40 MW battery: SSO screening and NESO compliance.
Twelve inverters on a 33 kV collector network, connecting in Great Britain under a Large BEGA. NESO wanted evidence of oscillation stability and fault performance. We delivered both from one PSCAD model, in one package.
- No.
- 03
- Client
- Confidential developer
- Plant
- 40 MW BESS, 12 inverters, 33 kV
- Agreement
- Large BEGA, Great Britain
- Studies
- SSO (Tests 3.2.1 and 3.3), FRT, FFCI
- Tools
- PSCAD/EMTDC v5, MHI DFScan
- Result
- Stable in all three modes
Every controller that shapes the oscillation response.
NESO wants the submitted model to include every controller that affects the plant, the power plant controller and its delays included. Pick a part to see where it sits.
Swipe to see it all
Phase-locked loop
Tracks the grid voltage angle so the current controller has a reference to work from. Its bandwidth sets the lowest interaction mode: too fast can excite SSO on a weak grid, too slow loses sync in a disturbance. We modelled it as it is in the inverter firmware.
- SRF PLL
- Filter stages
- PI regulator
- Key for SSO
Voltage and reactive power control
Holds terminal voltage by adjusting reactive power, and passes a reactive-current reference to the inner loop. It shapes the impedance in the 1 to 20 Hz range. Voltage-control mode is the NESO baseline; Q and power-factor modes were checked where they apply.
- V/Q droop
- Q limits
- Voltage-control baseline
Active power control
Turns the power setpoint into an active-current reference. A battery runs it both charging and discharging, with different gains, which is why NESO asks for export and import cases.
- Export +40 MW
- Import −40 MW
- DC link included
Power plant controller (PPC)
The site-level controller that sends P and Q targets to every inverter. NESO requires it in the model with its cycle time and communication delays, because they shape the response below 5 Hz.
- Site setpoints
- Cycle time
- Comms delays
Collector network and transformers
Twelve inverters feed the 33 kV point of connection through cable feeders and six three-winding transformers. To keep run times practical we split the model into seven PSCAD sub-projects with PNI, then checked the impedance at the PoC had not changed.
- 6 × 8.4 MVA
- 33/0.66/0.66 kV
- D0y11y11
- 7 PNI sub-projects
Grid equivalent
A Thévenin source at the minimum short-circuit level: the weakest grid, and the worst case for SSO. Its impedance is the grid half of the Nyquist test.
- SCL 849 MVA
- X/R 21.69
- Minimum SCL
Point of connection and impedance scan
Measures P, Q, voltage, angle and frequency at the 33 kV busbar. During the scan, the voltages and currents there are transformed to give the plant and grid impedances in the dq0 frame.
- P · Q · V · f
- DFScan injection
- dq0 3×3 matrix
Four fault types, export and import.
The plant is a Type C power park module under EREC G99. Each fault lasts 140 ms with 10% retained voltage at the POI, run at +40 MW export and −40 MW import at maximum leading power factor. The same model and fault module give both the FRT and the FFCI evidence.
Three-phase
140 ms · 10% retained · ±40 MW
Phase to phase
140 ms · 10% retained · ±40 MW
Two phases to earth
140 ms · 10% retained · ±40 MW
One phase to earth
140 ms · 10% retained · ±40 MW
Phases A · B · C, schematic. Measured at the POI: P, Q, line and phase voltages, and positive-sequence active and reactive current.
The two tests NESO asks for on a Large BEGA.
- 01
Test 3.2.1 · Voltage injection
A 1% voltage ripple at each frequency from 1 to 100 Hz, up to five at a time to save run time. PoC voltage is compared with and without the plant. A ratio below 1 means the plant damps that frequency.
No growing oscillations - 02
Test 3.3 · Impedance scan
MHI’s DFScan injects 0.5% positive- and negative-sequence signals from 1 to 500 Hz and returns the full dq0 impedance of plant and grid. Their product, the minor-loop gain, is checked with the Nyquist criterion.
No encirclement of −1
Pick a test to see the result.
Mapped to the NESO IBR oscillation guidance.
For a Large BEGA the guidance requires Tests 3.2.1 and 3.3. Fault performance is evidenced separately against the GB Grid Code.
FRT and FFCI were evidenced against the GB Grid Code with the same model, so the SSO and fault results form one consistent package. The PSCAD archive and FSOUT data make it simple to extend if NESO asks for more.
- 01Met
NESO-compliant EMT software and model
PSCAD/EMTDC v5 at a fixed 10 µs step, using the same model submitted to NESO.
- 02Met
All controllers, including the PPC
Inner current loop, V/Q, active power, PLL, and the PPC with its cycle time and delays.
- 03Met
Test 3.2.1, 1 to 100 Hz
1% injections in 1 Hz steps, multi-tone and single tone, in all three modes.
- 04Met
Test 3.3, 1 to 500 Hz
DFScan, NESO’s recommended tool, at 0.5%. Plant and grid dq0 matrices saved in FSOUT.
- 05Met
PNI partitioning with no change at the PoC
Seven sub-projects for run time, with an impedance reconciliation showing zero net change.
- 06Met
Export, minimum export and import
+40 MW, 0 MW and −40 MW, in voltage-control mode at minimum short-circuit level.
- 07Met
Nyquist stability shown
None of the three eigenloci encircle −1 in any mode, including a 0 to 3 Hz check.
What the client received.
PSCAD project archive
The runnable model and project files behind every study.
Scenario matrix and assumptions
Every case, mode and condition run, and the assumptions behind them.
FRT and FFCI results
Ride-through and reactive-current results for each fault case.
Impedance scan data
Plant and grid dq0 datasets in FSOUT, with Bode and Nyquist plots.
Compliance report
Results mapped, case by case, to the NESO requirements.
Review support and handover
Answers to NESO or DNO queries, and notes so the model can be re-run later.
Connecting an inverter-based plant in Great Britain?
Send us the scheme rating, connection voltage, plant type, deadline and the model you have, and we will scope the work.