, so scenarios can be swept without rewiring the network.

  • Transient mitigation

    The 8.5 MW motor inrush and ±5 pu torque swing decay within ~0.7 s; staging the ten chargers behind comparator-driven breakers keeps the 3 MW → 4.2 MW step clean and oscillation-free.